The expression and biological function of chemokine CXCL12 and receptor CXCR4/CXCR7 in placenta accreta spectrum disorders

Abstract Objectives Investigation of mechanism related to excessive invasion of trophoblast cells in placenta accreta spectrum disorders (PAS) provides more strategies and ideas for clinical diagnosis and treatment. Materials and Methods Blood and placental samples were collected from included patients. The distribution and expression of CXCL12, CXCR4 and CXCR7 proteins in the paraffin of placental tissue in the included cases were analysed, and we analyse the downstream pathways or key proteins involved in cell invasion. Results Firstly, our results determined that CXCL12 and CXCR4/CXCR7 were increased in extravillous trophoblastic cell (CXCL12: P < .001; CXCR4: P < .001; CXCR7: P < .001), and the expression levels were closely related to the invasion depth of trophoblastic cells. Secondly, CXCL12 has the potential to become a biochemical indicator of PAS since the high expression of placental trophoblast CXCL12 may be an important source of blood CXCL12. Using lentivirus‐mediated RNA interference and overexpression assay, it was found that both chemokine CXCL12 and receptor CXCR4/CXCR7 are associated with regulation of trophoblast cell proliferation, migration and invasion. Further results proved that through the activating the phosphorylation and increasing the expression of MLC and AKT proteins in the Rho/rock, PI3K/AKT signalling pathway, CXCL12, CXCR4 and CXCR7 could up‐regulate the expression of RhoA, Rac1 and Cdc42 proteins to promote the migration and invasion of extravillous trophoblastic cell and ultimately formate the placenta accrete compare to the normal placenta. Conclusions Our research proved that trophoblasts may contribute to a PAS‐associated increase in CXCL12 levels in maternal blood. CXCL12 is not only associated with biological roles of PAS, but may also be potential for prediction of PAS.


| INTRODUC TI ON
Placenta accreta spectrum disorders (PAS) is characterized by excessive invasion of the chorionic villi in the myometrium, resulting in severe haemorrhage during and after delivery. PAS is one of the main causes of perinatal emergency hysterectomy and maternal death. At present, studies on the pathogenesis of PAS mainly focus on the myometrial scar, 1 the loss or abnormal function of decidua 2 and the abnormal angiogenesis at the invasive site 3,4 but these could only partially explain PAS. The specific molecular mechanism underlying over-invasion phenomenon of PAS remains unclear.
Chemokine is a kind of small molecule protein with chemotaxis among members of cytokine superfamily, the molecular weight of which is approximately 8-10 kD, and it promotes the cell migration. 5 At present, more than 50 kinds of low molecular weight chemokines have been found. According to the number and spatial sequence of the N-terminal semi-desinine residues, chemokines can be categorized for four families: CXC, CC, CX3C and C family. 6 The expression of chemokine proteins is selectively induced on the surface of target cells through binding to chemokine receptors. 7 Basically, a chemokine may bind to multiple receptors, and a chemokine receptor can also recognize multiple chemokines, which constitutes a complex network and widely participates in cellular immunity, growth and development, inflammation and other physiological functions. 8 The consequence of CXCL12 with its receptor CXCR4 or CXCR7 has become one of the hotspots in the study of the development of various malignant tumours. 9 One of the important roles of CXCL12-CXCR4/CXCR7 is to regulate the adhesion, metastasis, colonization, angiogenesis and proliferation of tumour cells to endothelial cells. 10 Given the biological behaviour of trophoblasts similar to cancer cells, it has been indicated that CXCL12 and its receptors also mediated trophoblasts differentiation, invasion, and proliferation. In terms of the maternal-foetal interface of early pregnancy, CXCL12 and its receptors have attracted much attention to regulate the balance between trophoblast cells and decidual cells. 11 CXCL12-mediated regulation is associated with trophoblast cells proliferation, differentiation, invasion and uterine spiral arterial remodelling in early pregnancy. 12 Abnormal regulation of CXCL12 and its receptors gives rise to chemical placental diseases (IPD), such as abortion, foetal growth and development restriction, and preeclampsia. 13 The function of CXCL12 with its receptors varies in different stages of pregnancy and different diseases, and the functional role of CXCL12-CXCR4/ CXCR7 in PAS remains to be determined. Our study therefore preformed clinical research on PAS of trophoblast cells excessive invasion and function analysis of EVT cells in vitro, seeking to provide theoretical basis and new direction for the diagnosis and treatment of PAS.

| Patients
The study was approved by the ethics committee of the first af-

| Collection and preservation of clinical blood samples and placental samples
Before pregnant women received hormone therapy or blood transfusion, 4-6 mL venous EDTA-blood was collected and centrifuged (110 xg) at 4°C for 15 min. According to the instruction of CXCL12 kit, the blood was centrifuged (11 000 xg) for another 10 min at 2-8°C to completely clear away the platelets in the blood plasma.
The supernatant was collected and stored at −80°C for storage.
Placental samples, including uterus, partial excision or excised implanted lesion, were collected and fixed with 10% formalin immediately, for further dehydration by gradient ethanol, paraffin embedding.

| RNA extraction and Quantitative real-time PCR (qRT-PCR)
Total RNA was extracted by using TRIzol-Reagent (Invitrogen). cDNA was transcribed from mRNA by High-Capacity cDNA Reverse Transcription kits (Applied Biosystems, Foster City, CA). qPCR was conducted by using ABI PRISM 7700 System and FastStart Universal SYBR Green Master kit (Roche). β-actin was used as an internal reference gene. All primers used in this study was purchased from Sangon Biotech (Shanghai, China). The relative expression level was calculated using the 2 -ΔΔCt method.
Primers were listed in Table 1. Experiments were carried out in triplicates.

| Immunohistochemical staining of placental tissue
The sections were incubated at 65°C about 2 hour, and dewaxed in xylene I, II for 10 minutes. After being treated with gradient concentrations of 100%, 95%, 85% and 70% ethanol for 3 minutes, the sections were immersed in boiled 0.01M citrate buffer (pH = 6.0) for 2 minutes,

| The expressions of plasma CXCL12, CXCR4 and CXCR7 via enzyme-linked immunosorbent assay (ELISA)
The standard solution was prepared to make the standard curve, and the standard curve was drawn using CurveExpert 1.4 software for sample quantification according to the introduction of the kit (Invitrogen, Carlsbad, CA, USA). In brief, the 96-well plate coated with the corresponding antibodies was taken, added with the standard solution or diluted plasma samples, and affixed with the sealing membrane, followed by incubation at 37°C for 90 minutes. After the liquid in the plate was discarded and dried, the biotin-labelled antibody [anti-CXCL12, CXCR4 or CXCR7 antibody] was added, and the plate was sealed with the sealing membrane, followed by incubation at 37°C for 60 minutes. After the plate was washed with washing solution for 4 times (3 minutes per time), ABC working solution was added, and the plate was sealed, followed by incubation at 37°C for 30 minutes. After the plate was washed again with washing solution for 4 times, tetramethyl benzidine developing solution was added, and the plate was sealed, followed by incubation at 37°C for 20 minutes at dark. Then, TMB stop buffer was added and mixed evenly. The absorbance value at 450 nm was detected using a microplate reader and substituted into the standard curve to calculate the concentrations of plasma CXCL12, CXCR4 and CXCR7 in each group.

| Extraction of total protein from cells and Western blot
Cells in logarithmic growth period (about 80% to 90%) were taken, supernatant was discarded, and sediment was resuspended with precooled PBS for washing. RIPA cell lysate with 1 mol/L PMSF was added. About 200 μL lysate (including protease and phosphatase inhibitor) was added into each well of the 6-well plate at a ratio of  The results were imaged by the gel imaging analysis system. Image J software was used to analyse the grey value of the target bands.

| Colony formation assay
Colony formation assay was used to evaluate the cell formation ability. 2 × 10 3 cells were plated into each well of a 6-well plate and incubated for 10 to 14 days. Then, the plate was gently washed and stained with crystal violet. The cell colonies were observed and counted under an inverted microscope.

| Cell proliferation experiment
The cells were divided into five groups: Blank control group (HTR-8/

| Wound healing assay
For scratch wound healing assay, cells at concentration of 2 × 10 5 cells/mL were cultured in serum-free medium for 24 hours, at 37℃ with 5% CO 2 and wounded with pipette tips. Fresh medium was replaced. The wound closing procedure was observed for 48 hours.
Experiments were repeated three times. The mean distance between cells was calculated with Image J software.

| Transwell assay
The cells of each group were cultured to 60%-80% confluency and were trypsinized.

| Cell apoptosis assay
The cells of each group were cultured to 60%-80% confluency and were digested with 0.25% trypsin. About 500 μL binding Buffer was added to suspend cells, followed by 1μL Annexin V-PE for mixing.
The Aimexin V-PE was detected by flow cytometry (FACScan®; BD Biosciences), at the wavelength of 488 nm and emission wavelength of 578 nm. Experiments were repeated three times.

| Patient characteristics
The comparison between PAS and normal control groups (each group n = 33) was performed. There were no significant differences in age (P = .957), gestational age (P = .160), birthweight (P = .626), gravidity (P = .211), parity (P = .769), and the number of prior caesarean sections (P = .378), but the proportion of vaginal bleeding (P = .021) and caesarean hysterectomy (P = .053) in the case group were higher than that in the control (Table 2).

| The expression of CXCL12, CXCR4 and CXCR7 in PAS and normal pregnant placental tissues
Immunohistochemical staining was used to compare the expressions of CXCL12, CXCR4/CXCR7 in the internal or external trophoblast cells and placenta implant sites that were determined during surgery and histopathological section ( Figure 1, Table 3). Yellow colour appeared in membrane or cytoplasm was considered as positive cells.
Moreover, Western blot analysis also confirmed the up-regulation of CXCL12, CXCR4 and CXCR7 in PAS. It was further subdivided into the types of placental accreta, placental increta and placental percreta, and immunohistochemical staining was utilized to detect the expressional differences of CXCL12 and CXCR4/CXCR7 proteins in trophoblast cells of all types (Table 4). Results revealed that there were significant differences in CXCL12, CXCR4 and CXCR7 among the three subgroups of percreta, increta and accreta. The pairwise comparison results showed that the expressions of CXCR4 and CXCR7 of pregnant women in the percreta group were higher than those in the increta group (CXCR4: P = .047; CXCR7: P = .023), and CXCL12, CXCR4 and CXCR7 scores of pregnant women in the percreta group were significantly higher than those in the accreta group (CXCL12: P = .007; CXCR4: P = .001; CXCR7: P = .005).

| The expression of plasma CXCL12, CXCR4 and CXCR7 in PAS and normal pregnant women
The plasma concentrations of CXCL12, CXCR4 and CXCR7 were in accordance with normal distribution (Figure 2A-B). Analysis result revealed that the plasma CXCL12 level in the PAS group was significantly increased (3.144 ± 0.701 ng/mL) than that in normal control group (2.207 ± 0.442 ng/mL) (P < .001) ( Figure 2B, Table 5). We further divided the PAS into the placenta accreta, placenta increta and placenta percreta groups. We found that the level of plasma CXCL12 placenta percreta group (3.476 ± 0.500 ng/mL) was increased compared with that in the placenta increta group (3.047 ± 0.643 ng/mL) and placenta accreta group (2.927 ± 0.899 ng/mL), but there was no

TA B L E 2
Clinical characteristics of the study population significant difference (P = .188) ( Figure 2B, Table 6

| Establishment of human trophoblastic HTR-8/SVneo cell line with inhibition or overexpression of CXCL12, CXCR4 and CXCR7 proteins
The transfected HTR-8/SVneo cells were observed under inverted microscope, the CXCL12 overexpression or interference plasmid was inserted green fluorescence reporter, and CXCR4 and CXCR7 overexpression or interference plasmid was added red fluorescence reporter ( Figure S3). At the same time, HTR-8/SVneo cells transfected with blank or scramble plasmid were used as negative controls ( Figure S3).
To detect the efficiency of transfection, RT-qPCR was used to verify the expression of CXCL12, CXCR4 and CXCR7 mRNA in each cell line after interference or overexpression. Results showed that the levels of CXCL12, CXCR4 and CXCR7 were significantly decreased in cells treated with shCXCL12, shCXCR4 and shCXCR7, respectively (P < .05) ( Figure 3A). And the expressions of CXCL12, CXCR4 and CXCR7 in were significantly increased after overexpression with OE-CXCL12, OE-CXCR4 and OE-CXCR7, respectively (P < .05) ( Figure 3B).

| CXCL12, CXCR4 and CXCR7 genes promote cell proliferation of HTR-8/SVneo
To explore the function of CXCL12, CXCR4 and CXCR7 in cell pro-

| CXCL12, CXCR4 and CXCR7 genes promote cell migration and invasion of HTR-8/SVneo
We evaluated the function of CXCL12, CXCR4 and CXCR7 in cell migration and invasion and carried out cell scratch assay and transwell assay. The cell scratch assay suggested that the cell migration distance of HTR-8/SVneo cells was significantly reduced in silencing   group (P < .05) ( Figure 5A). Moreover, the cell migration ability of HTR-8/SVneo cells was enhanced after overexpression of CXCL12, CXCR4 and CXCR7 genes (P < .05) ( Figure 5B). Similarly, the transwell assay showed that invasion ability of HTR-8/SVneo cells was significantly reduced in silenced group (P < .05) ( Figure 5C), but significantly increased in overexpression group (P < .05) ( Figure 5D).

TA B L E 4 Staining intensity scores by depth of invasion
The above data suggested that CXCL12, CXCR4 and CXCR7 genes promote cell migration and invasion of HTR-8/SVneo.

| CXCL12, CXCR4 and CXCR7 have no obvious effect on cell apoptosis of HTR-8/SVneo
Flow cytometry showed no significant difference of percentage of apoptosis between silencing groups (P > .05, Figure 6A, C) or overexpression groups (P > .05, Figure 6B, D), which demonstrated that CXCL12, CXCR4 and CXCR7 genes have no obvious effect on cell apoptosis of HTR-8/SVneo.

| CXCL12, CXCR4 and CXCR7 regulate cell invasion via in Rho/rock, PI3K/AKT signalling pathways
To further investigate the functional mechanism of CXCL12, CXCR4 and CXCR7, we analysed the downstream pathways or key proteins Western blot results showed that the protein expression of RhoA, Rac1 and Cdc42 was significantly inhibited by transfection of shCXCL12, shCXCR4 and shCXCR7 (P < .05, Figure 7A). Meanwhile, the protein expression of Rock, AKT, p-AKT, MLC and p-MLC were also was significantly decreased (P < .05, Figure 7C). However, the protein expressions of RhoA, Rac1, Cdc42, Rock, AKT, p-AKT, MLC and p-MLC were all significantly increased in OE-CXCL12, OE-CXCR4 and OE-CXCR7 groups (P < .05, Figure 7B, E). No significant change of ratios of p-AKT/AKT and p-MLC/MLC was found among different groups (P > .05, Figure 7D, F).
To further validate the potential mechanism of CXCL12, CXCR4 and CXCR7 on cell invasion though Rho/rock, PI3K/AKT signalling pathway, we used Y-27632 (Rock inhibitor) and NVP-BEZ235 (PI3K inhibitor) to block signalling pathways, respectively and detected the cell invasive ability by transwell invasion experiments. Results suggested that after Rock or PI3K was inactivated, the cell invasive ability of wild-type and CXCL12, CXCR4 and CXCR7 overexpressing groups was significantly impaired (P < .05, Figure 8A, B) Taken together, these results corporately indicated that CXCL12, CXCR4 and CXCR7 regulated the HTR-8/SVNEO cell invasion through Rho/rock, PI3K/AKT signalling pathway.

| D ISCUSS I ON
The chemokine CXCL12 and the receptor CXCR4/CXCR7 are highly expressed in trophoblast cells due to PAS, and the expression intensity of CXCL12, CXCR4 and CXCR7 proteins is closely related to the invasion depth of trophoblast cells, which indicates promote the secretion of CXCL12, and subsequently stimulate the migration and invasion of trophoblast cells. 14 Therefore, some researchers proposed that PAS was prone to occur when the placenta was attached to the lower part or the scar of the uterus, possibly due to local hypoxic concentration. 15,16 In this study, we found that CXCL12-CXCR4/CXCR7 were closely associated with the excessive invasion of PAS trophoblast cells.
Recently, many studies have been actively identifying specific biochemical markers of placenta accreta. The early diagnosis of PAS in pregnant women can minimize the complications risk. It has proved that blood β-HCG mRNA concentration was higher in PAS group than that placenta previa group, 17 and several studies suggested that the human placental lactogen (hPL) mRNA was completely derived from the placenta trophoblasts and can predict intrapartal bleeding. 18,19 But these studies did not further analyse the exact invasion of depth. 20 We found that the pathological diagnosis of PAS maternal blood CXCL12 levels was significantly increased based on the data from gestational age paired case-control study, and blood CXCL12 levels was positively correlated with placental trophoblast CXCL12 protein expression. Therefore, it demonstrated that the high expression of placental trophoblastic chemokine CXCL12 might be originated from blood CXCL12.
This suggests that blood CXCL12 has the potential to become a predictive biochemical markers of PAS, and our data indicate the possibility of blood CXCL12 for distinguishing invasive depth of trophoblast cells.
The CXCl12-CXCR4/CXCR7 axis plays an important role in many angiogenesis related diseases, and it has been reported that changes in chemokines or their receptors can be detected in circulating blood. Studies have also reported elevated CXCL12 levels in the placenta and circulating blood of women with preeclampsia. 21 Kim et al dynamically monitored the serum CXCL12 levels of 100 patients undergoing cardiopulmonary bypass before surgery, during myocardial ischaemia, reperfusion and after surgery, and found that CXCL12 level was significantly increased during myocardial ischaemia, but had no significant effect during cardiac reperfusion, which means that CXCL12 serum levels are inversely correlated with organ dysfunction. 22 In addition, Nadimi et al revealed that serum levels of CXCL12 and CXCL10 were significantly correlated with the severity of coronary artery occlusion. 23 Matsuoka et al collected peripheral blood of 206 patients before angiography and detected CXCL12 level in plasma, and present that CXCL12 level in patients with ischaemic stroke was higher than that in patients without stroke. 24 In this study, we found that  16 And the trophoblast migration and invasion ability changed accordingly to different CXCL12 concentrations. 28 Above results suggested that CXCLl2 can promote the migration of trophoblasts to the aponeurosis and blood vessels, the invasion of placenta, uterine-placental vascular remodelling, immune tolerance and complete a series of pregnancy physiological processes, [29][30][31] In this study, we found  Figure 9). In this regard, the chemokine CXCL12 and its receptors CXCR4/CXCR7 become potential therapeutic targets.
The block of the interaction of CXCL12 and its receptors CXCR4/ CXCR7, or the deactivation of the downstream Rho/Rock and PI3K/ AKT signalling pathway, might be a new approach to approach to the treatment of PAS.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.