Targeting colony stimulating factor-1 receptor signalling to treat ectopic pregnancy

1–2% of pregnancies are ectopic, the majority implanting in the Fallopian tube. A single, systemic dose of methotrexate, a DNA-synthesis (S phase) inhibitor, has been used since 1991 for outpatient treatment of women with stable EP. However, methotrexate has limited clinical and cost effectiveness, restricting its use to 25–30% of these women. There is an unmet need for better medical treatment for EP. Colony stimulating factor-1 (CSF-1) promotes placentation and creates a pro-inflammatory environment that is fundamental for the maintenance of a normal pregnancy. We hypothesised that CSF-1 is also involved in the placentation and maintenance of an EP. Herein, we demonstrate the immunolocalisation of the CSF-1 receptor (CSF-1R) as well as its ligand (CSF-1) in immortalised first trimester trophoblast cells. We show that a specific CSF-1R kinase inhibitor, GW2580, abolishes CSF-1 induced trophoblast cell proliferation and migration and can be cytotoxic. We then demonstrate the expression of CSF-1R and CSF-1 in the cytotrophoblast and syncytiotrophoblast within ectopic implantation sites from women with EP. Our data suggests that CSF-1 is involved in the survival and proliferation of trophoblast cells in EP. This suggests that pharmacological disruption of CSF-1/CSF-1R signaling axis could be the basis of a new therapeutic for EP.


CSF-1R and CSF-1 are expressed in immortalised first trimester trophoblast cells. The immor-
talised first trimester trophoblast cell line, SW.71, expresses CSF-1 protein (Fig. 1A) and it is predominantly localised to the cell cytoplasm (Fig. 1B). As well as the CSF-1 the SW.71 cells also express its receptor CSF-1R (Fig. 1C). The expression of CSF-1R protein was mainly localised to the cell membrane of the SW.71 cells (Fig. 1D). These trophoblasts can both make and respond to CSF-1.
CSF-1R antagonism reduces proliferation. As the SW.71 cell line is able to respond to CSF-1 we examined the effects of exogenous CSF-1 on proliferation. CSF-1 caused a dose-dependent increase in cell proliferation (p < 0.0001; Fig. 2A). The maximal effect was seen at a concentration of 100 ng/ml and this concentration was used in subsequent studies. As the SW.71 cell line can also make CSF-1 we examined the effect of the CSF-1 antagonist, GW2580, on proliferation in a dose finding study. After 48 h there was no effect of 5 µM GW2580 but concentrations of 10 µM (p < 0.0001), 20 μM (p < 0.0001) and 40 μM (p < 0.0001) reduced proliferation, when compared to control, in a dose dependent manner (Fig. 2B). The proliferative response to exogenous CSF-1 was blocked by both 20 μM (p < 0.0001) and 40 μM (p < 0.0001) GW2580 (Fig. 2C). We therefore focussed on both  CSF-1R antagonism can promote cell death. In order to determine whether the cell death was associated with the anti-proliferation effects of CSF-1 antagonism we examined cell health using the ApoTox-Glo triplex assay. Interestingly while 20 µM GW2580 reduced SW.71 proliferation and also could block the prolif- www.nature.com/scientificreports/ erative effects of exogenous CSF-1 it was not associated with a reduction in cell viability (Fig. 2D). However, at the 40 µM dose there was a reduction in viability in the absence (p < 0.05) and presence (p < 0.01) of exogenous CSF-1 (Fig. 2D). There was increased cytotoxicity with 40 µM GW2580 (p < 0.05) and this was not overcome by the addition of CSF-1 (p < 0.05; Fig. 2E). CSF-1 promotes cell survival in immortalised first trimester trophoblast cells.

CSF-1R antagonism decreases cellular migration.
We then analysed the effects of GW2580 on cellular migration using a scratch assay. We first optimised the time-point which should be used to study cellular migration, taking measurements of scratch opening (gap between the cell fronts caused by scratch) at 18, 24 and 48 h. We found that differential results could be seen at the 18 h time-point only (Fig. 2F), as after 24 and 48 h of incubation cells tend to fill in the gap of the scratch, even in the presence of inhibitor (data not shown). Exposure to CSF-1 stimulated the migration of the trophoblast cells and scratch width was reduced as compared to vehicle (p < 0.001; Fig. 2G). Exposure to CSF-1 20 µM GW2580 had no effect on cell migration while exposure to 40 µM GW2580 inhibited cell migration, and scratch width was greater, when compared to vehicle control (p < 0.001). However, both 20 µM and 40 µM concentrations inhibited trophoblast migration in the presence of CSF-1, (p < 0.0001; Fig. 2G). CSF-1 increases cell migration in immortalised first trimester trophoblast cells.

CSF1-R and CSF-1 are expressed in trophoblasts within tubal ectopic implantation sites.
We examined the tubal EP implantation site in women (Fig. 3A). IF CSF-1 has effects within the trophoblasts in tubal ectopic pregnancy they would be expected to express the CSF-1 receptor. We were able to localise the CSF-1R to the tubal implantation site (Fig. 3B,C). We demonstrated the expression of CSF-1 and CSF-1R using immunohistochemistry (n = 5). CSF-1 ligand could be seen in both the syncytiotrophoblast (multinuclear tissue layer of the placenta which is adjoining to the maternal circulation) and cytotrophoblast (trophoblasts cells which are found underlying syncytiotrophoblasts that subsequently differentiate into syncytiotrophoblast) at the implantation site ( Fig. 3D-G). We found that CSF-1R is expressed by the syncytiotrophoblasts with strongest staining on the outer membrane of the syncytiotrophoblast facing the maternal circulation ( Fig. 3H-K). CSF-1R was also expressed by cytotrophoblasts ( Fig. 3H-K). The CSF-1 signalling system is present within trophoblasts in tubal ectopic implantation sites in women.

Discussion
Placental tissue of a normal intrauterine pregnancy is known to express high levels of CSF-1 and CSF-1R which are thought to play an important role in embryo implantation and development 19 . We therefore hypothesised that the CSF-1/CSF-1R system could be targeted for the resolution of EP. We first demonstrated that CSF-1 and CSF-1R are expressed in immortalised first trimester trophoblast cells. Then, we showed that GW2580, a specific CSF-1R antagonist, can decrease cell proliferation and migration of trophoblast cells, even following exposure to the CSF-1 ligand, and at pharmacological doses is cytotoxic. Finally, we demonstrated that CSF-1 and CSF-1R are expressed in the cytotrophoblast and syncytiotrophoblast at tubal ectopic implantation sites collected from women with EP. We began with the knowledge that the trophoblast of an intrauterine pregnancy has by far the highest expression of CSF-1 and CSF-1R compared with any non-macrophage-lineage cell types in humans (biogps.gnf.org). The placenta relies heavily on CSF-1 signalling 20 which has been shown to increase the proliferation of placentaderived cells, stimulated the differentiation of cytotrophoblast in-vitro and regulates pre-implantation blastocyst cells division 21,22 . In addition, CSF-1 signalling has been shown to stimulate extravillous trophoblast cell proliferation 12 . The full role of CSF-1 during abnormal embryo implantation and in pregnancy complications is still being investigated. Low CSF-1 serum levels are associated with recurrent miscarriages 23 and CSF-1 deficient mice demonstrated decreased embryo development 24 . Our data confirmed that CSF-1 promotes proliferation and migration of SW.71 trophoblast cells. These data are in agreement with previous findings suggesting the critical role of CSF-1 in embryo growth. Furthermore, specific inhibitor of CSF-1R, GW2580 negated the stimulatory effects of CSF-1 and could lead to cell death. We are aware that redundance of actions have been reported in the CSF family but we did not perform similar dynamic tests with other CSFs, as negative controls, to confirm the potential specificity of CSF1 action on our model. Nevertheless, our data collectively suggests that inhibiting CSF-1/CSF-1R signalling negatively affects placental cell growth, migration and survival and thus could be a potential approach to treat EP.
GW2580 selectively inhibits the tyrosine kinase domain of CSF-1R 25 and is shown to be effective in targeting tumour invasion when used in isolation, and potentiates its efficacy when used in combination with indoximod therapy 26 . Recently, engineered bi-specific inhibitors have been developed that simultaneously target the unique combination of CSF-1 and integrins 27 , these dual-specific proteins could bind to and inhibit both CSF-1 and αvβ3 integrin and have shown superior therapeutic potential, as compared to monospecific protein therapeutics. Our recent published work has pointed a role of endometrial receptivity marker and cell adhesion protein, integrin β1 (ITGB1), in the embryo attachment within FT, as a result of past chlamydial infection 28 . We can speculate that CSF-1 and ITGB1 may function synergistically in promoting embryo implantation and growth within the FT. Taken together, our current and previous studies 29, 30 provide a rationale to develop strategies of using a combination therapy approach to treat EP. Since we have demonstrated a role of CSF-1 and ITGB1 and in EP, further investigations are warranted to develop such bi-specific inhibitors that could pave the way to discover more efficient and specific drugs to treat EP. We also believe that it would be of interest to investigate the differences in action of CSF1 on ectopic compared to eutopic pregnancies.
In conclusion, we have demonstrated that CSF-1/CSF-1R system is present at the ectopic embryo implantation site and that CSF-1 promotes trophoblast proliferation, migration and survival in-vitro. We suggest that targeting Scientific Reports | (2020) 10:15638 | https://doi.org/10.1038/s41598-020-72785-y www.nature.com/scientificreports/ the CSF-1 system using receptor antagonists may have a novel therapeutic role in the medical management of EP. Targeted drug delivery systems, based on nanoparticles, may be additionally exploited as a new strategy for better therapeutic efficiency, along with reduction of side effects. Further in-vivo investigations are therefore warranted to decipher the role of CSF-1/CSF-1R in development of tubal EP and to determine the efficacy of CSF-1 antagonism for the treatment of EP.

Methods
All experiments were performed in accordance with relevant guidelines and regulations.
Cell culture. Swan Immunocytochemistry. Immunocytochemistry (ICC) was performed to demonstrate that SW.71 cells express the CSF-1 and CSF-1R. SW.71 cells were cultured on positively charged coverslips in 6-well plates until ~ 70% confluent. They were fixed using 4% paraformaldehyde for 10 min followed by methanol on ice for Tissue resources. This study was approved by the Scotland A Research Ethics Committee (LREC 04/ S1103/20) and written informed consent was taken from all study participants. FT ectopic implantation sites (n = 5) were collected from women aged 18-45 years undergoing laparoscopic salpingectomy for the treatment of tubal EP. Biopsies were transported on ice to the laboratory, fixed in 10% neutral buffered formalin overnight at 4 °C, stored in 70% ethanol, and then wax embedded for further immunohistochemical analysis.

Immunohistochemistry.
To determine the expression of CSF-1 and CSF-1R in the trophoblast invading the FT in an EP, immunohistochemistry (IHC) was performed, as previously described 34 . Briefly, 5 μm sections of ectopic implantation site were deparaffinised in xylene, rehydrated in ethanol and processed for antigen retrieval using citrate buffer in a pressure cooker for 20 min. Sections were quenched for endogeneous peroxidase in 3% H 2 O 2 and non-specific binding sites were blocked using horse serum. Sections were incubated with primary antibodies (1 in 500) against CSF-1 (anti-CSF1, Santa Cruz Biotechnology) and CSF-1R (anti-GMCSFRα, Santa Cruz Biotechnology) overnight at 4 °C. For control sections, isotype control immunogloblins were used at the same concentration. Positive immunostaining was revealed by application of goat anti-mouse peroxidase (Dako, Ely, UK) secondary antibody followed by, signal amplification using tyramide signal amplification kit (PerkinElmer) and mounted with VECTASHIELD with Dapi (Vector Laboratories). Samples were analysed with a confocal microscope using ZEN 2009 software (https:// www. softp edia. com/ get/ Multi media/ Graph ic/ Graph ic-Viewe rs/ ZEN-2009-Light-Editi on. shtml).
Statistical analysis. IHC and ICC data was analysed descriptively. Results from in-vitro assays (MTS, automated countess, ApoTox-Glo and scratch assay) are presented as the mean ± SEM and are analysed using oneway ANOVA followed by Dunn's multiple comparisons test. All statistical analysis was performed with the use of GraphPad Prism 6.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.