Novel Endometrial Cancer Models Using Sensitive Metastasis Tracing for CXCR4-Targeted Therapy in Advanced Disease

Abstract

Advanced endometrial cancer (EC) lacks therapy, thus, there is a need for novel treatment targets. CXCR4 overexpression is associated with a poor prognosis in several cancers, whereas its inhibition prevents metastases. We assessed CXCR4 expression in EC in women by using IHC. Orthotopic models were generated with transendometrial implantation of CXCR4-transduced EC cells. After in vitro evaluation of the CXCR4-targeted T22-GFP-H6 nanocarrier, subcutaneous EC models were used to study its uptake in tumor and normal organs. Of the women, 91% overexpressed CXCR4, making them candidates for CXCR4-targeted therapies. Thus, we developed CXCR4⁺ EC mouse models to improve metastagenesis compared to current models and to use them to develop novel CXCR4-targeted therapies for unresponsive EC. It showed enhanced dissemination, especially in the lungs and liver, and displayed 100% metastasis penetrance at all clinically relevant sites with anti-hVimentin IHC, improving detection sensitivity. Regarding the CXCR4-targeted nanocarrier, 60% accumulated in the SC tumor; therefore, selectively targeting CXCR4⁺ cancer cells, without toxicity in non-tumor organs. Our CXCR4⁺ EC models will allow testing of novel CXCR4-targeted drugs and development of nanomedicines derived from T22-GFP-H6 to deliver drugs to CXCR4⁺ cells in advanced EC. This novel approach provides a therapeutic option for women with metastatic, high risk or recurrent EC that have a dismal prognosis and lack effective therapies.

Keywords: CXCR4-targeted nanoparticles; advanced endometrial cancer; animal model; metastasis; orthotopic model.

Figure 1

CXCR4 expression in tissue microarrays obtained from endometrial cancer patients. (A) CXCR4 is overexpressed in tumor tissue (n = 79), as compared to adjacent healthy endometrial tissue (n = 27), that shows almost undetectable CXCR4 expression levels by immunohistochemistry (IHC) (Mann–Whitney test, *** p = 0.000; mean ± s.e.m). (B) CXCR4 protein expression after IHC is negligible in healthy endometrial tissue, whereas it is highly overexpressed and mostly located in the cell membrane of tumor endometrial tissue. Bar: 20 µm. NT: non-tumor, healthy tissue; T: tumor tissue.

Figure 2

Procedure to orthotopically inoculate human cell lines in NSG mice to obtain an aggressive endometrial cancer model that metastasizes in all clinically relevant organs. (A) Representative photographs of the taken procedure, starting by showing uterus exposure (1), followed by right horn ligature without clamping the arterial irrigation system (2 and 3) and intraluminal transmyometrial injection of 10⁶ CXCR4⁺ Luciferase⁺ AN3CA or CXCR4^- Luciferase⁺ AN3CA cells suspended on culture medium using a Hamilton syringe. (B) Representative image of Luciferase bioluminiscence emission in a mouse imlpanted with Luciferase⁺ AN3CA cells, indicating the correct engraftment of tumor cells inside the uterus. (C) Representative images of relevant organs bearing metastases after necropsy and their ex vivo emission of bioluminiscence, registered by IVIS Spectrum.

Figure 3

Dissemination pattern of the xenograft orthotopic model of advanced endometrial cancer generated in NSG mice and comparison of dissemination between models derived from CXCR4⁺ and CXCR4⁻AN3CA cells. Histology of liver, lungs and lymph nodes with possible metastatic colonization after hematoxilin-eosin (H&E) staining of tissue sections or using immunohistochemical staining for anti-human CXCR4 or anti-human vimentin. Note the dramatically higher sensitivity of the anti-human vimentin IHC staining, to spot single or clustered cancer cells in the mouse tissues as compared to anti-CXCR4 IHC or H&E-stained tissues. Bar: 100 µm.

Figure 4

In vitro CXCR4-dependent internalization and lack of cytotoxicity of T22-GFP-H6 nanocarrier in human endometrial cancer cell line AN3CA. (A) T22-GFP-H6 internalization in CXCR4⁺ AN3CA cells at 1, 6 and 24 h and different concentrations, expressed as percentage of GFP⁺ cells. (B) Blockage of T22-GFP-H6 internalization at 6 h, measured by flow cytometry, in CXCR4^- AN3CA cells, CXCR4⁺ AN3CA cells and CXCR4⁺ AN3CA cells after 1h pretreatment with 1 µM of the CXCR4 antagonist AMD3100 (Mann-Whitney test; * p < 0.05, ** p < 0.01; n = 3; mean ± s.e.m). (C) T22-GFP-H6 cytotoxicity on CXCR4⁺ AN3CA cells as measured by XTT viability test at 48 h. (n = 3; mean ± s.e.m).

Figure 5

In vivo biodistribution and toxicity assessment of nanocarrier T22-GFP-H6 in a subcutaneous mouse model derived from CXCR4+ AN3CA cells. (A) Representative images of fluorescence emitted by the nanocarrier after 2, 5 and 24 h after intravenous injection of 200 µg of T22-GFP-H6. (B) Area under the curve representation of fluorescence emitted over time by tumor and non-tumor tissues, and their respective percentage of nanocarrier uptake out of the total fluorescence emitted by T22-GFP-H6 in all tissues (n = 4/group; mean ± s.e.m). (C) Hematoxylin-eosin staining of tumor and non-tumor organs 48 h after administration of the nanocarrier. Bar: 100 µm.