Orchestrating NK and T cells via tri-specific nano-antibodies for synergistic antitumor immunity

The functions of natural killer (NK) and T cells in innate and adaptive immunity, as well as their functions in tumor eradication, are complementary and intertwined. Here we show that utilization of multi-specific antibodies or nano-antibodies capable of simultaneously targeting both NK and T cells could be a valuable approach in cancer immunotherapy. Here, we introduce a tri-specific Nano-Antibody (Tri-NAb), generated by immobilizing three types of monoclonal antibodies (mAbs), using an optimized albumin/polyester composite nanoparticle conjugated with anti-Fc antibody. This Tri-NAb, targeting PDL1, 4-1BB, and NKG2A (or TIGIT) simultaneously, effectively binds to NK and CD8+ T cells, triggering their activation and proliferation, while facilitating their interaction with tumor cells, thereby inducing efficient tumor killing. Importantly, the antitumor efficacy of Tri-NAb is validated in multiple models, including patient-derived tumor organoids and humanized mice, highlighting the translational potential of NK and T cell co-targeting.

6.In addition, a recent study demonstrated that the intratumoral administration of NK cells along with anti-NKG2A blockade mediated the anti-tumor response through cDC1 and CD8 T cells, but not endogenous NK cells (PMID: 37782273).One would expect that CD8 T cells dominate the antitumor response mediated by Tri-NAbs despite triggering NK cells as well, given the low prevalence of NK cells within the TME of MC38 cells.Studies using RAG1 deficient mice (lack of B and T cells) could help to unveil the anti-tumor response of Tri-Nab dependent on NK cells.If better anti-tumor response by NK is proven, would the authors expect NK cell-mediated immunogenic cell death to increase CD8 T cell anti-tumor response in your experimental settings?7. How do the authors reconcile the data shown in supplementary figure 20 and 22?In supplementary figure 22, authors show how NK and CD8 T cells reduce the percentage of tumor cells alive upon treatment (higher cytotoxicity), whereas in supplementary figure 20, it looks like the mAb treatment only halters tumor growth as the percentage of survival cells stays at 120 over time.
8. Can the authors show a FMO or biological control for anti-NK1.1 in Supplementary figure 25 or Figure 6f?Also, according to material and methods, a CD3 mAb was used in the mAb cocktail, thus it would be more appropriate gate first in CD3-cells to identify NK cells within the TME given that NKT cells do express NK1.1 as well.In addition, to identify CD8 and CD4 T cells, a prior gating on CD3 T cells is recommended.9. Could the authors display the data shown in figure 6e and f as total number of CD8 T cells o NK cells per gram of tumor tissue?Given the marked reduction of tumor growth shown in figure 6b  and c, provide the data this way would be a better way to determine if more effector cells are present in the TME upon treatment 10.In the discussion, a more detailed comparison of the data obtained in this manuscript with previous publication that target PD-L1, NKG2A, 41BB and /or TIGIT should be mention, given special attention to those strategies that combine stimulation and blockade of suppression and/or those strategies that simultaneously trigger three molecules like PMI 36071464 for example.Emphasize in why this approach is better than the others.
11.The authors need to mention in the figure legends and/or in the material and methods section how many times each experiment was done and if each figure panel it represents one, or it is representative of more independent experiments.This information is important to strengthen the reproducibility of the data.
Minor comments 1.The order and number of the Figures and supplementary figures should be mentioned in the text sequentially.Supplementary figure 17 is mentioned after Supplementary figure 13 (page 8 line 21), thus, it should be rewritten as Supplementary figure 14 instead of 17, and the rest of the figures named according to this change.Same for supplementary figure 21 and 22.
2. An introductory sentence should be added in page 13 line 20 to explain how the transcriptomic analysis was done.
3. Supplementary Figure 30 should not be detailed in the conclusion section.I would suggest moving this data that shows a safety profile for tri-Nab prior to describe the data of figure 7.
Reviewer #2 (Remarks to the Author): The authors have developed nanoparticles coated with 3 antibodies targeting NKG2A, 41BB and PDL1, that they have called nano-antibodies or Tri-NAbs.They have shown that these TriNAbs exert anti-tumor effects both in vitro and in vivo in mouse tumor models.They have also shown that Tri-NAbs coated with antibodies targeting human NKG2A, 41BB and PDL1 exert anti-tumoral effects in human tumor organoids and in humanized mouse models.
Although this represents an interesting technologial development with a promosing therapeutic potential, the study lacks a number of specificity controls and mechanistic analyses to warrent publication of the manuscript in its current form: * The assays via flow cytometry, confocal imaging and ELISA in Figure 3 lack specificity controls.For example, do NP with anti-NKG2A and anti-41BB without anti-PDL1 still bind to the tumor cells?Can an excess of free antibodies block the binding of the nanoparticles to the cells and/or block the observed effects of the Tri-NAbs on NK/CD8 proliferation and binding to tumor cells?Also in the colon cancer organoids in Figure 7, such specificity controls are required.
This study describes a new nanotechnology to conjugate anti-Fc antibody to the nanocore APCN and then add a combination of immunomodulatory antibodies, anto-PDL1, anti-4-1BB, and anti-NKG2A (or TIGIT) to form a therapeutic complex.Authors presented extensive biophysical characterization of the complex, its in vitro activity in activating NK and Cd8 T cells, and bringing together NK and CD8 T cells to interact with tumor cells.Authors also presented impressive in vivo therapeutic data.In vivo safety profile was studies through serum biochemical characterization and organ histology and deemed to be safe.The study presented a novel technology to combine multiple antibodies for therapy.Conceptually is novel.Extensive in vitro and in vivo data was demonstrated for efficacy, mechanisms of action, and safety.However, there are several concerns that need to be addressed before can be accepted for publication.

Critique:
1.The tile can be better worded for "Nano-Antibody" not to be confused with "nanobody" 2. Does the complex induce CRS in human PBMC?Invitro stimulation alone with TCR activation and in vivo data of serum cytokine would be informative.This is a safety concern with 4-1BB being a co-stimulatory receptor on T cells and NK cells.3. What is ratio of the different antibody in the Tri-Nab being controlled to a consistency in each experiment?Is there a method to test the occupancy of each of the three antibodies?This has to be taken into consideration for the translational potential if the technology will be taken into use in the clinic.4. Is the APCN biodegradable?How stable is Tri-NAb?What is the molecular weight of the Tri-Nab?
Are there any histological evidence that the Tri-Nab penetrate into tumor tissues? 5.It is good that the authors presented PD-L1 expression in IFNg stimulated tumor cells in vitro, which is a given.Is there any data to demonstrate PD-L1 expression in tumors in vivo by IHC or IF to support the MOA of the Tri-Nab in vivo? 6. Supplement Figure 25

Responses to the reviewers' concerns
We greatly appreciate the valuable comments of the three reviewers, helping us improve the quality of the manuscript.For clarity, the reviewers' comments are in black while our point-to-point answers are marked in blue.Additionally, the changes were highlighted (blue) in the revised manuscript and Supplementary Information for the convenience of reviewers.

Reviewer #1 (Remarks to the Author)
General Comments: Ye and collaborators show in this manuscript the therapeutic efficacy of Tri-specific nano-antibodies to improve CD8/NK cell-mediated anti-tumor responses.The use of novel albumin/polyester nanoparticles to improve biocompatibility, NP half-life and stability is interesting.In addition, the fact that this system allows the delivery of multiple mAbs to trigger both tumor cell death and the modulation of effector cells provides a very versatile strategy with potential clinical translation.Overall, the study is well done and the results obtained support the findings.However, some questions in terms of dosage, experimental replication and mechanism of action need to be addressed.

Response:
We express our gratitude to the reviewer for recognizing our work and sincerely appreciate his/her insightful suggestions.We have elaborated on the concerns as shown below and hope the modifications to the manuscript will meet his/her criteria.
Comment 1: It is not clear what is the actual dosage of Tri-NAbs that is being used in each experimental setting and if these dosages are equivalent to the dosage used for controls of free mAbs.In the material and methods section, the authors mention that the APCNs are mixed with the three mAbs at 1:1:1 (w/w), but what is the actual dosage of Tri-Nabs used after incubation?For in vitro experiments, free 5-10 μg/mL of each mAb is used as control, but is this dose equivalent to the amount of mAb present in the Tri-NAbs?In figure 3A it looks that the amount used of free mAb and tri-Nab binds to a similar degree to recombinant PD-L1, CD137 and NKG2A, however, the actual amount of Tri-NAb and free mAbs are not mentioned.In the material and methods´section (page 25 line 3), it states that a series of concentrations of free mAb (αPDL1, α4-1BB, 4 or αNKG2A) or Tri-NAb were added to 5 μg/mL recombinant proteins coated plates, however, the display on Figure 3A is optical density (for binding efficiency) against recombinant proteins concentration (mM) and not antibody/NAb concentrations.Is that correct?

Response:
We sincerely apologize for the oversight in accurately describing the actual dosage of Tri-NAb.The equal probabilities of three mAbs binding to APCN@NA were confirmed at the outset of our study (data was not provided in the original manuscript).Tri-NAb was prepared by mixing APCN@NA with AF647-labeled αPDL1, AF488-labeled α4-1BB, and AF555-labeled αNKG2A at a weight ratio of 1:1:1 at 4℃ for 24 hours.The resulting Tri-NAb was separated through high-speed centrifugation, while the exact quantity of unbound mAbs remaining in the supernatant was quantified using high-performance liquid chromatography (HPLC), thereby enabling us to calculate the binding efficiency of each individual mAb to APCN@NA.As illustrated in Fig. R1 and Supplementary Fig. 16a, when the weight ratio between APCN@NA and total mAbs was 1:1, the binding efficiencies of αPDL1, α4-1BB, and αNKG2A by APCN@NA were 89.5%, 91.6% and 91.8%, respectively.Based on these findings, it can be inferred that the final ratio of three mAbs on Tri-NAb was approximately 1:1:1.In subsequent experiments involving Tri-NAb preparation, unbound mAbs were removed and the term "dosage of Tri-NAb" refers to the combined quantity of all three mAbs present on Tri-Ab with an equal mass ratio unless otherwise specified.In the revised manuscript, we have revised relevant descriptions regarding the dosage of Tri-NAb.Additionally, we thank the reviewer for pointing out the error in Fig. 3a.The x-axis should represent a range of concentrations pertaining to free mAbs (αPDL1, α4-1BB, or αNKG2A) or Tri-NAb rather than recombinant protein concentration (mM).We have meticulously corrected this in the revised manuscript.
Comment 2: In addition, in figure 6 the authors checked different Tri-Nab doses (3.75, 6.25 and 7.5mg/mL of different concentrations for anti-PDL1/41BB/NKG2A used to generate the Tri-NAbs).They said that in Figure 6i they halved the dose of anti-CD137 to 6.5 mg/kg (page 15 line 5), however, according with the data showed in figure 6, the maximal dose used for anti-CD137 to generate tri-NAbs in this figure was 2.5 mg/kg.Was higher dose used in previous experiments, because the dose used for free mAbs was 2.5 mg/kg each?Please clarify it.
Response: In all other animal studies, we administered a dosage of 2.5 mg/kg for each mAb, resulting in a total dose of 7.5 mg/kg.However, in the experiment depicted in Fig. 6i, mice were treated with a total Tri-NAb dosage of 3.75 mg/kg, equivalent to a dosage of 1.25 mg/kg for each mAb; or alternatively, a total Tri-NAb dosage of 6.25 mg/kg comprising of 2.5 mg/kg αPDL1, 1.25 mg/kg α4-1BB, and another 2.5 mg/kg αNKG2A; or finally, a total Tri-NAb dosage of 7.5 mg/kg consisting of equal dosages (2.5mg/kg) for each mAb.We have included a detailed description in the revised Fig. 6i and in the Methods section of the revised manuscript.Response: Thanks for the insightful comment.In the original manuscript, we demonstrated that APCN@NA exhibited comparable binding affinities towards Rat IgG2a or Rat IgG2b mAbs (Fig. 2m-n).Furthermore, we have shown αPDL1 (Rat IgG2b subtype), α4-1BB (Rat IgG2a subtype), and αNKG2A (Rat IgG2a subtype) bind to APCN@NA with high efficiencies of 89.5%, 91.6%, and 91.8%, respectively, indicating a nearly equal affinity of APCN@NA towards these three mAbs (Fig. R1 and Supplementary Fig. 16a).Based on these findings, it can be inferred that the distribution of the three mAbs on an APCN@NA is random.Therefore, in the schematic diagram, we depict an equal number of all three mAbs attached to one APCN@NA.
We express our gratitude to the reviewer for highlighting the inadequacy of our schematic depiction.Undoubtedly, the augmented interaction between tumor cells and NK/T cells mediated, as well as the enhanced activation of NK/T cells, is not achieved by a single Tri-NAb alone but rather through the synergistic action of multiple Tri-NAb.Considering the stochastic distribution of 4-1BB and NKG2A on NK/T cells, it is plausible that α4-1BB and αNKG2A from one Tri-NAb may simultaneously recognize and engage their respective targets; however, it is also feasible that only α4-1BB or αNKG2A from one Tri-NAb interacts with its target.We observed that Tri-NAb mediated the formation of immune synapses between NK/T cells and tumor cells (Fig. R2), however, discerning whether synapse formation is attributed to simultaneous attachment of 4-1BB and NKG2A to their corresponding mAbs from one or more Tri-NAb remains challenging; we suspect that both possibilities exist.To avoid any potential misunderstanding, we have revised related schematic diagrams and added more descriptions in the revised manuscript.NPαPDL1+α4-1BB, NPαPDL1+αNKG2A, NPα4-1BB +αNKG2A, or Tri-NAb, where the concentration of αPDL1, α4-1BB, or αNKG2A was 10 μg/mL respectively, and the concentration of mAbs was 30 μg/mL.After co-incubation for 24 h, NK or CD8 + T cells were collected and subjected to western blot using Jess Automated Western Blot System (ProteinSimple, Bio-Techne).
Comment 4: PD-L1 is also highly upregulated on both T and NK cells (PMID: 32117218, PMID: 35417187, PMID: 32152508).Did they author evaluated if the PD-L1 NP could bind to the effector cells?If so, did the binding was similar/better/worse than the binding to PDL1 expressed on tumor cells?Could this binding lead to effector cell fratricide?
Response: Thank you for your insightful comment.Flow cytometry analysis revealed a minimal binding of NPαPD-L1 to NK or CD8 + T cells, which was significantly overshadowed by its remarkable affinity for tumor cells (Fig. R4).Per the reviewer's suggestion, we investigated to determine whether Tri-NAb induces fratricide of NK/T cells themselves.To accomplish this, we incubated NK or/and CD8 + T with Tri-NAb and propidium iodide (for labeling apoptotic cells), and observed cell apoptosis utilizing an advanced high-content assay system.As illustrated in Supplementary Video 1, no evidence of fratricidal behavior among NK cells was observed following treatment with Tri-NAb.Similarly, CD8 + cells did not exhibit any such tendencies as shown in Supplementary Video 2. Furthermore, even when NK and CD8 + T cells were co-incubated together, there was still no indication of fratricide occurring between these two cell populations as demonstrated in Supplementary Video 3. The negligible occurrence of effector cell fratricide may be attributed to the relatively weak cellular interaction facilitated by Tri-NAb and their inherent self-protection mechanism.We have included related description and experimental details in the revised manuscript.Comment 5: In Figure 3g-h, it is quite surprising that no NK cells or T cells are seen in the isotype or free tri-mAbs cultures after having added 2.5x10 5 NK and CD8 T cells because binding through other NK/T cell receptors could be possible (like CEACAM, Fas, TRAIL, NKG2D, etc) as MC38 and B16 can express ligands for these pathways specially after IFNγ treatment (PMID: 22143889, PMID: 35840162).
It is also surprising to see barely no NK cells and T cells in the NP anti-PDL1/41BB and NK anti-PDL1/NKG2A controls, respectively, since 41BB and NKG2A are expressed in both cell types as well.Please explain.
Response: Prior to imaging, we utilized pre-cooled 1×PBS to wash the cells several times to diminish the weak contact between NK/T cells and tumor cells facilitated by other ligands.However, nanoparticles-treated NK/T cells exhibited enhanced interactions with tumor cells growing on the petri dishes, rendering them less susceptible to being washed away.Through extensive observation of various fields and quantification of co-localized regions between NK/T cells and tumor cells, it was observed that Tri-NAbs-treated group displayed a significantly augmented abundance of tightly adhered NK/T cells to tumor cells (as illustrated in Fig. 3i).Response: Thanks to the reviewer for the insightful comments.Per the reviewer's suggestion, we conducted experiments to evaluate the involvement of CD8 + T and NK cells in Tri-NAb-mediated suppression of tumor growth by selectively depleting these cell populations using specific antibodies (Fig. R5a).As depicted in Fig. R5b-d, depletion of CD8 + T cells significantly impeded the remarkable antitumor efficacy of Tri-NAb, as evidenced by the tumor growth curves; similarly, depletion of NK cells had a substantial impact on the antitumor effect of Tri-NAb, particularly on the complete remission rate (CR).As anticipated, simultaneous depletion of both CD8 + T and NK cells mediated by anti-CD8 and anti-NK1.1 antibodies resulted in a complete loss of the antitumor potency exhibited by Tri-NAb.These findings underscored the crucial roles played by both CD8 + T and NK cells in mediating the anticancer response elicited by Tri-NAbs, with CD8 + T cells being predominant.Related data was also included in Supplementary Fig. 32 in the revised Supplementary Information.remarkable absence of T cells alongside a significant augmentation in the proportion of NK cells in the peripheral blood of Rag1-deficient mice (Fig. R6a).Subsequently, we established a subcutaneous MC38 tumor model and administered Tri-NAb treatment according to our experimental design (Fig. R6b).Although there was mild control over colon cancer growth in Rag1-deficient mice following Tri-NAb treatment (Fig. R6c), it was not as profoundly suppressed as observed in immunocompetent C57BL/6 mice (Fig. 4f).These findings suggest that the antitumor response elicited by Tri-NAb is moderately reliant on NK cells, with T cells play a more crucial role.
The revised manuscript has provided the corresponding data as Supplementary Fig.
Fig. 18, the fourth and fifth groups).As anticipated, treatment with NPα4-1BB+αNKG2A also markedly promoted NK/T cell proliferation similar to other NPs immobilizing α4-1BB (Fig. R11 and revised Supplementary Fig. 19).However, through CLSM observation, it was evident that the ability of NPα4-1BB+αNKG2A to enhance interactions between NK/T cells and tumor cells was notably weaker compared to Tri-NAb (Fig. R12b and revised Supplementary Fig.  Comment 2: For the in vivo experiments in mice using the anti-mouse and the anti-human tri-NAbs, it is important to know how much of the effect is due to the tumor targeting properties of the Tri-NAbs as opposed to inherent properties of the NPs, by including a control group with NPs that have only anti-anti-NKG2A and anti-41BB without anti-PDL1.
Response: Thanks for the insightful comment.Per the reviewer's suggestion, we incorporated the NPα4-1BB+αNKG2A group and repeated the antitumor experiment using the B16-F10 melanoma model.Tumor volumes were monitored throughout treatment while tumor-infiltrating NK and CD8 + T cells were analyzed at the end of treatment.
Response: We express our gratitude to the reviewer for recognizing our work and sincerely appreciate his/her insightful suggestions.We have elaborated on the concerns as shown below and hope the modifications to the manuscript will meet his/her criteria.
Comment 1: The tile can be better worded for "Nano-Antibody" not to be confused with "nanobody".
Response: Thanks to the reviewer for the thoughtful suggestion.We have revised it in the revised manuscript.
Comment 2: Does the complex induce CRS in human PBMC?In vitro stimulation alone with TCR activation and in vivo data of serum cytokine would be informative.This is a safety concern with 4-1BB being a co-stimulatory receptor on T cells and NK cells.

Response:
The reviewer posed a profoundly significant inquiry.As the cross-linking of 4-1BB, either by binding to 4-1BBL or agonist antibody, delivers a costimulatory signal for the activation and proliferation of NK/T-cells, as well as the release of various cytokines.Therefore, the biosafety of Tri-NAb emerged as a pivotal concern in our study.It is indeed gratifying that no serious adverse events were observed during Tri-NAb treatment; this was substantiated by negligible weight loss throughout the course of treatment (revised Supplementary Fig. 30 and Fig. 36) and corroborated by post-treatment serum biochemical and histological analyses (revised Supplementary Fig. 31).Per the reviewer's suggestion, we conducted additional investigations safety profile of Tri-NAb by evaluating its potential to induce cytokine release syndrome (CRS), characterized by excessive secretion of inflammatory cytokines.Following murine melanoma treatment, peripheral blood samples were collected to measure levels of specific inflammatory cytokines (including IL-10, TNF-α, IL-6, and IL-1β).As depicted in Fig. R14 and revised Supplementary Fig. 37, the levels of these serum cytokines in Tri-NAb-treated mice did not significantly differ from those in untreated mice; thus, indicating that at the administered dose (2.5 mg/kg per mAb), Tri-NAb did not elicit CRS in vivo.The favorable biosafety and the absence of CRS may be attributed to the relatively low dosage of α4-1BB administered in our study (for antitumor therapy, the dosage of α4-1BB was 2.5 mg/kg, less than 60 μg per mouse).Although the investigation of CRS occurrence following Tri-NAb administration was not performed on human PBMC, partly due to the high cost of human cytokine ELISA kits and NCG mice, we hope that the aforementioned data can alleviate reviewers' concerns regarding the safety of Tri-NAb.
In case these findings fail to assuage his/her concerns, we are prepared to further supplement with relevant data utilizing human PBMC models and express our gratitude for the reviewer's understanding.Comment 3: What is ratio of the different antibody in the Tri-Nab being controlled to a consistency in each experiment?Is there a method to test the occupancy of each of the three antibodies?This has to be taken into consideration for the translational potential if the technology will be taken into use in the clinic.
Response: Thanks for the valuable comments.The equal probabilities of three mAbs binding to APCN@NA were confirmed at the outset of our study (data was not provided in the original manuscript).Tri-NAb was prepared by mixing APCN@NA with AF647-labeled αPDL1, AF488-labeled α4-1BB, and AF555-labeled αNKG2A at a weight ratio of 1:1:1 at 4℃ for 24 hours.The resulting Tri-NAb was effectively separated through high-speed centrifugation, while the exact quantity of unbound mAbs remaining in the supernatant was quantified HPLC, thereby enabling us to calculate the binding efficiency of each mAb to APCN@NA.As illustrated in Fig. R1 and Supplementary Fig. 16a, when the weight ratio between APCN@NA and total mAbs was 1:1, the binding efficiencies of αPDL1, α4-1BB, and αNKG2A by APCN@NA were 89.5%, 91.6% and 91.8%, respectively.Based on these findings, it can be inferred that the final ratio of three mAbs on Tri-NAb was approximately 1:1:1.
In the revised manuscript, we have revised relevant descriptions regarding the dosage of Tri-NAb.Response: The APCN is composed of serum albumin and polylactic acid, both of which possess excellent biocompatibility and biodegradability properties.Serum albumin can be easily metabolized by the body, while polylactic acid is a commonly used biodegradable polymer material in the field of biomedicine (PMID: 31021482, PMID: 31681741).Based on this information, it is reasonable to presume that APCN could undergo degradation in vivo.Furthermore, we observed less than 6.0% release of αPDL1, α4-1BB, and αNKG2Afrom Tri-NAb in PBS buffer over a period of 15 days, confirming its excellent stability (Fig. R15 and revised Supplementary Fig. 16b).
Through asymmetric flow field-flow fractionation (AF4) and multi-angle light scattering (MALS) techniques provided by Wyatt Technology Corp., we measured the weight-average molecular weight of Tri-NAb to be 1.5×10 8 (±4.0%) g/mol with a radius of gyration (Rg) measuring at 54.1 (±4.1%) nm; these values were larger compared to those obtained for APCN with an Rg value of 37.4 (±2.4%) nm.
-NK1.1 separates NK very well as a clear population.Current gating does not represent typical NK cell population and not convincing that the population is NK cell.Authors need to present a more convincing gating data, such as CD3.Vs.NK1.1 to define NK population.7. Why in vitro data beautifully demonstrated the Tri-Nab brings tumor cells to simultaneously interact with NK and CD8 T cells, it would be more powerful to demonstrate the same interaction in tumors with multiplex IHC/IF.8. Supplement Fig 30b, what are the authors intending to demonstrate?

Comment 3 :
Figure 21) and not that NKG2A blockade and CD137 stimulation is accomplished by different Tri-NAbs?Have the authors checked the immunological synapse of NK and/or T cells after tri-NAbs to determine if both molecules are attached to their corresponding mAbs simultaneously from one or more nanoparticles?Determine the level of SHP1, ERK/NFkB phosphorilation of Tri-NAbs-treated NK/T cells after PD-L1+Qa1b+ tumor cells exposure in comparison with single and bi-NAbs-treated NK/T cells would also help to see if both pathways are activated at higher degree if the three mAb are present when compared to control, single and double NPs.
Fig. R2.Representative CLSM images showing the formation of immune synapses between NK/T cells and tumor cells.NK or CD8 + T cells were pretreated with Tri-NAb for activation and then incubated with the tumor cells, followed by fixation and staining with AF488-phalloidine.

Comment 6 :
Depletion studies to determine the role of CD8 T cells and NK cells should be done to evaluate the importance of each immune cell type efficacy in the anti-tumor response mediated by Tri-NAbs using single (anti-CD8b or anti-NK1.1)and double depletion.

Fig
Fig. R5.Both CD8 + T and NK cells played crucial roles in Tri-NAb-mediated tumor growth suppression.(a) Representative scatter plots of flow cytometry showing the proportion of CD8 + T/NK cells in the peripheral blood of mice after antibody-mediated depletion.(b) A subcutaneous MC38 tumor model was established in male C57BL/6 mice.According to the experimental protocol, CD8 + T and NK cells were depleted via intraperitoneal injection of 125 μg anti-CD8 mAb or/and 150 μg anti-NK1.1 mAb on specific days.Tri-NAb was administrated every three days for three repeats since the 8th day post-tumor inoculation; the injection dose of αPDL1, α4-1BB, and αNKG2A was 2.5 mg/kg.Average (c) and individual (d) tumor growth curves of MC38 tumors after being treated with Tri-NAb with or without T cells or/and NK cells depletion.The statistical data are presented as the mean ± s.d.(n = 5).S Statistical significance was calculated by one-way ANOVA with Tukey's post hoc test.****P < 0.0001.
Fig. R6.Tri-NAb mildly controlled tumor progression in Rag1-deficient mice.(a) Representative flow cytometry scatter plots show the proportion of CD8 + T and NK cells in the peripheral blood of immunocompetent C57BL/6 mice and Rag1-deficient mice.(b) Experimental scheme of subcutaneous MC38 tumor model in female Rag1-deficient mice.Tri-NAb with equivalent doses of αPDL1, α4-1BB, and αNKG2A (2.5 mg/kg each) was intravenously (i.v.) administered via the tail on days 8, 11, and 14 post MC38 tumor inoculation.(c) Tumor growth curves of MC38 tumors.The statistical data are presented as the mean ± s.d.(n = 5).Statistical significance was calculated via paired t-test with two-tailed.***P < 0.001.
Fig. R8.Tri-NAb effectively inhibited the progression of mouse melanoma.(a) Experimental scheme of subcutaneous B16-F10 (4.0×10 5 ) mouse melanoma model in male C57BL/6 mice.Different formulations with equivalent doses of αPDL1, α4-1BB, and αNKG2A (2.5 mg/kg each) were intravenously (i.v.) administered via the tail on days 6, 9, and 12 after B16-F10 tumor inoculation.Individual (b) and average (c) tumor growth kinetics in different formulations.Growth curves represent means ± s.d.(n = 6).Representative scatter plots of flow cytometry showing the number of NK cells as a percentage of CD3 -cell population (d) or CD8 + T cells as a percentage of CD3 + cell population (g) in the tumor after different treatments as indicated.Quantitative results of the number of NK (e) or CD8 + T cells (h) as a percentage of the total CD45 + cell population in the tumor.Quantitative results of the total number of NK (f) or CD8 + T cells (i) per gram of tumor tissue.n.s., not significant.FACS data are presented as means ± s.d.(n = 4).Statistical significance was calculated by one-way ANOVA with Tukey's post hoc test.*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.
Tri-NAb can bind PDL1-positive tumor cells more efficiently, thereby enhancing their accumulation within tumors and prolonging their residence time; secondly, Tri-NAb substantially augments interactions between NK/T cells and tumor cells, consequently increasing granzyme B and perforin release by NK/T cells for enhanced killing of neighboring tumor cells.Revised Fig. 6 and Supplementary Fig. 34 containing relevant data have been included in the revised manuscript.Ⅲ. NP αPDL1+α4-1BB Ⅳ. NP αPDL1+αNKG2A Ⅱ. Tri-mAbs Ⅴ. NP α4-1BB+αNKG2A Ⅵ.

Comment 3 :
Fig. R8.Tri-NAb effectively inhibited the progression of mouse melanoma.(a) Experimental scheme of subcutaneous B16-F10 (4.0×10 5 ) mouse melanoma model in male C57BL/6 mice.Different formulations with equivalent doses of αPDL1, α4-1BB, and αNKG2A (2.5 mg/kg each) were intravenously (i.v.) administered via the tail on days 6, 9, and 12 after B16-F10 tumor inoculation.Individual (b) and average (c) tumor growth kinetics in different formulations.Growth curves represent means ± s.d.(n = 6).Representative scatter plots of flow cytometry showing the number of NK cells as a percentage of CD3 -cell population (d) or CD8 + T cells as a percentage of CD3 + cell population (g) in the tumor after , depletion of CD8 + T cells significantly impeded the remarkable antitumor efficacy of Tri-NAb, as evidenced by the tumor growth curves; similarly, depletion of NK cells had a substantial impact on the antitumor effect of Tri-NAb, particularly on the complete remission rate (CR).As anticipated, simultaneous depletion of both CD8 + T and NK cells mediated by anti-CD8 and anti-NK1.1 antibodies resulted in a complete loss of the antitumor potency exhibited by Tri-NAb.These findings underscored the crucial roles played by both CD8 + T and NK cells in mediating the anticancer response elicited by Tri-NAbs, with CD8 + T cells being predominant.Related data was also included in Supplementary Fig.32in the revised Supplementary Information.
Fig. R5.Both CD8 + T and NK cells played crucial roles in Tri-NAb-mediated tumor growth suppression.(a) Representative scatter plots of flow cytometry showing the proportion of CD8 + T/NK cells in the peripheral blood of mice after antibody-mediated depletion.(b) A subcutaneous MC38 tumor model was established in male C57BL/6 mice.According to the experimental protocol, CD8 + T and NK cells were depleted via intraperitoneal injection of 125 μg anti-CD8 mAb or/and 150 μg anti-NK1.1 mAb on specific days.Tri-NAb was administrated every three days for three repeats since the 8th day post-tumor inoculation; the injection dose of αPDL1, α4-1BB, and αNKG2A was 2.5 mg/kg.Average (c) and individual (d) tumor growth curves of MC38 tumors after being treated with Tri-NAb with or without T cells or/and NK cells depletion.The statistical data are presented as the mean ± s.d.(n = 5).S Statistical significance was calculated by one-way ANOVA with Tukey's post hoc test.****P < 0.0001.
Fig. R14.The level of serum inflammatory cytokines.The levels of IL-10 (a), TNF-α (b), Fig. R1.The binding efficiencies of αPDL1, α4-1BB, and αNKG2A by APCN@NA, when the weight ratio between APCN@NA and total mAbs was 1:1.Data are presented as means ± Fig. R15.Tri-NAb had good stability in vitro.Tri-NAb was prepared by mixing APCN@NA with AF647 labeled αPDL1, AF488 labeled α4-1BB and AF555 labeled αNKG2A at 4℃ for 24 h.Purified Tri-NAb was obtained by high-speed centrifugation and incubated in PBS buffer at 37°C.The released free antibodies were collected by high-speed centrifugation and quantified by fluorescence on days 0, 5, 10, and 15.Data are presented as means ± s.d.(n = 3).

Fig. R17 .
Fig. R17.Representative FACS histograms showing the expression of PDL1 in tumor tissue.