Anti mouse cd8α

E.G7-OVA cells (2 × 10⁵ cells/mouse, ATCC, Manassas, VA, USA) were injected subcutaneously into the right flank of C57BL/6 mice (6 weeks old, female, CLEA Inc.). On days 7, 10, 14, and 21, mice were divided into 4 groups, and saline, OVA (100 μg/mouse), OVA-TLR7a (100 μg/mouse for OVA; 20 μg/mouse for Imiquimod), or Gd₂O₃-OVA-TLR7a (100 μg/mouse for OVA; 20 μg/mouse for Imiquimod; 1 mg/mouse for Gd₂O₃ nanotubes) were injected into their left flanks. The tumor volume was calculated by 1/2 × longest dimension × (perpendicular dimension)². Mouse survival rate was calculated on the basis of tumor size < 15 mm. Splenocytes were collected, stained with anti-mouse CD4, anti-mouse CD8α, anti-mouse IFNγ, and anti-mouse TNFα antibodies (BioLegend, San Diego, CA, USA), and analyzed using a FACSAria cell cytometer (BD Biosciences).

H&E staining was performed on cryosections (10 µm) of different tissues, including normal heart, lung, liver, and spleen, as well as lung metastases-bearing lungs. For immunohistochemical staining of tumor-infiltrating lymphocytes, cryosections (10 µm) of 4T1-luc lung metastases-bearing lung tissues obtained from the above treatment group on day 22 (24 h after the last inhalation) were immunostained with anti-mouse CD8α (1:500; BioLegend) or anti-mouse FoxP3 antibody (1:500; BioLegend), followed by HRP-conjugated goat anti-rat secondary antibody (1:500; Jackson Immuno). The sections were then developed with DAB Kits (3,3′-diaminobenzidine; Vector Laboratories) and counterstained by hematoxylin.

To evaluate which immune cells were required to confer the observed antitumor effect, one day before tumor resection and hydrogel implantation, NK cells and CD8⁺ T cells in mice were depleted by intraperitoneal injection of depleting antibodies (200 μg/mouse) every 3 days for five times in total. The antibodies used for depletion were anti-mouse NK-1.1 (catalog no. 108760; Biolegend) and anti-mouse CD8α (catalog no. 100764; Biolegend).