Dysfunctional endothelial cells directly stimulate cancer inflammation and metastasis

Abstract

Although the influence of context-dependent endothelial cell (EC) regulation of vascular disease and repair is well-established, the privileged roles ECs play as paracrine regulators of tumor progression has only recently become appreciated. We hypothesized that if the same endothelial physiology governs vascular and cancer biology then EC control in cancer should follow endothelial regulation of vascular health. Healthy ECs promote vascular repair and inhibit tumor invasiveness and metastasis. Dysfunctional ECs have the opposite effects in vascular disease, and we now ask if dysfunctionally activated ECs will promote cancer cell inflammatory signaling and aggressive properties. Indeed, while factors released from quiescent ECs induce balanced inflammatory signaling, correlating with decreased proliferation and invasiveness, factors released from dysfunctional ECs robustly activated NF-κB and STAT3 signaling within cancer cells, correlating with increased in vitro invasiveness and decreased proliferation and survival. Furthermore, matrix-embedded dysfunctional ECs stimulated intratumoral pro-inflammatory signaling and spontaneous metastasis, while simultaneously slowing primary tumor growth, when implanted adjacent to Lewis lung carcinoma tumors. These studies may broaden our appreciation of the roles of endothelial function and dysfunction, increase understanding and control of the tumor microenvironment, and facilitate optimization of anti-angiogenic and vascular-modifying therapies in cancer and other diseases.

Keywords: angiogenesis; endothelium; inflammation; metastasis.

Figure 1. The in vitro “dysfunctional” EC (DEC) phenotype includes dysregulated proliferation, tube formation, high permeability, and avid monocyte binding

(A) Growth curve of endothelial cells cultured under intact (“EC”) or dysfunctional (“DEC”) in vitro conditions. (B) Tube length per field of ECs and DECs after 2 days of culture. (C) Permeability to the passage of FITC-dextran of confluent EC or DEC monolayers after 4 days of culture. (D) Adhesion of THP-1 monocytic leukemia cells to confluent EC or DEC monolayers after 4 days of culture. * p<0.05 versus EC by t test. n = 4 per condition.

Figure 2. The DEC phenotype manifests in increased dysfunctional indices, reduced quiescent indices, altered and inflamed morphology, and highly pro-inflammatory and pro-thrombotic expression profiles

(A) Western blot of whole cell lysates of ECs and DECs, with quantification for DEC lysates relative to EC lysates. (B) Immunofluorescent staining for actin (fluorescent phalloidin), perlecan, and NF-κB p65 in ECs and DECs. Nuclei are labeled in blue (DAPI). (C) qRT-PCR array analysis of endothelial inflammatory, thrombotic, and quiescent differentiation genes. (D) Cytokine dot blot of EC and DEC secretions. * p<0.05 by t test. n = 3 per condition.

Figure 3. ECs inhibit cancer cell proliferation, but pro-inflammatory DECs induce cancer cell death

(A) MTT assay of the A549, NCI-H520, and HOP62 lung carcinoma cells after 4 days of culture in either unconditioned (control), EC-conditioned, or DEC-conditioned media. (B) BrdU incorporation of the same cells during hours 24–48 of culture in the same sets of conditioned media. (C) Caspase-3/7 activity in A549, NCI-H520, and HOP62 cells, and Western blot of full and cleaved PARP in A549 cells, after 4 hours of culture. * p<0.05 versus control, ⁺ p<0.05 versus EC by t test. n = 3 per condition.

Figure 4. ECs control lung cancer inflammatory signaling and reduce invasiveness, whereas DECs robustly stimulate lung cancer inflammatory signaling and stimulate invasiveness

(A) Time course of STAT3 (p-STAT3) and NF-κB pathway activity (p-P65 and IκBα) induced by unconditioned (control), EC-conditioned, or DEC-conditioned media as assayed by Western blot of whole cell lysates of A549 cells. Quantification, relative to control, is shown above each representative band. (B) In vitro chemoinvasion index of A549 cells after 4 days of culture in the same sets of conditioned media. (C) Immunofluorescent nuclear localization of NF-κB p65 in A549 cells after 4 days of culture in the same sets of conditioned media. (D) Correlation between effects of EC or DEC secretions on A549 invasiveness and NF-κB p65 nuclear localization. * p<0.05 versus control, ⁺ p<0.05 versus EC by t test. n = 3 per condition.

Figure 5. Adjacent D-MEECs cause increased spontaneous metastases and in parallel slow the net growth of primary tumors

(A) Tumor volumes estimated by caliper measurements of explanted Lewis lung carcinoma tumors. (B) Percent of Ki67-positive nuclei of thresholded immunofluorescent primary tumor cryosection images. (C) Numbers of host (murine) cleaved caspase 3 events per 10X field of immunofluorescent primary tumor cryosections. (D) Percent of NF-κB p65-positive nuclei of immunofluorescent primary tumor cryosections. (E) Table showing fractions of mice with macroscopic regional (posterior cervical) metastasis and with pan-lobar macroscopic lung metastases 2 weeks after primary tumor resection. * p<0.05 versus control by t test. ⁺ p<0.05 versus MEEC by t test. ^{^} p<0.05 by proportion z-test. n = 6 per condition.