Intraperitoneal Programming of CAR Macrophages for Cancer Immunotherapy

Peritoneal metastasis remains difficult to treat, and many patients are not candidates for approaches such as cytoreductive surgery with hyperthermic intraperitoneal chemotherapy. In a 2025 Nature Communications article, Gu et al. investigate a different strategy: deliver mRNA lipid nanoparticles (mRNA-LNPs) into the peritoneal cavity to program local macrophages as chimeric antigen receptor macrophages (CAR-Ms). The study combines CAR design comparisons with immune profiling to ask which intracellular signaling domains can best direct macrophage activity and how engineered macrophages affect the broader tumor microenvironment. Read the reference study.

Study Background and Research Question

Macrophages are abundant immune cells in the peritoneal environment, including in ascites associated with peritoneal metastasis. Yet their presence does not ensure effective tumor control: the tumor microenvironment can suppress antitumor immunity and limit the activity of immune cells. CAR engineering offers a way to direct immune-cell recognition toward selected tumor antigens, but CAR designs developed for one immune-cell type cannot simply be assumed to produce the same functions in another. Macrophages have distinct signaling and effector biology, making the choice of intracellular domains an important design question.

The authors focus on two linked problems. First, which CAR intracellular domain (ICD) configurations can empower macrophages to respond effectively? Second, can mRNA-LNP delivery program those cells in situ, avoiding the need to isolate, engineer, and return macrophages as an ex vivo cell product? The work therefore addresses both CAR-M signaling design and a local delivery route for a challenging disease setting. It does not establish that this strategy is ready for clinical use; rather, it provides preclinical evidence and a framework for further development.

Key Innovation from the Reference Study

The central innovation is the combination of local mRNA-LNP delivery with a systematic screen of CAR architectures. Gu et al. evaluate 36 CAR formats in macrophages and identify designs incorporating CD3ζ and TLR4 intracellular domains as particularly effective in the study’s experimental context. This comparison is meaningful because it treats the macrophage signaling module as a variable to optimize, rather than assuming that a CAR configuration established in another immune-cell lineage will work equivalently in macrophages.

The delivery concept is also notable. Intraperitoneal administration is intended to program macrophages in the compartment where peritoneal tumors grow. This provides a potential route to concentrate an immune-engineering approach at the disease site. The study describes this as in situ programming; it should not be interpreted as proof that delivery is restricted exclusively to macrophages or that all clinically relevant tumors would receive uniform exposure. Those questions depend on biodistribution, cell targeting, and disease context.

In the lead designs, the reported outcome extends beyond CAR-M activation alone. The authors describe adaptive immune activation, changes to the tumor microenvironment, and improved antitumor activity when CAR-M treatment is combined with PD-1/PD-L1 checkpoint therapy. The result suggests that CAR-Ms may function both as targeted effectors and as regulators of wider immune responses.

Methods and Experimental Design Insights

The study’s design links molecular engineering to tissue-level immune analysis. The investigators first compare a panel of 36 CAR formats in macrophages, then use macrophage-targeted mRNA-LNPs for intraperitoneal programming. The supplied article information identifies the CD3ζ–TLR4 intracellular-domain design as a lead configuration, but does not provide dosing, administration schedules, particle characteristics, or detailed antigen-specific assay parameters. These values should be taken from the final article and its methods and supplementary information before attempting to reproduce the work.

Single-cell RNA sequencing (scRNA-seq) provides a higher-resolution view of how treatment is associated with changes in the tumor microenvironment. This approach can characterize cell populations and gene-expression states across immune cells, complementing bulk measurements that can conceal changes in less abundant subsets. In this study, the authors use scRNA-seq to examine immune remodeling and report an increase in TCF1⁺PD-1⁺ progenitor-exhausted CD8⁺ T cells, often termed Tpex.

The authors also investigate signaling and phenotype. They report that CAR-Ms retain a proinflammatory phenotype while increasing MHC-I and PD-L1 expression, in association with perturbation of NF-κB pathways. These observations help connect CAR design to both macrophage state and adaptive immunity. However, expression changes and single-cell associations do not alone establish the full causal sequence; mechanistic experiments and functional validation are needed to determine which changes drive therapeutic activity.

Protocol Parameters

• Delivery route: Intraperitoneal mRNA-LNP delivery is the study’s defining approach for in situ macrophage programming; the supplied article information does not specify a dose or schedule.
• CAR design screen: The authors compare 36 CAR formats and highlight designs with CD3ζ and TLR4 intracellular domains. Consult the full methods for construct details and screening conditions.
• Immune profiling: scRNA-seq is used to characterize tumor-microenvironment changes, including CD8⁺ T-cell states; the available summary does not give sequencing or analysis parameters.
• Checkpoint combination: The study reports synergy with PD-1/PD-L1 therapy, but no combination dose or timing is specified in the supplied information. Do not infer a treatment schedule from the abstract-level findings.

Core Findings and Why They Matter

The first main finding is that CAR intracellular architecture matters in macrophages. The reported performance of CD3ζ–TLR4 designs supports the premise that signaling modules can be tailored to macrophage biology. It also gives later studies a reason to compare defined CAR domains systematically, rather than testing only a single construct. The result remains specific to the tested constructs and models; it does not establish a universally optimal CAR-M design.

Second, the study connects engineered macrophages with adaptive immune responses. The reported increase in TCF1⁺PD-1⁺ Tpex cells is relevant because progenitor-exhausted CD8⁺ T cells are a distinct population from terminally exhausted cells and may retain capacity to contribute to ongoing responses. The finding is consistent with CAR-M treatment reshaping the immune environment, but it does not by itself show that these cells convert into durable tumor control or predict patient outcomes.

Third, the reported synergy with PD-1/PD-L1 therapy supports testing combinations that address more than one part of the immune response. CAR-Ms may provide antigen-directed activity and inflammatory cues, while checkpoint blockade may relieve inhibitory signaling. At the same time, the study reports increased PD-L1 alongside proinflammatory features. That pattern illustrates a potential feedback response, not a contradiction: immune activation can coexist with inhibitory pathways that may constrain it. The combination finding is therefore a rationale for further testing, not a clinical treatment recommendation.

Finally, the reported MHC-I and PD-L1 changes, together with the NF-κB pathway observations, offer mechanistic leads for understanding how CAR-Ms influence neighboring immune cells. These results broaden the focus from direct tumor-cell engagement to the regulatory effects of engineered macrophages within the tumor microenvironment.

Comparison with Existing Internal Articles

The internal overview In Situ mRNA-LNP Programming of CAR Macrophages for Peritoneal Cancer covers the same study concept and is useful as a companion introduction to local delivery and CAR-M design. Here, the emphasis is on interpreting the study’s domain comparison, single-cell findings, and limits on what can be concluded from the reported mechanisms. For assay context rather than a description of methods used in this paper, D-Luciferin Sodium Salt for CAR-M Assays discusses reporter-based measurements that researchers may encounter in separate CAR-M experiments.

Limitations and Transferability

The available study summary does not provide quantitative effect sizes, detailed model descriptions, dosing, safety results, or the full CAR and LNP characterization. Conclusions about efficacy magnitude, reproducibility, or the relative performance of individual constructs therefore require review of the complete publication and supplementary data. The article is also identified as an unedited version in the supplied material, so readers should check the final publication for any revisions.

Transfer to human disease will depend on factors that are not resolved by the reported findings alone: the distribution and uptake of mRNA-LNPs in relevant cell populations, the persistence and function of programmed macrophages, tumor antigen heterogeneity, and local immune suppression. The peritoneal setting may offer practical advantages for local administration, but it differs from other solid-tumor sites. Likewise, a reported increase in Tpex cells and synergy with checkpoint therapy in experimental systems should not be treated as evidence of clinical benefit. Safety, dosing, manufacturing consistency, and the consequences of inducing inflammatory and inhibitory signals together remain important areas for evaluation.

Research Support Resources

Why this cross-domain matters, maturity, and limitations

The CAR-M study is centered on immune engineering, not luciferase imaging, and the supplied information does not report use of a luciferase reporter or luciferin. In a separate experiment, a firefly luciferase reporter can provide a measurable signal when paired with its substrate, supporting an ATP-dependent bioluminescence assay or cell viability and metabolism monitoring. Such a readout is a proxy influenced by reporter expression, ATP availability, and assay conditions; it does not directly measure CAR-M function or establish the immune mechanisms described above. For researchers planning an independent reporter workflow, D-Luciferin sodium salt (SKU B8311) is a firefly luciferase substrate that can support bioluminescence imaging in oncology research and cellular energy metabolism assessment. Its use should be matched to the reporter system and validated controls, rather than inferred as part of the reference study.