Valrubicin-loaded immunoliposomes targeting antigens on immunosuppressive cells to circumvent resistance to cancer immunotherapy
Aleksandra Georgievski, Noémie Blanc, Mélanie Bruchard, Cassandre Pignol, Pierre-Simon Bellaye, Carmen Garrido, Ronan Quéré
TL;DR
Researchers developed valrubicin-loaded immunoliposomes that target and reduce immunosuppressive cells in tumors, improving cancer immunotherapy effectiveness.
Contribution
The study introduces a novel nanoparticle therapy that targets multiple immunosuppressive cell antigens to enhance anti-PD-1 treatment in resistant cancer models.
Findings
Val-ILs reduce antigen expression on immunosuppressive cells like macrophages and T cells in the tumor microenvironment.
Val-ILs enhance anti-PD-1 therapy efficacy in both responsive and resistant mouse cancer models.
Val-ILs increase tumor-infiltrating lymphocytes and reprogram macrophages toward an anti-tumor phenotype.
Abstract
We develop valrubicin-loaded immunoliposomes (Val-ILs), a nanoparticle-based therapy designed to target immunosuppressive cells that promote immune evasion in cancer. In vivo screening following intravenous administration in mice identifies nine relevant surface targets, including known immunoregulatory markers (LAG-3 and VEGFR2) and not-well-characterized candidates (CD11b, CD64, TIM1, CD200R3, CD204, CD49b, and SIGLEC-F). Within the tumor microenvironment, Val-ILs treatment broadly reduces the expression of these antigens on immunosuppressive populations, including tumor-associated macrophages, myeloid-derived suppressor cells, regulatory T cells, and T helper 17 cells, as well as on innate anti-tumor cells such as tumor-associated natural killer cells and tumor-infiltrating dendritic cells. Across four murine cancer models, two responsive (T and B lymphomas) and two resistant…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsNanoparticle-Based Drug Delivery · Immunotherapy and Immune Responses · RNA Interference and Gene Delivery
Introduction
Cancer immunotherapy represents one of the most significant advances in oncology in recent years. Immune checkpoint inhibitors (ICIs), which activate the immune system to target cancer cells, have demonstrated success in various tumor types.1^,^2^,^3^,^4 However, for most patients, ICIs alone are insufficient to control tumor progression and a significant proportion remain unresponsive to treatment, this resistance being partly due to the suppression of cytotoxic T lymphocyte (T-Ly) activation thereby enabling immune evasion.1^,^2^,^3^,^4 A key determinant of ICIs’ response is the presence of targetable immune checkpoints, such as programmed cell death protein 1 (PD-1) and its ligand (PD-L1), cytotoxic T lymphocyte-associated protein 4 (CTLA-4), as well as LAG-3 (CD223).5^,^6^,^7^,^8^,^9^,^10^,^11^,^12 Certain tumor types, such as triple-negative breast cancer (TNBC)13 and non-small cell lung cancer,14^,^15 are inherently resistant due to their poor immunogenicity. Consequently, combining ICIs with other therapeutic modalities, including chemotherapy or specific inhibitors,16^,^17^,^18 radiotherapy,19 anti-angiogenic agents,20^,^21 or most often dual immunotherapy,4^,^6^,^11^,^22 has shown promise in overcoming resistance mechanisms. The limited efficacy of ICIs can be attributed to various factors, including reduced tumor antigen expression and the presence of immune cells in the tumor microenvironment (TME),23 including myeloid-derived suppressor cells (MDSCs),24^,^25^,^26 tumor-associated macrophages (TAMs),27^,^28^,^29 tumor-associated neutrophils,30^,^31 tumor-infiltrating dendritic cells (DCs) (tDCs),32^,^33 tumor-associated natural killer (NK) cells (TaNKs),34 regulatory T cells (Tregs), and T helper 17 (Th17) cells.35^,^36 Given this complexity, the immune cell composition of the TME is increasingly recognized as a predictive biomarker for immunotherapy response.1^,^2^,^3 Mouse models of cancer serve as valuable tools for identifying strategies and therapeutic combinations to enhance ICIs efficacy.16^,^17^,^18^,^19^,^37^,^38^,^39^,^40^,^41^,^42^,^43^,^44^,^45 The convergence of cancer immunotherapy, nanotechnology, bioengineering, and drug delivery offers a promising avenue for innovation, as these disciplines have evolved separately while maintaining strong complementary potential. Notably, nanotechnology can significantly enhance cancer immunotherapy by facilitating precise drug delivery and reducing unintended side effects.46^,^47^,^48
We developed valrubicin-loaded immunoliposome (Val-ILs) nanoparticles (NPs) as a therapy. Valrubicin, a hydrophobic analog of doxorubicin, readily integrates into lipophilic carriers. When functionalized with specific antibodies, Val-ILs can selectively recognize target cell populations within heterogeneous mixtures and induce targeted cell death.49 We demonstrated that, following intravenous (i.v.) administration in tumor-bearing mice, Val-ILs accumulated in the spleen and the tumor-draining lymph nodes (TDLNs), underscoring their potential to target immunosuppressive cells within these organs that contribute to tumor immune evasion. To optimize this approach, we functionalized the surface of Val-ILs with antibodies targeting over 30 antigens expressed by immunosuppressive cells, ultimately identifying nine promising candidates, including not-well-characterized targets and previously known markers lacking therapeutic options. In this study, we evaluated the therapeutic potential of these NPs in different cancer models and demonstrated that they can overcome resistance to anti-PD-1 (αPD-1) therapy.
Results
Val-ILs targeting immunosuppressive cells in the spleen and the TDLNs affect T lymphoma growth
We successfully produced Val-ILs exhibiting an average diameter of 110–120 nm (Figures 1A and 1B). Our screening focused on cell surface markers expressed by immune cells, including those found on immunosuppressive cells. We generated Val-ILs loaded with antibodies against 31 antigens and tested their efficacy in the EL4 mouse T lymphoma model, which is highly immunogenic and sensitive to ICIs.38^,^39 Immunocompetent C57BL/6 mice were subcutaneously (s.c.) implanted with 10^6^ EL4 lymphoma cells. When the tumor volumes reached 50 mm^3^, mice were i.v. injected with 10^12^ NPs into the tail vein, on days 6 and 9. Among the tested formulations, nine Val-ILs targeting CD11b, CD223 (LAG3), CD365 (TIM1), CD200R3, CD204, CD49b, CD309 (VEGFR2), CD64, and CD170 (SIGLEC-F) showed significant tumor growth inhibition when tumor volumes were assessed in vivo on day 12 (Figure 1C; Table S1).Figure 1. Identification of Val-ILs affecting T lymphoma growth(A) Transmission electron microscopy imaging showing a representative image of Val-ILs. Magnifications, ×94K (left) and ×310K (right); scale bars: 100 nm (left) and 50 nm (right).(B) Representative quantification and size of Val-ILs measured by NTA.(C) C57BL/6 mice were s.c. injected with 10^6^ EL4 cells, into the side of the body. When tumors reached 50 mm^3^, mice were treated with Val-ILs i.v. injected on days 6 and 9 (10^12^ NPs). A screening of 31 NPs loaded with different antibodies identified nine NPs that exhibited a strong tumor volume reduction by day 12 (pink bars).(D) Mice were treated with Val-ILs (10^12^ NPs) on days 6, 9, and 12. The efficacy of the nine individual Val-ILs, as well as the combined Val-ILs targeting the nine identified antigens (Val-ILs-Combo), was evaluated on day 12 (left panel) and over time (right panel), n = 4 mice per group, two independent experiments performed.(E) Mice were i.v. injected with trackable NPs labeled with the PKH67 green fluorescent dye, n = 4 mice. The proportion and the number of PKH67^+^ NPs detected by FC in different tissues, after 18 h. PKH67^+^ NPs were detected in the spleen, the lymph nodes, and the bone marrow, whereas undetected (ud) in the blood, the thymus, the kidney, and the tumor. UV, unilamellar vesicle.(F) Mice were i.v. injected with Val-ILs labeled with PKH67. Cells were extracted 18 h later from various tissues and analyzed by FC, n = 5 mice per group. Data showing that trackable NPs bound to immunosuppressive cells in the spleen and the TDLNs. MFI, mean fluorescence intensity.Data are shown as means ± SD, p values are compared to Val-ILs-IgG or between the different groups and are calculated using one-way ANOVA with Tukey’s multiple comparisons test. See also Figure S1.
To investigate the combination effects of treatment with Val-ILs functionalized with these nine antigens, NPs were first individually loaded with each antibody and then pooled to generate the Val-ILs-Combo. Nanoparticle tracking analysis (NTA), liquid chromatography-tandem mass spectrometry (LC-MS/MS), and flow cytometry (FC) were employed to evaluate production, valrubicin loading, and antibody conjugation (Figures S1A–S1E). As a control, Val-ILs-IgG conjugated with IgG isotype antibodies lacking specific functionalization were used. Val-ILs-IgG reached the endpoint of 1,500 mm^3^ by day 12. While mice treated with individual formulations of Val-ILs survived a few days longer, those treated with Val-ILs-Combo took twice as long to reach the endpoint criteria (Figure 1D).
To assess their biodistribution, Val-ILs were labeled with PKH67, a green fluorescent dye that integrates specifically into lipid bilayers. PKH67^+^ unilamellar vesicles were detected 18 h after i.v. injection, in the spleen, the bone marrow, and the TDLNs (Figures 1E and S1F). Using FC, we observed a major tropism of PKH67^+^ Val-ILs toward immunosuppressive cell populations, including MDSCs, macrophages, Th17 cells, and Tregs, in the spleen and the TDLNs, and due to the limited penetration of Val-ILs into the TME, no appreciable binding to intra-tumor immune cells was observed (Figure 1F).
In conclusion, Val-ILs-Combo, which target multiple antigens expressed by immunosuppressive cells, effectively inhibited T lymphoma growth in vivo.
Val-ILs modulate additional immunosuppressive cell populations during T lymphoma progression
We subsequently assessed the effect of Val-ILs-Combo on immunosuppressive cell populations involved in tumor growth in the spleen, as well as on the immune cells infiltrating the tumors. Using FC and unsupervised analysis methods with uniform manifold approximation and projection (UMAP), we observed several distinct clusters of immune cells in the TME, on day 12, when tumor volumes reached 1,500 mm^3^ (Figure 2A). The nine antigens targeted by Val-ILs-Combo were predominantly expressed on tumor-associated myeloid cells (Figure 2B). UMAP analysis further revealed that, while certain antigens could be attributed to specific immune cell populations, others were expressed across multiple clusters (Figure S2A).Figure 2. Val-ILs affect immunosuppressive cells and T lymphoma growthC57BL/6 mice were s.c. injected with 10^6^ EL4 cells into the body side. When tumors reached 50 mm^3^, mice were i.v. injected with Val-ILs (10^12^ NPs) on days 6 and 9. Analyses were performed on day 12 (excepted for L).(A) UMAP data frame and heatmap following FC showing expression levels of the different populations of immune cells in the TME of T lymphoma-bearing mice. Mean normalized data of n = 7 untreated mice.(B) UMAP data frame and heatmap following FC showing expression levels of the nine antigens targeted by Val-ILs-Combo among immune cells detected in the TME, mean normalized data of n = 7 untreated mice.(C) Treatment with NPs reduced the tumor volumes measured in vivo. Image of tumors isolated ex vivo.(D) NPs affected the repartition of immune cells in the TME.(E) NPs reduced the number of immunosuppressive cells in the TME.(F) NPs affected the repartition of immune cells in the spleen.(G) NPs reduced the number of immunosuppressive cells in the spleen.(H) MFI measured by FC on TAMs in the TME and macrophages in the spleen, to identify M1-like and M2-like phenotypes.(I) FC on tDCs in the TME and DCs in the spleen, to identify the percentage of activated MHC II^+^ DCs.(J) FC on T-Ly in the TDLNs showing an increase in these populations.(K) Val-ILs-Combo affected the amount of Th17 and Tregs in the TDLNs.(L) Tumor growth volumes measured in vivo, following i.v. injection of Val-ILs-Combo and/or αPD-1 (200 μg), n = 8 mice per group.(M) FC plots and heatmaps showing the expression levels of the nine antigens targeted by Val-ILs-Combo on immunosuppressive cells, in the TME; mean normalized data of n = 5 mice.(N) Histogram showing the number of antigens targeted by the NPs and found repressed (p < 0.05) on various immune cell populations.The number of mice used per group is indicated on the figure; data are shown as means ± SD, p values are compared to Val-ILs-IgG and are calculated using two-tailed unpaired Student’s t test. See also Figure S2.
Next, when T lymphoma tumor volumes reached 50 mm^3^, mice were treated with two i.v. injections of 10^12^ NPs, on days 6 and 9, which significantly slowed tumor growth (Figure 2C). FC on day 12 showed a significant reduction in TaNKs, tDCs, myeloid cells, and CD4^+^ CD223^+^ T-Ly in the TME (Figure 2D). FC analysis also revealed a significant decrease in immunosuppressive cell populations in the TME, including Th17 cells, Tregs, MDSCs, and TAMs (Figure 2E).
Given the pronounced presence of Val-ILs in the spleen, we investigated the impact of the treatment on systemic immune cell populations. FC revealed that the same immunosuppressive populations were also significantly modulated in the spleen. Specifically, a reduction was observed in CD11b^+^ myeloid cells, DCs, NK cells, and CD4^+^ CD223^+^ T-Ly (Figure 2F). This was accompanied by a significant decrease in Th17 cells, Tregs, MDSCs, and macrophages (Figure 2G).
Additionally, the treatment with Val-ILs-Combo led to changes in the polarization of TAMs and splenic macrophages, as assessed by an increased proportion of M1-like cells (Figure 2H). We also observed a significant increase in major histocompatibility complex class II (MHC II) expression on both tDCs in the TME and splenic DCs (Figure 2I), suggesting enhanced antigen-presenting potential. Furthermore, in the TDLNs, we observed a significant increase in CD4^+^ and CD8^+^ T-Ly (Figure 2J) and a decrease in Th17 cells and Tregs (Figure 2K). This accumulation was accompanied by an increased expression of activation markers, including PD1, CD69, KI67, granzyme B (GZMB), and tumor necrosis factor alpha (TNFα), in tumor-infiltrating lymphocytes (TILs) within the TME. Elevated levels of some of these markers were also observed in TDLNs (Figure S2B). Following Val-ILs treatment, we noted a marked reduction in NK cells and DCs in both the TME and the spleen (Figures 2D and 2F). However, NPs treatment did not affect conventional DCs (cDC1 or cDC2 subsets) in either the TME or the spleen, whereas non-cDC populations were altered. Interestingly, a decrease in macrophage-like CD11c-high cells was detected exclusively in the TDLNs (Figure S2C). Moreover, NPs treatment did not alter the expression of proliferation, activation, or exhaustion markers on NK cells (Figure S2D).
Following intraperitoneal (i.p.) injection of αPD-1, 75% of the mice exhibited complete tumor regression with no detectable tumors over a 30-day observation period, while 25% developed tumors. Moreover, when mice were treated with Val-ILs-Combo in combination with αPD-1, none of the mice developed lymphoma (Figure 2L). When mean fluorescence intensity of the antigens targeted by Val-ILs-Combo was assessed by FC in the TME, the expression of the nine antigens was markedly decreased on innate immune cells, such as TaNKs and tDCs (Figure S2E), as well as on immunosuppressive cells, such as Th17 cells, Tregs, TAMs, and MDSCs (Figure 2M). However, few dysregulations were detected on TILs, such as B lymphocyte (B-Ly) and CD4^+^ and CD8^+^ T-Ly (Figure S2F). To better quantify the extent of the targeted modulation across immune subsets, we calculated the number of dysregulated markers per population (p < 0.05). Most of the affected markers were observed in the myeloid and immunosuppressive cell compartments (Figure 2N).
In conclusion, these findings underscore the therapeutic potential of Val-ILs-Combo in modulating the immune landscape and inducing tumor regression in a preclinical mouse model of T lymphoma. These results suggest that the therapeutic efficacy of Val-ILs-Combo is largely driven by the selective elimination of immune cells expressing high levels of the nine targeted antigens, particularly those contributing to the immunosuppressive TME.
Val-ILs impact B lymphoma progression and alter the expression of the targeted antigens on immune cells
To confirm the broader applicability of NPs, we evaluated their therapeutic efficacy in the A20 murine model to study B lymphoma,39^,^40 and s.c. transplanted immunocompetent BALB/c mice with 10^6^ syngeneic A20 cells. We then investigated the immune infiltration in the TME, using FC on day 15, when tumor volumes reached 1,000 mm^3^. Specific markers and UMAP unsupervised analysis methods were used to analyze the immune cell populations, revealing several distinct clusters. When evaluating the expression of the nine antigens targeted by Val-ILs-Combo, we found that most of the targets were predominantly expressed on a main population that contains TaNKs and tDCs cells (Figure 3A). Next, when B lymphoma tumor volumes reached 50 mm^3^, mice were treated via three i.v. injections of 10^12^ NPs on days 6, 9, and 12. Treatment with Val-ILs-Combo led to an important tumor growth reduction, with a tumor volume that never reached over 400 mm^3^ and 50% of tumor-free mice (Figure 3B). FC on day 15 showed an increase in CD4^+^ and CD8^+^ T-Ly in the TME (Figure 3C). A reduction among CD19^+^ cells was detected, likely reflecting the loss of A20 cells since they also express this marker. Despite the increased infiltration of T-Ly, only a few TILs activation markers were significantly upregulated in the TME (only CD69 and GZMB), but several markers were found activated in TDLNs (Figure S3A). UMAP analysis showed a marked reduction in myeloid populations, including tDCs, CD11b^+^ myeloid cells, and TaNKs in the TME (Figure 3C). Similar trends were observed in the spleen, with an increase in T-Ly and a decrease in DCs and CD11b^+^ myeloid cells (Figure 3D). The treatment also led to a reduction in Th17 cells, Tregs, MDSCs, TAMs, or macrophages in the TME (Figure 3E) and the spleen (Figure 3F). Importantly, FC also confirmed a significant reduction in the proportion of innate anti-tumor immune cells expressing the nine targeted antigens, including TaNKs and tDCs, while having minimal impact on T- and B-Ly (Figures 3G and S3B).Figure 3. Val-ILs affect immunosuppressive cells and B lymphoma growthBALB/c mice were s.c. injected with 10^6^ A20 cells. When tumors reached 50 mm^3^, mice were i.v. injected with Val-ILs (10^12^ NPs) on days 6, 9, and 12. Analyses were performed on day 15.(A) UMAP data frame and heatmap following FC showing the expression levels of the nine antigens targeted by Val-ILs-Combo among immune cells detected in the TME. Mean normalized data of n = 8 untreated mice.(B) NPs reduced the tumor volumes measured in vivo. Image of tumors isolated ex vivo, n = 8 mice per group, four did not develop tumors (black cross).(C) UMAP data showing that Val-ILs-Combo affected the repartition of immune cells in the TME.(D) NPs affected the repartition of immune cells in the spleen.(E and F) Val-ILs-Combo reduced the amount of Th17, Tregs, MDSCs, TAMs, or macrophages, in the TME (E) and the spleen (F).(G) Histogram showing the number of antigens targeted by the NPs and found repressed on the cell surface of various immune cell populations.(H) FC on TAMs in the TME and macrophages in the spleen, to identify M1-like and M2-like phenotypes.(I) FC on tDCs in the TME and DCs in the spleen, to identify the percentage of activated MHC II^+^ DCs.(J) FC on CD4^+^ and CD8^+^ T-Ly in the TDLNs showing an increase in these populations.(K) FC showing that Val-ILs-Combo reduced the amount of Th17 and Tregs in the TDLNs.(L) Tumor growth volumes measured in vivo following the injection of Val-ILs-Combo and/or αPD-1 (200 μg). Image of tumors isolated ex vivo, n = 8 mice per group.The number of mice used per group is indicated on the figure, data are shown as means ± SD, and p values are compared to Val-ILs-IgG and are calculated using a two-tailed unpaired Student’s t test. See also Figure S3.
Additionally, treatment with Val-ILs-Combo led to changes in the polarization of TAMs, as assessed by an increase in the proportion of M1-like TAMs and a decrease in M2-like TAMs in the TME. These modifications were also observed in the spleen (Figure 3H). Immune activation was further supported by an increase in tDCs and splenic DCs expressing MHC II (Figure 3I), as well as the increase in CD4^+^ and CD8^+^ T-Ly (Figure 3J), and reduced number of Th17 cells and Tregs in TDLNs (Figure 3K).
Following i.p. injection of αPD-1 at the onset of tumor development, 75% of the mice exhibited complete tumor regression, with no detectable tumors on day 15, while 25% of mice developed tumors. Moreover, when mice were treated with Val-ILs-Combo in combination with αPD-1, none of the mice developed B lymphoma (Figure 3L).
In conclusion, these findings highlight the therapeutic potential of Val-ILs-Combo in reshaping the immune landscape of B lymphoma in the TME, primarily by reducing several populations of immunosuppressive cells, modulating the polarization of TAMs toward an M1-like phenotype, enhancing DCs function, and increasing the number of TILs in the TME, ultimately leading to tumor regression.
Val-ILs modulate immunosuppressive cell populations in a breast cancer model
To further explore the potential of these Val-ILs, we investigated their therapeutic impact in a murine TNBC model. The 4T1 model, known for its highly immunosuppressive TME, is a robust preclinical model for studying aggressive, ICIs-resistant breast cancer and for identifying combination therapies.16^,^17^,^18^,^19^,^41^,^42 10^6^ syngeneic 4T1 cells were transplanted through orthotopic intraductal (i.d.) injection in the mammary gland of female BALB/c mice. Using FC and UMAP unsupervised analysis methods, we investigated the immune infiltration in the TME of breast tumors on day 15, when tumor volumes reached 1,000 mm^3^. After evaluating the expression of the nine antigens targeted by Val-ILs-Combo, we found that the immune populations exhibited heterogeneous distribution, highlighting the complexity of the TME in this model (Figure 4A).Figure 4. Val-ILs affect immunosuppressive cells in a mouse model of breast cancer10^6^ 4T1 cells were transplanted through i.d. injection into the mammary gland of female BALB/c mice. When tumors reached 50 mm^3^, mice were i.v. injected with Val-ILs (10^12^ NPs) on days 6, 9, and 12. Analyses were performed on day 15.(A) UMAP data frame and heatmap following FC showing the expression levels of the nine antigens targeted by the NPs among immune cells detected in the TME. Mean normalized data of n = 5 untreated mice.(B) NPs reduced the tumor volumes measured in vivo. Image of tumors isolated ex vivo, four did not develop tumors (black cross).(C) UMAP data showing that Val-ILs-Combo affected the repartition of immune cells in the TME.(D) NPs affected the repartition of immune cells in the spleen.(E and F) Val-ILs-Combo reduced the amount of Th17, Tregs, MDSCs, TAMs, or macrophages, in the TME (E) and the spleen (F).(G) Antigens targeted by the NPs found repressed on the cell surface of various immune cell populations.(H) FC on TAMs in the TME and macrophages in the spleen, to identify M1-like and M2-like phenotypes.(I) FC on tDCs in the TME and DCs in the spleen, to identify the percentage of activated MHC II^+^ DCs.(J) FC on CD4^+^ and CD8^+^ T-Ly in the TDLNs showing an increase in these populations.(K) FC showing that Val-ILs-Combo affected the amount of Th17 and Tregs in the TDLNs.The number of mice used per group is indicated on the figure, data are shown as means ± SD, and p values are compared to Val-ILs-IgG and are calculated using two-tailed unpaired Student’s t test. See also Figure S4.
When the tumor volumes reached 50 mm^3^, mice were treated with Val-ILs-Combo i.v. injected on days 6, 9, and 12. NPs led to an important tumor growth reduction (Figure 4B). FC showed a decrease in CD11b^+^ CD11c^+^ tDCs and an increase in T- and B-Ly in the TME, attesting to an immune response. We observed a reduced number of CD4^+^ CD223^+^ T-Ly in the TME (Figure 4C). In the spleen, we detected an increase in T-Ly and CD11c^+^ DCs (Figure 4D). Moreover, treatment led to a strong reduction in key immunosuppressive subsets, in both the TME and the spleen, including Th17 cells, Tregs, MDSCs, TAMs, and macrophages (Figures 4E and 4F). Importantly, FC confirmed a significant reduction in the expression of the nine antigens on these immunosuppressive cells, as well as on TaNKs and tDCs, while having minimal impact on T- and B-Ly (Figures 4G and S4A).
Polarization of TAMs shifted in favor of anti-tumor M1-like macrophages, while pro-tumor M2-like cells were significantly diminished in the TME, and this was also the case in the spleen (Figure 4H). This immune activation was supported by an increased presence of MHC II^+^ DCs in both the TME and the spleen (Figure 4I), as well as an accumulation of CD4^+^ and CD8^+^ T-Ly in the TDLNs (Figure 4J). This was accompanied by a reduction in Th17 cells and Tregs (Figure 4K) and an activation of CD4^+^ T-Ly in the TDLNs (Figure S4B).
In conclusion, these findings highlight the therapeutic potential of Val-ILs-Combo in reshaping the immune landscape in the mouse breast cancer 4T1 model, primarily by depleting immunosuppressive populations such as TAMs, MDSCs, Tregs, and Th17 cells; repolarizing TAMs toward an M1-like phenotype; enhancing DCs function; and promoting TILs in the TME, ultimately contributing to tumor regression. Given that this model is highly resistant to αPD-1,16^,^17 we decided to study the combined administration of Val-ILs-Combo associated with αPD-1.
Val-ILs and αPD-1 enhance anti-tumor immunity, reduce breast cancer growth, and limit metastasis
When tumors reached a volume of 50 mm^3^, mice were treated on days 6, 9, and 12 with i.v. injection of Val-ILs-Combo and/or i.p. injection of αPD-1. We observed that this combination overcame immunotherapy resistance, as evidenced by a tumor volume reduction over time (Figure 5A). By day 15, breast tumors were excised from the mice. Remarkably, while all untreated or αPD-1-treated mice developed tumors, 50% of the mice treated with Val-ILs-Combo were tumor-free. The combination therapy of Val-ILs-Combo and αPD-1 resulted in 62.5% of mice being cured, with the remaining mice exhibiting a reduced tumor volume (Figure 5B).Figure 5. Val-ILs and αPD-1 affect the growth and spread of breast tumor cells10^6^ 4T1 cells were transplanted through i.d. injection into the mammary gland of female BALB/c mice. When tumors reached 50 mm^3^, mice were i.v. injected with Val-ILs (10^12^ NPs) and i.p. injected with αPD-1 (200 μg) on days 6, 9, and 12. Analyses were performed on day 15 (A–E).(A) Tumor growth volumes measured in vivo.(B) Image of tumors isolated ex vivo from the mammary gland of mice (two independent in vivo experiments). Val-ILs-Combo was efficient in reducing the tumor volumes, and the combination of Val-ILs-Combo and αPD-1 was even more efficient. Mice without breast tumors are identified with a black cross.(C) FC on TAM cells in the TME and macrophages in the spleen, to identify M1-like and M2-like phenotypes.(D) Images of mice measured with in vivo imaging system technology showing the detection of bioluminescent 4T1 cells in the mammary gland, following treatment with Val-ILs-Combo and/or αPD-1.(E) FC was used to detect metastatic CD44^+^ GFP^+^ 4T1 cells spreading to distant organs following treatment. Examples of FC plots and a heatmap displaying the mean percentages. Undetected (ud) indicates the absence of detectable tumor 4T1 cells.(F) On day 30, tumor-free mice were rechallenged with a secondary injection of 10^6^ 4T1 cells into the contralateral mammary gland. Tumor growth was monitored to evaluate the establishment of a protective anti-tumor memory.The number of mice used per group is indicated on the figure, data are shown as means ± SD, and p values are compared to Val-ILs-IgG + αIgG or between the different groups and are calculated using one-way ANOVA with Tukey’s multiple comparisons test. See also Figure S5.
The combination of Val-ILs-Combo and αPD-1 demonstrated efficacy in repolarizing TAMs in the TME, as well as macrophages in the spleen toward an M1-like phenotype (Figure 5C). Compared to untreated control mice or those treated with αPD-1 or Val-ILs-Combo alone, the combination therapy significantly increased the percentage of CD4^+^ and CD8^+^ T-Ly in the TME (Figure S5A). We also observed a reduction in the expression of exhaustion markers (PD1, LAG3, and TIM3) among T-Ly in the TME of tumor-bearing mice (Figure S5B).
We generated 4T1 cells expressing luciferase and GFP through lentiviral infection and transplanted these cells into mice. When tumors reached 50 mm^3^, mice were treated with i.v. injection of Val-ILs-Combo and/or i.p. injection of αPD-1. 4T1 cells spontaneously metastasize to distant organs, such as the lung.16^,^17^,^19 Treatment with Val-ILs-Combo markedly reduced the number of KI67^+^ proliferative cells in lung tissues isolated on day 15 (Figure S5C). These findings indicate that NPs importantly altered the proliferation of metastatic cells in the lung. By day 15, in vivo bioluminescence imaging showed that 4T1 cells remained primarily localized at the primary tumor site in the mammary gland, with no detectable bioluminescent signal observed in distant organs. This suggests that metastatic cell numbers were below the detection threshold for bioluminescence in other parts of the body (Figure 5D). To more precisely quantify the extent of breast cancer cell spread, we performed FC on ex vivo isolated organs to detect GFP^+^ 4T1 cells. In the untreated control group, cancer cells were mostly detected in the lung, and to some extent also found in other organs (Figure S5D). While αPD-1 monotherapy had no significant effect on metastatic spread compared to untreated controls, treatment with Val-ILs-Combo alone significantly reduced GFP^+^ 4T1 dissemination into several organs. The combination therapy with Val-ILs-Combo and αPD-1 was even more effective, as demonstrated by the low number of GFP^+^ cancer cells in the spleen, the bone marrow, the blood, and the lung of cured mice (Figures 5E, S5E, and S5F).
To evaluate the establishment of a durable anti-tumor immune memory, cured mice were rechallenged on day 30 with an orthotopic injection of GFP^+^ 4T1 cells into the contralateral mammary gland. Notably, treatment with Val-ILs-Combo and αPD-1 completely prevented tumor development in 66% of the mice that had been cured following treatment (Figure 5F), indicating the induction of a protective immune memory capable of preventing tumor recurrence.
In conclusion, the combination therapy with Val-ILs-Combo and αPD-1 significantly impeded breast cancer development in the mammary gland, as well as the spread of metastatic breast cancer cells into further organs. This association efficacy is correlated with an increased polarization of TAMs in the TME and macrophages in the spleen from an M2-like to an M1-like phenotype, an increased presence of T-Ly in the TME, a reduced expression of exhaustion markers on these T-Ly, and the establishment of a protective immune memory.
In a lung cancer model, Val-ILs and αPD-1 enhance anti-tumor immunity, reduce cancer growth, and limit metastasis
We next investigated the NPs’ therapeutic impact in a Lewis lung carcinoma-1 (LLC1) tumor model resistant to several ICIs.18^,^43^,^44^,^45 To monitor tumor burden, we generated LLC1 cells expressing luciferase and GFP through lentiviral infection. A total of 10^6^ syngeneic LLC1 cells were i.v. injected into the tail vein of C57BL/6 mice that were treated, on days 6, 9, 12, and 15, with Val-ILs-Combo and/or αPD-1. We used FC to detect GFP-expressing LLC1 cells on day 28, which corresponds to the experimental endpoint when untreated mice exhibit respiratory distress. Whereas Val-ILs-Combo alone partially inhibited the presence of GFP^+^ tumor cells, αPD-1 monotherapy was completely ineffective. Importantly, the combined administration of Val-ILs-Combo and αPD-1 was very efficient, since 62.5% of the mice had less than 0.5% GFP^+^ tumor cells in the lung (Figure 6A), with a 17-fold reduction in proliferating KI67^+^ cells, as assessed by immunohistochemistry on the lung (Figure 6B). These findings indicate that Val-ILs-Combo, combined with αPD-1, may suppress the proliferation of LLC1 cells in the lung. In the untreated control group, GFP^+^ cancer cells were mostly detected in the lung, the liver, the kidney, and the blood (Figure 6C). αPD-1 monotherapy or Val-ILs-Combo administered alone had a low effect on metastatic spread compared to untreated controls. In sharp contrast, the combined administration of Val-ILs-Combo and αPD-1 was very efficient, as demonstrated by the important reduction in the detection of GFP^+^ cancer cells, in the lung, the kidney, the liver, and the blood, compared to αPD-1 monotherapy (Figure 6D). Mice treated with Val-ILs-Combo and αPD-1 exhibited a prolonged survival, with more than half of the treated mice remaining tumor-free for over 55 days (Figure S6). These results demonstrate that the combination therapy confers durable anti-tumor efficacy in the aggressive LLC1 model.Figure 6. Val-ILs and αPD-1 affect the growth and spread of the lung LLC1 tumor cellsMice were i.v. injected with 10^6^ LLC1 cells (day 0), i.v. injected with Val-ILs (10^12^ NPs), and i.p. injected with αPD-1 (200 μg) on days 6, 9, 12, and 15. Analyses were performed on day 28.(A) 62.5% of the mice treated with Val-ILs-Combo + αPD-1 did not show any expansion of cancer cells in the lung, while all mice treated with Val-ILs-Combo or αPD-1 alone showed an expansion of cancer cells. FC plots of GFP^+^ cancer cells detected in the lung.(B) Immunohistochemistry on lung sections, showing KI67 staining (brown-colored cells). Representative images and quantification of the proliferative KI67^+^ cells; 3 slides analyzed per mouse. Scale bars, 100 μm.(C and D) Distribution of GFP^+^ LLC1 cells detected by FC in different tissues of untreated mice (C) and in the different groups following treatments (D).The number of mice used per group is indicated on the figure, data are shown as means ± SD, and p values are compared to Val-ILs-IgG + αIgG or between the different groups and are calculated using one-way ANOVA with Tukey’s multiple comparisons test. See also Figure S6.
Using FC and UMAP unsupervised analysis, we next examined the immune infiltration in the lung TME on day 28. Our analysis revealed several distinct clusters of immune cells in the lung. We found that most of the targets were predominantly expressed on a main population containing CD11b^+^ cells (Figure 7A). FC demonstrated that combining treatments of Val-ILs-Combo and αPD-1 significantly increased the infiltration of T-Ly and CD11b^+^ myeloid cells in the lung TME. We furthermore detected a decrease among CD4^+^ CD223^+^ T-Ly (Figure 7B). When we assessed immune cells in the spleen, we observed that treatment with Val-ILs-Combo and αPD-1 significantly reduced DCs, while increasing CD4^+^ T-Ly, NK cells, and B-Ly. We also observed a decrease among CD4^+^ CD223^+^ T-Ly (Figure 7C). This combination therapy also led to a robust depletion of key immunosuppressive populations, in both the TME and the spleen (Figures 7D and 7E), including Th17 cells, Tregs, MDSCs, TAMs, or macrophages. We confirmed that the nine antigens targeted by the NPs were markedly repressed on these subsets of immunosuppressive cells, as well as on innate anti-tumor immune cells, including TaNKs and tDCs, while having minimal impact on T- and B-Ly (Figures 7F and S7A). While there was a reduction in the number of TAMs in the TME, we also noted that they shifted toward an M1-like pro-inflammatory phenotype (Figure 7G). The enhanced presence of MHC II^+^ DCs, in the TME and the spleen (Figure 7H) and the accumulation of T-Ly in the TDLNs (Figure 7I) support immune activation. Interestingly, we also characterized in the TDLNs a significant activation of T-Ly, as assessed by an increased expression of CD69, CD25, KI67, and TNFα (Figure S7B). In parallel, a significant reduction of both Th17 cells and Tregs (Figure 7J) was observed in the TDLNs, further supporting a shift toward an immunostimulatory environment.Figure 7. Val-ILs and αPD-1 affect immunosuppressive cells in the lung LLC1 tumor modelMice were i.v. injected with 10^6^ LLC1 cells (day 0), i.v. injected with Val-ILs (10^12^ NPs), and i.p. injected with αPD-1 (200 μg) on days 6, 9, 12, and 15. Analyses were performed on day 28.(A) UMAP data frame and heatmap following FC showing expression levels of the nine antigens targeted by Val-ILs-Combo among immune cells detected in the lung TME. Mean normalized data of n = 6 untreated mice.(B) NPs affected the number of immune cells in the lung TME.(C) NPs affected the repartition of immune cells in the spleen.(D and E) Val-ILs-Combo reduced the amount of Th17, Tregs, MDSCs, TAMs, or macrophages, in the TME (D) and the spleen (E).(F) Histogram showing the number of antigens targeted by the NPs and found repressed on the cell surface of various immune cell populations.(G) FC on TAMs in the TME and macrophages in the spleen, to identify M1-like and M2-like phenotypes.(H) FC on tDCs in the TME and DCs in the spleen, to identify the percentage of activated MHC II^+^ DCs.(I) FC on CD4^+^ and CD8^+^ T-Ly in the TDLNs showing an increase in these populations.(J) FC data showing that NPs reduced the amount of Th17 and Tregs in the TDLNs.The number of mice used per group is indicated on the figure, data are shown as means ± SD, and p values are compared to Val-ILs-IgG + αIgG and are calculated using one-way ANOVA with Tukey’s multiple comparisons test. See also Figure S7.
In conclusion, in the LLC1 model, Val-ILs-Combo alone showed partial efficacy by reducing immunosuppressive cells and moderately decreasing the proportion of GFP^+^ cells in organs. Co-administration with αPD-1 overcame this limitation, while this combination markedly reduced cancer cell dissemination in the lung and multiple distant metastatic sites. This synergistic combination is correlated with an important decrease in the immunosuppressive populations, such as TAMs, MDSCs, Tregs, and Th17 cells, repolarizing TAMs in the TME and macrophages in the spleen toward an M1-like phenotype, enhancing DCs function, and promoting T- and B-Ly in the TME.
Discussion
The NPs developed in this study were designed to target nine antigens predominantly expressed on immunosuppressive cell populations, including TAMs, MDSCs, Th17 cells, and Tregs. Because these NPs delivered the lipophilic prodrug valrubicin, treatment consistently reduced the abundance of these immunosuppressive populations across all four cancer models. While Val-ILs downregulated the expression of the nine markers on both M1-like and M2-like TAMs, they preferentially depleted M2-like TAMs (Figure S8). This finding suggests that the expression of these markers is functionally relevant for M2-like TAMs, whereas M1-like TAMs were spared, which explains the observed increase across the four cancer models, in the proportion of the pro-inflammatory anti-tumor M1-like TAMs. Val-ILs treatment exerted variable effects on innate anti-tumor immune cells, sometimes leaving their abundance unchanged, while in other cases increasing or decreasing it. Nevertheless, in all models, we consistently observed a broad reduction in the expression of the nine Val-ILs-targeted antigens, on both TaNKs and tDCs in the TME. Although these populations are generally considered as anti-tumor cells, some other subpopulations can also promote tumor growth.50^,^51^,^52 These findings therefore raise the possibility that the specific markers targeted by Val-ILs in our study may be associated with pro-tumor functions. We also consistently observed a significant increase in MHC II expression on tDCs in the TME, suggesting enhanced antigen-presenting potential. Finally, anti-tumor immune cells such as T- and B-Ly that barely expressed these antigens remained less affected, and we consistently observed an increased activation of CD4^+^ and CD8^+^ TILs and a reduction in specific exhaustion markers. These observations confirm that the NPs selectively target immunosuppressive cells without impairing key immune populations involved in anti-tumor responses.
Although lymphoma mouse models are responsive to ICIs, their combination with other treatments is often necessary to improve their effectiveness.38^,^39^,^40 In our study, we demonstrated that Val-ILs were effective in treating EL4 and A20 lymphomas, and this efficiency was related to a reduction in the proportion of immunosuppressive cells expressing high levels of the nine antigens targeted by the NPs. While 25% of the mice treated with αPD-1 alone developed T or B lymphomas, the combined treatment with NPs enhanced the success of αPD-1, as none of the treated mice developed lymphoma.
The 4T1 breast16^,^17^,^18^,^19^,^41^,^42 and LLC1 lung18^,^43^,^44^,^45 mouse cancer models are known for their aggressiveness, high risk of metastasis, and poor response to ICIs. In our study, we demonstrated that combining NPs with αPD-1 effectively inhibited tumor development in a substantial proportion of mice. While all mice treated with αPD-1 alone developed tumors, cancer progression was markedly attenuated in those receiving the combination therapy. Notably, in both models, this co-treatment significantly reduced metastatic spread to multiple organs, addressing a limitation of ICIs, which often show incomplete efficacy in preventing metastasis.53^,^54^,^55
The amount of valrubicin in NPs is extremely low, making the in vivo biodistribution by LC-MS/MS difficult to assess. Therefore, alternative techniques, such as radiolabeled NPs, are needed to study Val-ILs biodistribution and pharmacokinetics. Drug release from the NPs also needs to be further clarified to assess potential cytotoxic effects in vivo. In all four cancer models, NPs treatment did not affect the mice body weight over time (Figure S9). Furthermore, non-cancerous mice, receiving four NPs injections during the first 10 days, exhibited no long-term alterations in blood parameters, hematopoietic compartments in the bone marrow and the spleen, body weight, or organ histopathology after 30 days (Figure S10). Thus, while valrubicin is toxic upon i.v. administration, it can be safely delivered through Val-ILs. Antibody decoration can enhance targeting, but nonspecific protein corona formation may alter biodistribution and mask epitopes. In an ex vivo experiment, we observed that plasma did not impair the ability of Val-ILs to recognize and eliminate immunosuppressive cells, suggesting that potential nonspecific protein adsorption did not disrupt antigen recognition (Figure S11). These NPs preferentially bound to repressive cell populations within the spleen and the lymph nodes, likely due to antibody-mediated targeting. This selective tropism aligns with the high density of immunosuppressive cells in these organs, which facilitates recognition by Val-ILs conjugated with the nine specific antibodies.
The nine targeted antigens were expressed exclusively by immune cells and not by tumor cells. Consistent with this, Val-ILs had no direct cytotoxic effects on cancer cells in any of the four tested models (Figure S12). NPs screening validated two established therapeutic targets, CD223 (LAG-3)6^,^7^,^8^,^9^,^10^,^11^,^12 and CD309 (VEGFR2),20^,^21 whereas other immunotherapy targets (e.g., CTLA-4, PD-L1, and PD-1) were ineffective. This distinction among known targets may reflect the mechanism of Val-ILs, which eliminate cells directly rather than solely blocking receptor signaling. Notably, this screening also identified seven not-well-characterized targets: CD11b, CD365 (TIM1), CD200R3, CD204, CD49b, CD64, and CD170 (SIGLEC-F) that had not yet been fully recognized as key regulators of immune cell function and for which therapeutic options are lacking, making them promising candidates for further investigation (Table S1). Importantly, all murine targets identified in this study are also expressed across multiple human immune cell populations (Figure S13).
Our work identified Val-ILs as a promising technology due to their ease of preparation and their ability to reveal poorly characterized immunosuppressive markers. This opens avenues for therapeutic strategies and offers a potential approach to overcome resistance in hard-to-treat cancers. The discovery of these targets may also support the development of next-generation antibody-drug conjugates (ADCs). NPs reduce systemic toxicity by encapsulating therapeutic agents and enabling precise targeting, thereby limiting drug exposure to healthy tissues and offering a distinct advantage over ADCs or free-drug formulations. This also represents a crucial step toward clinical translation. Additionally, NPs can easily encapsulate lipophilic agents and allow surface functionalization with one or multiple targeting antibodies for versatile, multi-antigen targeting. By simultaneously targeting nine antigens, we demonstrated that Val-ILs can specifically modulate several pro-tumor immune cells and reshape the immune landscape, representing an efficient therapeutic approach distinct from previous ILs strategies that directly target cancer cells.56^,^57 Overall, our findings indicate that Val-ILs hold substantial preclinical promise with potential for clinical translation in cancer treatment.
Limitations of the study
This study identified nine targets, but additional candidates may emerge through the Val-ILs approach and broader screening. It remains to be determined whether targeting all nine antigens with Val-ILs provides superior efficacy compared with combinations directed against two or several of these antigens. Moreover, some targets are still poorly characterized in murine and human hematopoietic or immune cells, and their expression by suppressor populations is not yet well defined. For translational relevance, further prospective studies will be required to investigate these targets in greater detail.
We successfully tracked Val-ILs by incorporating a dye into the formulation and demonstrated via FC that i.v.-injected Val-ILs preferentially bind immunosuppressive cells in the spleen and the lymph nodes. While informative, FC on dissociated tissues may underestimate overall organ exposure to valrubicin and is susceptible to processing-related losses. Consequently, pharmacokinetics, biodistribution, and drug release kinetics under physiological conditions warrant further investigation. In particular, following i.v. injection, the composition of the protein corona on NPs requires deeper examination due to its potential impact on cell-targeting specificity and efficiency, drug release behavior, and overall pharmacokinetics.
A deeper characterization of the immune response to Val-ILs therapy is also needed. Relevant approaches could include cross-presentation assays, activation of naive antigen-specific T cells, interferon signature analysis, cytokine and chemokine profiling, evaluation of checkpoint and coinhibitory ligands on myeloid cells, as well as spatial methods such as multiplexed immunohistochemistry to assess effector cell proximity to tumor nests and myeloid niches.
Resource availability
Lead contact
Further information and requests for resources and data should be directed to and will be fulfilled by the lead contact, Dr. Ronan Quéré ([email protected]).
Materials availability
There are restrictions on the availability of Val-ILs NPs due to the lack of an external centralized repository for their distribution and the need to maintain internal stock. Val-ILs NPs generated in this study will be made available upon request; however, processing and shipping fees may apply, and a completed materials transfer agreement may be required.
Data and code availability
- •All data reported in this paper will be shared by the lead contact upon request.
- •This paper does not report any original code.
- •Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.
Acknowledgments
This work was supported by grants from the Fondation d'Entreprise Bristol Myers Squibb pour la Recherche en Immuno-Oncologie to R.Q., the Conférence de Coordination Interrégionale Est (CCIR Est) de la Ligue contre le Cancer to R.Q., and the 2024 prize from the Ligue contre le Cancer, comité de Côte-d’Or, awarded to R.Q. This work has been supported by a fellowship from the LipSTIC LabEx (ANR-11-LABX-0021) and the Région Bourgogne-Franche-Comté, as well as by the INTHERAPI, SFRI INTEGRATE Graduate School (ANR-20-SFRI-0008) and the Région Bourgogne-Franche-Comté. The team of C.G. holds the label d’excellence Ligue Nationale contre le Cancer and is funded by the Fonds Européen de Développement Régional (FEDER) and the Fondation Ruban Rose. We thank Valérie Saint-Giorgio and Sébastien Lapipe from the animal housing facility at the Université Bourgogne Europe. We also thank the members of the UMS58 BioSanD facilities at the Université Bourgogne Europe for their valuable technical support and insightful suggestions. UMS58 receives support from the Région Bourgogne-Franche-Comté. The authors are grateful to Laure Avoscan from the DImaCell imaging facility for conducting the TEM experiment, as well as to Hélène Choubley and Jean-Paul Pais de Barros from the DiviOmics platform for performing the LC-MS/MS study. The authors thank the staff of the ImaFlow facility for their contributions to immunohistochemistry and flow cytometry support, and in particular Audrey Geissler for her work on the immunohistochemistry.
Author contributions
A.G. led and performed most of the experiments. A.G. and R.Q. analyzed the data, designed the study, wrote the first draft of the manuscript, and prepared the figures. N.B. helped with the ex vivo experiments, flow cytometry studies, and manuscript editing. M.B. contributed to data discussion and manuscript editing. C.P. helped with the ex vivo experiments. P.-S.B. helped with the bioluminescence study. C.G. contributed to discussion and manuscript editing. R.Q. contributed to funding acquisition and project supervision and wrote and edited the revised manuscript. All authors have read and approved the final version of the manuscript.
Declaration of interests
R.Q. and A.G. are inventors on an international patent (WO2025202213), covering lipid nanoparticles loaded with anti-tumor agents and functionalized to target immunosuppressive cells.
STAR★Methods
Key resources table
REAGENT or RESOURCESOURCEIDENTIFIERAntibodiesPE-CF594 anti-Mouse CD69 (Clone H1.2F3)BD BiosciencesCat# 562455; RRID: AB_11154217PerCP anti-Mouse CD8a (Clone 53–6.7)BD BiosciencesCat# 553036; RRID: AB_394573BUV395 anti-Mouse CD45 (Clone 30-F11)BD BiosciencesCat# 564279; RRID: AB_2651134BV421 anti-Mouse CD11b (Clone M1/70)BD BiosciencesCat# 562605; RRID: AB_11152949BV421 anti-Mouse CD45 (Clone 30-F11)BD BiosciencesCat# 563890; RRID: AB_2651151BV605 anti-Mouse/Rat/Human Ki67 (Clone B56)BD BiosciencesCat# 567122; RRID: AB_2916454APC-R700 anti-Mouse CD11b (Clone M1/70)BD BiosciencesCat# 564985; RRID: AB_2739033BV480 anti-Mouse CD49b (Integrin α2 chain) (Clone HMα2)BD BiosciencesCat# 746355; RRID: AB_2743674BV605 anti-Mouse Siglec-F (Clone E50-2440)BD BiosciencesCat# 740388; RRID: AB_2740118BV650 anti-Mouse CD44 (Clone IM7)BD BiosciencesCat# 740455; RRID: AB_2740182BUV395 anti-Mouse CD223 (Clone C9B7W)BD BiosciencesCat# 745693; RRID: AB_2743177BUV805 anti-Mouse CD3ε (Clone 145-2C11)BD BiosciencesCat# 749276; RRID: AB_2873651APC anti-Mouse CD80 (Clone16-10A1)BD BiosciencesCat# 560016; RRID: AB_1645212Alexa Fluor 700 anti-Mouse CD11c (Clone HL3)BD BiosciencesCat# 560583; RRID: AB_1727421FITC anti-Mouse CD25 (Clone 3C7)BD BiosciencesCat# 558689; RRID: AB_1645241APC-Cy7 anti-Mouse CD279 (PD-1) (Clone 29F.1A12)BioLegendCat# 135224; RRID: AB_2563523BV510 anti-Mouse TNFα (Clone MP6-XT22)BioLegendCat# 506339; RRID: AB_2563127BV650 anti-Mouse CD223 (LAG3) (Clone C9B7W)BioLegendCat# 125227; RRID: AB_2687209BV711 anti-Mouse CD206 (MMR) (Clone C068C2)BioLegendCat# 141727; RRID: AB_2565822BV785 anti-Mouse CD366 (TIM-3) (Clone RMT3-23)BioLegendCat# 119725; RRID: AB_2716066BV785 anti-Mouse NK1.1 (Clone PK136)BioLegendCat# 108749; RRID: AB_2564304PE-Cy7 anti-Mouse F4/80 (Clone BM8)BioLegendCat# 123114; RRID: AB_893478BV711 anti-Mouse Il-17 (Clone TC11-19H10.1)BioLegendCat# 506921; RRID: AB_2125011PerCP-Cy5.5 anti-Mouse CD127 (Clone A7R34)eBioscienceCat# 45-1271-82; RRID: AB_1106998PerCP-eFluor710 anti-Mouse CD172 (SIRPα) (Clone P84)eBioscienceCat# 46-1721-82; RRID: AB_10804639APC anti-Mouse TIM-1 (Clone REA692)Miltenyi BiotecCat# 130-110-427; RRID: AB_2654157APC-Vio770 anti-Mouse CD11c (Clone REA754)Miltenyi BiotecCat# 130-110-841; RRID: AB_2654715PE anti-Mouse CD200R3 (Clone REA128)Miltenyi BiotecCat# 130-102-486; RRID: AB_2656120PE anti-Mouse/Rat/Human Granzyme B (Clone REA226)Miltenyi BiotecCat# 130-116-486; RRID: AB_2727564PerCP-Vio700 anti-Mouse CD11b (Clone REA592)Miltenyi BiotecCat# 130-113-809; RRID: AB_2751174PE-Vio615 anti-Mouse CD309 (VEGFR-2) (Clone REA1116)Miltenyi BiotecCat# 130-119-435; RRID: AB_2732937PE-Vio770 anti-Mouse CD64 (Clone REA286)Miltenyi BiotecCat# 130-119-659; RRID: AB_2751786VioBlue anti-Mouse CD204 (Clone REA148)Miltenyi BiotecCat# 130-103-019; RRID: AB_2656174VioBright B515 anti-Mouse CD19 (Clone REA749)Miltenyi BiotecCat# 130-112-040; RRID: AB_2751063VioBright FITC anti-Mouse CD25 (clone 7D4)Miltenyi BiotecCat# 130-119-660; RRID: AB_2751787VioBright R720 anti-Mouse CD19 (Clone REA749)Miltenyi BiotecCat# 130-128-560; RRID: AB_2921892VioBlue anti-Mouse CD4 (Clone REA604)Miltenyi BiotecCat# 130-118-696; RRID: AB_2733942VioGreen anti-Mouse CD4 (Clone REA604)Miltenyi BiotecCat# 130-118-693; RRID: AB_2734087PerCP-Vio700 anti-Mouse Gr-1 (Clone RB6-8C5)Miltenyi BiotecCat# 130-102-171; RRID: AB_2659870PE anti-Mouse CD64 (Clone REA286)Miltenyi BiotecCat# 130-118-684; RRID: AB_2751553PE anti-Mouse CD25 (Clone REA568)Miltenyi BiotecCat# 130-120-696; RRID: AB_2752168APC anti-Mouse CD44 (Clone REA664)Miltenyi BiotecCat# 130-119-121; RRID: AB_2751628VioGreen anti-Mouse MHC Class II (Clone REA813)Miltenyi BiotecCat# 130-112-395; RRID: AB_2652910FITC anti-Mouse F4/80 (Clone REA126)Miltenyi BiotecCat# 130-117-509; RRID: AB_2727970APC-Vio770 anti-Mouse XCR1 (Clone REA707)Miltenyi BiotecCat# 130-111-187; RRID: AB_2654291BV421 anti-Mouse CD45 (Clone 30-F11)BD BiosciencesCat# 563890; RRID: AB_2651151APC-R700 anti-Mouse CD11b (Clone M1/70)BD BiosciencesCat# 564985; RRID: AB_2739033BV480 anti-Mouse CD49b (Integrin α2 chain) (Clone HMα2)BD BiosciencesCat# 746355; RRID: AB_2743674BV650 anti-Mouse CD223 (LAG3) (Clone C9B7W)BD BiosciencesCat# 740560; RRID: AB_2740261RB545 anti-Mouse Siglec-H (Clone 440c)BD BiosciencesCat# 756214; RRID: AB_3688452BV605 anti-Mouse CD274 (PD-L1, B7-H1) (Clone10F.9G2)BioLegendCat# 124321; RRID: AB_2563635APC-Cy7 anti-Mouse CD279 (PD-1) (Clone 29F.1A12)BioLegendCat# 135224; RRID: AB_2563523BV711 anti-Mouse CD206 (MMR) (Clone C068C2)BioLegendCat# 141727; RRID: AB_2565822BV605 anti-Mouse CD274 (Clone 10F.9G2)BioLegendCat# 124321; RRID: AB_2563635FITC anti-Mouse CD182 (Clone SA044G4)BioLegendCat# 149310; RRID: AB_2566148PE-Cy7 anti-Mouse Il-17 (Clone TC11-19H10.1)BioLegendCat# 506921; RRID: AB_2125011PE anti-Mouse CD152 (Clone UC10-4B9)Miltenyi BiotecCat# 130-102-570; RRID: AB_2655252PE anti-Mouse CD200R3 (Clone REA128)Miltenyi BiotecCat# 130-102-486; RRID: AB_2656120PE anti-Mouse CD204 (Clone REA148)Miltenyi BiotecCat# 130-120-811; RRID: AB_2784059PE anti-Mouse CD309 (VEGFR-2) (Clone REA1116)Miltenyi BiotecCat# 130-119-433; RRID: AB_2733730PE anti-Mouse CD366 (TIM-3) (Clone REA602)Miltenyi BiotecCat# 130-118-691; RRID: AB_2733707PE anti-Mouse CD64 (Clone REA286)Miltenyi BiotecCat# 130-118-684; RRID: AB_2751553PE anti-Mouse Ly-6G (Clone REA526)Miltenyi BiotecCat# 130-123-780; RRID: AB_2811549PE anti-Mouse Siglec-F (Clone REA798)Miltenyi BiotecCat# 130-112-332; RRID: AB_2653439PE anti-Mouse TIM-1(Clone REA692)Miltenyi BiotecCat# 130-110-325; RRID: AB_2654156PE anti-Mouse CD115 (Clone REA827)Miltenyi BiotecCat# 130-112-828; RRID: AB_2654552APC anti-Mouse Ly6C (Clone REA796)Miltenyi BiotecCat# 130-111-917; RRID: AB_2652804VioBright FITC anti-Mouse CD25 (clone 7D4)Miltenyi BiotecCat# 130-119-660; RRID: AB_2751787PE anti-Mouse CD36 (Clone REA1184)Miltenyi BiotecCat# 130-122-090; RRID: AB_2784160PE anti-Mouse JAML (Clone REA862)Miltenyi BiotecCat# 130-114-679; RRID: AB_2726744PE anti-Mouse CD141 (Clone REA964)Miltenyi BiotecCat# 130-116-094; RRID: AB_2727343PE anti-Mouse CD24 (Clone REA742)Miltenyi BiotecCat# 130-110-826; RRID: AB_2656538PerCP-Vio 700 anti-Mouse Gr-1 (Clone RB6-8C5)Miltenyi BiotecCat# 130-102-171; RRID: AB_2659870PE-Vio 770 anti-Mouse Hsp-70(Clone REA349)Miltenyi BiotecCat# 130-125-096; RRID: AB_2857739PE anti-Mouse CD244.2 (Clone REA388)Miltenyi BiotecCat# 130-105-988; RRID: AB_2656629PE anti-Mouse CD31 (Clone REA784)Miltenyi BiotecCat# 130-111-540; RRID: AB_2657296PE anti-Mouse F4/80 (Clone REA126)Miltenyi BiotecCat# 130-116-499; RRID: AB_2727574PE anti-Mouse CD123 (Clone REA114)Miltenyi BiotecCat# 130-102-869; RRID: AB_2654774APC anti-Mouse CD44 (Clone REA664)Miltenyi BiotecCat# 130-119-121; RRID: AB_2751628APC-R700 anti-Mouse IgG2b, κ (Clone R35-38)BD BiosciencesCat# 564984; RRID: AB_2869634BV650 anti-Mouse IgG1, κ (Clone R3-34)BD BiosciencesCat# 563848; RRID: AB_2869526PE anti-mouse IgG1 (Clone REA1017)Miltenyi BiotecCat# 130-117-098; RRID: AB_2733864APC anti-mouse IgG1 (Clone REA1017)Miltenyi BiotecCat# 130-117-099; RRID: AB_2733425Rabbit anti-Mouse/Rat/Human Ki-67 antibody (Clone SP6)Thermo Fisher ScientificCat# 11365083; RRID: AB_10979488Rat anti-mouse PD-1 (Clone RMP1-14)BioXCellCat# BX-BE0146; RRID: AB_10949053Rat IgG2a isotype control (Clone 2A3)BioXCellCat# BX-BE0089; RRID: AB_1107769Chemicals, peptides, and recombinant proteinsFixable Viability Stain 440UVBD BiosciencesCat# 566332; RRID: AB_2869748Fixable Viability Stain 700BD BiosciencesCat# 564997; RRID: AB_2869637Collagenase Type CLS-1SERLABOCat# WOLS04194Collagenase Type CLS-2SERLABOCat# WOLS04174Collagenase Type CLS-4SERLABOCat# WOLS04186DNase IMerckCat# 04536282001L-α-phosphatidylcholineMerckCat# 840054PCholesterolMerckCat# C8667DSPE-PEG-square (2000)MerckCat# 880136PDSPE-PEG-NHS (5000)SINOPEGCat# 06030500706ValrubicinMerckCat# SML2516DMSOMerckCat# D5879ChloroformMerckCat# 288306GlycineEuromedexCat# 26-128-6405-CD-luciferinMerckCat# L9504Lipofectamine 2000Thermo Fisher ScientificCat# 11668027DMEM mediumMerckCat# D6429RPMI-1640 mediumDominique DutscherCat# L0500-500Penicillin-Streptomycin-Amphotericin (PSA)Pan BiotechCat# P06-07300Fetal Bovine Serum (FBS)Dominique DutscherCat# 500105N1Nβ-mercaptoethanolThermo Fisher ScientificCat# 21985023Hank’s Balanced Salts Solution (HBSS)Dominique DutscherCat# L0611-500Trypsin-EDTADominique DutscherCat# X0930-100Phosphate-buffered salineDominique DutscherCat# X0515-500Pyruvate sodiumThermo Fisher ScientificCat# 113600704-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES)Thermo Fisher ScientificCat# 15630080L-glutamineThermo Fisher ScientificCat# 25030149Minimal Essential Medium (MEM)Thermo Fisher ScientificCat# 11140035Critical commercial assaysCytofix/Cytoperm Fixation/Permeabilization KitBD BiosciencesCat# 554714HRP anti-rabbit IgG polymer detection kitVector LaboratoriesCat# MP-7401Vector NovaRED substrate kitVector LaboratoriesCat# SK-4800PKH67 Green Fluorescent Cell Linker Mini Kit for General Cell Membrane LabelingMerckCat# MINI67MycoAlert Detection KitLonza-BioscienceCat# LT0710% buffered formalin (CellStor Pot)CellpathCat# BAF-6000-08AExperimental models: Cell linesMurine T lymphoma EL4 cellsATCCCat# TIB-39Murine B lymphoma A20 cellsATCCCat# TIB-208Murine breast cancer 4T1 cellsATCCCat# CRL-2539Murine Lewis lung carcinoma LLC1 (LL/2) cellsATCCCat# CRL-1642Human embryonic kidney (HEK) 293T cellsATCCCat# CRL-3216Experimental models: Organisms/strainsMouse: C57BL/6EnvigoN/AMouse: BALB/cJanvierN/ARecombinant DNApCCLc-MNDU3-Luciferase-PGK-EGFP-WPRE plasmid vectorAddgeneCat# 89608; RRID: Addgene_89608PAX2 plasmid vectorAddgeneCat# 12260; RRID: Addgene_12260pCMV-VSV-G plasmid vectorAddgeneCat# 8454; RRID: Addgene_8454Software and algorithmsFlowJo (v. 10.8.1)Tree Starhttps://www.flowjo.comUMAP pluginTree Starhttps://www.flowjo.comFlowSOM pluginTree Starhttps://www.flowjo.comGraphPad Prism (v. 10.2.3)GraphPadhttps://www.graphpad.comLiving Image software (v. 4.0)PerkinElmerhttps://www.perkinelmer.comZen software (v. 3.9)Zeisshttps://www.zeiss.com/microscopyBioRenderBioRenderhttps://www.biorender.comGIMP (v. 3.0.4)GIMPhttps://www.gimp.orgOtherVal-ILs nanoparticlesThis paperN/A300 kDa dialysis membranesThermo Fisher ScientificCat# 11550970Amicon Centrifugal Filter Unit 100 kDa cutoffMilliporeCat# UFC910096AxioScope microscopeZeisshttps://www.zeiss.com/microscopyAurora flow cytometerCytekhttps://cytekbio.com/pages/auroraVideodropMyriade Labhttps://www.myriadelab.comHemocytometer (Vet ABC+).SCILhttps://www.scilvet.fr/produits/diagnostic-de-laboratoire/hematologie/scil-vet-abc-plusNanoSight NS300Malvern Panalyticalhttps://www.malvernpanalytical.com
Experimental model and study participant details
Establishment of in vivo cancer models
In this study, we utilized the murine T lymphoma EL4 cells (ATCC, TIB-39, C57BL/6 strain), the murine B lymphoma A20 cells (ATCC, TIB-208, BALB/c strain), the murine breast cancer 4T1 cells (ATCC, CRL-2539, BALB/c strain) and the murine Lewis lung carcinoma LLC1 (LL/2) cells (ATCC, CRL-1642, C57BL/6 strain). EL4, 4T1 and LLC1 cells were cultured in DMEM medium (Merck), while A20 cells were cultured in RPMI-1640 medium (Dominique Dutscher) supplemented with 0.05 mM 2-mercaptoethanol (Thermo Fisher Scientific). All media were supplemented with 10% FBS (Dominique Dutscher) and PSA (Pan Biotech). For 4T1 and LLC1 cells, cell dissociation with trypsin-EDTA (Dominique Dutscher) was used in HBSS (Dominique Dutscher) less than 2 h before the in vivo injection. A large stock of low passage frozen cells was prepared. Cells were further authenticated by morphology, growth and pattern of metastasis in vivo and routinely screened for mycoplasma contamination every six months, using the MycoAlert Detection Kit (Lonza-Bioscience). 10^6^ EL4 or A20 cells suspended in 100 μL of physiological saline solution were subcutaneously (s.c.) injected into the flanks of C57BL/6 mice (Envigo) or BALB/c mice (Janvier), respectively. For the 4T1 breast cancer model, 10^6^ cells in 100 μL of physiological saline solution were administered via intraductal (i.d.) injection into the mammary gland of female BALB/c mice (Janvier). For the challenge study, 4T1 cells were injected in the contralateral mammary gland. For the LLC1 lung carcinoma model, 10^6^ LLC1 cells in 300 μL of physiological saline solution were i.v. injected into C57BL/6 mice (Envigo). We developed cancer models with stable expression of luciferase and GFP by infecting EL4, LLC1 and 4T1 cell lines with lentivirus produced in the human embryonic kidney (HEK) 293T cells (ATCC, CRL-3216), following transduction with Lipofectamine 2000 (Thermo Fisher Scientific) using the pCCLc-MNDU3-Luciferase-PGK-EGFP-WPRE plasmid vector, along with the PAX2 and pCMV-VSV-G plasmid vectors (Addgene). After 48 h, viral supernatants were collected, filtered, and applied to cancer cells. Three days later, lentivirally transduced GFP^+^ cells were sorted using a FACSAria III cell sorter (BD Biosciences), cultured in vitro, then transplanted in vivo into C57BL/6 or BALB/c mice. All animal experiments were approved by the institutional review board of the animal ethics committee at the Université Bourgogne Europe (Dijon, France), and by the French ministry of higher education and research, under references APAFIS #39295-2022111415299765-v3, #44570-2023090410159399-v5, #49842-2024061213454031-v4 and #53837-2025022110398352-v4.
Method details
ILs loaded with valrubicin
Following a previously described protocol,49 L-α-phosphatidylcholine (Merck), cholesterol (Merck), DSPE-PEG-square (2000) (Merck) and DSPE-PEG-NHS (5000) (SINOPEG) were dissolved in chloroform (Merck) at a molar ratio of 100:37:1.5:0.2, to a final volume of 26 μL. To this lipid mixture, 26 nmol of valrubicin (Merck), reconstituted in DMSO (Merck), was added. Chloroform was evaporated under a nitrogen stream to form a thin lipid film in a round-bottom flask by removing the organic solvent. The lipid film was then hydrated with 1 mL of filtered 1× PBS (Dominique Dutscher), preheated to 65°C. After 10 min of incubation at 65°C, the mixture was vortexed for 1 min and stabilized for 30 min at the same temperature. The vesicles were subsequently sonicated for five cycles (30 seconds on/off) at an amplitude of 20% to form UV, which were stabilized for an additional 30 min at 65°C. NPs were then conjugated with 2 μg of an antibody, defined in key resources table. We also generated Val-ILs-IgG without specific antibody functionalization, conjugated with isotype IgG antibodies. The reaction was halted by the addition of 1 mM glycine (Euromedex). Purification of Val-ILs was carried out through two consecutive dialyses in one liter of 1× PBS using 300 kDa dialysis membranes (Thermo Fisher Scientific). The first dialysis lasted 2 h, followed by an overnight dialysis. Following this process, we generated individual Val-ILs, each loaded with a specific antibody, and verified antibody loading by FC. For the Val-ILs-Combo, the individual preparations were pooled and concentrated using an Amicon centrifugal filter unit (100 kDa cutoff, Millipore) through centrifugation at 7,000 g for 40 min at 4°C. We prepared Val-ILs labeled with the PKH67 green fluorescent dye (Merck) using a previously described protocol.49 The size and concentration of Val-ILs were determined by NTA using a NanoSight NS300 (Malvern Panalytical) or a Videodrop (Myriade Lab). The concentration of valrubicin in Val-ILs was quantified by LC-MS/MS, by the DiviOmics platform (UMS58 BioSanD) at the Université Bourgogne Europe, as previously described.49 Transmission electron microscopy (TEM) was performed following previously described protocol,49 with imaging conducted at the DImaCell facility at the Université Bourgogne Europe.
In vivo treatment studies in cancer mouse models
Male and female mice were randomly assigned to experimental groups, shaved, and treated intravenously (i.v.) via a tail vein injection with Val-ILs at a dose of 10^12^ NPs (approximately 100 μL of the concentrated preparation at 10^13^ NPs/mL) in 300 μL of physiological saline solution. For EL4, A20 and 4T1 models, tumor treatments started when tumors reached 50 mm^3^. Val-ILs were i.v. administered at the time points indicated in the figures as well as in the figure legends. No blinding methods were employed during the injections. In all models, co-treatment with an αPD-1 antibody (BioXCell) or an isotype control antibody was conducted alongside Val-ILs injections. αPD-1 was administered intraperitoneally (i.p.) at a dose of 200 μg per mouse in 300 μL of physiological saline solution, at the time points indicated in the figures as well as in the figure legends. Tumor growth was monitored in all experimental groups using caliper measurements, and mouse body weight was recorded concurrently. Volume of the tumor was determined by the formula measuring elliptical shape, length × width2/2. A difference in tumor volume measurement can be observed in vivo and ex vivo, which is due to the adipose tissue as well as the skin and the dermis. This difference in tumor volume measurement is particularly observed for the 4T1 model, in the mammary gland. Val-ILs labeled with the PKH67 green fluorescent dye (Merck) were i.v. injected into mice. UV were isolated from multiple organs ex vivo, 18 h post injection, and PKH67^+^ Val-ILs were detected by FC, using a previously described protocol.49 We also assessed by FC the binding of PKH67^+^ Val-ILs on further immune cell populations in the TME, the spleen and the TDLNs. The percentage of GFP^+^ cancer cells was determined by FC in several organs isolated ex vivo. Expression of the nine markers was characterized ex vivo in the TME by FC, on GFP^+^ tumor cells and CD45^+^ GFP^−^ immune cells. For the 4T1 model, following anesthesia with isoflurane, animals were imaged 15 min after the i.p. injection of D-luciferin (Merck) at a dose of 150 mg/kg body weight. Imaging was acquired using the in vivo imaging system (IVIS, Lumina III system). Images of bioluminescence were analyzed with Living Image software (PerkinElmer) at the Imathera platform (Center Georges-François Leclerc-Unicancer, Dijon).
KI67 immunohistochemistry in lung tissues
Immunohistochemistry was performed by the ImaFlow platform (UMS58 BioSanD) at the Université Bourgogne Europe. Lung tissues were collected and fixed in 10% buffered formalin (CellStor Pot, Cellpath) for 48 h, processed and embedded in paraffin. Thick sections were cut from paraffin-embedded samples. We used an anti-KI67 antibody (1:100, clone SP6, Thermo Fisher Scientific), with a secondary anti-rabbit antibody, conjugated with Horseradish Peroxidase (ImmPRESS HRP anti-rabbit IgG polymer detection kit, Vector Laboratories) and substrate (Vector NovaRED substrate kit, Vector laboratories). Images were acquired using an AxioScope microscope (Zeiss) and processed with Zen software (Zeiss).
Safety evaluation of Val-ILs in non-tumor-bearing mice
Mice that were not transplanted with cancer cells received four injections of Val-ILs on days 0, 3, 6 and 9, and organs were harvested on day 30. Peripheral blood was collected via tail vein sampling, and hematopoietic cells as well as blood parameters were quantified using a hemocytometer (Vet ABC+, SCIL) (Figure S10A). Flow cytometry analysis was assessed on hematopoietic cell populations, within the spleen and the bone marrow (Figure S10B). Mice body weight was monitored (Figure S10C). Organs were collected and fixed in 10% buffered formalin (CellStor Pot, Cellpath) for 48 h, processed and embedded in paraffin. Thick sections were cut from paraffin-embedded samples. For histopathological analysis, tissues were stained with hematoxylin and eosin (H&E). Images were acquired using an AxioScope microscope (Zeiss) and processed with Zen software (Zeiss). Histopathological analysis was carried out by the ImaFlow platform (UMS58 BioSanD) at the Université Bourgogne Europe (Figure S10D).
Functional preservation of antibody-mediated targeting of Val-ILs assessed by cell-binding assay
To assess whether plasma influences Val-ILs activity following i.v. injection in mice, Val-ILs were incubated with plasma isolated from the blood (Figure S11). Mouse plasma was obtained by cardiac puncture, followed by centrifugation of the whole blood at 800 g for 30 min at room temperature. Plasma was used either untreated or after heat inactivation (55°C, 1 h). Val-ILs were incubated for 1 h with 1× PBS, plasma, or heat-inactivated plasma. Subsequently, immune cells isolated from the spleens and the lymph nodes of mice were treated with 4,000 Val-ILs per cell under each condition. After 72 h of culture in RPMI-1664 medium, supplemented with 10% FBS, 1% PSA, 1× MEM (Thermo Fisher Scientific), 2 mM L-glutamine (Thermo Fisher Scientific), 10 mM HEPES (Thermo Fisher Scientific), 1 mM pyruvate sodium (Thermo Fisher Scientific) and 5.5 nM β-mercaptoethanol (Thermo Fisher Scientific), FC was performed to evaluate Val-ILs-induced cell death on immune cells.
Val-ILs exert no direct cytotoxic effects on cancer cells
To demonstrate that Val-ILs exert no direct cytotoxic effects on the four cancer models tested in this study (Figure S12), we followed a previously described in vitro protocol.49
Expression of antigens targeted by Val-ILs on ortholog human immune cells
The nine murine targets identified through Val-ILs screening each have corresponding human orthologs (Figure S13). To assess their expression in human immune cells, expression values across multiple human immune cell populations were retrieved from the Immune Cell Atlas database.
Flow cytometry
To perform FC on immune cells, mice were euthanized when untreated control group tumors reached the endpoint volume. Spleens were mechanically dissociated, purified through 70 μm separation filters, and treated with a hemolysis solution. The resulting cells were washed with 1× PBS and centrifuged at 500 g. We isolated the inguinal TDLNs (EL4, A20, 4T1) or the tracheobronchial TDLNs (LLC1) that were mechanically dissociated, purified through 70 μm separation filters, and centrifuged at 500 g. Tumors (EL4, A20, 4T1) and lung (LLC1) were cut into 3 to 4 mm pieces using a sterile scalpel and placed in 2.5 mL of dissociation buffer containing 60 U/mL Collagenase Type 1, 30 U/mL Collagenase Type 2, 60 U/mL Collagenase Type 4 (SERLABO) and 25 μg/mL DNase I (Merck) in filtered 1× PBS. The tumor-dissociation mixture was incubated at 37°C with agitation for 45 min. Following incubation, the resulting cell suspension was purified through 70 μm separation filters and centrifuged at 500 g. Antibodies used to analyze immune cell markers are listed in the key resources table. Cell viability was assessed using Fixable Viability Stains (FVS440UV, FVS700, diluted 1:1000, BD Biosciences). For cancer models transplanted with cells expressing GFP, organs were isolated ex vivo and presence of GFP^+^ cancer cells were detected by FC. Cell subsets were analyzed using the Aurora flow cytometer (Cytek) and data were processed using FlowJo software (TreeStar). Flow cytometry analysis were conducted at the ImaFlow plateform (UMS58 BioSanD) at the Université Bourgogne Europe. For dimensionality reduction and clustering of high-parameter datasets, the Uniform Manifold Approximation and Projection (UMAP) and Flow self-organizing maps (FlowSOM) plugins within FlowJo were employed to visualize data in a two-dimensional space. The FC gating strategies are shown in Figure S14.
Figure design and creation
The graphical abstract was created using BioRender. Mouse icons included in the figures were designed using GIMP (v3.0.4).
Quantification and statistical analysis
All data were expressed as means ± standard deviation (SD). Differences between the two groups were assessed with the two-tailed unpaired Student’s t test. The one-way ANOVA with Tukey’s multiple comparison test was used to assess differences between more than two groups. Differences in Kaplan-Meier survival curves were analyzed using the Log Rank (Mantel-Cox) test. No statistical methods were used to predetermine the sample size. No animal exclusion criteria were applied. Mice were randomly allocated to experimental groups. The variance was similar between the groups that were statistically compared. Statistics were performed using GraphPad Prism (GraphPad), where significance is indicated in the figures. No statistic is shown when the p value is >0.05. The exact numbers of mice (n) for each experiment are reported directly on the figures, or in the figure legends. The statistical analyses applied are reported in the figure legends.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Bruni D.Angell H.K.Galon J.The immune contexture and Immunoscore in cancer prognosis and therapeutic efficacy Nat. Rev. Cancer 20202066268010.1038/s 41568-020-0285-732753728 · doi ↗ · pubmed ↗
- 2Sharma P.Hu-Lieskovan S.Wargo J.A.Ribas A.Primary, Adaptive, and Acquired Resistance to Cancer Immunotherapy Cell 168201770772310.1016/j.cell.2017.01.01728187290 PMC 5391692 · doi ↗ · pubmed ↗
- 3Ren D.Hua Y.Yu B.Ye X.He Z.Li C.Wang J.Mo Y.Wei X.Chen Y.Predictive biomarkers and mechanisms underlying resistance to PD 1/PD-L 1 blockade cancer immunotherapy Mol. Cancer 1920201910.1186/s 12943-020-1144-632000802 PMC 6993488 · doi ↗ · pubmed ↗
- 4Sharma P.Goswami S.Raychaudhuri D.Siddiqui B.A.Singh P.Nagarajan A.Liu J.Subudhi S.K.Poon C.Gant K.L.Immune checkpoint therapy-current perspectives and future directions Cell 18620231652166910.1016/j.cell.2023.03.00637059068 · doi ↗ · pubmed ↗
- 5Andrews L.P.Yano H.Vignali D.A.A.Inhibitory receptors and ligands beyond PD-1, PD-L 1 and CTLA-4: breakthroughs or backups Nat. Immunol.2020191425143410.1038/s 41590-019-0512-031611702 · doi ↗ · pubmed ↗
- 6Andrews L.P.Butler S.C.Cui J.Cillo A.R.Cardello C.Liu C.Brunazzi E.A.Baessler A.Xie B.Kunning S.R.LAG-3 and PD-1 synergize on CD 8(+) T cells to drive T cell exhaustion and hinder autocrine IFN-gamma-dependent anti-tumor immunity Cell 187202443554372.e 2210.1016/j.cell.2024.07.01639121848 PMC 11323044 · doi ↗ · pubmed ↗
- 7Andrews L.P.Cillo A.R.Karapetyan L.Kirkwood J.M.Workman C.J.Vignali D.A.A.Molecular Pathways and Mechanisms of LAG 3 in Cancer Therapy Clin. Cancer Res.2820225030503910.1158/1078-0432.CCR-21-239035579997 PMC 9669281 · doi ↗ · pubmed ↗
- 8Aggarwal V.Workman C.J.Vignali D.A.A.LAG-3 as the third checkpoint inhibitor Nat. Immunol.2420231415142210.1038/s 41590-023-01569-z 37488429 PMC 11144386 · doi ↗ · pubmed ↗