Lithium inhibits tumor lymphangiogenesis and metastasis through the inhibition of TGFBIp expression in cancer cells

Abstract

Metastasis is the main cause of mortality in cancer patients. Although there are many anti-cancer drugs targeting tumor growth, anti-metastatic agents are rarely developed. Angiogenesis and lymphangiogenesis are crucial for cancer progression; in particular, lymphangiogenesis is pivotal for metastasis in cancer. Here we report that lithium inhibits colon cancer metastasis by blocking lymphangiogenesis. Lithium reduces the expression of transforming growth factor-β-induced protein (TGFBIp) in colon cancer cells by inhibiting Smad3 phosphorylation via GSK3β inactivation. Moreover, lithium inhibits lymphatic endothelial cell migration, which is increased upon TGFBIp expression in tumor cells. Lithium had no significant effect on SW620 tumor growth in vitro and in vivo; however, it inhibited lymphangiogenesis in tumors. In tumor xenografts model, lithium was found to prevent metastasis to the lungs, liver, and lymph nodes by inhibiting TGFBIp-induced tumor lymphangiogenesis. Collectively, our findings demonstrate a novel role of lithium in the inhibition of colon cancer metastasis by blocking TGFBIp expression, and thereby TGFBIp-induced lymphangiogenesis, in primary tumors.

Figures

Figure 1. Lithium carbonate had no effect on tumor growth.

SW620 cells were injected subcutaneously into the midline of the backs of NOS-SCID mice (n = 5 per group). After tumor cell inoculation, mice were randomized to receive either 125 mg/kg lithium carbonate (Li2CO3) in 100 μL deionized water (d.H2O), or 100 μL d.H2O alone, two times per week by oral gavage, from the time of tumor inoculation. Tumors were removed after 62 days, excised, and serially sectioned. (a) Representative tumors. (b,c) Analysis of tumor growth rates and weights. (d) TGFBIp protein levels in tumors derived from lithium- or control-injected mice were measured by western blotting.

Figure 2. Lithium-carbonate reduces tumor lymphangiogenesis in colon cancer.

(a) Margins of tumor sections were stained for infiltrating LECs using anti-LYVE1 and anti-CD31 antibodies. Nuclei were stained with DAPI (blue fluorescence). Arrows indicate lymphatic vessels. Yellow dotted lines point to the margins of the tumor. Images were viewed using an Olympus IX81-ZDC microscope with a LUCPL FLN 10×/1.0 NA lens. (b) Central regions of tumor sections were stained for infiltrating LECs using anti-LYVE1 and anti-CD31 antibodies. Nuclei were stained with DAPI (blue fluorescence). Arrows indicate lymphatic vessels. Images were viewed by using an Olympus IX81-ZDC microscope with a LUCPL FLN 10×/1.0 or 2.0 NA lens. (c) Quantitative assessment of LYVE1- and CD31-positive lymphatic vessels per field for each tumor section. Data are presented as the mean ± S.E. **P < 0.01 vs. control group.

Figure 3. Lithium carbonate suppresses tumor metastases to lung and liver.

(a,c) Images of lung and liver metastases of mice injected with SW620 cells, and hematoxylin and eosin staining of frozen sections of lung and liver tissues isolated from mice subcutaneously injected with SW620 cells. (b,d) The number of metastatic nodules per lung or liver was quantified by microscopic inspection. **P < 0.01 vs. control group. Black dotted lines indicate metastatic tumors in the mouse lung or liver. Data are presented as the mean ± s.e. **P < 0.01 vs. control group.

Figure 4. Lithium carbonate suppresses CCR7+ tumor metastases to lung and liver.

(a,b) CCR7 staining of frozen sections of lung and liver tissues isolated from mice subcutaneously injected with SW620 cells. Arrows indicate CCR7-positive metastatic tumors in the mouse lung or liver.

Figure 5. Lithium carbonate suppresses tumor metastasis to the lymph nodes.

(a) Lateral axillary lymph node morphology of tumor-bearing groups (n = 5 per group). (b,d) Hematoxylin and eosin, and CCR7 staining of frozen sections of lateral axillary lymph node tissues isolated from mice subcutaneously injected with SW620 cells. (c) The incidence of metastasis per lateral axillary lymph node was quantified by microscopic analysis. Data are presented as the mean ± s.e. **P < 0.01 vs. control group.