Fig 1: MAGE-A4 is expressed in human NSCLC and is associated with loss of PTEN.(A) MAGE-A4 expression in NSCLC by mRNA from TCGA dataset, n = 989. Fragments per kilobase million ≥1 was considered positive. (B) Representative H&E of human NSCLC and serial sections show MAGE-A4pos staining (IHC) in lung adenocarcinoma (LUAD) and in lung squamous cell carcinoma (LUSC); 100× (top); scale bar, 150 μm (insets). (C) Incidence of MAGE-A4 in NSCLC array of 23 NSCLC tumors and cellular localization of MAGE-A4 as determined by IHC detection. LUSC/mixed includes LUSC, adenosquamous (mixed features), and large-cell carcinoma subtypes. Total MAGE-A4pos in tumor cases (left graph) and all MAGE-A4pos by subcellular localization (right side of graph). (D) The Cancer Genome Atlas (TCGA) data of 1023 cases shows MAGE-A4 expression in human NSCLC is correlated with decreased PTEN, as indicated by homozygous (hom) loss of both copies, or heterozygous (het) loss of PTEN, compared with intact, nonmutated PTEN, having ≥2 copies (wt/gain). Association of MAGEA4 with PTEN loss comparing both loss groups with wt/gain by t test (P = 1 × 10−8).
Fig 2: MAGE-A4 indirectly induces CXCL12 in endothelial cells to retain PCs by CXCL12-CXCR4 axis.(A) CXCL12 concentration from whole-lung homogenate from 7- to 8-month-old mice by ELISA. Data from three independent experiments combined, n = 7 to 21 per group. (B) Featureplot of whole lung shows colocalization of Cxcl12 with Pecam1. (C) Dotplot of Cxcl12 in endothelial (Pecam1+) cells. (D) Schematic of assay design, using human umbilical vein endothelial cells (HUVEC) cocultured with human NSCLC MAGE-A4k/o or MAGE-A4pos H1299 cells for 24 hours. (E) CXCL12 gene expression quantified using qPCR. Two independent experiments combined, n = 3 to 4 per group. Significance determined by one-way ANOVA with Dunnett’s correction or Student’s t test. *P < 0.05, **P < 0.01, and ***P < 0.001; ns = not significant.
Fig 3: MAGE-A4pos human NSCLC exhibits increased PCs and excluded T cells.IMC was used to evaluate a microarray of 24 individual NSCLC samples. (A and B) Overlay of gated CD19+ CD138+ CD38+ PCs onto images showing PCs (red) in (A) MAGE-A4neg and (B) colocalization of MAGE-A4 and E-Cadherin (pink) with PCs (red) in MAGE-A4pos. (C) CD19+ CD138+ PCs quantified as relative abundance of total immune cells. Total immune cells are the sum of non–B cell CD45+, B cells, and PCs. (D) Representative IMC staining of CD4 and CD8 T cells in MAGE-A4neg and MAGE-A4pos tumors showing T cell exclusion in MAGE-A4pos NSCLC. (E to H) Mean cell density of T cells in a 200 μm diameter around each tumor cell. Cell density calculation schematic (inset). Mean cell density quantification of (E) CD8 T cells, (F) CD4 T cells, (G) PD-1+ CD8, and (H) PD-1+ CD4 T cells. (I and J) TCGA data evaluated for PCs (PRDM1/IRF4/CD138) and MAGEA4. (I) PCs are significantly enriched in human MAGE-A4pos NSCLC. (J) Human NSCLC with high enrichment of PCs are positively correlated with MAGE-A4pos. Significance determined by Welch’s or Student’s t test for normally distributed data and Wilcoxon rank-sum test for non-normal data. *P < 0.05 and **P < 0.01.
Fig 4: MAGE-A4 role in NSCLC development by immune modulation.MAGE-A4 promotes tumorigenesis and metastasis. MAGE-A4 stimulates endothelial cells to increase production of CXCL12, which recruits and retains cognate CXCR4+ PCs, expressing TGF-β and IL-10 and producing IgA. These immunosuppressive PCs inhibit antigen-presenting cells, including macrophages and cross-presenting CD103+ dendritic cells. Immunosuppressive PCs also inhibit the infiltration and activation of all T cell subsets. While CD163+ TAMs are increased in the MAG4/Pt tumor microenvironment, these are driven by B cells, and the absence of both immune cell types does not affect tumor burden. In contrast, PCs promote tumor progression.
Fig 5: De novo hMAGE-A4 expression in the lung causes invasive LUAD in the absence of Pten.(A) Breeding scheme to generate mouse models and controls: CCSPWT hMAGE-A4LSL/Ptenf/f (Cre-MAG4/Pt), CCSPiCre Ptenf/f (Pt), CCSPiCre hMAGE-A4LSL (MAG4), and CCSPiCre hMAGE-A4LSL/Ptenf/f (MAG4/Pt). (B) Representative H&E-stained images of Cre-MAG4/Pt, Pt, MAG4, and MAG4/Pt lung from 6- to 7-month-old mice demonstrate tumors in MAG4/Pt but not in MAG4 or Pt alone. Arrows indicate areas with loss of basement membrane and invasion. The scale bar is 300 μm; the inset scale bar is 150 μm. (C) Tumor incidence among groups of mice at indicated ages as determined by histology. n = 8 to 27 per group. (D) Representative IHC of MAGE-A4 expression in all four genotypes at 6 to 7 months of age. Black arrows in MAG4/Pt indicate areas of invasion with loss of basement membrane. The scale bar is 300 μm (left panel) and 75 μm (right insets). (E) H&E (top) and respective IHC of MAGE-A4 (bottom) of Cre-MAG4/Pt and MAG4/Pt spleen at 12 months of age; scale bar, 75 μm. (F) Summary of MAG4/Pt tumors metastasize to spleen, intestine, and cervical lymph nodes by age as determined from d-luciferin screening. (G) Detection of cgb1a1 in 12 to 13 months in Cre-MAG4/Pt versus MAG4/Pt spleen by quantitative polymerase chain reaction (qPCR), n = 4 per group, one experiment. Student’s t test, *P < 0.05
Supplier Page from Abcam for Recombinant Human MAGEA4 protein (His tag N-Terminus)