Differences in lymphocyte subsets of spiral artery associated with impaired vascular remodeling in hypertensive disorders of pregnancy

Abstract

Hypertensive disorders of pregnancy (HDP) is assumed to be triggered by incomplete remodeling of the decidual spiral arteries. We examined the lymphocyte subsets and identified immune differences in the spiral arteries between HDP and normal pregnancies, and between remodeled and non-remodeled arteries such as the muscular artery and acute atherosis. One hundred seventy-three patients with HDP diagnosed at our hospital between 2008 and 2017 were included. Patients with a normal course of delivery and patients with premature birth after 34 weeks without pathological changes and abnormalities in the pregnancy course were included in the control group. The infiltrating lymphocytes were evaluated using immunohistochemistry. Acute atherosis was observed in 56 patients (32.4%) of all HDP patients. In patients that presented HDP with acute atherosis, the density of CD8⁺ T cells and CD4⁺ T cells was higher in the acute atherosis than those in the non-remodeled muscular artery. CD4⁺ T cells were more abundant in the non-remodeled muscular artery compared to the remodeled artery. In control patients, the density of CD4⁺ T cells was higher in the non-remodeled muscular artery than that in the remodeled artery; there was no difference in the density of CD56⁺ natural killer cells. There was no difference in the density of CD3⁺ T cells and CD56⁺ natural killer cells in the remodeled artery between patients presenting HDP with acute atherosis and control patients. These immune differences may cause changes in local cytokine balance around the spiral artery, contributing to the development of HDP.

Introduction

Hypertensive disorders of pregnancy (HDP) occur in approximately 10% of all pregnancies and is a major cause of maternal and perinatal mortality and morbidity [1]. HDP is a multisystem pregnancy disorder characterized by a variable degree of placental malperfusion, with the release of soluble factors that cause vascular endothelial injury into the circulation [2]. These soluble anti-angiogenic factors such as soluble fms-like tyrosine kinase-1 (sFlt-1) and soluble endoglin, secreted in excess by stressed syncytiotrophoblast induce vascular endothelial injury leading to hypertension and multiorgan dysfunction [3,4,5]. Poor vascular remodeling decreases maternal blood supply, causing hypoxic and ischemic conditions in placental tissue and increased production of soluble factors by syncytiotrophoblasts [6]. The only known cure for HDP is pregnancy termination.

In normal early pregnancy, decidual spiral arteries are transformed from muscular arteries to distended calibers, replacing invading extravillous trophoblasts (EVTs) [7, 8]. EVT originates at the tip of the anchoring villi, and EVT has migrated beyond the confines of the villous tree [9]. Spiral artery remodeling ensures the delivery of a high volume of maternal blood to the placental intervillous space, which plays an important role in human placentation. During this process, maternal-fetal immune adaptations occur to avoid an immune attack against the semi-allogenic fetus [10, 11]. Maternal immune alterations are critical for successful pregnancy. Decidual natural killer (dNK) cells regulate EVT invasion and spiral artery remodeling [12, 13]. Furthermore, T cell differentiation is biased towards immune-tolerant cell types such as T helper 2 (Th2) and regulatory T cells (Treg) during pregnancy [14,15,16].

The mechanism underlying impaired spiral artery remodeling remains unclear. However, some maternal-fetal immunogenic maladaptations may impair EVT invasion into the maternal spiral arteries, resulting in poor vascular remodeling and a predisposition to development of HDP [6]. The balance is shifted toward T helper 1 (Th1) cells in preeclampsia compared with that in normal pregnancy [17]. Moreover, systemic inflammation present in preeclampsia may be associated with an imbalance of the T-helper type 17 (th17)/Treg axis, and patients with preeclampsia exhibit a decreased number of Treg cells, whereas Th17 cells are increased [18, 19]. These findings suggest differences in the systemic immune state between HDP and normal pregnancies. In contrast, whether the local immune response around the spiral arteries is related to vascular remodeling remains unclear. Clarifying which lymphocyte subsets are present around spiral arteries during successful as well as unsuccessful vascular remodeling will lead to an understanding of the microenvironment around these arteries. A histological technique that allows morphological identification is useful for evaluating lymphocyte subsets around the spiral arteries. However, histological manifestations of decidual arteriopathy and ischemic changes in the placenta have not been observed in all HDP cases [20, 21]. Additionally, mathematical modeling has revealed that spiral artery transformation has only a relatively minor effect in maternal if blood flow throughout the placenta, and if blood flow is likely to be significantly impaired only when there is a secondary atherotic change in the spiral arteries [22, 23].

In the current study, we aimed to investigate the lymphocyte subsets surrounding the spiral arteries and compared them between remodeled and non-remodeled arteries, including acute atherosis and muscular arteries. HDP cases with acute atherosis, which implies secondary atherotic changes, were defined as those that developed due to impaired spiral artery remodeling. We extracted acute atherosis from all HDP cases and then analyzed it.

Materials and methods

Cases

This study was approved by the medical ethics committee at the Shinshu University School of Medicine, Japan (No. 4488). Clinical information was collected from 173 patients diagnosed with HDP at the Department of Obstetrics of Shinshu University Hospital, between 2008 and 2017. The inclusion criteria for pregnant women with HDP were as follows: (1) singleton gestation, (2) absence of systemic lupus erythematosus or antiphospholipid antibody syndrome, and (3) undergoing regular medical examinations. Fifteen healthy controls were recruited among consecutive pregnant women who delivered at the same hospital and during the same period. The control patients meet one of the following criteria: (1) women with normal pregnancy and delivery, (2) patients who underwent caesarean section due to a previous surgery, or (3) patients with premature delivery after 34 weeks, no pathological abnormalities, and almost normal course during pregnancy. These patients exclude those with premature rupture of membranes or inflammatory diseases such as chorioamnionitis.

HDP was defined as the presence of hypertension during pregnancy. Hypertension was diagnosed as systolic pressure ≥ 140 mmHg or diastolic pressure ≥ 90 mmHg, measured at two time points with at least a 4-hour interval. Severe hypertension was defined as systolic pressure ≥ 160 mmHg or diastolic pressure ≥ 110 mmHg. Proteinuria was diagnosed as at least 300 mg of urine protein in a 24-h urine sample. Severe proteinuria was defined as 2 g or above of urine protein in a 24-h urine sample. Early-onset HDP was defined as hypertension developing before 34 weeks of gestation, and late-onset HDP as hypertension developing after 34 weeks gestation [23, 24]. The participants were divided into two groups according to fetal weight using birth weight reference values by gestational age, and small for gestational age (SGA) was defined as less than the 10th percentile [25]. Placental weights were divided based on the 10th percentile of the singleton placental weights standard [26].

Pathologic evaluation

Histopathological slides of placental tissues from patients with HDP and control patients were retrieved from the pathology files at the Department of Laboratory Medicine of Shinshu University Hospital. All the tissue samples were fixed in 10% neutral-buffered formalin and embedded in paraffin. Tissue sections (3-µm-thick) were prepared for hematoxylin and eosin (HE) staining, and immunohistochemical studies.

We classified the spiral arteries into the following three groups: (1) acute atherosis (AA), (2) non-remodeled muscular artery (non-RA), and (3) remodeled artery (RA; Fig. 1a-c). The AA group was characterized by preserved endothelial cells, accumulation of foamy macrophages, and mural fibrin deposition. The Non-RA group displayed intact vascular muscle cells layers and endothelial cells. The RA group was characterized by a distended caliber and exhibited a complete loss of vascular muscle cells. The vascular wall was replaced with an often-continuous layer of fibrinoid and frequent EVTs. To classify the spiral arteries described above, we identified a spirally coiled muscular vessel in the decidua using HE staining, which was conserved throughout the serial sections for immunohistochemical analyses. Additionally, we examined the frequency of distal villous hypoplasia (DVH) and increased syncytial knots (Fig. 1d, e). These findings were defined according to the criteria of the Amsterdam Placental Workshop Group Consensus Statement 2016 [27].

Fig. 1

Pathological evaluation of placenta stained with hematoxylin and eosin. (a) Acute atherosis. Preserved endothelial cells, accumulation of foamy macrophages, and mural fibrin deposition in spiral artery. (b) Non remodeled muscular artery. Intact vascular muscle cells layers and endothelial cells are observed. (c) Remodeled artery. Spiral artery was distended caliber and exhibited a complete loss of vascular muscle and replaced with a layer of fibrinoid. (d) Distal villous hypoplasia. Small or thin and elongated terminal villi and increased intervillous space. (e) Increased syncytial knots. Syncytial knots are recognized as cluster of syncytiotrophoblastic nuclei. Syncytial knots on greater than 33% of villi are defined as increased

Immunohistochemistry

The sections were incubated in 3% H₂O₂ for 10 min at room temperature to quench endogenous peroxidase activity. Antigen retrieval was performed by heating the sections in 10 mM ethylenediamine tetraacetic acid (EDTA) buffer (pH 8.0) in a microwave oven at 600 W for 30 min. Lymphocytes infiltrating the spiral arteries were immunophenotyped by immunohistochemistry using the following mouse monoclonal antibodies: anti-CD3 (clone LN10; Novocastra, Newcastle UK), anti-CD8 (clone 1A5; Novocastra), and anti-CD56 (clone CD564; Monosan, Uden, Netherlands). To detect the endovascular lining cells (the endothelium and trophoblast), immunohistochemistry was performed using mouse monoclonal antibodies against CD31 (clone JC/70A; Dako, Glostrup, Denmark), and cytokeratin 7 (clone OV-TL; Dako). CD31 staining was used to identify vascular endothelial cells, and CK7 staining was used to identify trophoblasts. Incubation with primary antibodies was performed overnight at 4°C, and subsequent signal development was performed using the immunoenzyme polymer method (Histofine Simple Stain MAX PO Multi; Nichirei, Tokyo, Japan) with 3,3’-diaminobenzidine as the chromogen (Fig. 2).

Fig. 2

Histopathology and immunostaining of spiral arteries. Hematoxylin and eosin staining, immunostaining of CD3, CD8, CD56. Many CD3+, CD8 + T cells infiltrated the vascular wall in acute atherosis. The number of CD3+, CD8 + T cells and CD56 + NK cells is low on the vascular wall in remodeled artery

Evaluation of infiltrating lymphocytes

The number of infiltrating lymphocytes was evaluated using immunohistochemistry. Slides were independently examined by a single pathologist (A.T.), and the number of immunopositive infiltrating lymphocytes within the vascular wall was counted. The vascular wall was defined as the interval from the luminal surface to the decidual border of the spiral artery. We estimated the CD4⁺ lymphocyte concentration based on the results of CD3 and CD8 staining, as described in a previous publication [28]. As the CD8 stained slide was a serial section of the slide stained with CD3, the presumptive CD4⁺ lymphocyte population were estimated by subtracting the number of CD8⁺ cells from that of CD3⁺ cells. The number of endovascular lining cells was evaluated by immunohistochemistry; endothelial cells were identified by CD31 and EVT by CK7. The lymphocyte density was defined as the number of lymphocytes within the spiral artery vascular wall divided by the number of endovascular lining cells (lymphocytes/endovascular lining cells; Fig. 3).

Fig. 3

Schematic representation of the pre-defined areas for the quantification of lymphocytes within the vascular wall of spiral arteries. Examples of the applied area for a decidual spiral artery section with a non-remodeled muscular artery stained for CD3. The area surrounded by the blue line represents the vascular wall, from the luminal surface to the decidual border of the artery. The blue dots represent the nuclei of endovascular lining cells, and their number was counted. The lymphocyte count within the vascular wall was measured, and lymphocyte density was calculated as lymphocytes per endovascular lining cell

Statistical analyses

Statistical analyses of the clinical data were performed using the Mann-Whitney U test, Fisher’s exact test, and chi-squared test. The Wilcoxon signed-rank test was used to compare the lymphocyte density in the spiral arteries in the same cases. The Mann-Whitney U test was used to compare the density of lymphocytes in the spiral arteries between patients with HDP /AA and control patients. Analyses were performed using the IBM SPSS Statistics software (version 26; IBM Corp., Armonk, NY, USA). A p value < 0.05 was considered significant.

Results

Clinical findings

HDP was diagnosed in 173 patients, and AA was observed in 56 of the patients (32.4%). The maternal characteristics and perinatal outcomes are summarized in Table 1. Patients with HDP and AA exhibited early onset and lower gestational age at delivery, birth weight, and Apgar score at 1 min than those without AA (p < 0.001, Table 1). In addition, patients with HDP and AA had frequent severe proteinuria, severe hypertension, and SGA infants (p < 0.01, Table 1). The frequency of multiparity was also significantly higher in these patients (p < 0.01, Table 1). Maternal age, maternal body mass index, diabetes mellitus, use of assisted reproductive technology, sex of the infant, blood pH in the umbilical artery, and Apgar score at 5 min were not significantly different between patients with HDP and AA and those without AA. In contrast, patients with HDP and AA exhibited lower gestational age at delivery, birth weight and Apgar score at 1 min, and higher maternal body mass index than healthy controls (p < 0.05, Table 1). Patients with HDP and AA had a higher number of SGA infants compared with control patients (p < 0.001, Table 1). Maternal age, parity, diabetes mellitus, use of assisted reproductive technology, placental weight, sex of the infant, and blood pH in the umbilical artery were not significantly different between patients with HDP and AA and control patients.

Table 1 Maternal characteristics and perinatal outcomes

Histopathological findings

We compared AA and non-RA (Fig. 4a) and non-RA and RA (Fig. 4b) within HDP cases with acute atherosis. We also compared non-RA and RA within the control cases (Fig. 4c). In addition, we compared HDP cases with acute atherosis and control cases in each of the non-RA (Fig. 4d) and RA (Fig. 4e) groups. In patients with HDP and AA, the density of CD3⁺ T cells, CD8⁺ T cells and CD4⁺ T cells in the AA group was higher than those in the non-RA group (p < 0.05, Fig. 4a). The density of CD56⁺ NK cells was not significantly different between the AA and non-RA groups (Fig. 4a). The density of CD3⁺ T cells, CD8⁺ T cells, CD4⁺ T cells and CD56⁺ NK cells in the non-RA group was higher than those in the RA group (p < 0.05, Fig. 4b). In control patients, the density of CD4⁺ T cells was higher in the non-RA compared to the RA group (p < 0.05). The density of CD8⁺ T cells and CD56⁺ NK cells was not significantly different between RA and non-RA groups (Fig. 4c). The density of CD56⁺ NK cells in the non-RA group was higher in patients with HDP and AA than in control patients (p < 0.05, Fig. 4d). The density of CD3⁺ T cells, including CD8 and CD4, around non-RA was not significantly different between patients of HDP with AA and control patients (Fig. 4d). The density of CD3⁺ T cells, CD4⁺ T cells, and CD56⁺ NK cells in the RA group was not significantly different between patients of HDP with AA and control patients (Fig. 4e).

Fig. 4

The density of lymphocytes in spiral arteries (the ratio of immune positive lymphocytes to endovascular lining cells). Spiral artery was classified into three groups, acute atherosis, non-remodeled muscular artery, and remodeled artery. (a) In cases of HDP with AA. Box-plot graphs show that the density of CD3+, CD8+, CD4 + T cells in AA is significantly higher than those in non-RA. The density of CD56 + NK cells is no significantly different AA and non-RA. (b) In cases of HDP with AA, comparison of non-RA and RA. Box-plot graphs show that the density of CD3+, CD8+, CD4 + T cells and CD56 + NK cells in non-RA is significantly higher than those around RA. (c) In control cases. Box plot graph show that the density of CD3+, CD4 + T cells in non-RA is significantly higher than those in RA. The density of CD8 + T cells and CD56 + NK cells is not significantly different between those arteries. (d) Comparison of HDP with AA cases and control cases in non-RA. Box plot graphs show that significantly different in only density of CD56 + NK cells. (e) Comparison of HDP with AA cases and control cases in RA. Box plot graphs show that the density of CD3 + T cells and CD56 + NK cells is no significantly different between HDP with AA and control cases

The placental tissues of patients with HDP and AA exhibited DVH and increased syncytial knots more frequently than those without AA (Table 2).

Table 2 Histopathological findings of HDP patients. Analysis by fisher exact test

Discussion

We found that patient with HDP and AA frequently had early onset, lower gestational age, and more severe conditions than control patients. Moreover, the placental tissue from patients with HDP and AA frequently exhibited DVH and increased syncytial knots, reflecting placental hypoxia/ischemia.

These findings are consistent with previous reports that perfusion of the placenta from the mother is impaired only when there are secondary atherotic changes in the spiral arteries [22] and that fetal growth restriction (FGR) is more frequent in patients with HDP and AA [29]. The validity of the case selection in this study was confirmed.

In patients with HDP and AA, the density of CD56⁺ NK cells in the AA and non-RA groups was higher than that in the RA group. In contrast, the density of CD56⁺ NK cells was not significantly different between the RA and non-RA groups in control patients. The density of CD56⁺ NK cells in the non-RA group was significantly higher in patients with HDP and AA compared to control patients. These findings suggest that an increase in CD56⁺ NK cells in the spiral arteries is involved in impaired vascular remodeling. In patients with HDP, the AA group exhibited a higher number of CD8⁺ and CD4⁺ T cells compared with the non-RA group. The density of CD56⁺ NK cells did not differ significantly between the AA and non-RA groups. It is presumed that CD4⁺ and CD8⁺ T cells, rather than CD56⁺ NK cells, are associated with the development of AA in spiral arteries. Furthermore, both in patients with HDP and AA and in control patients, the density of CD4⁺ T cells was higher in non-RA than in RA. This finding suggests that CD4⁺ T cells play a role in spiral artery remodeling. Therefore, several immune cells have been suggested to be involved in spiral artery remodeling, impaired vascular remodeling, and AA development.

During pregnancy, NK cells represent 50–70% of the decidual lymphocytes. In contrast to peripheral blood NK cells, decidual NK cells release various cytokines/chemokines that induce immune tolerance, trophoblast invasion, and vascular remodeling [13]. Using a bioinformatics approach, Rabaglino et al. reported that insufficient or defective maturation of the endometrial and decidual NK cells during the secretory phase and early pregnancy preceded the development of preeclampsia [30]. William et al. evaluated the major leukocyte populations (CD3⁺ and CD8⁺ T cells, CD14⁺ macrophages, and CD56⁺ NK cells) in the decidua from normal pregnancies and pregnancies complicated with preeclampsia or FGR [31]. They demonstrated that CD3⁺ T cells, CD8⁺ T cells, CD14⁺ macrophages, and CD56⁺ NK cells decreased in preeclampsia decidua, and only CD56⁺ NK cells decreased in FGR decidua. Their findings suggest that the function and abundance of decidual NK cells may alter the local cytokine balance and affect spiral artery remodeling [30, 31]. The evaluation of immune cells in the spiral artery in this study differed from those of previous studies in terms of site, and result in term of increase/decrease in the number of NK cells were also different. However, previous findings are in agreement with our results in that there was a change in the number of NK cells and that NK cells were associated with spiral artery remodeling.

Johnsen et al. examined the lymphocyte subsets CD3, CD8, FoxP3, and CD56 around AA in preeclampsia and normotensive pregnancies. They found that CD3⁺/CD8^–/FoxP3^– T cells were associated with AA and that AA was associated with an increased intramural CD8⁺ T cell concentration, but the number of cells was low. They suggested that non-regulatory CD4⁺ T cells might be involved in the development of AA in the decidual spiral arteries [28]. These results are partly consistent with our observation that CD4⁺ T cells are involved in AA formation, but differ from our results in that the density of CD8⁺ T cells around AA was higher than that around other arteries. We conclude that CD8⁺ T cells contribute to AA development. Stallmach et al. reported a higher frequency of decidual CD8⁺ T cells in preeclampsia complicated by FGR than in normal pregnancies [32]. In our study, the proportions of HDP cases complicated by FGR were 39.3% in patients with HDP and AA, and 7.7% in patients with HDP without AA. As cases of HDP with AA and preeclampsia complicated by FGR are presumed to overlap, it is suggested that the formation of AA is associated with an increase in CD8⁺ T cells. CD8⁺ T cells play a role not only in the antigen-dependent acquired immune system but also in the innate immune system, which is activated in response to cytokines [33]. Therefore, it is necessary to examine the phenotypes of CD8⁺ T cells around the spiral arteries.

CD4⁺ T cells are activated in secondary lymphoid tissues and differentiate into effector T cells. CD4⁺ effector T cells are characterized by the production of cytokines and are divided into the Th1, Th2, Th17, Thf, and Treg subsets [34]. Ma et al. demonstrated that invading endovascular EVT secreted transforming growth factor-beta1 (TGF-β1) which promoted naïve CD4⁺ T cell differentiation into immunosuppressive Treg, thereby developing immune tolerance along the placental maternal circulation [35].

This result obtained confirm our hypothesis that CD4⁺ T cells play a role in spiral artery remodeling. However, Johnsen et al. observed that Treg cells are generally scarce or absent in the decidua, including around the spiral arteries [28]. Therefore, the differentiation of CD4⁺ T cells around spiral arteries requires examination.

The strengths of our study are that the morphological identification of the spiral arteries allowed for the evaluation of local immune cells, and rigorous evaluation of the association with impaired spiral artery remodeling was accomplished by the identification of patients with HDP and AA. The limitations of our study are the relatively small sample size and the fact that the results were based on placental tissue obtained from full-term pregnancies. Our findings may not reflect spiral artery remodeling during early pregnancy. In addition, appropriate adjustment for gestational age was not performed in the control group.

Conclusions

We found differences between HDP and normal pregnancies in the density of lymphocyte subsets in the spiral arteries with successful and unsuccessful vascular remodeling. Our findings suggest that CD4⁺ T cells may play a role in spiral artery remodeling during pregnancy, that increased CD56⁺ NK cells are associated with impaired vascular remodeling, and that the formation of AA involves CD4⁺ and CD8⁺ T cells. These immune differences may cause changes in the local cytokine balance around the spiral artery, contributing to HDP development.