Studying the effect of compressive solid stress on macrophage phenotype and function in the context of glioblastoma

Introduction: : Glioblastoma (GBM) is one of the most aggressive forms of brain cancer, with an overall survival of less than two years. One barrier to therapeutic efficacy is the extremely immunosuppressive GBM tumor microenvironment (TME). Myeloid cells, such as macrophages, accumulate in GBM tumors and can comprise up to half of the tumor bulk. Macrophage phenotype typically ranges from pro-inflammatory (M1-like) to anti-inflammatory (M2-like). In GBM, they adopt a pathological, pro-tumor phenotype that thwarts anti-tumor immune responses. Macrophages also directly contribute to GBM stemness, aggressiveness, and treatment resistance. Macrophage abundance is therefore associated with worse patient prognosis. The GBM TME also features solid stress, which arises as the tumor expands and presses against the surrounding tissue. Macrophages are highly mechanosensitive and able to change phenotype based on physical cues. Because macrophages are capable of generating solid stress, as established in our previous work, there may be a feedback loop between macrophages and compression in the TME. However, the response of macrophages to chronic compressive solid stress in the context of tumor growth has not been explored. Here, we use a custom-built compression system to interrogate the macrophage response to uniaxial compressive solid stress as seen in GBM. This will help to identify new macrophage-targeting therapeutic strategies to normalize the GBM immune microenvironment and lower the immunosuppressive barrier for immunotherapies.

Materials and

Methods: : Murine RAW 264.7 macrophages were seeded in a transwell cell culture insert and compression was applied with a 3D printed weight. A media and gas-equilibrated 1% agarose cushion between the cells and the weight served as a media reservoir and prevented direct contact between the cells and the rigid weight. The weight was designed to apply 0.14 kPa of uniaxial compressive stress uniformly across the transwell surface. After compression, cells were fixed for staining and imaging, or else lysed for RNA sequencing and RT-qPCR analyses. We performed fluorescence lifetime imaging on fixed, unstained cells to determine the cells’ metabolic state. The coenzyme NAD(P)H is the predominant autofluorescent species in cells, and its fluorescence lifetime changes depending on whether the cell relies more on glycolysis versus oxidative phosphorylation. We also assessed changes in cell size and morphology based on fluorescent images. To determine phagocytic capacity, we removed the weight and agarose after 24 hours of compression and incubated cells with fluorescent microbeads, quantifying bead uptake after 2 hours. We also the Greiss assay to determine the concentration of nitrite, an indicator of nitric oxide production, in media of cells that were compressed for 48 hours. To determine the transcriptional response to compression, we performed bulk sequencing on compressed and control cells as well as RT-qPCR for specific markers of interest.

Results, Conclusions, and Discussions:: Macrophages show a robust response to compression, though not all metrics of polarization fall on the same end of the traditional pro- versus anti-inflammatory spectrum. Compression caused an increase in average cell fluorescence lifetime, indicating a shift towards oxidative phosphorylation (Fig. 1A,B). This is consistent with an M2-like, anti-inflammatory phenotype. Increased macrophage size under compression is a typical characteristic of M1-like macrophages (Fig. 1C). However, the more irregular cell shape and increased phagocytosis is more commonly associated with the M2-like phenotype (Fig. 1 D-F). Compressed macrophages also produced more nitric oxide, a hallmark of pro-inflammatory M1-like polarization (Fig. 1G). We observed a trending upregulation of both Nos2 and Arg1 (M1 and M2 markers, respectively) under varying magnitudes of compression, varying time under compression, and under simultaneous compression and hypoxia. RNA sequencing analysis confirmed this result, as compressed macrophages were enriched for both M1 and M2 gene signatures (Fig. 2).

One consistent trend was an enrichment of glioma-associated and pan-tumor-associated macrophage signatures (Fig. 2 D,E). These upregulated signatures were all associated with worse patient outcomes. The one tumor-associated macrophage signature that was down-regulated was a pro-inflammatory macrophage gene set associated with better patient response to immunotherapy. The upregulated biological and molecular processes under compression were largely related to chemotaxis and cytokine signaling. Visualizing this process in a protein-protein interaction map highlighted IL-1β as one of the central hubs. IL-1β participates in a feedback loop in several human cancers, promoting tumor immune escape and worse patient outcomes.

In conclusion, we have demonstrated that mechanical compression alone can drive macrophages towards a phenotype resembling those seen in glioblastoma and other cancers. We previously established that macrophages themselves can generate solid stress. This work provides evidence to close the feedback loop between macrophages and solid stress, wherein macrophages may both generate and respond to solid stress in a manner that contributes to immune suppression and tumor progression. This cycle represents a promising therapeutic target, as reducing solid stress, macrophage sensing and response to it, or macrophage contribution to it, could break the cycle and normalize the tumor immune microenvironment.

Acknowledgements and/or References (Optional):: This work was funded by NIH/NIGMS. We also thank the Notre Dame Genomics and Bioinformatics Core Facility.