Lymphatic Endothelial Markers and Tumor Lymphangiogenesis Assessment in Human Breast Cancer

Abstract

Metastasis via lymphatic vessels or blood vessels is the leading cause of death for breast cancer, and lymphangiogenesis and angiogenesis are critical prerequisites for the tumor invasion-metastasis cascade. The research progress for tumor lymphangiogenesis has tended to lag behind that for angiogenesis due to the lack of specific markers. With the discovery of lymphatic endothelial cell (LEC) markers, growing evidence demonstrates that the LEC plays an active role in lymphatic formation and remodeling, tumor cell growth, invasion and intravasation, tumor-microenvironment remodeling, and antitumor immunity. However, some studies have drawn controversial conclusions due to the variation in the LEC markers and lymphangiogenesis assessments used. In this study, we review recent findings on tumor lymphangiogenesis, the most commonly used LEC markers, and parameters for lymphangiogenesis assessments, such as the lymphatic vessel density and lymphatic vessel invasion in human breast cancer. An in-depth understanding of tumor lymphangiogenesis and LEC markers can help to illustrate the mechanisms and distinct roles of lymphangiogenesis in breast cancer progression, which will help in exploring novel potential predictive biomarkers and therapeutic targets for breast cancer.

Keywords: breast cancer; lymphangiogenesis; lymphatic endothelial cell; lymphatic vessel density; lymphatic vessel invasion; markers.

Figure 2

Lymphangiogenesis and lymphatic endothelial cellular origins in breast cancer. (a) LEC precursors are polarized by the specific expression of PROX-1 in nuclei; then, the expression of VEGFR-3, LYVE-1, and podoplanin on the cell membrane is up-regulated and the expression of VEC-specific genes is down-regulated. (b) Among multiple signaling axes, VEGF-C/VEGF-D/VEGFR-3 is the central axis promoting LEC proliferation, migration, and survival. Podoplanin-expressing TAMs [42] and the CCL21/CCR7 chemokine axis [44] might mediate VEGF-C secretion by tumor cells and stimulate lymphangiogenesis. Additionally, tumor cells expressing lysyl oxidase-like protein 2 might enhance lymphangiogenesis via stimulating VEGF-C and CXCL12 secretion by TAFs [46] and activating the CXCR4/CXCL12-LYVE-1 axis [45]. The majority of tumor LECs sprout from pre-existing lymphatic vessels; a few LECs were found to originate from bone marrow-derived cell progenitors, such as M-LECPs [36] and TEMs [49].

Figure 1

Tree-shaped lymphatics are composed of lymphatic capillaries, initial collecting lymphatics, afferent and efferent lymphatics, and the thoracic duct. Lymphatic capillaries are expanded blind-ended irregular lumens lined with a single layer of oak-leaf-shaped LECs without pericytes or SMCs, and anchored to the surrounding tissue via anchoring filaments. There is no BM around the capillaries. Overlapping flaps between adjacent oak-leaf-shaped LECs are anchored on the sides by discontinuous button-like junctions. CD31 is located at the tip of flaps, where there is a lack of buttons. By contrast, collecting lymphatics are lined with spindle-shaped LECs anchored by zipper-like junctions and covered with continuous BM and SMCs. Gaps at overlapping flaps in lymphatic capillaries are lymph inlets, while valves in collecting lymphatics are backflow-prevention structures.

Figure 3

Immunoreactivity of lymphatic vessel endothelial hyaluronan receptor-1 (LYVE-1), D2-40, and prospero homeobox gene protein 1 (PROX-1) in breast cancer tissues. D2-40 showed higher sensitivity in identifying lymphatic endothelial cells (LECs) than anti-LYVE-1 antibody. The anti-LYVE-1 antibody showed weak or no immunoreactivity on lymphatics (red arrows) that were positive for D2-40 (black arrows) in periphery normal tissue ((a,d), 100×), peritumoral stroma ((b,e), 400×), and intratumoral tissue ((c,f), 200×). D2-40-positive stainings were localized on membrane of LECs (black arrows) ((g), 200×). PROX-1-positive stainings were localized in nuclei of LECs (yellow arrows) ((h), 200×). The lymphatic vessel seemed to be negative in PROX-1 staining (red arrow), because the nuclei of the lymphatic vessel were not present at this section ((h), 200×). Protocols for immunohistochemistry stainings were stated in Appendix A and the antibodies used were shown in Table A1.

Figure 4

Double D2-40/CD31 immunofluorescence labeling was utilized to visualize distribution pattern and morphology of tumor-associated lymphatic vessels in breast cancer tissues. (a) Continuous D2-40-positive and discontinuous CD31-positive lymph vessels are mainly present in peritumoral area, but are negligible in intratumoral area (50×). (b) D2-40-positive large dilated lymphatic vessels in peritumoral area (white arrows, 200×). (c) D2-40-positive collapsed vessels (white arrows) and long narrow lymphatic vessel (yellow arrow, 400×) in intratumoral areas. (d) D2-40-positive lymphatic endothelial bundle (white arrow, 400×) in intratumoral areas. (e) D2-40-positive tumor-associated lymphatic endothelial cells (red arrows, 400×) in intratumoral areas. Protocols for immunofluorescence and immunohistochemistry stainings were stated in Appendix A and the antibodies used were shown in Table A1.

Figure 5

D2-40/CD31 immunohistochemistry staining was utilized to distinguish between lymphatic vessel invasion (LVI) and clefts, blood vessels invasion, or ductal carcinoma in situ in breast cancer tissues. (a) Clefts in HE (red arrows). (b) CD31-negative clefts (blue arrows) and CD31-positive blood vessels (green arrows). (c) D2-40-negative clefts (black arrows). (d) LVI in HE staining (red arrow). (e) Discontinuous CD31-positive LVI (blue arrow) and continuous CD31-positive blood vessels (green arrows). (f) Continuous D2-40-positive LVI (black arrow) and D2-40-positive myoepithelial cells (red stars). (g) Ductal carcinoma in situ (DCIS) in HE staining (red arrow). (h) CD31-negative DCIS (blue arrow) and CD31-positive blood vessels (green arrows). (i) D2-40-positive myoepithelial layer of the DCIS (red stars). (200×). Protocols for HE and immunohistochemistry stainings were stated in Appendix A and the antibodies used were shown in Table A1.