Wild-type male C57BL/6 mice (10–12 weeks old; n = 6 from three independent cohorts; Jackson Laboratories, Bar Harbor, ME) were used. Animals were maintained under standard housing conditions with ad libitum access to food and water. All procedures were approved by the Augusta University Institutional Animal Care and Use Committee [IACUC #2011-0062] and conducted in accordance with the NIH Guide for the Care and Use of Laboratory Animals. The study was adhered to the ARRIVE guidelines for re-porting animal research (ARRIVE 2.0 Essential 10 guidelines).

The luciferase-expressing GL261 line (GL261-Luc2) used in this study was obtained from the Georgia Cancer Center Cell Repository. This line is derived from the parental GL261 glioma cell line (NCI/ATCC origin) and stably transduced with firefly luciferase, allowing reliable in vivo bioluminescence imaging. The GL261 orthotopic model was selected as it represents an immunocompetent, syngeneic system that recapitulates key features of human glioblastoma, including an immunosuppressive microenvironment and limited immune cell infiltration, while allowing investigation of immune cell trafficking without confounding allograft rejection. GL261-Luc2 cells were cultured under standard conditions and stereotactically implanted into the right striatum (3 × 10⁴ cells in 3 µL PBS) of C57BL/6 mice as described previously (1). Mice were anesthetized using 3% isoflurane for induction and maintained with 1.5–2% isoflurane throughout surgical procedures. Tumor establishment and progression were verified by in vivo bioluminescence imaging (IVIS Spectrum). At the endpoint, animals were euthanized by CO₂ inhalation followed by cervical dislocation, in accordance with the American Vet-erinary Medical Association (AVMA) Guidelines for the Euthanasia of Animals.

Bone marrow cells were harvested from donor C57BL/6 mice as described previously (19). Lineage-negative (Lin^-) ILC2s were isolated using magnetic negative selection (BioLegend antibodies against CD3ϵ, CD5, CD19, LY6G, NK1.1, CD11b, CD11c, and Ter119) followed by Streptavidin MicroBeads (Miltenyi Biotec). Purified ILC2s were ex-panded for six days in RPMI-1640 medium supplemented with 10 ng/mL recombinant IL-7 and IL-33 (R&D Systems). ILC2 identity was confirmed by flow cytometry (Lin^-CD45⁺Sca-1⁺ST2⁺GATA3⁺CD25⁺ICOS⁺c-Kit⁺IL-5⁺IL-13⁺). Prior to adoptive transfer, activated ILC2s were labeled with 5 µM carboxyfluorescein succinimidyl ester (CFSE; Invitrogen) for 10 minutes at 37 °C, following established protocols (18–20). CFSE labeling provides a stable, heritable fluorescent marker that allows definitive tracking of trans-ferred cells in recipient animals without requiring re-staining for phenotypic markers (20–22). To eliminate the possibility of free CFSE contributing to tissue fluorescence, la-beled ILC2s were washed three times with complete medium until the supernatant ex-hibited no measurable fluorescence. CFSE is an amine-reactive, intracellularly retained dye and does not dissociate once bound. Labeled ILC2s were intravenously injected via the tail vein (3 × 10⁶ cells/mouse) on day 9 post-tumor implantation, a time point ensuring established tumor growth. The cell dose was adapted from established ILC2 adoptive transfer protocols (23) and adjusted to ensure sufficient cell numbers for reliable detection and tracking in the CNS while avoiding potential adverse effects from excessive cell infusion. Control tumor-bearing mice received PBS injections. In flow cytometric analysis of recipient tissues, CD45⁺CFSE⁺ events represent transferred ILC2s, as CFSE was applied exclusively to the donor population. During downstream analyses, CFSE⁺ events were only considered ILC2s if they also expressed the expected immune phenotype (Lin^-CD45⁺Sca-1⁺ST2⁺GATA3⁺). No CFSE⁺ events were detected in the brains of PBS-injected control mice, confirming signal specificity. This methodology is consistent with standard practice in adoptive immune cell transfer studies (18–20).

Bioluminescence imaging confirmed tumor growth, while fluorescence microscopy visualized CFSE⁺ ILC2s in the brain, meninges and peripheral organs (spleen, lung). Single-cell suspensions from these tissues were analyzed for CFSE⁺ ILC2s and other ILC subsets using a NovoCyte Quanteon flow cytometer (Agilent Technologies). Data were processed with FlowJo software (V10) as described previously (8). Fluorescence quanti-fication was performed as mean pixel intensity (MFI) per microscope field. For each mouse, five non-overlapping regions were imaged at identical exposure settings, and the CFSE fluorescence channel was extracted and background-subtracted using ImageJ. The average pixel intensity per field per animal was used for statistical comparisons.

Tumor-associated photon emission was quantified using Living Image 4.7 software. For each mouse, a fixed-size region of interest (ROI) was drawn over the tumor, and an identical ROI was placed over a non-tumor brain region to obtain background signal. Background-subtracted radiance (photons/sec) was used for all comparisons. All imaging sessions were performed using identical exposure settings. Tumor-bearing mice included n = 6 per condition, derived from three independent experimental cohorts.

All data were analyzed using GraphPad Prism (version 9). Frequencies of CFSE⁺ ILC2s in brain, meninges, spleen, and lung were compared using one-way ANOVA followed by Tukey’s post-hoc test. Quantification of CFSE fluorescence intensity in tissue sections was based on mean pixel intensity per field and analyzed using one-way ANOVA. For tumor bioluminescence, background-subtracted radiance values were compared between groups using unpaired t-tests, whereas longitudinal comparisons within the same animals used paired t-tests. All values are reported as mean ± SEM, and statistical significance was defined as p < 0.05.

Bioluminescence imaging confirmed reliable establishment of orthotopic GBM tumors by day 9 post-implantation, providing a stable model for subsequent ILC2 transfer (Figure 1A). At this stage, CFSE-labeled bone marrow–derived ILC2s were intravenously administered to tumor-bearing mice to assess their migratory behavior and CNS entry.

At 96 hours post-transfer, fluorescence imaging revealed distinct CFSE⁺ signals within the meninges, brain, and tumor parenchyma (Figure 1B). Flow cytometric analysis confirmed that these infiltrating cells retained canonical ILC2 markers (Lin^-CD45⁺Sca-1⁺ST2⁺GATA3⁺), identifying them as a discrete population within the glioblastoma microenvironment (Figure 1C). This represents, to our knowledge, the first direct indication that adoptively transferred ILC2s can access the CNS and localize within glioblastoma tissue following systemic delivery. While these findings demonstrate CNS entry, the precise route of infiltration, whether via direct BBB crossing, meningeal pathways, or perivascular access, cannot be definitively determined from the current data. To ensure that the CFSE⁺ signals represented intact ILC2s rather than free dye, we implemented multiple verification steps. CFSE binds covalently to intracellular amines and does not dissociate once incorporated; labeled ILC2s were therefore washed repeatedly until no detectable fluorescence remained in the supernatant. The CFSE⁺ events identified in brain and tumor tissue also co-expressed canonical ILC2 markers (Lin^-CD45⁺Sca-1⁺ST2⁺GATA3⁺), a phenotype that cannot arise from free dye (Figure 1D). In tissue sections, CFSE⁺ structures displayed discrete, nucleated, cell-like morphology rather than diffuse extracellular staining, and no CFSE signal was observed in PBS-injected tumor-bearing controls. Together, these findings support that the CFSE⁺ signals reliably reflect transferred ILC2s.

Beyond the brain, CFSE⁺ ILC2s were also detected in peripheral tissues, including spleen and lung (Figure 2A), suggesting preserved systemic migratory potential. However, their frequency was markedly higher in the meninges and tumor regions (Figure 2B), suggesting a degree of preferential accumulation within CNS compartments. These findings indicate that ex vivo–expanded ILC2s maintain the capacity to traffic through the circulation and traverse the blood–brain barrier under the inflammatory conditions associated with GBM.

Despite successful homing and infiltration, ILC2 transfer did not significantly alter tumor progression as determined by longitudinal bioluminescence imaging (Figure 3A). Individual tumor radiance measurements at days 9 and 13 post-implantation showed comparable growth trajectories between ILC2-treated and control mice (Figure 3B), with no significant difference in tumor burden at either timepoint. These findings suggest that unmodified ILC2s alone are insufficient to exert antitumor activity within the GBM microenvironment. Nevertheless, their observed ability to enter and persist within intracranial tumors support a proof-of-concept for using ILC2s as a deliverable cell platform. Future efforts will focus on enhancing their therapeutic potential through cytokine priming, checkpoint modulation, or genetic engineering to improve tumor specificity and efficacy.

As a proof-of-concept study, several limitations should be acknowledged. The sample size (n = 6 mice from three independent cohorts) established biological feasibility but limits definitive conclusions regarding ILC2 frequency, persistence, or functional heterogeneity within the tumor microenvironment. The short post-transfer observation window (96 hours) restricted assessment of long-term persistence and downstream immunological consequences.

A critical gap is the absence of functional readouts. We did not measure in situ cytokine production (IL-5, IL-13), evaluate effects on tumor-associated immune populations, or assess macrophage polarization. While transferred cells retained canonical ILC2 markers, whether they remained functionally competent, capable of cytokine secretion, immune cell recruitment, or tissue remodeling, was not determined. Additionally, ILC2s are functionally heterogeneous and can exhibit plasticity, with subsets potentially promoting immunosuppression (via type 2 cytokines and M2 polarization) or supporting antitumor immunity. The absence of subset-specific characterization limits understanding of whether transferred ILC2s maintained uniform phenotypes or underwent functional diversification within the tumor.

The transferred ILC2s were unmodified, likely contributing to the absence of antitumor activity. We used a single syngeneic model (GL261); generalizability to other glioblastoma models or human GBM remains undetermined. The molecular mechanisms governing ILC2 CNS entry and tumor interactions remain undefined. These limitations underscore the exploratory nature of this work and highlight critical areas for future investigation employing single-cell approaches and functional profiling to delineate ILC2 phenotypes and therapeutic potential.