Engineering nano‐clustered multivalent agonists to cross‐link TNF receptors for cancer therapy

Tumor necrosis factor receptors (TNFRs) are promising targets for cancer therapy. However, activating their downstream signaling requires cross‐linking of TNFRs. Herein, to devise strong agonists of TNFRs, ligands targeting TNFRs, such as OX40L and tumor necrosis factor‐related apoptosis‐inducing ligand (TRAIL), were fused with a multivalent protein scaffold (MV) to prepare multivalent agonists for cross‐linking TNFRs. The nano‐clustered multivalent‐OX40L (MV‐OX40L) and MV‐TRAIL could promote T cell activation and directly induce tumor cell apoptosis. Moreover, to develop a universal nano‐adaptor for the rapid preparation of multivalent agonists of different TNFRs, the Fc receptor that could immobilize antibodies was fused with MV to prepare MV‐FcR, which could multimerize commercial agonist antibodies targeting TNFRs, such as anti‐OX40 antibody (αOX40). Simply incubating αOX40 with MV‐FcR could prepare MV‐αOX40 to enhance its antitumor efficacy. In addition, MV‐FcR could multimerize with other therapeutic antibodies, such as anti‐PD‐L1 antibody, to enhance their valency. This study provides a promising strategy for engineering multivalent antitumor protein drugs.


INTRODUCTION
Tumor necrosis factor receptors (TNFRs) are involved in the signaling pathways of cell survival, differentiation, inflammatory response, and apoptosis.They are closely relevant to tumorigenesis and are important targets for cancer therapy, such as tumor necrosis factor receptor superfamily member 4 (TNFRSF4, also known as OX40) and TNFRSF10B (also known as death receptor 5, DR5). [1,2][8] Numerous studies on developing ligands or antibodies targeting TNFRs have been conducted for cancer therapy.However, some candidates have failed to show therapeutic benefits in clinical trials, which was because TNFRs required to be cross-linked to trigger effective activation. [9]Multimerized OX40 ligand (OX40L), but not its monomer, can induce the aggregation of OX40, resulting in T cell activation and enhancing the cytotoxicity of T cells. [10]Moreover, the therapeutic effects of antibodies targeting OX40 can be further enhanced by multimerization. [11]imilarly, the multimerized DR5 ligand tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) or antibodies induce cross-linking of DR5 strongly activate the downstream apoptosis signal pathway. [12]urrent strategies for multimerizing ligands or antibodies to induce TNFR cross-linking are mainly achieved by tandem expression of ligands or antibody derivatives.The tetravalent DR5 agonist antibody shows stronger downstream signaling activation and induces tumor-specific cell death. [13]Alternatively, modifying the Fc fragments of antibodies for Fc receptor-mediated conjugation achieves enhanced agonism and effector functions of TNFRs. [14,15][18][19] Nanoplatforms have also been reported to prepare multimerized antibodies using Fcspecific anti-immunoglobulin G (IgG) antibody-conjugated polystyrene nanoparticles. [20,21]n this study, we engineered an innovative series of multimeric protein drugs based on a multivalent protein scaffold (MV) for cross-linking TNFRs to enhance the antitumor response of T cells or induce tumor cell apoptosis (Scheme 1).MV was constructed by engineering IgG Fc to imitate immunoglobulin M (IgM) multimerization.Briefly, IgG Fc was engineered by adding the tailpiece of IgM to the C-terminal for hexamerization and introducing a cysteine mutation for inner disulfide bond formation. [22,23]hen, multivalent-OX40L (MV-OX40L) targeting OX40 was developed by adding OX40L to the N-terminal of the MV monomer for promoting T cell activation.MV-TRAIL was developed to induce tumor cell apoptosis.In addition, a universal antibody binding adaptor MV-FcR, which could bind the Fc fragment of IgG antibodies to multimerize commercial agonist antibodies, was devised by fusing FcR to the Nterminal of the MV monomer.The prepared MV-FcR could effectively bind anti-OX40 antibodies (αOX40) to induce multimerization and form MV-αOX40, which achieved more effective activation of T cells.Furthermore, MV-FcR was applied to the multimerization of antibody targeting programmed death-ligand 1 (PD-L1) to prepare MV-αPD-L1, which effectively relieved T cell inhibition and enhanced cancer therapy.In summary, this study provides a promising strategy for preparing multivalent or multimeric protein drugs to enhance cancer therapy.

Preparation and characterization of multivalent protein scaffold (MV)
The MV monomer was engineered from the Fc fragment of IgG.[24][25][26] Specifically, the IgM tailpiece (TP) sequence was fused to the C-terminal of IgG Fc, and residues L309 and H310 of IgG Fc were replaced with C and L, respectively, to produce mut IgG Fc-TP.The mut IgG Fc-TP sequence was cloned into the plasmid under the control of the elongation factor 1 alpha (EF1α) promoter for MV expression (Figure S1).The engineered IgG Fc fragment was composed of a homodimer of mut IgG Fc-TP, and six engineered IgG Fc fragments comprised MV (Figure 1A).For MV production, CHO-K1 cells were transduced with lentivirus containing the mut IgG Fc-TP sequence.Similar to IgM production, mut IgG Fc-TP was assembled into homologous dimerized monomers and further hexamerize to form MV in cells.MV was secreted from cells controlled by a signal sequence.MV was purified from the cell culture supernatant using Protein G agarose, which had a high affinity for the IgG Fc region in the MV monomer.
We then characterized the properties of MV.The diameter of the MV was 50.4 nm, and the zeta potential was −4.8 mV (Figure 1B,C).Transmission electron microscope (TEM) revealed that the MV exhibited a hexagonal snowflakelike structure, with the boundaries of each MV carefully delineated for enhanced clarity (Figure 1D).To investigate the composition and molecular weight of the MV protein, MV treated with or without β-mercaptoethanol (β-ME) was separated using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) followed by Western blot-ting (WB) using the anti-IgG antibody to identify the MV monomer or MV.With β-ME treatment, the disulfide bonds between and within the MV monomers would be disrupted and broken up into smaller components mut IgG Fc-TP with a calculated molecular weight of 28.6 kDa.Without β-ME treatment, MV would remain intact with a predicted molecular weight of 343.2 kDa.As shown in the WB results in Figure 1E,F, MV without β-ME treatment showed a single band at approximately 300 kDa, which was consistent with the predicted molecular weight of intact MV.MV treated with β-ME displayed a single band at approximately 30 kDa, indicating that MV was composed of mut IgG Fc-TP.These results suggest that MV comprised multiple monomers and corresponded to our design.
After proving that MV was successfully constructed in the above experiments, we explored whether MV was effectively enriched in tumors.Nanoscale drugs can passively accumulate in the tumor region through the enhanced permeability and retention effect. [27]To examine the biodistribution of MV after systematic administration, Cy5-labeled MV (MV Cy5 ) was intraperitoneally injected into mice bearing CT26 colorectal cancer.A significant Cy5 fluorescence signal was observed in tumors 12 h after administration (Figure 1G).The quantitative analysis of the Cy5 intensity from different organs showed that MV Cy5 mainly accumulated in tumors (Figure 1H,I).These results suggested that MV was enriched in tumors after circulation and had the potential to be used for tumor treatment.Thus, MV was used to prepare multivalent or multimeric protein drugs for cancer therapy.

MV-OX40L activates T cells for cancer therapy
TNFR cross-linking is important for the effective activation of the TNFR signaling pathway, including OX40L-OX40.The interaction between OX40L and OX40 can activate the classical NF-κB1 pathway or the nonclassical NF-κB2 pathway, PI3K/PKB, and NFAT pathway, inducing proliferation of T cells and enhancement of cytokine expression. [28,29]X40L-OX40 interaction can also reduce the expression of CTLA-4 and FOXP3 to enhance the activation of effector T cells and inhibit the function of regulatory T cells. [30]To effectively cross-link OX40, OX40L was added to the Nterminal of mut IgG Fc-TP, which contained the monomer to construct MV-OX40L (Figure 2A).The MV-OX40L was designed to directly improve T cell functions (Figure 2B).We subsequently used size-exclusion chromatography (SEC) but not SDS-PAGE to measure because of the high molecular weight of MV-OX40L, which exhibited close to the predicted value of 613.2 kDa (Figure 2C).We characterized the properties of MV-OX40L.The diameter of the MV-OX40L was 60.1 nm, and its zeta potential was −4.7 mV (Figure 2D,E).
We then investigated whether MV-OX40L could effectively bind to T cells.Activated T cells were incubated with Cy5-labeled MV-OX40L (MV-OX40L Cy5 ) for 2 h.Confocal laser scan microscopy (CLSM) revealed the localization of MV-OX40L on the surface of T cells following treatment (Figure 2F).Furthermore, flow cytometric analysis demonstrated that 92.5% of T cells exhibited binding with MV-OX40L, significantly higher compared to the negative control and free OX40L groups (Figure 2G).
We next examined whether the effective binding of MV-OX40L could promote T cell proliferation and activation.Anti-CD3 antibody-stimulated T cells were incubated with MV-OX40L and demonstrated an increase in the expression of the proliferation molecule Ki-67 in CD3 + T cells from 10.9% to 72.0% after treatment (Figure 2H).The expression of the T cell activation marker CD69 was also upregulated (Figure 2I).Moreover, expression of interleukin-2 (IL-2) and interferon gamma (IFN-γ) mRNA in MV-OX40L-treated T cells was increased by 5.8-and 2.1-fold, respectively, compared with that in control T cells (Figure 2J,K).These results demonstrated that MV-OX40L could effectively bind to T cells and induce their efficient proliferation and activation.
Encouraged by the in vitro activation of T cells via MV-OX40L, we proceeded to assess the in vivo fate and antitumor efficacy of MV-OX40L.Flow cytometric analysis revealed that MV-OX40L efficiently bound to T cells in the bloodstream following intraperitoneal injection (Figure 2L).Furthermore, we observed significant accumulation of MV-OX40L in tumor tissues after circulation, while minor accumulation was detected in the kidney, liver, and lung (Figure 2M and Figure S2).After three intraperitoneal injections of MV-OX40L, the growth of CT26 colorectal tumors was significantly inhibited compared with that of the phosphate buffer saline (PBS) control group (Figure 2N).In addition, 50% of mice treated with MV-OX40L sur-vived > 102 days, whereas none of the mice in the PBS group survived > 72 days (Figure 2O).These results suggested that MV-OX40L could effectively activate T cells to inhibit tumor growth and prolong survival in colorectal tumor models.

MV-TRAIL triggers tumor cell apoptosis for cancer therapy
DR5 is also a cell surface receptor of TNFR that interacts with its ligand tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) and induces apoptosis.DR5 is highly expressed in different cells in tumors, including tumor, tumor endothelial, and immunosuppressive cells, but shows low expression in normal organs and tissues, allowing its ligand TRAIL to selectively kill tumor cells without damaging normal tissue cells. [7,31]Similar to OX40, DR5 needs to be aggregated to trigger downstream signaling.Therefore, we constructed an MV-based multivalent TRAIL protein drug to directly trigger tumor cell apoptosis.To enhance the function of TRAIL, trimeric TRAIL was fused to the N-terminal of mut IgG Fc-TP to construct MV-TRAIL (Figure 3A).The MV-TRAIL was devised to directly induce tumor cell apoptosis (Figure 3B).We also used SEC to measure the molecular weight range of MV-TRAIL.However, due to the huge molecular weight of 1566 kDa of MV-TRAIL, it showed much larger than the available commercial protein marker of 670 kDa.And, there was only one main peak was observed, which was consistent with the expected result (Figure 3C).We characterized the properties of MV-TRAIL.The diameter of the MV-TRAIL was 67.2 nm, and its zeta potential was −0.9 mV (Figure 3D,E).
To evaluate whether MV-TRAIL could effectively induce tumor cell apoptosis, Annexin V and propidium iodide were used for apoptosis detection.4T1 breast tumor cells treated with MV-TRAIL displayed apoptosis after 2 h of incubation (Figure 3F).The apoptosis process was also confirmed by continuous observation using CLSM.Annexin V and propidium iodide were added to the 4T1 tumor cells treated with MV-TRAIL to indicate the apoptosis status, and 4T1 tumor cells exhibited a gradual apoptotic process from 60 to 360 min (Figure 3G).In addition, tumor cells were labeled with both Annexin V and propidium iodide at 360 min, which suggested that tumor cells were necrotic at this time.
Furthermore, we detected variations in intracellular protein levels related to the apoptosis signal pathway after MV-TRAIL treatment.As shown in Figure 3H, MV-TRAIL induced cleavage of caspase-8, Bid, and PARP, which could activate the extrinsic apoptotic pathway.Meanwhile, MV-TRAIL triggered the activation and phosphorylation of c-Jun N-terminal kinases (JNK), which could trigger the intrinsic pathway of apoptosis. [32,33]These results indicated that MV-TRAIL could effectively kill tumor cells by inducing apoptosis.
We then investigated whether MV-TRAIL could be used for cancer therapy.Flow cytometric analysis revealed that MV-TRAIL treatment could effectively induce tumor cell apoptosis (33.4%) after intraperitoneal injection (Figure 3I).Tumor growth was significantly inhibited after three intraperitoneal injections of MV-TRAIL in mice bearing 4T1 breast cancer (Figure 3J), and the survival time of mouse models was significantly prolonged.In the MV-TRAILtreated group, 50% of mice survived >40 days, whereas mice in the PBS group did not survive >36 days (Figure 3K).These results demonstrated the potential of MV-TRAIL for cancer therapy.

MV-FcR induces multimerization of therapeutic antibodies to enhance antibody valency and efficacy
Inspired by the MV-based multimerization of protein drugs, we developed a universal nano-adaptor to bind and multimerize commercial therapeutic antibodies by engineering MV for enhancing cancer immunotherapy.
Given that most therapeutic antibodies are IgG, we added the Fc receptor FcγRI (FcR) that could bind IgG Fc to the N-terminal of mut IgG Fc-TP to engineer MV for antibody multimerization. [34]The engineered MV with FcR for IgG binding was denoted as MV-FcR (Figure 4A).Incubating therapeutic antibodies with MV-FcR could multimerize and enhance the valency of the antibodies (Figure 4B).The calculated molecular weight of MV-FcR was 825.6 kDa, which was also larger than the protein markers (670 kDa) and was confirmed by SEC analysis (Figure 4C).
We then investigated whether MV-FcR could bind therapeutic antibodies such as αOX40.MV-FcR was incubated with αOX40, and the antibody-binding ability of MV-FcR was calculated by measuring MV-FcR binding on the plate.The dissociation constant K D value was 32.93 nM (Figure 4D), indicating that MV-FcR could effectively bind therapeutic antibodies to induce their multimerization.We performed characterization of the diameter and zeta potential for MV-FcR, MV-αOX40, and MV-αPD-L1.The diameter of MV-FcR increased from 63.7 nm to 97.6 nm and 98.1 nm after incubation with αOX40 and αPD-L1, respectively (Figure S3).
Next, we examined whether the αOX40 multimerization mediated by MV-FcR could enhance the antigen binding ability of αOX40.The potential OX40 antigen binding abilities of αOX40 and MV-αOX40 were detected using enzymelinked immunosorbent assay (ELISA).The variation of the antigen binding ability after MV-FcR-mediated multimerization was analyzed by detecting the amount of effectively bound αOX40.MV-αOX40 has a higher valency and potential to bind more OX40 antigens (Figure 4E).These results demonstrated that MV-FcR could effectively bind αOX40 to construct MV-αOX40 with enhanced valency.
We then investigated whether the constructed MV-αOX40 could promote T cell function.MV-αOX40 could effectively bind to the surface of T cells after incubation (Figure 4F).Furthermore, MV-αOX40 treatment increased the expression of proliferation molecule Ki-67 and activation marker CD69 in CD3 + T cells from 56.7% to 74.1% and 50.5% to 56.3%, respectively, compared with that in the αOX40 treatment (Figure 4G,H).In addition, the expression of IL-2 and IFN-γ was also upregulated (Figure 4I,J).

MV-αOX40 enhances cancer therapy by promoting T cell functions
Having confirmed that MV-αOX40 could effectively bind and activate T cells, we investigated the therapeutic effi-ciency of MV-αOX40 for cancer therapy.Mice bearing CT26 colorectal cancer were intraperitoneally injected with MV-αOX40 to evaluate the therapeutic efficacy (Figure 5A).After three injections, the function of T cells in tumors was measured using flow cytometry, including the expression of antitumor functional cytokines IFN-γ, Perforin, and Granzyme B (GranB) as well as that of cytokines IL-2 and IL-4, which maintain the proliferation and survival of T cells.The expression of IFN-γ, Perforin, GranB, IL-2, and IL-4 significantly increased after MV-αOX40 treatment (Figure 5B), suggesting that MV-αOX40 could enhance T cell activation and improve T cell function in vivo.The in vivo biodistribution data demonstrated that MV-αOX40 effectively accumulated in tumor tissues following intraperitoneal injection (Figure 5C,D).Meanwhile, the tumor growth of mice bearing CT26 colorectal cancer was inhibited after treatment with MV-αOX40 (Figure 5E).We also noticed that αOX40 alone could inhibit tumor growth compared with that in the PBS control group, but the MV-αOX40 treatment could significantly prolong mouse survival time compared with that following αOX40 treatment (Figure 5F).These results suggested that MV-αOX40 could effectively inhibit tumor growth and prolong survival by activating T cells and enhancing T cell functions.

MV-αPD-L1 enhances cancer therapy by relieving the T cell inhibition
We then investigated whether MV-FcR could induce multimerization of other antitumor therapeutic antibodies.Several anti-PD-L1 antibodies (αPD-L1) that can block PD-1/PD-L1 signal pathways to enhance T cell responses have been approved for cancer therapy, including that of skin, bladder, and breast tumors, and consequently, we applied MV-FcR to promote the efficacy of αPD-L1 in cancer therapy.

CONCLUSIONS
In summary, we engineered a series of MV-based nanoclustered agonist protein drugs, including MV-OX40L and MV-TRAIL, to cross-link TNFRs.We proved that MV-OX40L and MV-TRAIL could efficiently promote T cell activation and induce tumor cell apoptosis.In addition, we further devised a universal nano-adaptor MV-FcR that could cross-link TNFR targets by simply incubating and binding with commercial agonist antibodies.The results showed that MV-αOX40 prepared by incubating MV-FcR with αOX40 displayed enhanced antitumor efficacy.MV-FcR could additionally be applied to multimerize other therapeutic antibodies to enhance their valence.Overall, this study describes a strategy for cross-linking TNFR targets to enhance cancer therapy and provides an approach for the development of multivalent or multimeric protein drugs.

Preparation and characterization of MV
MV was prepared by engineering the IgG Fc to contain six engineered IgG Fc. [22][23][24][25][26] IgG Fc was engineered by introducing a tailpiece derived from IgM for hexamerization and L309C and H310L mutations to stabilize the hexameric structure.The sequence was denoted mut IgG Fc-TP and was cloned into the lentiviral plasmid pCDH-EF1 (Addgene, 72266).The constructed pCDH-EF1-mut IgG Fc-TP plasmid was used to transfect 293T cells together with psPAX2 (Addgene, 12260) and pMD2.G (Addgene, 12259) plasmids for lentivirus packaging.Then, lentiviruses containing the mut IgG Fc-TP sequence were transduced into CHO-K1 cells to prepare stable transfection cell lines for MV production.MV proteins were purified from the culture supernatants of CHO-K1 stable transfection cells using Protein G agarose.The size distribution and zeta potential of MV were characterized using Nano ZSE system (Malvern, UK).The morphology of MV was detected using TEM (Ther-moFisher, USA).The composition and molecular weight of MV was analyzed by WB using anti-IgG antibody (Abcam, UK).The protein transfer durations of MV treated without or with the reducing agent β-ME were 30 and 200 min, respectively.

4.4
Preparation and characterization of MV-OX40L, MV-TRAIL, and MV-FcR OX40L, TRAIL, or FcR sequences were inserted at the Nterminal of the mut IgG Fc sequence in the MV expression plasmid.The MV-OX40L, MV-TRAIL, and MV-FcR stable expression cell lines were constructed via lentivirus transduction, respectively.Cell culture supernatants of these stable expression cell lines were collected for preparing multivalent protein drugs.These multivalent protein drugs were purified via affinity chromatography using Protein G agarose.The molecular weight range was detected using SEC.
For flow cytometry analysis, these T cells were stained with anti-CD3 antibody and were analyzed with FACSCelesta (BD Biosciences, USA).
To evaluate the efficiency of MV-OX40L binding to T cells in vivo, 100 μg MV-OX40L was intraperitoneally injected into mice bearing CT26 tumors.Blood samples were collected 12 h after injection.Then, the blood samples were stained with anti-CD45 and anti-CD3 antibodies before being analyzed with FACSCelesta.

T cell proliferation and activation
Purified mouse T cells were cultured with 2 μg/mL anti-CD3 agonist antibody and 10 μg/mL MV-OX40L or MV-αOX40 for 72 h.For flow cytometry analysis, these T cells were stained with anti-CD45, anti-CD3, and anti-CD69 antibodies to label the cell surface markers and treated with the cell nuclear membrane penetration kit for labeling with the anti-Ki-67 antibody.Then, these T cells were analyzed with FACSCelesta.For quantitative real-time polymerase chain reaction (qRT-PCR) analysis, these T cells were further treated with RNAiso Plus (TaKaRa, Japan) to extract the total RNA.Reverse transcription was conducted using the Prime-Script RT reagent Kit with gDNA Eraser (TaKaRa, Japan).qRT-PCR analysis was performed using FastStart Universal SYBR Green Master Mix (Roche, Switzerland).

Tumor cell apoptosis
To evaluate the MV-TRAIL efficacy to induce tumor cell apoptosis, 2 × 10 5 4T1 cells were cultured for 12 h and then treated with 100 ng MV-TRAIL.For flow cytometry analysis, 4T1 tumor cells were collected 2 h after treatment, stained with Annexin V and propidium iodide, and then analyzed with FACSCelesta.For WB analysis, 4T1 tumor cells were collected after 12 h and lysed with RIPA buffer.After SDS-PAGE electrophoresis, the samples were incubated with the primary antibody (anti-Caspase 8, anti-Bid, anti-PARP, anti-p-JNK, anti-JNK, or anti-β-tubulin).Then, samples were incubated with horseradish peroxidase (HRP)labeled second antibodies and HRP substrates were used for chemiluminescent imaging.
To detect the apoptotic cells within tumor tissues, 20 mg/kg MV-TRAIL was intraperitoneally injected into mice bearing 4T1 tumors every other day for a total of three injections.The tumor tissues were collected after treatment and dissociated into single-cell suspensions.The apoptotic cells were stained with Annexin V and propidium iodide, and analyzed with FACSCelesta.

MV-FcR binding therapeutic antibodies
The OX40 antigen was dissolved and diluted to 5 μg/mL using PBS.A volume of 50 μL (5 μg/mL) OX40 was added to the MaxiSorp flat-bottom plate (Nunc, Denmark) per well to coat the plate at 4 • C for 12 h.After washing the plate twice with PBS, 200 μL 5% BSA was added to the Max-iSorp flat-bottom plate per well for 2 h to block the remaining protein-binding sites.For detecting the capability of MV-FcR binding therapeutic antibodies, 50 μL 20 μg/mL αOX40 was added per well for 2 h.After washing with PBS, MV-FcR with different concentrations was added and MV-FcR binding on the plate was subsequently measured.To detect the enhancement of valency via MV-FcR-induced therapeutic antibody aggregation, MV-FcR and αOX40 were mixed 1:1 in mass ratio at 4 • C for 2 h to obtain MV-αOX40.Then, αOX40 or MV-αOX40 with different concentrations was added to the plate for 2 h.The potential antigen binding capabilities of αOX40 or MV-αOX40 were analyzed by detecting the amount of αOX40 bound on the plate.

Organ distribution of MV and MV-based multivalent proteins
MV, MV-OX40L, and MV-αOX40 were labeled with Cy5 N-hydroxysuccinimide (NHS) ester to prepare MV Cy5 , MV-OX40L Cy5 , and MV-αOX40 Cy5 .Then, 100 μg Cy5-labeled MV Cy5 , MV-OX40L Cy5 , or MV-αOX40 Cy5 was intraperitoneally injected into mice bearing CT26 tumors.The distribution of MV Cy5 , MV-OX40L Cy5 , and MV-αOX40 Cy5 in mouse models was captured 12 h later using an in vivo imaging system (PerkinElmer, USA), and the fluorescence intensity of dissected tumors and organs was quantified by Living Image software for analysis.

4.10
Mouse tumor models C57BL/6 and BALB/c mice were purchased from Hunan Silaikejingda Experimental Animal Co., Ltd.All animals were maintained in a specific pathogen-free facility and received care in compliance with the Guide for the Care and Use of Laboratory Animals.All animal procedures and experiments were approved by the Animal Care and Use Committee of South China University of Technology.
To construct mouse colorectal cancer models, 1 × 10 6 CT26 cells were subcutaneously inoculated into the right flanks of female BALB/c mice.To construct the mouse breast cancer models, 5 × 10 5 4T1 cells were inoculated into the mammary fat pads of female BALB/c mice.To construct mouse melanoma models, 5 × 10 5 B16-F10 cells were subcutaneously inoculated into the right flanks of female C57BL/6 mice.When the tumor volume was ∼100 mm 3 , the multimeric protein drugs were intraperitoneally administered every other day for a total of three injections.The doses of MV-OX40L, MV-TRAIL, αOX40, MV-αOX40, αPD-L1, and MV-αPD-L1 were 10, 20, 1, 5, 2, and 10 mg/kg, respectively.The tumor diameters were measured using a digital caliper, and tumor volume was calculated using the following formula: tumor volume = length × width 2 × 0.5.

S C H E M E 1
Cross-linking tumor necrosis factor receptors (TNFRs) by engineering nano-clustered multivalent agonists to enhance cancer therapy.Nano-clustered multivalent agonists were engineered by adding ligands or antibodies targeting TNFRs to the N-terminal of the MV monomer.In contrast to the monomer of the ligand or antibody, MV drugs triggered strong activation signals of TNFRs.MV-OX40L and MV-TRAIL promoted T cell activation and induced tumor cell apoptosis, respectively.MV-FcR was constructed to develop a universal nano-adaptor to bind and multimerize commercial agonist or antagonist antibodies for enhancing their antitumor responses.

F I G U R E 1
Preparation of MV for cancer therapy.(A) Schematic illustration of the mechanism for preparing MV. (B and C) Size distribution (B) and zeta potential (C) of MV. (D) Morphology of MV.Left: representative transmission electron microscope (TEM) images of MV.Right: MV with delineated boundaries.Scale bar: 50 nm.(E and F) Western blotting (WB) analysis of MV treated without (E) or with (F) the reducing agent β-mercaptoethanol (β-ME).(G) In vivo biodistribution image of MV Cy5 at 12 h after intraperitoneal injection.(H and I) Organ distribution image (H) and quantification (I) of MV Cy5 at 12 h after intraperitoneal injection.Bar graphs are shown as the mean ± standard deviation (SD).Data were analyzed by one-way analysis of variance (ANOVA) with Tukey's multiple comparison test (I).***p < 0.001.

F
I G U R E 2 MV-OX40L enhances T cell proliferation and activation for colorectal cancer therapy.(A) Schematic of MV-OX40L construction.(B) Schematic of MV-OX40L improving T cell functions.(C) Size-exclusion chromatography (SEC) traces of MV-OX40L (blue) and marker (gray).(D and E) Size distribution (D) and zeta potential (E) of MV-OX40L.(F) Confocal laser scan microscopy (CLSM) images of MV-OX40L Cy5 (blue) binding to T cells (red).(G) Flow cytometric analysis of MV-OX40L Cy5 binding to T cells.(H and I) Flow cytometric analysis of T cell proliferation (H) and T cell activation (I) after MV-OX40L treatment.(J and K) qRT-PCR analysis of mRNA expression of T cell cytokine IL-2 (J) and interferon gamma (IFN-γ) (K) after MV-OX40L treatment.(L) Flow cytometric analysis of MV-OX40L Cy5 binding to T cells in blood after intraperitoneal injection.(M) Tumor distribution image of MV-OX40L Cy5 at 12 h after intraperitoneal injection.(N and O) Growth curves (N) and survival curves (O) of mice bearing CT26 colorectal cancer after intraperitoneal injection of MV-OX40L (n = 9-10 mice per group).Bar graphs are shown as the mean ± SD.Data were analyzed by Student's unpaired t test (I-L) or the log-rank (Mantel-Cox) test (O).*p < 0.05; **p < 0.01; ***p < 0.001.F I G U R E 3 MV-TRAIL directly induces tumor cell apoptosis for breast cancer therapy.(A) Schematic of MV-TRAIL construction.(B) Schematic diagram of MV-TRAIL killing tumor cells.(C) Size-exclusion chromatography (SEC) traces of MV-TRAIL (green) and marker (gray).(D and E) Size distribution (D) and zeta potential (E) of MV-TRAIL.(F) Percentage of 4T1 cell apoptosis at 2 h after MV-TRAIL treatment.(G) Images of apoptosis process of 4T1 cells after treated with MV-TRAIL.(H) Western blotting (WB) analysis of apoptosis-related signal proteins in MV-TRAIL-treated 4T1 cells.(I) Flow cytometric analysis of apoptotic tumor cells after intraperitoneal injection of MV-TRAIL.(J and K) Growth curves (J) and survival curves (K) of mice bearing 4T1 breast cancer after intraperitoneal injection of MV-TRAIL (n = 9-10 mice per group).Bar graphs are shown as the mean ± SD.Data were analyzed by Student's unpaired t test (I) or the log-rank (Mantel-Cox) test (K).***p < 0.001.

F I G U R E 4
MV-FcR mediates therapeutic antibody multimerization and enhances valency.(A) Schematic of MV-FcR construction.(B) Schematic of MV-FcR multimerizing therapeutic antibodies.(C) Size-exclusion chromatography (SEC) traces of MV-FcR (orange) and marker (gray).(D) The capability of MV-FcR binding antibodies.(E) The capability of MV-αOX40 (dark purple) and αOX40 (light purple) binding OX40 antigen.(F) Confocal laser scan microscopy (CLSM) images of MV-αOX40 Cy5 (purple) binding T cells (green).(G and H) Flow cytometric analysis of T cell proliferation (G) and T cell activation (H) after MV-αOX40 treatment.(I and J) qRT-PCR analysis of T cell cytokine IL-2 (I) and interferon gamma (IFN-γ) (J) expression after MV-αOX40 treatment.Bar graphs are shown as the mean ± SD.Data were analyzed by one-way ANOVA with Tukey's multiple comparison test (I and J).*p < 0.05; **p < 0.01; ***p < 0.001.

F
I G U R E 5 MV-αOX40 improves T cell function and effectively treats colorectal cancer.(A) Experimental design for analyzing the in vivo efficacy of MV-αOX40.(B) Flow cytometric analysis of T cell cytokine expression after intraperitoneal injection of MV-αOX40.(C and D) Organ distribution image (C) and quantification (D) of MV-αOX40 Cy5 at 12 h after intraperitoneal injection.(E and F) Growth curves (E) and survival curves (F) of mice bearing CT26 colorectal cancer after intraperitoneal injection of MV-αOX40 (n = 7 mice per group).Bar graphs are shown as the mean ± SD.Data were analyzed by one-way ANOVA with Tukey's multiple comparison test (D) or the log-rank (Mantel-Cox) test (F).*p < 0.05; ***p < 0.001.

F I G U R E 6
MV-αPD-L1 relieves T cell inhibition and effectively treats melanoma.(A) Experimental design for analyzing the in vivo efficacy of MV-αPD-L1.(B-F) Flow cytometric analysis of T cell proliferation, activation, and cytotoxicity after intraperitoneal injection of MV-αPD-L1.(G) Growth curves of mice bearing B16-F10 melanoma after intraperitoneal injection of MV-αPD-L1.(H) Tumor image of mice bearing B16-F10 melanoma after intraperitoneal injection of MV-αPD-L1.(I) Survival curves of mice bearing B16-F10 melanoma after intraperitoneal injection of MV-αPD-L1 (n = 7 mice per group).(J-L) Concentrations of inflammation cytokines in melanoma after intraperitoneal injection of MV-αPD-L1.Bar graphs are shown as the mean ± SD.Data were analyzed by one-way ANOVA with Tukey's multiple comparison test (B-F and J-L) or the log-rank (Mantel-Cox) test (I).*p < 0.05; **p < 0.01; ***p < 0.001.ns indicates no significant difference.