Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2026 Jan 2;12(1):eadu0292.
doi: 10.1126/sciadv.adu0292. Epub 2026 Jan 1.

Systemically inducing trained immunity overcomes solid tumors' immunosuppressive microenvironment

Bram Priem  1   2   3 , Lisa Willemsen  4 , Tom Anbergen  1   3 , Yuri van Elsas  1   3 , Jeroen Deckers  1   3 , David P Schrijver  5 , Stijn R J Hofstraat  5 , Iris V Messing  1 , Jazz Munitz  3 , Dionne Honing  1 , Geoffrey Prévot  3 , Yohana C Toner  1 , Judit Morla  3 , William Wang  3 , Anna Ranzenigo  3 , Isabella Sirchia  3 , Tomas Post  3   6 , Raphaël Duivenvoorden  3   6 , Ewelina Kluza  5 , Glenn A O Cremers  7 , Joost H C M Kreijtz  7 , Carlos Pérez-Medina  3   8 , Mandy M T van Leent  3   9 , Abraham J P Teunissen  3 , Roy van der Meel  5 , Yi Zhou  10 , Chris Glass  10 , Jordi Ochando  11 , Leo A B Joosten  12   13 , Zahi A Fayad  3 , Romana T Netea-Maier  1 , Mihai G Netea  1   14   15 , Menno P J de Winther  4 , Thijs J Beldman  1 , Arjan W Griffioen  2 , Willem J M Mulder  1   5
Affiliations

Systemically inducing trained immunity overcomes solid tumors' immunosuppressive microenvironment

Bram Priem et al. Sci Adv. .

Abstract

Hematopoietic bone marrow progenitors are increasingly implicated as an origin of immunosuppression in cancer. We have previously shown that trained immunity induction using nanomedicine potentiates checkpoint blockade therapy. Here, we studied how this approach's induction of trained immunity systemically overcomes the immunosuppressive tumor microenvironment. We found changes in the tumor microenvironment to mirror functional changes in the hematopoietic system in a melanoma mouse model. Single cell sequencing methods disclosed a shift in the tumor-associated macrophage population from immunosuppressive to antitumorigenic. Uniquely, a trained immunity and checkpoint blockade combination therapy mobilized natural killer cells which, in conjunction with the functional changes in the myeloid cell compartment, effectively activated T cells. Last, we established the effectiveness of our approach in mouse models of breast, lung, and pancreatic cancer. Collectively, our data show that the systemic induction of trained immunity rebalances the immune system for effective checkpoint blockade therapy.

PubMed Disclaimer

Conflict of interest statement

W.J.M.M., L.A.B.J., J.O., Z.A.F., and M.G.N. are scientific cofounders of and have equity in Trained Therapeutix Discovery. W.J.M.M. and Z.A.F. have consulting agreements with Trained Therapeutix Discovery. W.J.M.M. is the shareholder and CSO of Trained Therapeutix Discovery and founder, shareholder, and CTO of BioTrip. J.H.C.M.K., G.A.O.C., B.P., and T.J.B. have positions at Trained Therapeutix Discovery. All other authors declare that they have no competing interests.

Figures

Fig. 1.
Fig. 1.. MTP-HDL treatment induces a metabolic shift in myeloid cells.
(A) Overview of preceding study (9) in which we developed a bone marrow-avid and trained immunity inducing nanoparticle named MTP-HDL. Through extensive analysis of the bone marrow, we highlighted the various effects this particle elicits on myeloid progenitors in the bone marrow. We finalized the study by showing that MTP-HDL is an effective monotherapy to treat B16F10 melanoma in a C57BL/6 mouse model, and it potentiates CI therapy in this otherwise refractory model. (B) Schematic overview of treatment regimen. Mice were treated with either PBS, three injections of MTP-HDL (1.5 mg/kg), anti–CTLA-4 and anti–PD-1 at 200 μg each, or combination therapy of MTP-HDL and CIs. (C) Tumor growth in B16F10 inoculated C57BL/6 mice. Results show significant decrease in tumor growth of combination therapy compared to all other groups (n = 9/10 per group). (D) Qualitative FDG PET-CT image of a C57BL/6 mouse. (E) Gamma counting of the bone marrow/spleen posttreatment shows significant higher uptake in groups containing MTP-HDL (n = 6 to 7 per group). (F) Flow cytometry showing total number of cells per femur/spleen posttreatment (n = 7 to 9 per group). (G) Flow plots showing gating strategy to get GLUT-1 expression of monocytes and neutrophils. (H) GLUT-1 expression of bone marrow/spleen monocytes and neutrophils displays significantly higher expression of GLUT-1 in groups containing MTP-HDL (n = 6 to 8 per group). (I) Glycolytic capacity as measured by CENCAT flow cytometry of neutrophils and monocytes in bone marrow and spleen (n = 6 to 8 per group). For all panels, data are presented as means ± SD and means ± SEM for tumor growth experiments. For the tumor growth experiment, significance shown in the graph is that of the tumor growth rate. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. ns, not significant. APC, allophycocyanin; PE, phycoerythrin; %ID/g, percentage injected dose per gram (%ID/g); MFI, mean fluorescence intensity.
Fig. 2.
Fig. 2.. Transcriptional changes of immune cells shape the tumor microenvironment.
(A) Qualitative histology images of the tumor [hematoxylin and eosin (H&E), CD1bb staining, and CD3 staining], tumors excised day after last treatment with MTP-HDL. (B) Percentage of CD11b-positive cells in the core and the periphery shows no significant differences between tumor groups and a significant variation in different tumors (n = 5 per group). (C) GLUT-1 expression of tumor monocytes and neutrophils displays significantly higher expression of GLUT-1 in groups containing MTP-HDL (n = 5 to 7 per group). (D) Experimental overview of the scRNA-seq setup. (E) In total, 33 × 103 cells were used for analysis; each treatment group was n = 2, and each n consisted of the tumor-infiltrating leukocytes of five mice. (F) tSNE plot of all analyzed cells with cluster names. (G) Feature plots showing main cluster markers. (H) Plots displaying the population shift and the differentially expressed genes (DEGs) in either myeloid cells, lymphoid cells minus B cells, or B cells. Most of the DEGs [false discovery rate (FDR) < 0.05 and fold change (FC): 0.25 > FC < −0.25] as compared to PBS are present in the lymphoid cells minus B cells and myeloid group. SSC-A, side scatter area; pDC, plasmacytoid dendritic cell; cDC, conventional DC; NK, natural killer.
Fig. 3.
Fig. 3.. A pro-inflammatory shift in tumor-associated macrophages.
(A) This figure is mainly focused on myeloid cells as presented accompanied by a legend of the treatment groups. (B) Plot displaying the amount of neutrophils as a percentage of total leukocytes and the DEGs in this cluster as compared to PBS (FDR < 0.05 and fold change: 0.25 > FC < −0.25). (C) Volcano plot displaying up- and down-regulated genes of the neutrophil cluster when comparing PBS versus MTP + CI. (D) Plot displaying the amount of neutrophils as a percentage of total leukocytes and the DEGs in this cluster as compared to PBS, and Venn diagrams of up and down DEGs in all treatment arms compared to PBS (FDR < 0.05 and fold change: 0.25 > FC < −0.25). (E) Dot plot highlighting key pro- and anti-inflammatory genes in all treatment groups. (F) Heatmap displaying functional grouping of the DEGs in the macrophage cluster across all treatment groups. (G) tSNE plots displaying the bulk reclustering of the macrophage population identified by our original analysis. The main macrophage cluster is now divided into four subclusters Mac 1 to Mac 4. (H) Bar graphs displaying the distribution of macrophages in the different clusters between different treatment groups. (I) Volcano plot displaying up- and down-regulated genes of the Mac 1 cluster when comparing PBS versus MTP + CI. (J) Flow cytometry analysis of CD206 and inducible nitric oxide synthase (iNOS) in splenic macrophages. MFI of CD206 is significantly lower in all treatment groups as compared to PBS. iNOS is only significantly up-regulated in the combination therapy group (n = 8 per group). All genes mentioned are significant, defined as an adjusted P value below 0.05 and a fold change below −0.25 or above 0.25. For all panels, data are presented as means ± SD. **P < 0.01 and ***P < 0.001. ns, not significant, or significance is given.
Fig. 4.
Fig. 4.. NK cells are activated following MTP-HDL treatment.
(A) This figure is mainly focused on lymphoid cells excluding B cells, accompanied by a tSNE plot displaying lymphoid reclustering in bulk. (B) Plots displaying the population shift and the DEGs in CD4, CD8, and CD8 Tc1 (FDR < 0.05 and fold change: 0.25 > FC < −0.25). (C) Heatmap displaying functional grouping of the DEGs in the macrophage cluster across all treatment groups. (D) Bar graph displaying the lower presence of the RR T cell group after treatments containing MTP-HDL. (E) Heatmap of the genes expressed by rapid response T cell group described by Wang et al. (24). Several pro-inflammatory genes are highly expressed as compared to other clusters and quiescent markers Slfn1 and Slfn5. (F) Plot displaying the population shift and the DEGs in NK cells and a heatmap of cluster marker genes of NK cells. (G) Plot displaying the population shift and the DEGs in NK cells (FDR < 0.05 and fold change: 0.25 > FC < −0.25). (H) Violin plots of NK cell perforins Gzma, Gzmb, and Prf1 across different treatment groups. These genes are all differentially expressed in the combination treatment group as compared to PBS (FDR < 0.05). (I) Flow cytometry dot plot and bar graphs highlighting the amount of splenic NK cells and Granzyme B expression. Significantly higher amount of NK cells was present in the combination treatment group, and a significantly higher expression of Granzyme B was found in treatment groups containing MTP-HDL (n = 8 per group). All genes mentioned are significant, defined as an adjusted P value below 0.05 and a fold change below −0.25 or above 0.25. For all panels, data are presented as means ± SD and means ± SEM for tumor growth experiments. For the tumor growth experiment, significance shown in the graph is that of the tumor growth rate. *P < 0.05 and **P < 0.01.
Fig. 5.
Fig. 5.. Trained immunity induction enhances myeloid-lymphoid cross-talk.
(A) CellChat analysis found combination treatment to display a higher amount of cell-cell interactions compared to PBS or both monotherapy groups. (B) Circle plot displaying the changes in cell-cell interactions comparing combination therapy with PBS. (C) Top 10 highest probability unique interactions and the top unique ligand-receptor pairs found in PBS and combination therapy. Interactions present in the PBS group are mostly of an anti-inflammatory nature, whereas the combination therapy displays pro-inflammatory interactions. (D) Schematic overview of the binding of CD28 to CD80/86, and CTLA-4 to CD80/86. (E) CellPhoneDB analysis highlighting cell-cell interaction between myeloid cells and the CD8 Tc1 effector group. Treatment stops CTLA-4 and CD80/86 interaction. (F) Bar graphs displaying cytokine levels of tumor lysate. Most strikingly is the higher presence of IFN-γ and IL-2 in the combination treatment group (n = 8 per group). (G) Heatmap highlighting NK cell contribution of IFN-γ production based on their Ifny expression level. (H) Schematic overview of tumor growth analysis in NK-cell–depleted mice. Depletion was achieved with 250 μg of anti-NK1.1, and mice were treated with PBS, three injections of MTP-HDL at 1.5 mg/kg, or combination therapy of MTP-HDL and CIs. (I) Tumor growth curves of NK cell–depleted mice C57BL/6 mice inoculated with 1 × 105 B16F10 cells. Results show significant decrease in tumor growth of mice treated with MTP-HDL; however, no significant additive effect of CI therapy is observed (n = 10 per group). All genes mentioned are significant, defined as an adjusted P value below 0.05 and a fold change below −0.25 or above 0.25. For all panels, data are presented as means ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.
Fig. 6.
Fig. 6.. Translational tumor-agnostic immunotherapy.
(A) T cell suppression assay experimental overview. CFSE-stained naïve T cells stimulated with CD3/CD28 beads were cocultured with monocytes from either naïve mice, tumor-bearing mice, or tumor-bearing mice that have received MTP-HDL treatment. (B) Bar plots displaying the percentage of proliferating CD4 or CD8 cells after the above described treatment regimen. Monocytes from untreated tumor-bearing mice significantly hampered T cell proliferation of both CD4 and CD8 T cells, while monocytes from treated tumor-bearing mice regained their effectiveness (n = 6 per group). (C) Schematic overview of treatment regimen. Mice were treated with either PBS, three injections of MTP-HDL at 1.5 mg/kg, anti–CTLA-4 and anti–PD-1 at 200 μg each, or combination therapy of MTP-HDL and CIs. (D to F) Tumor growth curve of (i) C57BL/6 mice inoculated with 1 × 105 Lewis lung carcinoma (LLC) cells, (ii) Balb/c mice inoculated with 1 × 105 4T1 breast cancer cells, and (iii) C57BL/6 mice inoculated with 2 × 106 PANC-02 pancreatic cancer cells. (D) Both therapeutic strategies containing MTP-HDL reduced the tumor growth rate relative to PBS, with the combination therapy being superior over MTP-HDL monotherapy. (E) Both therapeutic strategies containing MTP-HDL reduced the tumor growth rate relative to PBS, with the combination therapy being superior over MTP-HDL monotherapy. (F) Both MTP-HDL treatments resulted in significant growth inhibition; yet, the difference in growth rate between the combination therapy and MTP-HDL monotherapy was not statistically significant. However, the tumor in the combination therapy was significantly smaller at the last day of experiment as compared to MTP-HDL monotherapy. For all panels, data are presented as means ± SD and means ± SEM for tumor growth experiments. For the tumor growth experiment, significance shown in the graph is that of the tumor growth rate. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. FITC, fluorescein isothiocyanate.
See this image and copyright information in PMC

References

    1. Gabrilovich D. I., Ostrand-Rosenberg S., Bronte V., Coordinated regulation of myeloid cells by tumours. Nat. Rev. Immunol. 12, 253–268 (2012). - PMC - PubMed
    1. Kumar V., Patel S., Tcyganov E., Gabrilovich D. I., The nature of myeloid-derived suppressor cells in the tumor microenvironment. Trends Immunol. 37, 208–220 (2016). - PMC - PubMed
    1. Rabold K., Zoodsma M., Grondman I., Kuijpers Y., Bremmers M., Jaeger M., Zhang B., Hobo W., Bonenkamp H. J., de Wilt J. H. W., Janssen M. J. R., Cornelissen L. A. M., I. C. H. van Engen-van Grunsven, Mulder W. J. M., Smit J. W. A., Adema G. J., Netea M. G., Li Y., Xu C. J., Netea-Maier R. T., Reprogramming of myeloid cells and their progenitors in patients with non-medullary thyroid carcinoma. Nat. Commun. 13, 6149 (2022). - PMC - PubMed
    1. Netea M. G., Joosten L. A., Latz E., Mills K. H., Natoli G., Stunnenberg H. G., O’Neill L. A., Xavier R. J., Trained immunity: A program of innate immune memory in health and disease. Science 352, aaf1098 (2016). - PMC - PubMed
    1. Netea M. G., Quintin J., van der Meer J. W. M., Trained immunity: A memory for innate host defense. Cell Host Microbe 9, 355–361 (2011). - PubMed

Substances