Ovulation releases G-CSF to induce peritoneal neutrophil influx and netosis, facilitating peritoneal seeding of high-grade serous carcinoma.

The data that support the findings of this study are available from the corresponding author [ [email protected] ] upon reasonable request.

Tang-Yuan Chu: Conceptualization, Funding acquisition, Supervision, Writing – original draft, Writing – review & editing. Pao-Chu Chen: Conceptualization, Investigation. Aye Aye Khine: Methodology. Ying-Hsi Chen: Investigation, Methodology. Sung-Chao Chu: Investigation. Hsuan-Shun Huang: Conceptualization, Funding acquisition, Methodology, Writing – original draft, Writing – review & editing.

Protocols for clinical specimen collection were conducted in accordance with the Declaration of Helsinki and were approved by the Institutional Review Board of Hualien Tzu Chi General Hospital (IRB112-028-A). Written informed consent was obtained from all participants before each collection. All mouse experiments were approved by the Animal Care and Use Committee of Tzu Chi General Hospital (Approval No. 111-64). For IVIS tumor imaging, mice were anesthetized with 4% isoflurane, and CO₂ inhalation was used for euthanasia. All procedures were conducted according to the applicable veterinary guidelines of the American Veterinary Medical Association (AVMA).

Not applicable.

This study was supported by grants from the National Health Research Institutes (NHRI -EX112-11216BI to T.Y.C.), National Science and Technology Council, (NSTC 112-2314-B-303-002-MY3 to T.Y.C., NSTC 112-2314-B-303 -011-MY3 to H.S.H. and NSTC 112-2811-B-303-007 to A.A.K.), Taiwan, ROC; Tzu Chi General Hospital (TCRD112-044 to H.S.H.), Hualien, Taiwan.

1) Influx of neutrophils to the peritoneal cavity before and after ovulation in mice 1) Influx of neutrophils to the peritoneal cavity before and after ovulation in mice Apart from infectious conditions such as bacterial peritonitis, neutrophils are rarely found in the peritoneal cavity [ 15 , 16 ]. To investigate whether ovulation induces an increase in neutrophils in the peritoneal cavity, we analyzed peritoneal lavage fluid cells in mice before and during ovulation. In female C57BL/6 mice under controlled ovulation, peritoneal cells were collected one hour prior to or at the expected time of ovulation (12 hours after hCG administration). Mice treated with PMSG but not hCG served as the non-ovulating control ( Fig. 1 a). As shown in Figs. 1 b-c, non-ovulating mice exhibited few Ly6G+ neutrophils (5.4 ± 1.5%, 3.4 ± 1.0 × 10 4 /ml) and abundant F4/80+ macrophages (22.4 ± 12.4%, 13.5 ± 6.5 × 10 4 /ml) in their lavage fluid. In contrast, pre-ovulatory and ovulating mice displayed a dramatic increase in neutrophils (17.9 ± 2.1% or 26.9 ± 15.8 × 10 4 /ml, and 22.5 ± 2.8% or 29.5 ± 3.6 × 10 4 /ml, respectively). These results indicate a significant influx of neutrophils into the peritoneal cavity before and during ovulation. 2) Influx of neutrophils to the peritoneal cavity in women after LH surge and during ovulation Fig. 1 Peritoneal neutrophil influx before and during ovulation. (a) Timeline of controlled ovulation and collection of peritoneal cells by lavage at one hour before ovulation (Pre-OV, -1h) and during ovulation (OV, 0h). Non-ovulation control lavage (Non-OV) was performed at 0 h in mice treated with PMSG but without hCG injection. (b, c) Flow cytometry analysis (b) and quantification (c) of peritoneal lavage cells across different ovulatory phases: non-OV, Pre-OV, and OV. The entire population analyzed (lower panel) was subjected to four-quadrant gating based on Ly6G and F4/80 markers. In the upper left (UL) quadrant, Ly6G(+)/F4/80(-) cells were identified as neutrophils, while Ly6G(-)/F4/80(+) cells in the lower right (LR) quadrant were macrophages. Double-positive cells in the upper right (UR) quadrant represented macrophages engaged with neutrophils, and the lower left (LL) quadrant consisted of double-negative cells, including lymphocytes and cell debris. ** P<0.01 compared to the non-ovulation. Fig 1 2) Influx of neutrophils to the peritoneal cavity in women after LH surge and during ovulation Peritoneal neutrophil influx before and during ovulation. (a) Timeline of controlled ovulation and collection of peritoneal cells by lavage at one hour before ovulation (Pre-OV, -1h) and during ovulation (OV, 0h). Non-ovulation control lavage (Non-OV) was performed at 0 h in mice treated with PMSG but without hCG injection. (b, c) Flow cytometry analysis (b) and quantification (c) of peritoneal lavage cells across different ovulatory phases: non-OV, Pre-OV, and OV. The entire population analyzed (lower panel) was subjected to four-quadrant gating based on Ly6G and F4/80 markers. In the upper left (UL) quadrant, Ly6G(+)/F4/80(-) cells were identified as neutrophils, while Ly6G(-)/F4/80(+) cells in the lower right (LR) quadrant were macrophages. Double-positive cells in the upper right (UR) quadrant represented macrophages engaged with neutrophils, and the lower left (LL) quadrant consisted of double-negative cells, including lymphocytes and cell debris. ** P<0.01 compared to the non-ovulation. The ovulation-associated influx of neutrophils into the peritoneal cavity was validated in women. As shown in Fig. 2 and Table 1 , peritoneal fluids (PFs) from postmenopausal women (n = 3) contained the lowest abundance of CD66b high /CD11b low neutrophils, representing 13.6 ± 2.8% of the PF cell population. In the follicular phase (n = 8), the proportion of neutrophils increased to 22 ± 7.5%. This rose dramatically to 39 ± 15.7% in pre-ovulatory PFs collected before follicle puncture, and declined to a non-ovulating level of 27.3 ± 7.6% ( Fig. 2 a-c and Table 1 ). In contrast, the populations of macrophages and lymphocytes did not show significant changes throughout the menstrual cycle (Supplementary Fig. 1). Importantly, neutrophils in the pre-ovulatory PF exhibited elevated levels of the NETosis enzyme PAD4, whereas neutrophils from the follicular phase PF expressed slight amounts ( Fig. 2 d). 3) Exposure to ovulatory FF induces neutrophil influx and NETosis Fig. 2 Human polymorphonuclear neutrophils influx to the peritoneal cavity before and after ovulation (a) Flow cytometry analysis diagrams of human peritoneal fluid (PF) cells, with CD11b low /CD66b high defined as neutrophils (P2) and CD11b high /CD66b low as monocytes/macrophages (P1). The double-negative CD11b/CD66b corresponds to lymphocytes in the P3 region and cell debris in the P4 region. (b) Changes in cell population percentages in PFs of postmenopausal women and normally ovulating women during the follicular phase, pre-ovulatory phase (Pre-OV, collected before follicle puncture), and luteal phase. (c) Dot plot data illustrating the percentage change of neutrophils in PFs across each phase. (d) Analysis of PF cells for the NETotic enzyme (PAD4) expression in each phase via western blot. *P<0.05, comparison as shown in the Figure. Fig 2 Table 1 Distribution of peritoneal cells in women at different ovulation phases. Table 1 ID Age Clinic Phase Ovulation phase Lym * % Ma * % Neu * % Total events Acquiring time Lym no. (10^4/ml) Ma no. (10^4/ml) Neu no. (10^4/ml) Cohort1 1 230051 42 Right ovarian teratoma Menopaused Nil 52.0 19.6 16.0 2500 1419 43.5 16.4 13.4 2 230075 65 Uterine myoma Menopaused Nil 30.1 48.7 14.4 3000 2698 15.9 25.8 7.6 3 230076 59 Uterine prolapse Menopaused Nil 23.3 59.4 10.5 3000 3332 10.0 25.4 4.5 4 230091 46 Uterine myoma Leuteal OV+4 14.9 29.0 39.5 3000 1916 11.1 21.6 29.4 5 230096 42 Uterine myoma Leuteal OV+7 18.1 32.6 29.0 3000 1497 17.3 31.1 27.6 6 230070 54 Uterine myoma Leuteal OV+12 31.8 31.2 25.7 3000 8148 5.6 5.5 4.5 7 230055 42 Broad ligament myoma Leuteal OV+23 12.3 58.6 21.4 10000 791 73.9 352.1 128.6 8 230053 52 Uterine myoma Leuteal OV+29 23.9 43.3 20.9 10000 4875 23.3 42.2 20.4 9 230065 53 Adenomyosis Follicular OV+15 19.0 52.6 22.3 3000 305 88.8 245.9 104.2 10 230062 44 Leiomyoma Follicular OV+15 27.7 23.8 34.7 3000 1259 31.4 27.0 39.3 11 230078 49 Adenomyosis Follicular OV+20 13.4 49.0 28.5 3000 1586 12.0 44.1 25.6 12 230136 49 Uterine myoma Follicular OV+21 63.0 11.5 13.3 3000 1742 51.6 9.4 10.9 13 230061 32 Uterine myoma Follicular OV+22 19.4 42.0 23.0 3000 1881 14.7 31.8 17.4 14 230073 46 Adenomyosis Follicular OV+23 18.8 55.2 12.0 3000 1873 14.3 42.0 9.1 15 230067 37 Uterine myoma Follicular OV+23 22.9 40.0 22.3 3000 1828 17.9 31.2 17.4 16 220023 46 Broad ligament and uterine myoma Pre-ovulatory OV+28 20.7 56.6 18.8 10000 362 271.6 742.9 246.7 Cohort2 1 230007 27 IVF * Follicular Pre-OV 39.0 24.0 21.4 1349 7281 3.4 2.1 1.9 2 230038 28 IVF Follicular Pre-OV 12.0 17.0 63.4 2500 2823 5.1 7.2 26.7 3 230054 40 IVF Follicular Pre-OV 20.5 34.2 37.4 2500 342 71.2 118.8 129.8 4 230056 32 IVF Follicular Pre-OV 13.8 52.4 26.4 3000 1517 13.0 49.3 24.8 5 230058 30 IVF Follicular Pre-OV 45.3 12.3 38.8 3000 580 111.4 30.2 95.4 6 230066 35 IVF Follicular Pre-OV 20.6 49.7 19.0 3000 1129 26.0 62.8 24.0 7 230069 46 IVF Follicular Pre-OV 25.5 4.6 51.0 3000 7926 4.6 0.8 9.2 8 230071 45 IVF Follicular Pre-OV 36.6 8.7 52.2 3000 400 130.5 31.0 186.1 9 240006 33 IVF Follicular Pre-OV 22.0 18.2 52.6 3000 125 250.9 207.6 600.0 10 210114 36 IVF Follicular Pre-OV 0.9 64.0 23.0 5000 1679 1.3 90.6 32.6 ⁎ Lym: lymphocyte; Ma: macrophage; Neu: Neutrophil; IVF: in vitro fertilization 3) Exposure to ovulatory FF induces neutrophil influx and NETosis Human polymorphonuclear neutrophils influx to the peritoneal cavity before and after ovulation (a) Flow cytometry analysis diagrams of human peritoneal fluid (PF) cells, with CD11b low /CD66b high defined as neutrophils (P2) and CD11b high /CD66b low as monocytes/macrophages (P1). The double-negative CD11b/CD66b corresponds to lymphocytes in the P3 region and cell debris in the P4 region. (b) Changes in cell population percentages in PFs of postmenopausal women and normally ovulating women during the follicular phase, pre-ovulatory phase (Pre-OV, collected before follicle puncture), and luteal phase. (c) Dot plot data illustrating the percentage change of neutrophils in PFs across each phase. (d) Analysis of PF cells for the NETotic enzyme (PAD4) expression in each phase via western blot. *P<0.05, comparison as shown in the Figure. Distribution of peritoneal cells in women at different ovulation phases. Lym: lymphocyte; Ma: macrophage; Neu: Neutrophil; IVF: in vitro fertilization Ovulation is characterized by an acute inflammatory response, marked by a significant accumulation of cytokines in the mature follicle prior to rupture [ 17 ]. To investigate the effects of follicular fluid (FF) from IVF patients, we injected 200 µl of 10% FF into the peritoneal cavity of mice and analyzed changes in peritoneal cell composition at various time points post-injection. In non-ovulating mice, neutrophil counts were minimal, averaging 3.9 ± 1.6 × 10 4 per cavity, which constituted 6.6 ± 1.6% of total cells. In contrast, the injection of FF led to a dramatic increase in neutrophil count, reaching 62.8 ± 13.5 × 10 4 per cavity (33.2 ± 0.1% of total cells) at 30 minutes. This count was significantly higher than that observed in mice treated with G-CSF, which had 45.8 ± 7.7 × 10 4 neutrophils per cavity (30.9 ± 1.0%). Control mice receiving saline showed a neutrophil count of only 4.5 ± 1.3 × 10 4 per cavity (8.7 ± 2.5% of total peritoneal cells) ( Figs. 3 a-c). Further investigations into the time course revealed that the recruited peritoneal neutrophils underwent NETosis, indicated by a rapid increase in PAD4 protein levels, peaking at 30 minutes and then declining by 1 h after FF injection ( Fig. 3 d). Notable increases in citrullinated histone H3 (Cit H3) and DNA extrusion were also observed ( Figs. 3 e, f). Additionally, human peripheral blood neutrophils demonstrated a similar response to FF exposure; after a 30-minute exposure to 10% FF, CD11b levels increased by 2.6-fold ( Fig. 3 g), indicating neutrophil activation[ 18 ]. NETosis was evident 1 h after exposure, accompanied by increases in PAD4, Cit H3, and NETotic DNA fibers ( Figs. 3 h-i). These findings suggest that ovulatory FF rapidly induces neutrophil influx and activates NETotic processes in the peritoneal cavity. 4) G-CSF in FF is responsible for the NET activation in a ROS/ NOX/PAD4-dependent manner Fig. 3 The ovulatory follicular fluid induces peritoneal neutrophil influx and NETosis. (a) Experimental timeline for intraperitoneal (IP) injection of 10% FF in 200 µl saline, G-CSF (600 pg in 200 µl saline), or saline, followed by peritoneal lavage in C57BL/6 mice. (b, c) Flow cytometry analysis (b) and quantification (c) of peritoneal cells 30 minutes after IP treatment. The entire cell events analysis (lower panel) was subjected to four quadrants of gating using neutrophil (Ly6G) and macrophage (F4/80) markers. (d) Time-course of PAD4 protein levels in mouse peritoneal cells post-FF IP injection. (g) CD11b and CD66b expression on human peripheral blood (PB) neutrophils after 30-minute exposure to FF. (e, f, h, i) Fluorescence microscopy images of extruded DNA (arrows) stained with DAPI, along with Western blot analysis of NETotic protein markers, PAD4 and citrullinated histone 3 (CitH3). Scale bar: 25 µm. Fig 3 4) G-CSF in FF is responsible for the NET activation in a ROS/ NOX/PAD4-dependent manner The ovulatory follicular fluid induces peritoneal neutrophil influx and NETosis. (a) Experimental timeline for intraperitoneal (IP) injection of 10% FF in 200 µl saline, G-CSF (600 pg in 200 µl saline), or saline, followed by peritoneal lavage in C57BL/6 mice. (b, c) Flow cytometry analysis (b) and quantification (c) of peritoneal cells 30 minutes after IP treatment. The entire cell events analysis (lower panel) was subjected to four quadrants of gating using neutrophil (Ly6G) and macrophage (F4/80) markers. (d) Time-course of PAD4 protein levels in mouse peritoneal cells post-FF IP injection. (g) CD11b and CD66b expression on human peripheral blood (PB) neutrophils after 30-minute exposure to FF. (e, f, h, i) Fluorescence microscopy images of extruded DNA (arrows) stained with DAPI, along with Western blot analysis of NETotic protein markers, PAD4 and citrullinated histone 3 (CitH3). Scale bar: 25 µm. To identify the factor in FF responsible for neutrophil recruitment and activation, we conducted a cytokine array analysis of 10 FF aspirates. This analysis revealed abundant levels of G-CSF, GRO, MCP-1, and IL-8, all of which are known to induce neutrophil activation [ 19 ] ( Fig. 4 a). Among these, only recombinant G-CSF was able to induce PAD4 expression in human neutrophils at concentrations relevant to those found in FF ( Fig. 4 b). Additionally, G-CSF elicited the highest levels of DNA extrusion ( Fig. 4 c). As an increase in intracellular reactive oxygen species (ROS) is an essential step for NETosis [ 20 ], we found that the antioxidant NAC effectively abolished DNA extrusion induced by either FF or G-CSF ( Fig. 4 c). Our analysis showed a significant increase in intracellular ROS, but not in mitochondrial ROS, following treatment with FF or G-CSF ( Figs. 4 d, e), suggesting that the source of ROS is mediated by NADPH oxidase (NOX). Indeed, the application of two NOX inhibitors, DPI and GTK137831, completely inhibited the elevated ROS levels in neutrophils induced by FF ( Fig. 4 f). Notably, treatment with G-CSF-depleted FF led to a substantial decrease in intracellular ROS levels in neutrophils ( Fig. 4 f). The roles of G-CSF and PAD4 in NET formation were confirmed by reduced PAD4, CitH3 in cells treated with G-CSF-depleted FF or the PAD4 inhibitor GSK484 ( Fig. 4 g), indicating that FF-derived G-CSF induces NETs via NOX-dependent ROS signaling. 5) G-CSF-induced NET confers cell-cell adhesion and anchorage-independent growth (AIG) of STIC-mimicked and HGSC cells Fig. 4 G-CSF in FF is responsible for the NET activation of neutrophils through the NOX-ROS-dependent PAD4 pathway. (a) Levels of NETosis-related cytokines in 10 FF aspirates. (b, c) Expression of PAD4 and extracellular DNA of human neutrophils after treatment with 10% FF with or without NAC antioxidant (200 μM), or equivalent level of cytokines as in FF treatment, as analyzed by Western blot and in Sytox Green-NETosis Assay. (d, e) Flow cytometry analysis to compare the change level of intracellular ROS and mitochondrial ROS in human neutrophils by 10% FF or 120 pg/ml G-CSF treatment. (f) Intracellular ROS levels of human neutrophils treated with vehicle medium, 10% FF, G-CSF, or 10% G-CSF-depleted FF (deG FF). NOX inhibitor DPI (60 nM) or GTK137831 (100 nM) was used in FF treatments. (g) Western blot analysis of the protein levels of PAD4 and cit H3 in human neutrophils treated with 10% FF or 10% G-CSF-depleted FF (deG FF). NETosis inhibitor (10 μM GSK 484) was used in FF treatments. *P<0.05, compared to untreated or vehicle medium control; #P<0.05, compared to their regent treatments (FF) without inhibitors. Fig 4 5) G-CSF-induced NET confers cell-cell adhesion and anchorage-independent growth (AIG) of STIC-mimicked and HGSC cells G-CSF in FF is responsible for the NET activation of neutrophils through the NOX-ROS-dependent PAD4 pathway. (a) Levels of NETosis-related cytokines in 10 FF aspirates. (b, c) Expression of PAD4 and extracellular DNA of human neutrophils after treatment with 10% FF with or without NAC antioxidant (200 μM), or equivalent level of cytokines as in FF treatment, as analyzed by Western blot and in Sytox Green-NETosis Assay. (d, e) Flow cytometry analysis to compare the change level of intracellular ROS and mitochondrial ROS in human neutrophils by 10% FF or 120 pg/ml G-CSF treatment. (f) Intracellular ROS levels of human neutrophils treated with vehicle medium, 10% FF, G-CSF, or 10% G-CSF-depleted FF (deG FF). NOX inhibitor DPI (60 nM) or GTK137831 (100 nM) was used in FF treatments. (g) Western blot analysis of the protein levels of PAD4 and cit H3 in human neutrophils treated with 10% FF or 10% G-CSF-depleted FF (deG FF). NETosis inhibitor (10 μM GSK 484) was used in FF treatments. *P<0.05, compared to untreated or vehicle medium control; #P<0.05, compared to their regent treatments (FF) without inhibitors. The epithelium at the fimbriated end of the fallopian tube is constantly exposed to peritoneal fluid (PF)[ 3 ]. We investigated whether the NETosis state in PF influences the malignant phenotypes of transforming FTE cells, specifically the STIC-mimicked FE25 4 and the high-grade serous carcinoma (HGSC) cell OVSAHO [ 21 ]. Following a short-term (1 hour) co-culture with neutrophils primed by different treatments, we assessed the transformation phenotypes using the AIG assay ( Fig. 5 a). Our results demonstrated that while FF alone increased AIG in both cell types [ 4 ], co-culturing with FF-primed (NETotic) human neutrophils further enhanced AIG ( Figs. 5 b, c). Additionally, cell-cell adhesion between FTE cells was significantly increased in the absence of DNase interference ( Figs. 5 d, e). In contrast, co-culturing with naïve neutrophils, neutrophils primed with G-CSF-depleted FF, or the addition of either a PAD4 inhibitor or DNase did not result in these phenotypic changes. Notably, neither naïve nor FF-primed neutrophils formed colonies in the AIG assay (data not shown). We also evaluated the conditioned medium from naïve and FF-primed neutrophils and found no enhancing effect on AIG ( Figs. 5 b, c). These findings suggest that FF-induced NET formation promotes cell-cell adhesion and AIG in STIC and HGSC cells. 6) FF-primed NETotic neutrophils exert paracrine effects to promote the attachment and invasion capacity of STIC-mimicked and HGSC cells Fig. 5 FF-GCSF/NETosis axis promotes anchoring independent growth (AIG) of STIC-mimicked and HGSC cells in a cell-autologous manner. (a) Upper panel: Diagram of the soft agar AIG assay for FE25 or OVSAHO cells after co-culturing with differently primed neutrophils, along with treatments during culture. Lower panel: Treatment groups for cell adhesion and matrigel invasion assays. (b, c) Representative images of AIG colonies (left) and quantification (right) of FE25 cells (B) and OVSAHO cells (c) after co-culture with naïve (Naïve), FF-primed (FF-P) neutrophils ± GSK484 or DNase, or G-CSF-depleted FF-primed (dG FF-P) neutrophils, or under the treatment of conditioned medium from the neutrophils. *P<0.05 or **P<0.01 compared to vehicle medium (Veh). ##P<0.01 or #P<0.05 compared to FF-primed co-culture. (d, e) Fluorescence microscopy of 2 μM calcein AM-stained (green) FE25 cells (d) or OVSAHO cells (e) co-cultured (for 1 hour) with FF-primed neutrophils (stained with DAPI, blue), with or without 1 U DNase I treatment. Scale bar: 20 µm. Fig 5 6) FF-primed NETotic neutrophils exert paracrine effects to promote the attachment and invasion capacity of STIC-mimicked and HGSC cells FF-GCSF/NETosis axis promotes anchoring independent growth (AIG) of STIC-mimicked and HGSC cells in a cell-autologous manner. (a) Upper panel: Diagram of the soft agar AIG assay for FE25 or OVSAHO cells after co-culturing with differently primed neutrophils, along with treatments during culture. Lower panel: Treatment groups for cell adhesion and matrigel invasion assays. (b, c) Representative images of AIG colonies (left) and quantification (right) of FE25 cells (B) and OVSAHO cells (c) after co-culture with naïve (Naïve), FF-primed (FF-P) neutrophils ± GSK484 or DNase, or G-CSF-depleted FF-primed (dG FF-P) neutrophils, or under the treatment of conditioned medium from the neutrophils. *P<0.05 or **P<0.01 compared to vehicle medium (Veh). ##P<0.01 or #P<0.05 compared to FF-primed co-culture. (d, e) Fluorescence microscopy of 2 μM calcein AM-stained (green) FE25 cells (d) or OVSAHO cells (e) co-cultured (for 1 hour) with FF-primed neutrophils (stained with DAPI, blue), with or without 1 U DNase I treatment. Scale bar: 20 µm. In addition to cell-cell adhesion, we assessed the impact of NETs on the ability of cells to attach to a matrix and invade the basement membrane. FE25 and OVSAHO cells showed minimal attachment to a petri dish coated with 0.8% agarose gel, and neither FF nor the conditioned medium from naïve neutrophils facilitated attachment. In contrast, the conditioned medium from FF-primed neutrophils significantly enhanced attachment, increasing it by 578-fold for FE25 cells and 80-fold for OVSAHO cells. When PAD4 inhibitor was used for priming or when a G-CSF-depleted FF was included, the increases were markedly reduced by 81% and 97% for FE25 cells, and by 63% and 59% for OVSAHO cells, respectively ( Figs. 6 a, b). Furthermore, the same conditioned medium from FF-primed neutrophils also enhanced Matrigel invasion capability of FE25 cells by 4.1-fold, with reductions observed when a PAD4 inhibitor or G-CSF-depleted FF was used ( Fig. 6 c). For OVSAHO cells, the increase was 3.0-fold, which could be completely abrogated by the PAD4 inhibitor or G-CSF depletion ( Fig. 6 d). Interestingly, while the conditioned medium from NET-primed neutrophils did not promote migration of FE25 and OVSAHO cells, FF treatment alone was able to enhance migration, as reported previously 4 (Supplementary Fig. 2). These results indicate a paracrine effect of neutrophil NETosis in promoting attachment and invasion phenotypes in STIC-mimicked and HGSC cells. 7) FF-induced NETosis promotes early peritoneal seeding of STIC-mimicked and HGSC cells Fig. 6 Conditioned medium released from FF G-CSF-induced NETosis promotes cell adhesion and Matrigel gel invasion of STIC-mimicked and HGSC cells. Representative pictures and the quantitation data of cell adhesion to 0.8% agarose-coated low attachment dish and of cell invasion through matrigel of STIC-mimicked cell FE25 (a, c) and HGSC cell OVSAHO (b, d) after treatment with control medium (Veh), 10% FF, and conditioned medium of naïve neutrophils (Naive), neutrophils primed with FF (FF-prime) ± PAD4 inhibitor (+GSK), or neutrophils primed with G-CSF-depleted FF (deG FF-prime). ** P<0.01, compared to Veh. # P<0.05, ## P<0.01, compared to FF-prime. Fig 6 7) FF-induced NETosis promotes early peritoneal seeding of STIC-mimicked and HGSC cells Conditioned medium released from FF G-CSF-induced NETosis promotes cell adhesion and Matrigel gel invasion of STIC-mimicked and HGSC cells. Representative pictures and the quantitation data of cell adhesion to 0.8% agarose-coated low attachment dish and of cell invasion through matrigel of STIC-mimicked cell FE25 (a, c) and HGSC cell OVSAHO (b, d) after treatment with control medium (Veh), 10% FF, and conditioned medium of naïve neutrophils (Naive), neutrophils primed with FF (FF-prime) ± PAD4 inhibitor (+GSK), or neutrophils primed with G-CSF-depleted FF (deG FF-prime). ** P<0.01, compared to Veh. # P<0.05, ## P<0.01, compared to FF-prime. The oncogenic consequences of NETosis in enhancing cell-cell adhesion, AIG, cell-matrix attachment, and Matrigel invasion of transforming FTE and HGSC cells were examined in vivo . By intraperitoneally co-injecting luciferase-expressing FE25 cells or OVSAHO cells with FF or FF combined with a PAD4 inhibitor in NSG mice, intraperitoneal growth was assessed two weeks later using IVIS imaging. As illustrated in Fig. 7 , co-injection with FF resulted in increased intraperitoneal signals by 6.4 folds for FE25 ( Fig. 7 a) and 4.5 folds for OVSAHO ( Fig. 7 b), respectively. The addition of the PAD4 inhibitor significantly diminished the enhancing effect of FF. These findings suggest that FF-induced intrabdominal NETosis plays a critical role in the peritoneal seeding and growth of STIC-mimicked and high-grade serous carcinoma (HGSC) cells in vivo . Fig. 7 FF-induced peritoneal NETosis promotes the peritoneal adhesion growth of high-grade serous carcinogenesis. (a) FE25-luc cells or (b) OVSAHO-luc cells (1 × 10 5 ) were injected intraperitoneally (IP) together with FF, with or without the PAD4 inhibitor (GSK, 10 µM). Luciferase activity was measured 2 weeks later using IVIS. *P<0.05, #P<0.05, comparison as shown in the Fig. Fig 7 FF-induced peritoneal NETosis promotes the peritoneal adhesion growth of high-grade serous carcinogenesis. (a) FE25-luc cells or (b) OVSAHO-luc cells (1 × 10 5 ) were injected intraperitoneally (IP) together with FF, with or without the PAD4 inhibitor (GSK, 10 µM). Luciferase activity was measured 2 weeks later using IVIS. *P<0.05, #P<0.05, comparison as shown in the Fig.

Peritoneal fluids (PFs) were collected in regularly menstruating women who underwent laparoscopic surgery due to non-malignant, non-infectious, and non-endometriosis causes as described previously [ 3 ]. Based on the interval of the menstrual cycle and the time since the most recent menstruation, samples were grouped into the follicular phase, luteal phase, and menopause. To avoid blood contamination, fluid in the cul-de-sac was aspirated before the surgical procedures once entering the peritoneal cavity. In women who entered the in vitro fertilization (IVF) program, PF was aspirated during the procedure of transvaginal oocyte retrieval, routinely at 36 h after hCG injection (or LH surge in the natural cycle). PF was aspirated from the cul-de-sac and then follicles were punctured to aspirate the FF with care to exclude blood and medium contamination. PFs were freshly subjected to flow cytometry after lysis of red blood cells. FF aspirates were centrifuged (1200g, 10 minutes) to remove cell debris and frozen at -80°C before use. This study utilized two human cell lines for in vitro analysis, immortalized fallopian tube epithelium (FTE) cells, FE25, and HGSC cells, OVSAHO. FE25 cells were established by transducing FTE cells with hTERT and HPV16 E6/E7, which inactivates p53/Rb[ 12 ]. Late-passage FE25 cells (p90∼100), which spontaneously acquired STIC-like traits including copy number variations and in vivo carcinogenic potential, were used as STIC-mimicked cells for this study[ 5 ]. Cells were maintained in MCDB105 and M199 media (Sigma) supplemented with 10% fetal bovine serum (FBS) and P/S. The human HGSC cells, OVSAHO, were cultured in RPMI-1640 medium with 10% FBS, 100 IU/mL of penicillin, and 100 μg/mL of streptomycin. Human polymorphonuclear neutrophils (PMNs) were isolated from 4 mL of peripheral blood layered over Lymphoprep™, followed by purification and culture in RPMI medium. C57BL/6 mice were purchased from the Taiwan National Laboratory Animal Center. The superovulation protocol of Fowler and Edwards[ 13 ] was adopted for the 7- to 8-week-old female mice. To induce estrus, mice were intraperitoneally injected with pregnant mare’s serum gonadotropin (PMSG, 5 IU) followed by human chorionic gonadotropin (hCG, 5 IU). Mice were euthanized via CO₂ inhalation either 1 h before or at the expected time of ovulation (12 hours post-hCG) for peritoneal lavage collection. PMSG-only treated mice served as non-ovulating controls. In separate experiments, mice received intraperitoneal injections of either 10% FF or G-CSF (600 pg) in 200 μL saline, and lavage fluid was collected at 30 minutes, 1 h, 2 hours, and 24 hours post-injection for analysis. We employed intraperitoneal (IP) xenograft mouse models to investigate the effect of NETosis on HGSC peritoneal growth and dissemination. In this model, NOD-scid IL-2Rγnull (NSG) mice (Jackson Laboratory) were injected intraperitoneally with 1 × 10 5 STIC-mimicked FE25-luc cells or OVSAHO-luc cells (transduced with the luciferase gene) in 200 μL of culture medium containing either vehicle saline, 10% FF, or FF supplemented with or without a NET inhibitor (10 μM, GSK484). Intraperitoneal cell growth was monitored using the In Vivo Imaging System (IVIS) following intraperitoneal injection of luciferin (3.75 mg/mouse) and anesthesia with 4% isoflurane. NETosis was assessed using Sytox Green staining [ 14 ] of extracellular DNA from human neutrophils (1 × 10⁵ cells/well) treated with 10% FF or FF cytokines (120 ng/mL G-CSF, 380 ng/mL GRO, 2 μg/mL MCP-1, 190 ng/mL IL-8) at FF-equivalent concentrations. After 2 hours at 37°C, micrococcal nuclease (500 mIU) was added for 20 minutes, followed by 5 mM EDTA to stop the reaction. Supernatants were collected by centrifugation, stained with 100 nM Sytox Green, and fluorescence was measured (Ex/Em: 485/535 nm). Flow cytometry was employed to analyze white blood cells from peripheral blood and peritoneal fluids or lavages. For mouse cells, we used anti-mouse Ly6G-PE (Cat# 127607) as a neutrophil marker and anti-mouse F4/80-FITC (Cat# 123107) as a macrophage marker. For human cells, we employed anti-human CD66b-APC (ab275586) as a neutrophil marker and anti-human CD11b-FITC (Cat# 101205) as a marker for both neutrophils and macrophages. For cellular ROS detection, 1 × 10⁵ live cells were treated with the test reagent for 30 minutes, followed by incubation with 2 µM DCFDA or MitoSOX for an additional 30 minutes. Cells were then washed, resuspended in PBS, and analyzed by flow cytometry. The protein isolated from cell lysates was mixed with an equal volume of 2 × Laemmli sample buffer and heated at 95°C for 5 min. SDS-PAGE electrophoresis running and then transferred onto a nitrocellulose membrane for detection with a specific antibody. After washing with Tris-buffered saline with 0.05% Tween, the membrane will be treated with appropriate horseradish peroxidase (HRP)–conjugated secondary antibodies and stained with the ECL western blot detection reagent (GE Healthcare, RPN2209). Major specific antibodies in this project are listed as follows: anti-PAD4 (MABE254, Merck Millipore), anti-citrulline-Histone H3 (A18298, Abclonal), and anti-GAPDH (#2118, Cell Signaling) Human peripheral blood neutrophils (1 × 10 7 ) were placed in 1 ml serum-free IMDM medium and primed for ∼1 h for the NETosis activation with either 10% FF with or without GSK484 inhibitor (10 μM), or with 10% G-CSF-depleted FF. Cells were sampled and stained with DAPI to visualize nuclear and NETotic phenotypes under microscopy, followed by Western blot analysis for further confirmation. The remaining cells were pelleted and resuspended in 1 mL fresh serum-free IMDM medium overnight at 37°C, and the conditioned medium was collected for further study. AIG assays were performed to assess transformative activity in FE25 and OVSAHO cells co-cultured with various neutrophils or their conditioned media ( Fig. 5 A). In the conditioned medium group, 3,000 FTE cells were directly seeded into a 0.4% top agarose gel, which was placed over a pre-coated 0.8% gel. In the co-culture group, FTE cells were initially co-cultured with differently primed neutrophils at a 3:10 ratio for 1 h, after which they were seeded into the top gel at the same density of 3,000 FTE cells. The neutrophils were primed for 1 h using either 10% follicular fluid (FF), 10% GCSF-depleted FF, or vehicle control before co-culture. For the two co-culture conditions involving primed neutrophils, the soft agar culture was supplemented with the priming agent (FF or GCSF-depleted FF), with or without the PAD4 inhibitor (10 μM GSK484) or 1 U DNaseI, on days 0 and 1. The culture medium was refreshed every two days to maintain humidity within the gel. After 2 weeks, spheroid colonies were fixed with 4% paraformaldehyde and stained with 0.5% crystal violet. Colonies larger than 50 μm were counted using a microscope. Six-well plates were pre-coated with 0.8% low-attachment agarose gel, and 5 × 10 4 cells of either FE25 or OVSAHO were seeded into the wells using serum-free medium, with or without the addition of 10% FF or 10% human neutrophil-conditioned medium. After an overnight incubation, suspended cells and medium were removed. The wells were then gently washed with 1 ml of PBS, and the cells adhered to the dish were fixed with 4% paraformaldehyde and stained with 0.5% crystal violet. The adhered cells were counted under a microscope. For the cell invasion assay, 24-well transwell plates with inserts coated with 8 µm basement membrane extract (Matrigel) were utilized. Briefly, 5 × 10 4 cells in serum-free cell medium were seeded into the upper chamber, while the lower chamber was filled with normal cell medium. Testing agents, including 10% FF or 10% neutrophil-conditioned medium, were added to the upper chamber. After 24 hours, cells that migrated to the lower chamber were fixed with 4% paraformaldehyde, stained with 0.5% crystal violet, and counted under a microscope. Statistical analysis was conducted using GraphPad Prism (version 5.0c) (GraphPad Software, San Diego, CA, USA), SPSS 19.0, or Excel. Differences between groups were assessed using unpaired Student's t-test, Kruskal-Wallis test, and one-way analysis of variance (ANOVA).

In summary, the ovulatory follicle produces G-CSF, which triggers neutrophil influx and activates ROS/NOX-dependent NET formation in the peritoneal cavity. Excessive NETosis contributes to the early peritoneal dissemination of STIC cells, positioning it as a potential target for prevention and early treatment. Meanwhile, the physiological defensive role of ovulation-induced NETosis warrants further investigation.

Neutrophils are typically scarce in the peritoneal cavity but can be rapidly recruited in response to exogenous insults [ 22 ]. By analyzing the peritoneal cells of women and mice at different phases of the ovulation cycle, we observed a significant increase in peritoneal neutrophils following the LH surge and during ovulation, with levels quickly returning to baseline during the luteal phase. Previous studies on peritoneal cells have often overlooked the ovulation cycle, particularly the ovulation period itself. This study represents the first evidence that neutrophils are recruited to the peritoneum during ovulation in both women and female mice. Notably, these recruited neutrophils are quickly activated for NETosis. We propose that the influx of neutrophils during the preovulatory phase serves an innate immunity function. During the fertile window, the estrogen-dominant hormonal profile temporarily alters the structure of the cervical mucus [ 23 ]. We suggest that a transient surge of neutrophils undergoing NETosis may act as a defense against opportunistic microorganisms during this fertile period. Additionally, previous studies have observed a similar dramatic recruitment of neutrophils in the ovarian follicle, occurring approximately 12 hours after hCG administration [ 24 ]. It is likely that the same G-CSF mechanism is responsible for recruiting neutrophils in both the follicle and peritoneum, thereby providing protection against ascending infections. In this study, we demonstrated that G-CSF from the FF is responsible for the recruitment and activation of neutrophils in the peritoneum. G-CSF is well known for its roles in the production, differentiation, mobilization, and function of neutrophils [ 25 , 26 ]. Traditionally, G-CSF is produced by fibroblasts and endothelial cells in the bone marrow stroma, as well as by monocytes and macrophages. Our finding that G-CSF is abundantly present in FF at levels as high as 123 ± 96 pg/ml suggests that ovarian follicles may serve as an important source of G-CSF during ovulation. This is consistent with observations that G-CSF levels were 2.6 times higher in sera from women at the time of ovulation compared to other phases of the menstrual cycle (28.0 ± 3.1 pg/ml vs. 10.7 ± 0.6 pg/ml) [ 27 ]. Additionally, Noël et al. found that polymorphonuclear leukocytes retrieved from the FF of IVF women and from the peripheral blood of women in the late follicular phase could induce the synthesis and secretion of G-CSF by granulosa cells[ 28 ]. Thus, it can be deduced that under the stimulation of infiltrating leukocytes, G-CSF is produced by granulosa cells and secreted into the follicular space, as well as into circulation through the leaky capillaries surrounding the granulosa layer [ 29 ]. An increasing body of evidence has revealed a protumorigenic role of NETs [ 9 , 30 , 31 ]. In ovarian cancer patients, an elevated neutrophil-to-lymphocyte ratio (NLR) in peripheral blood is associated with poor prognosis[ 32 ]. A retrospective analysis of clinical data and samples from 340 ovarian cancer patients, along with in vitro studies, further demonstrated that cancer cell-derived G-CSF induces NETosis and promotes peritoneal dissemination[ 33 ]. Similarly, in the mouse ID8 ovarian cancer model, cancer cell-secreted G-CSF promoted early neutrophil mobilization and NETosis, thereby facilitating omental metastasis [ 19 ]. In this study, we first discovered that ovulation regularly sources G-CSF, which induces neutrophil influx and NETosis in the peritoneal cavity. Notably, we found that this FF-induced NETosis contributes to the in vivo peritoneal seeding of STIC-mimicked cells and HGSC cells. Various cell phenotypes relevant to metastatic seeding, including cell-cell adhesion, AIG, cell-matrix attachment, and material gel invasion, were all enhanced. Furthermore, STIC-mimicked cells showed a greater dependence on NETosis compared to HGSC cells, suggesting that FF plays a critical role in supporting the intraperitoneal growth of exfoliated STIC cells ( Fig. 7 ). These findings suggest that ovulation-related NETosis may play a significant role in the early dissemination of cancer cells in the peritoneum before the development of overt metastasis. Women with STIC in the fallopian tubes, who undergo risk-reducing salpingo-oophorectomy due to BRCA mutations, face a 10.5% (95% CI, 6.2 to 17.2) risk of developing peritoneal carcinomatosis within five years, and a 27.5% (95% CI, 15.6 to 43.9) risk within ten years [ 34 ]. Notably, these women were free of intraperitoneal tumor involvement at the time of prophylactic surgery. This raises the question of how metastatic growth can occur when the original tumor source has been removed, especially in such a short time frame. Crum introduced the concept of "precursor escape," suggesting that precancerous cells may escape into the peritoneal cavity and evolve silently into malignancy [ 35 , 36 ]. However, this hypothesis lacks an oncogenic trigger to facilitate the transformation. Our study identified ovulation as a potential driving force. We have previously shown that ovulation-released FF contains a full spectrum of carcinogens capable of inducing malignant transformation in FTE cells[ 5 , 37 , 38 ]. The induced NETosis may serve as a critical mechanism in this oncogenic process, supporting the essential step of peritoneal seeding by precursor cells.

Epithelial ovarian cancers (EOC) are mostly diagnosed at advanced stages with peritoneal metastasis, presenting significant challenges for treatment. High-grade serous carcinoma (HGSC) is the most common and aggressive subtype of EOC[ 1 ]. Understanding the mechanisms of the peritoneal metastasis of the precursor STIC is crucial for improving patient prognosis. The fimbriated end of the fallopian tube is widely accepted as the origin of the majority of HGSCs. The fimbriae play a crucial role in capturing the oocyte and are inevitably exposed to carcinogens from ovulatory FF, which can lead to malignant transformation [ [2] , [3] , [4] ]. With continuous ovulation, FF exposure facilitates the progression of carcinogenesis from normal fallopian tubal epithelium (FTE) to the p53 signature, then to STIC, and ultimately to HGSC [ 5 ]. It is estimated that FF-exposed FTE takes approximately 10 years to develop the p53 signature, followed by an additional 14 years to progress to STIC[ 6 ]. Current estimates suggest that STIC arises before menopause and takes another 6–7 years to metastasize to the peritoneum and ovary, ultimately resulting in clinically detectable HGSC [ 6 , 7 ]. The way STIC metastasizes to the peritoneum is unique. It is thought that exfoliated intraepithelial carcinoma cells directly shed into the peritoneal cavity through the opening of the fallopian tube lumen at the fimbriae [ 8 ], allowing a seeding onto the visceral surface without tissue invasion. Nevertheless, the mechanism of disseminating STIC cells to the peritoneum, particularly the anoikis survival and attachment growth to the peritoneum, is largely unclear. There is an increasing body of evidence indicating that NETs play an important role in cancer progression and metastasis [ 9 ]. NETs are formed in dying neutrophils when nucleosomes unravel due to the citrullination of histones by peptidyl arginine deiminase 4 (PAD4) [ 10 ]. This loss of nucleosome integrity leads to the expulsion of DNA along with antimicrobial components, creating a web-like structure outside the cell membrane [ 11 ]. We hypothesized that the NETosis function of neutrophils occurs in the peritoneal cavity during ovulation and can be exploited by exfoliated cancer cells to facilitate their seeding and metastasis within the peritoneum. In this study, we demonstrated in both mouse and human models that high levels of neutrophil influx and NETosis occur in the peritoneal cavity after the LH surge and during ovulation. G-CSF derived from FF is responsible for the recruitment of neutrophils and the formation of NETs in a ROS/NOX-dependent manner. Importantly, this mechanism is exploited during the development of high-grade serous carcinoma (HGSC), facilitating the peritoneal seeding of STIC cells.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.