Notch4 is required for tumor onset and perfusion

Abstract

Background: Notch4 is a member of the Notch family of receptors that is primarily expressed in the vascular endothelial cells. Genetic deletion of Notch4 does not result in an overt phenotype in mice, thus the function of Notch4 remains poorly understood.

Methods: We examined the requirement for Notch4 in the development of breast cancer vasculature. Orthotopic transplantation of mouse mammary tumor cells wild type for Notch4 into Notch4 deficient hosts enabled us to delineate the contribution of host Notch4 independent of its function in the tumor cell compartment.

Results: Here, we show that Notch4 expression is required for tumor onset and early tumor perfusion in a mouse model of breast cancer. We found that Notch4 expression is upregulated in mouse and human mammary tumor vasculature. Moreover, host Notch4 deficiency delayed the onset of MMTV-PyMT tumors, wild type for Notch4, after transplantation. Vessel perfusion was decreased in tumors established in Notch4-deficient hosts. Unlike in inhibition of Notch1 or Dll4, vessel density and branching in tumors developed in Notch4-deficient mice were unchanged. However, final tumor size was similar between tumors grown in wild type and Notch4 null hosts.

Conclusion: Our results suggest a novel role for Notch4 in the establishment of tumor colonies and vessel perfusion of transplanted mammary tumors.

Figure 1

Notch4 expression is upregulated in the vasculature of mouse and human mammary tumors. (a) Notch4/CD31 immunostaining reveals an increase of Notch4 expression in the vasculature of mouse orthotopic MMTV-PyMT mammary tumors (vii-xii) compared to age and stage matched (virgin 8 weeks old) normal mouse mammary glands (i-vi). In panels vii-viii arrows indicate Notch4 low-expressing vessels; arrowheads indicate Notch4 positive vessels. Panels i to iii show low expression levels of Notch4 and CD31 in the vessels of normal mammary gland. Panels iv to vi show a detail of the staining of Notch4 and CD31 in the vessels of normal mammary gland. Panels vii-xii show Notch4 and CD31 staining in tumors without adjacent normal tissues in these images. Panels vii and ix show higher expression of Notch4 in vessels of the tumors than those in the normal gland (as compared to panels i and iii). Panels x (Notch4 staining), xi (CD31 staining) and xii (merged) show higher magnification of tumor vessels with Notch4 expression. (b) Notch4 staining in the tumor is mainly restricted to the vasculature. Orthotropic MMTV-PyMT mammary tumors at 3 weeks post transplant were immunostained for (i) Notch4, and (ii) CD31. Panel iii shows the merge of Notch4 and CD31. (c) Notch4 mRNA expression is increased in orthotopic tumors derived from transplanted MMTV-PyMT cells compared to normal mammary tissue as measured by quantitative RT-PCR. Data is represented as fold change of Notch4 expression in tumor over normal mammary glands, normalized to the expression of Tie2, VE-cadherin and CD31. Samples were analyzed in duplicate and error bars represent standard error of mean (s.e.m). (d) Notch4 immunostaining (brown stain) is increased in the vasculature of human infiltrating ductal carcinoma tissue (ii) compared to adjacent normal tissue (i). Arrows indicate Notch4 staining in the endothelium.

Figure 2

Vessel density is comparable in Notch4-deficient and wild type mouse mammary glands. (a) CD31 immunostaining (green) pattern is similar in wild type (i) and Notch4−/− (ii) mouse mammary glands. (b) Vessel density (represented as branch points/20X field in intact inguinal mammary fat pads) is similar in wild type and Notch4−/− mice. Bars represent average values and error bars represent s.e.m.

Figure 3

Host Notch4 deficiency delays onset of tumors derived from orthotopic transplantation of MMTV-PyMT cells. Incidence of palpable tumors in wild type and Notch4−/− mice at 1, 2 and 3 weeks post orthotopic transplantation of MMTV-PyMT tumor cells. Tumor onset is significantly delayed in Notch4−/− mice at the early time points.

Figure 4

Vascular perfusion is compromised in tumors developed in Notch4^−/−hosts whereas vessel density is unaffected. (a) CD31 immunostaining (green) and lectin-perfused (red) blood vessel patterns in orthotopic MMTV-PyMT tumors are similar in wild type (i) and Notch4^−/− (ii) hosts. (b) Vessel perfusion in MMTV-PyMT tumors is decreased in Notch4^−/− hosts compared to wild type hosts 9 days after orthotopic transplantation. (c and d) Vessel density, either measured as number of vessel segments (c) or as number of branch points (d) in MMTV-PyMT tumors is similar in wild type and Notch4^−/− hosts 9 days after orthotopic transplantation. (e) CD31 staining in tumors developed in wild type (i) and Notch4^−/− (ii) hosts 2 weeks after transplantation. (f) Vessel density in MMTV-PyMT tumors is similar in wild type and Notch4^−/− hosts 2 weeks after orthotopic transplantation. Bars represent average values and error bars represent s.e.m.

Figure 5

Tumor mass and vascular architecture are similar in tumors developed in wild type and Notch4^−/−hosts at later stages. (a) Tumor mass of MMTV-PyMT tumors at 3 weeks after orthotopic transplantation. Bars represent average values and error bars represent s.e.m. (b) CD31 staining highlighting tumor vascular architecture 3 weeks after orthotopic transplantation.