Figure 1. Castration-induced vascular damage is reversed by TNF signaling blockade. (A) Representative sections of CD31 immunoreactivity in Myc-CaP prostate tumors from four mice (M1–M4) castrated (upper panels) or castrated and treated with sTNFR2-Fc (lower panels). Scale bars are 100 µm. (B) Microvessel density (vessels per square millimeter). (C) Vessel cross-sectional area, in µm2; (D) Vessel perimeter length, in µm; (E) Vessel wall thickness, in µm. (B–E) Large bars, means of measures from the tumors of five vehicle-treated castrated mice and from four sTNFR2-Fc-treated castrated mice (open circles), small bars = SEM. * p < 0.05.

Figure 1. Castration-induced vascular damage is reversed by TNF signaling blockade. (A) Representative sections of CD31 immunoreactivity in Myc-CaP prostate tumors from four mice (M1–M4) castrated (upper panels) or castrated and treated with sTNFR2-Fc (lower panels). Scale bars are 100 µm. (B) Microvessel density (vessels per square millimeter). (C) Vessel cross-sectional area, in µm2; (D) Vessel perimeter length, in µm; (E) Vessel wall thickness, in µm. (B–E) Large bars, means of measures from the tumors of five vehicle-treated castrated mice and from four sTNFR2-Fc-treated castrated mice (open circles), small bars = SEM. * p < 0.05.

Figure 2. Castration-induced reduction of intratumoral blood flow in Myc-CaP tumor was dependent on TNF signaling. (A) Power Doppler (PD) images of subcutaneous Myc-CaP tumors pre-castration (Pre-Cx) of mice treated with PBS (Vehicle), tumors 1–4 (of 8 evaluable tumors). Left to right: Gray-scale ultrasound image (B-mode); PD pseudo-colored image to illustrate blood flow level (Doppler); Composite of PD image overlaid on the B-mode image (Composite). (B) PD images of tumors in panel A, one day after castration (Cx D1). (C) PD images of a second set of subcutaneous Myc-CaP tumors pre-castration of mice treated with sTNFR2-Fc, tumors 11, 12, 13, 15 (of 8 evaluable tumors). (D) PD images of tumors in panel C, one day after castration. (E) Mean PD signal (% vascularity, large bars) pre-castration, and at one and four days after castration from tumors in vehicle-treated (blue, D1 n = 8, D4 n = 7) and sTNFR2-Fc-treated mice (red, D1, D4 n = 8), PD signal of each tumor indicated by closed circles. (F) Waterfall plot of % change in vascularity in individual tumors at one day after castration. (G) Mean % change in paired measures of vascularity pre-castration versus D1 or D4 after castration. Columns and large bars are means and small bars are SEM. * p < 0.05, *** p < 0.001.

Figure 2. Castration-induced reduction of intratumoral blood flow in Myc-CaP tumor was dependent on TNF signaling. (A) Power Doppler (PD) images of subcutaneous Myc-CaP tumors pre-castration (Pre-Cx) of mice treated with PBS (Vehicle), tumors 1–4 (of 8 evaluable tumors). Left to right: Gray-scale ultrasound image (B-mode); PD pseudo-colored image to illustrate blood flow level (Doppler); Composite of PD image overlaid on the B-mode image (Composite). (B) PD images of tumors in panel A, one day after castration (Cx D1). (C) PD images of a second set of subcutaneous Myc-CaP tumors pre-castration of mice treated with sTNFR2-Fc, tumors 11, 12, 13, 15 (of 8 evaluable tumors). (D) PD images of tumors in panel C, one day after castration. (E) Mean PD signal (% vascularity, large bars) pre-castration, and at one and four days after castration from tumors in vehicle-treated (blue, D1 n = 8, D4 n = 7) and sTNFR2-Fc-treated mice (red, D1, D4 n = 8), PD signal of each tumor indicated by closed circles. (F) Waterfall plot of % change in vascularity in individual tumors at one day after castration. (G) Mean % change in paired measures of vascularity pre-castration versus D1 or D4 after castration. Columns and large bars are means and small bars are SEM. * p < 0.05, *** p < 0.001.

Figure 3. TNF signaling is necessary for castration-induced reduction of perfusion in Myc-CaP tumor. (A) Contrast-enhanced ultrasound (CE-US) images of subcutaneous Myc-CaP tumors pre-castration (Pre-Cx) of mice treated with PBS (Vehicle), tumors 1–4 (of 7 evaluable tumors) are shown. Left to right: Gray-scale ultrasound image (B-mode); contrast-mode image prior to contrast agent injection (Pre-contrast); contrast-mode image after injection at the peak enhancement of contrast (Post-contrast); pseudo-colored image of the change in contrast enhancement (Perfusion). (B) CE-US images of tumors 1–4 in panel A after one day castration (Cx D1). (C) CE-US images of a second set of subcutaneous Myc-CaP tumors pre-castration of mice treated with sTNFR2-Fc, tumors 8, 10, 11, 12 (of 7 evaluable tumors) are shown. (D) CE-US images of tumors in panel C, one day after castration, (E) Mean perfusion (large bars) pre-castration and post-castration in tumors in vehicle-treated (blue) and sTNFR2-Fc-treated (red) mice, perfusion signal of each tumor indicated by closed circles. (F) Waterfall plot of %change in perfusion in individual tumors (columns). (G) Average % change in perfusion pre-castration and post-castration. (E,G): Columns and large bars are means and small bars are SEM. * p < 0.05, ** p < 0.01.

Figure 3. TNF signaling is necessary for castration-induced reduction of perfusion in Myc-CaP tumor. (A) Contrast-enhanced ultrasound (CE-US) images of subcutaneous Myc-CaP tumors pre-castration (Pre-Cx) of mice treated with PBS (Vehicle), tumors 1–4 (of 7 evaluable tumors) are shown. Left to right: Gray-scale ultrasound image (B-mode); contrast-mode image prior to contrast agent injection (Pre-contrast); contrast-mode image after injection at the peak enhancement of contrast (Post-contrast); pseudo-colored image of the change in contrast enhancement (Perfusion). (B) CE-US images of tumors 1–4 in panel A after one day castration (Cx D1). (C) CE-US images of a second set of subcutaneous Myc-CaP tumors pre-castration of mice treated with sTNFR2-Fc, tumors 8, 10, 11, 12 (of 7 evaluable tumors) are shown. (D) CE-US images of tumors in panel C, one day after castration, (E) Mean perfusion (large bars) pre-castration and post-castration in tumors in vehicle-treated (blue) and sTNFR2-Fc-treated (red) mice, perfusion signal of each tumor indicated by closed circles. (F) Waterfall plot of %change in perfusion in individual tumors (columns). (G) Average % change in perfusion pre-castration and post-castration. (E,G): Columns and large bars are means and small bars are SEM. * p < 0.05, ** p < 0.01.

Figure 4. Castration-induced hypoxia in Myc-CaP tumor is reversed by TNF signaling blockade. (A,B) Photoacoustic images (PAI, pseudo-colored) and ultrasound (B-mode) of Myc-CaP subcutaneous tumors in four tumors from each group, pre- and post-castration from vehicle-treated mice (A) or sTNFR2-Fc-treated mice (B). (C) Mean PAI signal (%sO2, large bars) from tumors pre-castration (n = 16), and at one (n = 15) and four (n = 9) days post-castration in vehicle-treated (blue) or in sTNFR2-Fc-treated (red) mice. PAI signal from each tumor indicated by closed circles. (D) Waterfall plot of change in %sO2 in individual tumors at one day post-castration versus pre-Cx (columns). (E) Mean total hemoglobin (large bars) from tumors pre-castration (n = 16), and at one (n = 16), and at four (n = 14) days post-castration in vehicle-treated (blue) or sTNFR2-Fc-treated (red) mice. PAI signal of each tumor indicated by closed circles. (F) Mean change in paired measures of %sO2 pre-castration versus D1 or D4 post-castration. Mean (columns) and SEM (bars). * p < 0.05, *** p < 0.001.

Figure 4. Castration-induced hypoxia in Myc-CaP tumor is reversed by TNF signaling blockade. (A,B) Photoacoustic images (PAI, pseudo-colored) and ultrasound (B-mode) of Myc-CaP subcutaneous tumors in four tumors from each group, pre- and post-castration from vehicle-treated mice (A) or sTNFR2-Fc-treated mice (B). (C) Mean PAI signal (%sO2, large bars) from tumors pre-castration (n = 16), and at one (n = 15) and four (n = 9) days post-castration in vehicle-treated (blue) or in sTNFR2-Fc-treated (red) mice. PAI signal from each tumor indicated by closed circles. (D) Waterfall plot of change in %sO2 in individual tumors at one day post-castration versus pre-Cx (columns). (E) Mean total hemoglobin (large bars) from tumors pre-castration (n = 16), and at one (n = 16), and at four (n = 14) days post-castration in vehicle-treated (blue) or sTNFR2-Fc-treated (red) mice. PAI signal of each tumor indicated by closed circles. (F) Mean change in paired measures of %sO2 pre-castration versus D1 or D4 post-castration. Mean (columns) and SEM (bars). * p < 0.05, *** p < 0.001.

Figure 5. Castration-induced hypoxia is reversed by TNF blockade in prostate tumors of PbCre4 × Ptenfl/fl mice. (A,B) Photoacoustic images (PAI, pseudo-colored) and ultrasound (B-mode) of five tumors pre- and one day post-castration from vehicle-treated mice (A), or sTNFR2-Fc-treated mice (B). (C) Intra-tumoral mean PAI intensity (%sO2 Average) pre-castration, and at one and at four days post-castration in vehicle-treated (n = 8, 8, 7 respectively, blue) or in sTNFR2-Fc-treated (n = 6, 5, 4 respectively, red) mice. (D) Waterfall plot of change in intra-tumoral %sO2 one day after castration in mice treated with sTNFR2-Fc (red) or vehicle (blue). (E) Waterfall plot of change in intra-tumoral %sO2 four days after castration in mice treated with sTNFR2-Fc (red) or vehicle (blue). (F) Change in paired measures of intra-tumoral %sO2 pre-castration versus D1 or D4 after castration. Mean (columns or lines) and SEM (bars). * p < 0.05, ** p < 0.01.

Figure 5. Castration-induced hypoxia is reversed by TNF blockade in prostate tumors of PbCre4 × Ptenfl/fl mice. (A,B) Photoacoustic images (PAI, pseudo-colored) and ultrasound (B-mode) of five tumors pre- and one day post-castration from vehicle-treated mice (A), or sTNFR2-Fc-treated mice (B). (C) Intra-tumoral mean PAI intensity (%sO2 Average) pre-castration, and at one and at four days post-castration in vehicle-treated (n = 8, 8, 7 respectively, blue) or in sTNFR2-Fc-treated (n = 6, 5, 4 respectively, red) mice. (D) Waterfall plot of change in intra-tumoral %sO2 one day after castration in mice treated with sTNFR2-Fc (red) or vehicle (blue). (E) Waterfall plot of change in intra-tumoral %sO2 four days after castration in mice treated with sTNFR2-Fc (red) or vehicle (blue). (F) Change in paired measures of intra-tumoral %sO2 pre-castration versus D1 or D4 after castration. Mean (columns or lines) and SEM (bars). * p < 0.05, ** p < 0.01.

Figure 6. TNF in the absence of castration is insufficient to induce intra-tumoral hypoxia in Myc-CaP tumors. (A) Large bars, mean PAI signal (%sO2) from tumors pre-castration (n = 7), and at one day following TNF treatment (n = 7). PAI signals from individual tumors (open circles) and SEM (small bars). * p < 0.05. (B) Waterfall plot of %change in %sO2 in individual tumors (columns) after TNF treatment. Mean %change is dashed line.

Figure 6. TNF in the absence of castration is insufficient to induce intra-tumoral hypoxia in Myc-CaP tumors. (A) Large bars, mean PAI signal (%sO2) from tumors pre-castration (n = 7), and at one day following TNF treatment (n = 7). PAI signals from individual tumors (open circles) and SEM (small bars). * p < 0.05. (B) Waterfall plot of %change in %sO2 in individual tumors (columns) after TNF treatment. Mean %change is dashed line.

Figure 7. Proposed mechanism for TNF regulation of prostate regression. A proposed mechanism for prostate cancer regression following androgen deprivation therapy is illustrated, focusing on the contribution of events that are mediated by the tumor microvasculature. We propose that one of the earliest events is the activation of TNF signaling in endothelial cells, leading to increased permeability (‘leakiness’) and endothelial cell death. The damage to the endothelium leads to a cascade of vascular changes—reduced blood flow, ischemia due to reduced perfusion, and eventually transient tissue hypoxia—that likely enhances the death of the epithelial component of the tumor. A prediction of the model is that endothelial cell apoptosis precedes epithelial cell apoptosis, which has been noted in the literature (see text). There is limited support for the usual cell death consequence of hypoxia (namely necrosis or necroptosis), but there is a partial requirement for p53 in castration-induced regression, suggesting that p53-hypoxia signaling contributes to death receptor-mediated apoptosis of epithelial tumor cells.

Figure 7. Proposed mechanism for TNF regulation of prostate regression. A proposed mechanism for prostate cancer regression following androgen deprivation therapy is illustrated, focusing on the contribution of events that are mediated by the tumor microvasculature. We propose that one of the earliest events is the activation of TNF signaling in endothelial cells, leading to increased permeability (‘leakiness’) and endothelial cell death. The damage to the endothelium leads to a cascade of vascular changes—reduced blood flow, ischemia due to reduced perfusion, and eventually transient tissue hypoxia—that likely enhances the death of the epithelial component of the tumor. A prediction of the model is that endothelial cell apoptosis precedes epithelial cell apoptosis, which has been noted in the literature (see text). There is limited support for the usual cell death consequence of hypoxia (namely necrosis or necroptosis), but there is a partial requirement for p53 in castration-induced regression, suggesting that p53-hypoxia signaling contributes to death receptor-mediated apoptosis of epithelial tumor cells.