Figure 1.
CX3CL1 immunoreactivity in healthy and malignant ovaries.
(A) healthy ovary, CX3CL1 immunoreactivity in the OSE (i) and Fallopian tube (ii). (B) Serous (i) and mucinous (ii) benign ovarian epithelial tumors. (C) Serous (i) and mucinous (ii) borderline ovarian epithelial tumors. (D) Malignant epithelial ovarian tumors: mucinous (i), endometrioid (ii), clear-cell (iii) and serous (iv), CX3CL1 immunoreactivity in epithelial cells is confined to the cytoplasm, no staining in the nuclei of tumor cells. (E) Cytocentrifuged CD326− non epithelial (i) and CD326+ epithelial (ii) cells isolated from malignant ascites collected from a patient diagnosed with invasive EOC. CX3CL1 is detected in CD326+ cells and also in some CD326− cells. (F) Non-epithelial ovarian granulosa cell tumor, absence of CX3CL1 immunostaining. (A–F) Magnification x 40.
Figure 2.
Steady-state levels of CX3CL1 products in healthy and malignant ovaries.
(A) CX3CL1 mRNA was detected by conventional PCR at the expected size (387 bp) in representative specimens from 2 healthy ovaries, 5 cystadenomas (benign tumors), 2 borderline tumors, 5 adenocarcinomas (malignant tumors) and in the EOC-derived cell lines, SKOV3, OVCAR3 and BG1. The white vertical line separates lanes not run on the same gel. (B) CX3CL1 mRNA levels were quantified by real-time PCR and are expressed as CX3CL1 content normalized to that of ß-actin. The diagram shows the distribution of immunostaining scores versus the amount of CX3CL1 mRNAs normalized to those of ß-actin for 5 EOC samples. Each symbol represents an individual sample run in triplicate (mean value); Spearman's test, P<0.05, r = 0.88. (C) CX3CL1 immunoblots of total protein lysates from EOC specimens, from CD326+ epithelial cell-enriched malignant ascites samples, and from SKOV3, BG1 and OVCAR3 cell lines. The CX3CL1 protein is indicated as a ∼90 kDa band. ß-actin levels are shown for normalization. (D) Detection by ELISA of sCX3CL1 in the 24 h culture medium of BG1, OVCAR3 and SKOV3 cells (data are means ± SEM of three separate experiments); undetectable sCX3CL1 in the culture medium of HEK-293T cells (HEK), used as a negative control.
Figure 3.
Correlation of CX3CL1 and Ki-67 and GILZ in EOC.
(A and B) CX3CL1 and Ki-67 final scores (A, Spearman test, P<0.01, r = 0.38) and CX3CL1 and GILZ final scores (B, Spearman test, P<0.0001, r = 0.59) were positively correlated in 54 EOC specimens including serous (black squares) and non serous (white squares) samples. (C) Dendrogram generated by hierarchical agglomerative cluster analysis for the 54 EOC specimens studied, against age at diagnosis, FIGO stage, grade, Ki-67, GILZ, CX3CL1 and CXCL12 immunoreactivity levels. Two clusters are identified, with low (top) and high (bottom) levels of proliferation. Relevant specimens are labeled with numbers.
Table 1.
Correlations of CX3CL1, Ki-67 and GILZ immunoreactivity in EOC specimens.
Table 2.
Identification of high- and low-proliferation clusters.
Figure 4.
GILZ upregulation increases CX3CL1 levels in ovarian epithelial malignant cells.
(A) CX3CL1 PCR signal intensity in CTRL and pGILZ BG-1 cells was quantified by densitometry with normalization against the signal for ß-actin; results are expressed as CX3CL1/ß-actin ratios. One experiment representative out of three. (B) Immunostaining for CX3CL1 on CTRL and pGILZ BG1 cytocentrifuged cells. Magnification x 40. (C) Total cellular protein extracts of CTRL and pGILZ BG1 cells cultured with or without 100 ng/ml PMA for 24 h were analyzed by western blotting with a specific Ab recognizing the extracellular domain of CX3CL1. CX3CL1 levels were quantified by densitometry, with normalization against the signal for ß-actin; results are expressed as CX3CL1/ß-actin ratios. One experiment representative of three. (D) Histograms show the release of sCX3CL1 into the supernatant of cells treated or not with PMA, as measured by ELISA. Results are the means ± SEM of three independent experiments. *P<0.05, absence versus presence of PMA and #P<0.05, CTRL versus pGILZ BG1 cells (unpaired t test).
Figure 5.
Expression of CX3CR1 in ovarian epithelial malignant cells.
(A) Levels of CX3CR1 and of the pan-hematopoietic marker CD45 were determined by flow cytometry in 2 freshly dissociated samples of EOC specimens. Numbers indicate frequencies of CX3CR1+ CD45− cells. (B) Representative FACS profiles for CX3CR1 levels in BG1 cells. Left, surface expression of CX3CR1 under basal conditions; middle, surface expression of CX3CR1 in cells after acidic treatment; right, expression of CX3CR1 in permeabilized cells. Numbers indicate percentage of CX3CR1+ cells (mean ± SEM) for 3 independent experiments. (C) Histograms show the fluorescence intensity of CX3CR1 staining at the surface of CTRL BG1 cells cultured in the presence of 1% or 10% FBS, and of CTRL and pGILZ BG1 cells cultured in the presence of 1% FBS. Numbers indicate the percentage of CX3CR1+ cells for one representative experiment out of three.
Figure 6.
CX3CL1 promotes the proliferation of BG1 cells.
(A–C) Proliferation was measured by [3H]-thymidine incorporation (A) in CTRL cells incubated with or without 10 ng/ml rhCX3CL1 for 24 h, 9 independent experiments. Histograms represent means ± SEM, paired t test, **P<0.001; (B) in CTRL cells incubated with or without 10 ng/ml rhCX3CL1 for 24 h, in the presence or absence of 10 µg/ml CX3CR1 antagonist; each symbol represents an individual sample run in triplicate, lines represent mean values, * P<0.05, t test, one representative experiment out of 3; (C) in pGILZ cells with and without treatment with 10 µg/ml CX3CR1 antagonist replaced every 24 h for 72 h, each symbol represents an individual sample run in triplicates, lines represent mean values, * P<0.05, t test, one representative experiment out of 3. (D) Total cellular protein extracts of CTRL cells cultured in the presence and absence of 10 ng/ml rhCX3CL1 analyzed by western blotting with specific Abs. pAKT levels were quantified by densitometry, with normalization against the signal for total AKT. Results are expressed as pAKT/AKT ratios. One blot, representative of three carried out, is shown.
Table 3.
Impact of CX3CL1 on cell proliferation and pAKT content in GILZ-negative EOC specimens (scored 0).
Figure 7.
Impact of GILZ overexpression in xenografted tumors.
(A) pGILZ or CTRL BG1 cells (40×106 cells/ml) were injected subcutaneously into the right flanks of nude mice. Tumor size was measured every 5 days, for 35 days (N = 3 mice per group). Tumor volume [mm3] was calculated as follows: (length [mm]) × (width [mm])2×0.5.** P<0.001 (unpaired t test). (B) Total cellular protein extracts of xenografted tumors were analyzed by western blotting with specific Abs. CX3CL1 and GILZ levels were quantified by densitometry, with normalization against the signal for ß-actin; results are expressed as CX3CL1 or GILZ/ß-actin ratios. One blot representative of three carried out is shown. (C) Serial sections of pGILZ and CTRL xenografted tumors were stained for CX3CL1, GILZ and Ki-67. Negative control: no labeling was detected when each primary Ab was omitted. Magnification x 40.