A Synthetic Analog of Resveratrol Inhibits the Proangiogenic Response of Liver Sinusoidal Cells during Hepatic Metastasis

Abstract

We utilized Fas21, a resveratrol analog, to modulate the function of hepatic stellate cells (HSCs) and liver sinusoidal endothelial cells (LSECs) during the angiogenic phase of murine liver metastasis by B16 melanoma and 51b colorectal carcinoma. Preangiogenic micrometastases were treated with Fas21 (1 mg/kg/day) or vehicle during the development of intra-angiogenic tracts. Mice treated with Fas21 showed reduced liver tumor foci in both liver metastasis models. Micrometastases were classified immunohistochemically, as well as according to their position coordinates and connection to local microvasculature. The volume of liver occupied by sinusoidal-type foci, containing infiltrating angiogenic capillaries, decreased by ~50% in Fas21-treated mice compared to vehicle-treated ones in both tumor metastasis models. The volume of portal foci, containing peripheral neoangiogenesis within a discontinuous layer of myofibroblasts, was similar in all experimental groups in both tumor metastasis models, but displayed enhanced necrotic central areas devoid of angiogenesis following Fas21 treatment. As a result, sinusoidal tumors from mice treated with Fas21 showed a 50% reduction in desmin(+)/asma(+) HSCs and CD31(+) vessel density, and a 45% reduction in intrametastatic VEGF mRNA compared with sinusoidal tumors from vehicle-treated mice. Necrotic portal metastases increased 2-4-fold in treated mice. In vitro, Fas21 reduced VEGF secretion by HSCs and 51b cells dose-dependently. Additionally, HSCs migration in response to tumor soluble factors was dose-dependently diminished by Fas21, as was LSEC migration in response to HSCs and tumor soluble factors. Resveratrol analog Fas21 inhibits the proangiogenic response of HSCs and LSECs during the development of murine liver metastasis.

Keywords: Angiogenesis; Cancer; Hepatic stellate cell; Liver metastasis; Liver sinusoidal endothelial cell; Tumor microenvironment.

Fig. 1

Structural and electronic features of Fas21. Geometric parameters, partial charges, and solvent accessible surfaces of (A) resveratrol and (B) Fas21. Bond distances are given in Å. Dihedral angles of the central moieties (HC=CH and HC=N, respectively) with phenyl groups A and B are given in absolute value and deg. Solvent accessible surface areas (SASA) and polar surface areas (PSA) have been calculated with a probe radius of 1.4 Å (associated with water) and are given in Ų. Partial charges in the hydroxyl groups (in red) and the sp²-hybridized nitrogen atom (in blue) are given in atomic units.

Fig. 2

Effect of Fas21 administration on avascular micrometastases from B16M and 51bCC cells. Five days after cancer cell injection, Fas21 (1 mg/kg/day) or vehicles were administered via an intragastric tube for the following 7 days. Then, livers were processed for histological analyses. Metastatic foci were arbitrarily subdivided into groups attending to their average diameter, and the average density of each group was calculated in liver sections from Fas21-treated or vehicle-treated (untreated) animals. (A) B16M, (B) 51bCC. Histograms represent the average values of two independent experiments (n=30). *p<0.01 with respect to vehicle-injected mice. F21: Fas21.

Fig. 3

Effect of Fas21 on the occurrence of B16M or 51bCC micrometastasis types regarding their position coordinates in liver tissue and their connection to local microvasculature. Metastatic foci in their angiogenic stage, larger than 250 μm in diameter and stained with H/E or anti-SMA antibodies, were separated according to their sinusoidal-type or portal-type character, were counted under microscopy, and are expressed as foci per 100 mm³: (A) Sinusoidal-type 51bCC micrometastasis from untreated mouse associated with infiltrating capillaries. (B) Portal-type 51bCC micrometastasis from untreated mouse containing blood-filled caveoli. (C) Sinusoidal-type B16M micrometastasis from Fas21 mouse containing SMA-positive myofibroblasts (blue staining) associated with infiltrating angiogenic capillaries. (D) Portal-type B16M micrometastasis from Fas21-treated mice surrounded by a discontinuous layer of SMA-positive myofibroblasts (red staining) containing blood-filled caveoli. (E, F) Histograms on the number of sinusoidal and portal foci in B16M and 51bCC metastasized livers. P: portal-type metastatic foci; S: sinusoidal-type metastatic foci; AV: avascular metastatic foci; F21: Fas21. Scale bar: 100 μm. Yellow arrows: blood-filled caveoli. *p<0.05 versus untreated mice.

Fig. 4

Effect of Fas21 in intratumoral desmin expression and angiogenesis of B16M micrometastasis. Cryostat sections from B16M-bearing livers stained with anti-desmin (in red) and/or anti-CD31 (in green) antibodies: (A, B) desmin-expressing cells (A) or CD-31-expressing cells in non-necrotic tumor foci. (C, D) Intratumoral neoangiogenic tracts composed of CD31-expressing endothelial cells and associated desmin-expressing pericytes are connected to liver sinusoidal capillaries. (E-G) Density and length of desmin- or CD31-expressing vessel tracts in non-necrotic micrometastasis. F21: Fas21. T: tumor foci. S: hepatic sinusoids. White arrows: representative microvessel inside the tumor foci. Scale bar: 100 μm. *p<0.05 versus untreated mice.

Fig. 5

Effect of Fas21 on LSECs’ chemotactic migration induced by B16M cells and by tumor-activated HSCs: (A) HSCs were cultured on collagen-coated inserts in the presence of increasing concentrations of Fas21. Then, media was replaced by DMEM in the cell compartment and supernatants from B16M in the lower compartment. Migrated cells were stained with H/E and counted under the microscope. *p<0.05 versus control cells in basal medium number of migrated HSCs in basal medium: 15 ± 8 and ^#p<0.05 versus control cells in B16M-CM. (B) LSECs were cultured on collagen-coated inserts and either 10 μM of Fas21 or vehicle. Then, media was changed and replaced by basal media in the upper cell compartment and supernatants from B16M-treated HSCs in the lower compartment. Migrated cells were stained with H/E and counted under the microscope. *p<0.05, **p<0.01 and ***p<0.005 versus control LSECs in basal medium (10 ± 2 LSECs) and ^#p<0.05 versus control cells in B16M-CM. F21: Fas21.

Fig. 6

Fas21 reduces VEGF in the tumor microenvironment: (A) Quantitative RT-PCR analysis of VEGF mRNA in laser microdissected foci from murine liver metastasis by 51bCC. =VEGFCt/βactin Ct=35,71/34,70. (B) In vitro ELISA measurement of VEGF in 51bCC- and HSC-conditioned media after treatment or not with 10 μM of Fas21. *p<0.05 versus non-tumoral tissue. **p<0.05 versus micrometastases of 51bCC-bearing mice treated with Fas21. ^#p<0.05 versus untreated 51bCC cells. ^##p<0.05 versus untreated basal HSCs. ^###p<0.05 versus untreated tumor-activated HSCs.

Fig. 7

Comparison of the antimetastatic effects on B16M between Fas21 and resveratrol: (A) Mice were injected intrasplenically with B16M cells. Animals were treated intragastrically with 1 mg/kg/day of Fas21, resveratrol, or vehicle. On day 12, livers were analyzed immunohistochemically. (B) Comparison of the effects of Fas21 and resveratrol on the proliferative potential of HSC on B16 cells. Tumor cells were culture in the presence of HSCs soluble factors pre-treated with either resveratrol or Fas 21. Viability was measured after 48 h by Presto Blue. *p<0.01 with respect to B16 viability cultured in the presence of untreated HSCs.