Most surface antigens (e.g., HER2, EGFR) are shared with healthy tissues, leading to dose-limiting “on-target, off-tumor” toxicities (18). Furthermore, the vast majority of oncogenic drivers (e.g., p53, RAS) are intracellular and inaccessible to conventional antibodies (19, 20). To address the “undruggable” nature of these intracellular oncoproteins, the field has advanced toward targeting pMHC complexes (Figures 1B, C). This has bifurcated into two distinct structural approaches. (i) TCR Mimic (TCRm) Antibodies (Figure 1B). Researchers have utilized phage display and hybridoma technologies to engineer antibodies that mimic the specificity of TCRs (21). For instance, Li et al. highlighted the utility of chimeric HLA tetramers to refine the specificity of these TCRm antibodies, reducing cross-reactivity against the MHC backbone (22). Recent work by Hsiue et al. and Douglass et al. also demonstrated the generation of high-affinity “MANAbodies” (Mutation-Associated NeoAntigen antibodies) in a single-chain diabody (scDb) format (23, 24). These bispecific antibodies can distinguish single amino acid changes in mutant p53 (R175H) or RAS neoantigens presented on HLA, inducing potent T cell killing even at extremely low antigen densities (<10 copies/cell). Furthermore, many researchers utilized de novo generative models (e.g., RFdiffusion) to design “minibinders” against specific pMHC targets with high precision, bypassing the limitations of animal immunization or library screening (25–27). (ii) Immune-Mobilizing Monoclonal TCRs Against Cancer (ImmTAC) (Figure 1C). An alternative strategy employs affinity-matured soluble TCRs fused to anti-CD3 effectors (28). The landmark approval of tebentafusp for metastatic uveal melanoma, as detailed by Nathan et al., validated this class. Tebentafusp targets the gp100 peptide presented by HLA-A*02:01 with picomolar affinity, significantly exceeding the affinity of natural TCRs (29). This ultra-high affinity allows for the recruitment of T cells to tumors with low antigen expression, though it necessitates rigorous safety profiling to prevent cross-reactivity against healthy tissues.

Despite the potency of TCEs, the therapeutic window remains a critical bottleneck due to systemic toxicity. To decouple antitumor efficacy from systemic side effects, “masked” TCEs (also named probodies) have been developed (Figure 1D). As described by Boustany et al. and Cattaruzza et al., these masking peptides link to the antibody domains via protease-cleavable linkers (30, 31). In the circulation and healthy tissues, the mask prevents binding; however, upon entering the TME, dysregulated proteases cleave the linker, unmasking the TCE and restricting T cell activation specifically to the tumor site. This probody strategy has enabled the safe targeting of ubiquitous antigens like HER2 and EGFR in preclinical models with significantly widened therapeutic index (32).

Canonical TCEs rely solely on “Signal 1” (TCR/CD3 engagement), which in the absence of “Signal 2” (co-stimulation), can lead to the T cell anergy or activation-induced cell death (AICD). To recapitulate a physiological immunological synapse, novel trispecific antibodies have been engineered to engage a tumor antigen alongside both CD3 and the co-stimulatory receptor CD28 (Figure 1E). As demonstrated by Wu et al. and Seung et al., this “dual-signaling” approach significantly enhances T cell proliferation through co-stimulation. Crucially, these constructs lower the threshold for T cell activation, enabling the elimination of tumors with low antigen density that are typically resistant to standard TCEs (33, 34).

To mitigate tumor immune escape caused by antigenic drift and to enhance tumor specificity, the field has advanced toward dual-targeting strategies (Figure 1F). This approach employs TCEs designed to bind two epitopes on the same target or two distinct tumor-associated antigens (TAAs). Bacac et al. developed a bivalent Fab domain to a single carcinoembryonic antigen (CEA) and a single Fab domain to CD3, increasing avidity and facilitating the selective lysis of high-expressing tumor cells (35). An alternative approach utilizes the IgM-based bispecific antibody. Baliga et al. introduced a high avidity of IgM-based CD20×CD3 bispecific antibody (IGM-2323) to enhance T cell dependent killing with minimal cytokine release (36). Although dual-targeting engagers binding the same target effectively increase the avidity to tumor cells, it may also increase binding to healthy tissues that have low levels of the target antigen and cause toxicity. To address this, Carretero-Iglesia et al. described ISB 2001, a trispecific antibody targeting BCMA and CD38; and Roskopf et al. described 33-3–19 targeting CD19⁺CD33⁺ leukemia cells (37, 38). These molecules leverage “super-avidity” to induce cytotoxicity against tumor cells with low antigen density using two distinct TAAs that escape mono-targeting agents. While Shen et al. and Tapia-Galisteo et al. illustrated this using logic-gated “AND” designs (TriTCEs or TriTEs) (39, 40). These logic-gated TCEs require simultaneous binding to both antigens to trigger robust T cell activation, thereby sparing healthy tissues that express only one of the antigens. This dual-targeting modality is essential for overcoming intratumoral heterogeneity and preventing relapse due to single-antigen loss.

The first generation, illustrated in Figure 2A, focuses on the robust engagement of FcγRIIIA (CD16A). A paradigmatic example is AFM13, a tetravalent bispecific antibody targeting CD30 and CD16A which has been tested in phase II clinical trial (44). These molecules, often termed Innate Cell Engagers (ICEs), recruit NK cells (and macrophages) to the tumor site. By binding CD16A with high affinity, they trigger the phosphorylation of ITAMs that leads to degranulation and lysis. In addition to AFM13, AFM24 targeting EGFR and targeting CD123 have also been tested in clinical trials. While effective, relying solely on CD16A has vulnerabilities (45). CD16A is prone to proteolytic shedding by ADAM17 upon activation and is frequently downregulated in the immunosuppressive TME, potentially dampening efficacy over time (46).

To overcome the instability of CD16A expression, the second generation incorporates a second activating receptor, such as NKG2D, NKp46 or NKp30 (Figure 2B). For instance, NKG2D and CD16A dual activating NKCEs targeting HER2⁺, EGFR⁺, CD33⁺ or BCMA⁺ tumor cells have shown greater activation efficacy and entered clinical trials (47). Nevertheless, persistent engagement of NKG2D by its ligands induces receptor internalization and subsequent NK cell hypo responsiveness (48). Furthermore, the therapeutic potency of these bispecific antibodies is often compromised by the presence of soluble NKG2D ligands released by tumors via proteolytic shedding or exosomal secretion, which act as decoys to intercept the therapeutic agents (49). Gauthier et al. detailed a “trifunctional” NKCE co-engaging NKp46 and CD16A (50). NKp46 is a natural cytotoxicity receptor (NCR) that is constitutively expressed and, crucially, retained on tumor-infiltrating NK cells even when other receptors like NKG2D are downregulated. Expanding the repertoire of dual-targets, Pekar et al. developed a dual activating NKCE targeting NKp30 and CD16A, which shows synergistic cytotoxicity (51, 52). Although dual activating NKCEs induce synergistic activation, leading to full secretory and cytotoxic competence superior to single-receptor engagement, they do not fully address the issue of NK cell persistence in the absence of exogenous cytokines. In clinical, NK cells in patients are often short-lived or functionally exhausted due to cytokine deprivation in the TME.

Addressing the need for survival signals, the third generation integrates stimulatory cytokines, such as IL-15 or IL-2 variant (Figures 2C, D). Cytokine-armored engagers with single activating receptor (Figure 2C) are exemplified by 16-15-B7-H3 and 16-15–33 GTB-3550, incorporate an IL-15 moiety between the anti-CD16A and anti-TAA domains (53, 54). Mechanistically, this provides a “survival signal” directly at the immune synapse. The most advanced architecture (Figure 2D) represents a convergence of previous strategies, targeting dual activating receptors (CD16A⁺ NKp46) while providing an optimized cytokine signal (IL-2 variant) (55). These molecules utilize an IL-2 variant (IL-2v) engineered to lose binding affinity for the high-affinity IL-2Rα (CD25) subunit. By co-engaging NKp46 and CD16A, they ensure potent tumor lysis via synergistic signaling. Simultaneously, the IL-2v provides proliferation signals to effector NK cells (which express IL-2Rβγ) but strictly avoids stimulating CD25⁺ Regulatory T cells (Tregs). This mechanism selectively boosts the effector function while preventing the activation of immuno-suppressive populations.

The earliest iterations of T-DC co-engagers are designed to bypass the reliance on endogenous ligands for co-stimulation. Houtenbos et al. demonstrated a novel bispecific diabody targeting CD40 on DCs and CD28 on T cells (Figure 3A) (68). The mechanism here is twofold: agonistic binding to CD40 induces DC maturation and upregulation of B7 molecules, while simultaneous binding to CD28 provides the essential costimulatory signal to T cells. This approach is particularly relevant in acute myeloid leukemia (AML), where leukemic DCs often lack sufficient costimulatory capacity. Building upon this logic, Muik et al. described the DuoBody-CD40×4-1BB, an Fc-inert bispecific antibody (Figure 3B) (69). The primary mechanism of action here is conditional agonism, which addresses a critical safety concern associated with monospecific agonists. The stimulation of 4-1BB on T cells and CD40 on APCs occurs only when the bispecific antibody effectively cross-links both cells. This mutual dependency ensures that immune activation is restricted to sites where T cells and APCs interact, thereby widening the therapeutic window and enhancing the priming of tumor-specific CD8⁺ T cells.

While co-stimulation is vital, the TME requires the simultaneous alleviation of inhibitory signals. A pivotal finding by Liu et al. reshaped our understanding of PDL1 targeting (70). They demonstrated that the therapeutic efficacy of PDL1×CD3 bispecific antibodies (Figure 3C) stems not from bridging T cells to tumor cells, but from bridging T cells to PDL1⁺ DCs. Mechanistically, this interaction blocks the PD1-PDL1 inhibitory axis while simultaneously facilitating CD28 co-stimulation via the B7 molecules expressed on the DC surface. This “cis-interaction” on the DC surface leads to the rejuvenation of exhausted CD8⁺ T cells, inducing a durable memory response that conventional T-Tumor engagers fail to generate. Similarly, Sung et al. introduced ABL501, a bispecific antibody targeting LAG3 and PDL1 (Figure 3D) (71). This format addresses resistance mechanisms where T cells upregulate alternative checkpoints like LAG3. By simultaneously blocking PDL1 on DCs and LAG3 on T cells, ABL501 enhances T cell activation and, crucially, mitigates the suppressive effects of Tregs, which also express LAG3. These approaches represent a sophisticated integration of checkpoint blockade with physical cell-cell bridging. While the above strategies target DCs broadly, Shapir Itai et al. refined this approach with the Bispecific DC-T Cell Engager (BiCE), which pairs PD-1 targeting on T cells with CLEC9A, a marker exclusive to conventional type 1 dendritic cells (cDC1s) (Figure 3E) (72). The primary mechanism here is high-precision spatial orchestration: BiCE physically facilitates the interaction between cDC1s (critical for cross-presentation) and PD1⁺ T cells in draining lymph nodes and the TME This targeted engagement specifically promotes the expansion of progenitor exhausted T cells (T_PEX), a stem-like population essential for sustained antitumor immunity, which identified as the pivotal cellular targets governing the efficacy of immune checkpoint blockade. Despite these advancements, ensuring that T-DC interactions occur specifically in the vicinity of the tumor remains a challenge.

To achieve T-DC co-engagement within TME, Lin et al. introduced a breakthrough with Multimodal Targeting Chimeras (Multi-TACs, Figure 3F), facilitated by a triple orthogonal linker (T-Linker) chemical conjugation technology (15). The EGFR-CD3-PDL1 Multi-TAC moves beyond bispecific to tri-specific integration, forms a synthetic trimeric cellular complex (Tumor-T-DC) with targeting EGFR on tumor cells for localization, CD3 on T cells for engagement and activation, and PDL1 on DCs for engagement and checkpoint blockade. This spatial orchestration ensures that T cell priming by DCs happens in the immediate presence of tumor antigens, leading to robust cytotoxicity and minimizing off-target effects. The modularity of Multi-TACs allows for the incorporation of diverse immunomodulators, creating a programmable system for “immune management” within the TME.

T cells are potent effectors of adaptive immunity, their efficacy is often limited by tumor heterogeneity, specifically the downregulation of MHC-I molecules. While NK cells, acting as the “first responders” of innate immunity, operate independently of MHC restriction and can eliminate MHC-deficient tumor cells via “missing-self” recognition (73). By simultaneously harnessing these two cell types, co-engagers integrate the power of both T cells in adaptive immunity and NK cells in innate immunity, creating a synergistic attack that prevents tumor immune escape.

For T-NK cell co-engagers, Ye et al. utilized an albumin/polyester composite nanoparticle (APCN) to construct a Nanoparticle-based Tri-specific Nano-Antibody (Tri-NAb) (74). The nanoparticle surface is coated with anti-Fc antibodies, allowing for the directional immobilization of three distinct monoclonal antibodies (anti-PDL1 targeting tumor cells, anti-4-1BB targeting T cells and anti-NKG2A targeting NK cells) (Figure 4A). Mechanistically, it performs three simultaneous functions: it blocks the immune checkpoint PD-L1 on tumor cells; antagonizes the inhibitory receptor NKG2A to unleash cytotoxicity; and agonizes 4-1BB (CD137) to provide potent costimulatory signals. This spatiotemporal synchronization triggers robust activation and proliferation of both CD8⁺ T cells and NK cells, leading to synergistic tumor eradication. After that, Fan et al. optimized the nanoparticle-based platform through FcγR1-serum albumin fusion protein and hydrophobic poly(l-lactide) single-step assembly, achieving the one-step combination of three distinct monoclonal antibodies (anti-PDL1 targeting tumor cells, anti-PD1 targeting T cells and anti-NKG2A targeting NK cells) (75). Furthermore, Yu et al. adopted a yeast-based library and yeast surface display system for directed evolution of Staphylococcal enterotoxin B (SEB), then integrated a tumor-targeting nanobody (e.g., anti-Mesothelin) with an engineered high-affinity superantigen (SEB variant) and a computationally designed cytokine (Neo-2/15) through fusion expression (Figure 4B) (14). Unlike conventional bispecific antibodies that rely on anti-CD3 scFv, this construct uses a superantigen to cross-link MHC-II on APCs and specific Vβ chains on the TCR, inducing broad T cell activation. Crucially, the inclusion of the Neo-2/15 cytokine component allows the molecule to simultaneously bind and activate NK cells (which express IL-2/15R). This design promotes the expansion and survival of both T and NK lineages directly at the tumor site, overcoming the limitations of low-affinity superantigens. Additionally, Lameris et al. created a bispecific antibody based on variable domain of heavy chain of heavy chain (VHH) by fusing a CD1d-specific VHH with a Vδ2-TCR-specific VHH (Figure 4C) (76). This engager possesses trispecific properties despite being a bispecific molecule. By binding to CD1d on tumor cells, it not only recruits and activates Vγ9Vδ2 T cells but also engages Type 1 NKT cells, which recognize lipid antigens presented by CD1d. This dual recruitment effectively bridges innate-like T cell subsets and NKT cell subsets for tumor control.

Myeloid cells, particularly tumor-associated macrophages (TAMs), are another type responders of innate immunity through ADCC and cytokine secretion. However, in solid tumors, TAMs often adopt an immunosuppressive M2-like phenotype, erecting physical and metabolic barriers that exclude T cells (11). Consequently, T-myeloid cell co-engagers are clinically significant because they do not merely target malignant cells; but remodel the stromal architecture to convert “cold” immune-desert tumors into “hot” inflamed environments (77, 78).

For T-myeloid cell co-engagers, Scott et al. engineered bi-valent T cell engagers fusing CD3 scFv targeting T cells and CD206 VHH or folate receptor β scFv targeting M2-like TAMs. They further developed tri-valent T cell engagers, incorporating an additional anti-CD3 or anti-CD28 domain to enhance potency (Figure 4D) (79). These agents redirect endogenous T cells to physically engage and lyse M2-like TAMs. This selective depletion of the immunosuppressive stroma creates a pro-inflammatory window, indirectly facilitating antitumor immunity by removing the “brakes” on the TME. Von Locquenghien et al. utilized the distinct characterization of TAMs to remodel TME (80). They developed myeloid-targeted immunocytokines and natural killer (NK)/T cell enhancers (MiTEs), which targeted TREM2 (a myeloid checkpoint on TAMs) and was specifically activated by the tumor-associated protease MMP14 to release IL2 cytokines reacting with IL-2Rβγ on T and NK cells (Figure 4E). MiTEs employ a dual-action trans mechanism. First, the antibody antagonizes TREM2, reprogramming TAMs from a suppressive to a pro-inflammatory state. Second, high MMP14 levels on TAMs cleave the linker, unmasking IL-2 locally. This triggers robust proliferation and cytotoxicity in bystander CD8⁺ T cells and NK cells, achieving potent antitumor efficacy with minimal systemic toxicity.

T-Treg co-engagers function as “ratio-modulators” Their clinical significance lies in simultaneously expanding effector T cells while depleting Tregs, thereby mechanically dismantling the immunosuppressive barrier that limits T cell infiltration and function. Based on the study by Son et al., ABL112 represents a novel Fc-competent bispecific antibody designed to exploit the T-Treg axis (Figure 4F) (81). ABL112 is constructed by fusing an anti-TIGIT VHH to the N-terminus of a human IgG1 Fc region, with an anti-4-1BB scFv linked to the C-terminus. It blocks the TIGIT-CD155 interaction to restore T cell function while simultaneously inducing TIGIT-dependent clustering and activation of 4-1BB signaling on CD8⁺ T cells. Since intratumoral Tregs express significantly higher levels of 4-1BB and TIGIT compared to effectors, the Fc-competent design facilitates ADCC mediated by macrophages and NK cells, selectively eliminating Tregs.

Genetic fusion remains the most prolific approach for generating multispecific co-engagers (Figure 5A). These strategies are subdivided into two parts based on the presence or absence of an Fc region: IgG-like formats and non-IgG-like formats (82). Here introduces the representative genetic fusion strategies to construct multispecific co-engagers.

IgG-like formats provide outstanding half-life and stability, but face the primary challenge for preventing random association of heavy and light chains. Knobs-into-Holes (KiH) pioneered by Ridgway et al. engineer steric complementarity into the CH3 domains of the Fc region (83). By introducing a bulky amino acid (T366Y “knob”) into one heavy chain and a smaller residue (Y407T “hole”) into the other, heterodimerization is thermodynamically favored over homodimerization. Furthermore, scientists also use phage display technology to construct a more stable “3 + 1” pattern: with three amino acid mutations (T366S, L368A, and Y407V) form a recessed “holes” type. Duobody utilizes controlled Fab-arm exchange (cFAE). Separately expressed IgG1 molecules containing matched point mutations in the CH3 domain (F405L and K409R) are mixed under mild reducing conditions, driving the efficient recombination of half-molecules into stable bispecific heterodimers (84). To solve light chain mispairing problem, CrossMab swaps CH1 and CL domains within one arm to enforce correct light chain assembly while retaining the KiH Fc structure (85). Similar to CrossMab, Ortho-Fab utilizes orthogonal interfaces engineered via specific mutations at VH/VL or CH1/CL domains to electrostatically or sterically drive the correct pairing of light chains (86). Dual-Variable-Domain Ig (DVD-Ig) and IgG-scFv fuse the other variable domain to the N-terminus or C-terminus, avoiding complex pairing of heterodimers (87, 88).

Non-IgG-like are smaller, fragment-based formats lack an Fc region, offering superior tissue penetration but generally shorter half-lives. They could be directly fused and constructed through simple flexible linkers. BiTE is a tandem scFv format where two scFvs (typically anti-CD3 and antitumor antigen) are linked by a short flexible peptide (17). Diabody uses a short linker (too short for intramolecular pairing) to link a VH domain from one specificity with a VL domain from another on the same polypeptide chain, forcing intermolecular dimerization (89). Dual-Affinity Re-Targeting (DART) is an evolution of diabody (90). DART incorporates a C-terminal disulfide bridge between the two variable domain chains. This covalent linkage significantly enhances stability and reduces the formation of inactive homodimers compared to conventional diabodies. Tandem Diabody (TandAb) is a tetravalent, homodimeric evolution of diabody, formed by the head-to-tail association of two polypeptide chains, each containing four variable domains (91). Bi-nanobody is composed of VHHs which are ligated by flexible linkers (92). Dock-and-Lock (DNL) exploits the natural high-affinity dimerization and docking domain of protein kinase A (PKA) and the anchoring domain of A-kinase anchoring proteins (AKAP) to assemble multivalent complexes (93).

Besides genetic fusion, nanomedicines or assemblies have emerged as a versatile platform (Figure 5B). These supramolecular systems leverage the high surface area-to-volume ratio of nanoparticles to present multiple binding ligands for constructing multispecific co-engagers.

One prominent example is the multivalent bi-specific nanobioconjugate engager (mBiNE) (94). Yuan et al. leveraged carboxyl-functionalized polystyrene nanoparticles as a stable scaffold, employing EDC/NHS chemistry to covalently anchor antibodies via the formation of amide bonds between primary amines on the proteins and carboxyl groups on the nanoparticle surface. By adjusting the molar ratios of the input proteins during the conjugation reaction, the surface density and ratio of proteins can be fine-tuned to optimize avidity without altering the core nanoparticle properties. Ye et al. utilized a more biocompatible material, APCN coated with anti-Fc antibodies to construct a nanoparticle-based Tri-NAb via the directional immobilization of antibodies (74). Furthermore, this team optimized this platform through the single-step assembly of FcγR1-serum albumin fusion protein and hydrophobic poly(l-lactide), which is called the fusion protein/polymer-based nano-adaptor (FP-NA), achieving the one-step combination of antibodies (75). Guan et al. engineered the bispecific molecularly imprinted nanoimmunoblocker (bsMINIB) using molecularly imprinted polymers (MIPs) (95). This platform anchors specific N-terminal epitope peptide of proteins onto a silica nanoparticle substrate via boronate affinity. Subsequently, a silicate layer is polymerized around these templates using functional monomers (e.g., organosilanes). Upon removal of the peptide templates, artificial recognition cavities are left behind in the polymer matrix. This construction creates robust, synthetic receptors (“plastic antibodies”) with high affinity and specificity, avoiding the use of biological antibodies. Leveraging the natural complexity of cell membranes, Li et al. constructed the TCR nanovesicle antibody (TCR NV) using a top-down genetic engineering and membrane extrusion strategy (96). Jurkat T cells are genetically transduced to stably express tumor-specific TCRs alongside proteins scFvs on their plasma membrane. These engineered membranes are then harvested and mechanically extruded to form nanoscale vesicles. This approach preserves the natural topological orientation and transmembrane anchorage of the targeting proteins, while the vesicle lumen serves as a reservoir for loading small-molecule modulators, integrating biologic display with drug delivery in a single construct.

Chemical conjugations represent a highly modular platform for the construction of multiple immune cell co-engagers (97), employing synthetic chemistry to covalently assemble diverse binding moieties and functional payloads, including full-length antibodies, antigen-binding fragments, and immunomodulators (Figure 5C).

A representative example of this strategy is the synthesis of IgG-like bispecific antibodies using genetically encoded unnatural amino acids (UAA) (98). Kim et al. demonstrated the site-specific incorporation of p-acetylphenylalanine (pAcF) into the light chains of Fab fragments. These ketone-containing UAAs serve as bioorthogonal handles for conjugation with hydroxylamine-functionalized linkers bearing either an azide or a a bi-cyclo [2.1.0] nonyne (BCN) group. The two functionalized Fabs are then covalently coupled via a copper-free strain-promoted azide-alkyne cycloaddition (SPAAC) reaction. Combining enzymatic precision with synthetic chemistry, Wagner et al. employed a chemo-enzymatic approach utilizing sortase A-mediated transpeptidation to fuse two full-length anti-influenza IgG antibodies at their C-termini (99). Each antibody is first engineered with a sortase recognition motif (LPXTG) and then enzymatically labeled with either an azide or a dibenzocyclooctyne group (DBCO) group. The subsequent SPAAC reaction generates a stable, covalently linked bispecific IgG heterodimer. Expanding the functional complexity of chemical conjugates, Thoreau et al. utilized a pyridazinedione (PD) scaffold to functionally re-bridge the interchain disulfide bonds of Fab fragments while simultaneously introducing bioorthogonal handles (BCN or tetrazine) to construct IgG-like bispecific synthetic antibodies (SynAbs) (100). This team then applied this platform to assemble of a three-protein conjugate site-selectively, with the addition of the third bioorthogonal handles (azide) (101).

As an advancement in chemical conjugation strategies, a highly modular and programmable platform (Multimodal Targeting Chimera, Multi-TAC) developed by Lin et al. leverages a proprietary triple orthogonal linker (T-Linker) technology (Figure 6A) (15). This small-molecule scaffold incorporates a mono-glycine motif, an azide group, and a tetrazine group, facilitating the sequential assembly of three distinct therapeutic modules via sortase A-mediated transpeptidation, strain-promoted azide-alkyne cycloaddition (SPAAC) and inverse electron-demand Diels-Alder (IEDDA) (Figure 6B). Functionally, Multi-TACs are engineered to spatially orchestrate multiple immune pathways; thereby coordinating lymphocyte cell with myeloid cell activation with specifically within the tumor microenvironment (Figure 6C). Nevertheless, the practical application of this platform faces certain limitations. The reliance on multistep orthogonal reactions necessitates intricate chemical synthesis and purification processes. Furthermore, the in vivo pharmacokinetics of these synthetic constructs have not been exhaustively characterized. Despite these challenges, Multi-TACs offer a modular and programmable strategy for the rational design of multispecific agents, representing a promising avenue for future immunotherapeutic development.