Tumor galectin-1 mediates tumor growth and metastasis through regulation of T-cell apoptosis

Abstract

Galectin-1 (Gal-1), a carbohydrate-binding protein whose secretion is enhanced by hypoxia, promotes tumor aggressiveness by promoting angiogenesis and T-cell apoptosis. However, the importance of tumor versus host Gal-1 in tumor progression is undefined. Here we offer evidence that implicates tumor Gal-1 and its modulation of T-cell immunity in progression. Comparing Gal-1-deficient mice as hosts for Lewis lung carcinoma cells where Gal-1 levels were preserved or knocked down, we found that tumor Gal-1 was more critical than host Gal-1 in promoting tumor growth and spontaneous metastasis. Enhanced growth and metastasis associated with Gal-1 related to its immunomodulatory function, insofar as the benefits of Gal-1 expression to Lewis lung carcinoma growth were abolished in immunodeficient mice. In contrast, angiogenesis, as assessed by microvessel density count, was similar between tumors with divergent Gal-1 levels when examined at a comparable size. Our findings establish that tumor rather than host Gal-1 is responsible for mediating tumor progression through intratumoral immunomodulation, with broad implications in developing novel targeting strategies for Gal-1 in cancer.

Figure 1. Silencing Gal-1 expression in LLC cells delayed tumor growth and metastasis

Panel A: Average tumor volume over time for the 4 different groups of mice. The shGal-1 tumors were significantly smaller compared to Scr tumors regardless of mouse phenotype (p = 0.01). Tumor growth was impaired when mice (WT and Gal-1 null mice) were inoculated with shGal-1 cells. The Scr tumors in WT and Gal-1 null mice showed similar growth rate. Panel B upper: In-vitro growth curves showing similar growth rate between WT (wild-type parental), Scr and shGal-1 tumors. Panel B lower: Ki67 staining (200x magnification; scale bar = 100 μm) showing similar number of Ki-67 positive cells in excised Scr and shGal-1 tumors. Panel C: the upper graph shows the absolute number of spontaneous lung metastasis quantified with light microscopy. The lower table shows the average number of lung metastases/section for each mouse group. Note that Gal-l WT/Scr mice had the highest number of lung metastases (n = 12, 4.1 ± 0.4 mets/lung section) followed by Gal-1 null/Scr mice (n = 12, 2.3 ± 0.4 mets/lung section). None of the mice with shGal-1 tumors developed spontaneous lung metastasis (Gal-1 WT and Gal-1 null: n = 12, 0 ± 0 mets/lung section).

Figure 2. The effects of Gal-1 on tumor growth and metastasis are eliminated in immuno-compromised mouse models

Panel A: The average tumor volume over time of Scr and shGal-1 tumors grown in Nod.Cg-Prkdc^scid IL2rg^tm1Wjl/SzJ (Nod-Scid; n = 15), B6.Cg-Foxn1^nu/J (T cell^−/−; n = 15) and B6.129S2-Ign-6^tm1Cgn/J (B cell^−/−; n = 15) mice. Note the growth rate was similar for both tumor types. Panel B: Average number of spontaneous lung metastases for Scr (n = 5) and shGal-1 (n = 7) in Nod-Scid mice. The mice inoculated with Scr (19.8 ± 3.8 mets/lung section) and shGal-1 (17.4 ± 2.8 mets/lung section) tumors show similar number of lung metastasis.

Figure 3. Silencing Gal-1 expression in tumor cells enhanced lymphocyte survival

Panel A: CD3 staining (DAB detection, 200x magnification, scale bar = 50 μm) demonstrated greater intratumoral lymphocytes in the shGal-1 tumors than the Scr tumors grown in the Gal-1 WT host. Staining of mouse lymph nodes is used as a positive control. Panel B: colocalization of CD4 and CD8 (red) with TUNEL (green) was performed on Gal-1 WT mouse tumors (400x magnification, scale bar = 50 μm). Cell nucleus is labeled with DAPI (blue). The greatest amount of CD8 cell death is found in the Scr tumors (n = 4, 61.2 ± 8.3%) when compared to the shGal-1 tumors (n = 4, 15.3 ± 4.6%). Similarly, CD4 T-cell death in the Scr tumors (n = 4, 46.2± 7.5%) is greater than the shGal-1 tumors (n = 4, 3.9 ± 1.5%).

Figure 4. Pro-angiogenic effect of Gal-1 is related to the tumor volume

Panel A: Quantification of microvessel density count (MVD) based on CD31 staining were performed on WT mouse tumor tissues (200x magnification, scale = 100 μm). The graph shows the timepoints when tumors were collected to examine for MVD. The MVD was examined at 22 days after tumor implantation when tumor volume between the Scr and shGal-1 tumors showed the largest difference (panels A I and B I). At 22 days post-implantation, the MVD was significantly reduced in the shGal-1 tumor (n = 3, 34.0 ± 4.6 microvessels/mm²; tumor volume: 0.35 ± 0.15 cm³) when compared to the Scr tumors (n = 3, 62.0 ± 5.7 microvessels/mm², tumor volume: 0.85 ± 0.25 cm³). The MVD of tumors at collected at >25 days after tumor implantation when tumor volume was comparable between the two groups was also examined (panels A II and B II). There was no difference in MVD between the Scr (n = 3, 40.1 ± 2.6 microvessels/mm², tumor volume: 1.80 ± 0.25 cm³) and shGal-1 (n = 3, 40.9± 2.9 microvessels/mm², tumor volume: 1.46 ± 0.42 cm³) tumors. Finally, the MVD of Scr tumors collected when the tumor volume (0.36 ± 0.08 cm³, < 22 days) was comparable to that of shGal-1 tumor at day 22 was examined (panels A III and B III). The MVD of the smaller Scr tumors (n = 3, 33.5 ± 1.6 microvessels/mm²) is simlar to the MVD of shGal-1 at day 22 post implantation (34.0 ± 4.6 microvessels/mm²). Panel C: quantification of MVD in Scr and shGal-1 tumors grown in B cell^−/− and T-cell^−/− mice showed no significant difference between the Scr and shGal-1 tumors. The average MVD for Scr in the B cell^−/− host (n = 3, 43.3 ± 2.1 microvessels/mm²) and T-cell^−/− host (n = 3, 22.8 ± 1.3 microvessels/mm²) did not differ from their respective shGal-1 tumors in the B cell^−/− host (n = 3, 37.7 ± 3.7 microvessels/mm²) and T-cell^−/− host (n = 3, 23.9 ± 1.0 microvessels/mm²).