Immunological comparison of pregnant Dahl salt-sensitive and Sprague-Dawley rats commonly used to model characteristics of preeclampsia

Abstract

The pregnant Dahl salt-sensitive (S) rat is an established preclinical model of superimposed spontaneous preeclampsia characterized by exacerbated hypertension, increased urinary protein excretion, and increased fetal demise. Because of the underlying immune system dysfunction present in preeclamptic pregnancies in humans, we hypothesized that the pregnant Dahl S rat would also have an altered immune status. Immune system activation was assessed during late pregnancy in the Dahl S model and compared with healthy pregnant Sprague-Dawley (SD) rats subjected to either a sham procedure or a procedure to reduce uterine perfusion pressure (RUPP). Circulating immunoglobulin and cytokine levels were measured by enzyme-linked immunosorbent assay (ELISA) and Milliplex bead assay, respectively, and percentages of circulating, splenic, and placental immune cells were determined using flow cytometry. The pregnant Dahl S rat exhibited an increase in CD4⁺ T cells, and specifically TNFα⁺CD4⁺ T cells, in the spleen compared with virgin Dahl S rats. The Dahl also had increased neutrophils and decreased B cells in the peripheral blood as compared with Dahl virgin rats. SD rats that received the RUPP procedure had increases in circulating monocytes and increased IFN-ɣ⁺CD4⁺ splenic T cells. Together these findings suggest that dysregulated T cell activity is an important factor in both the pregnant Dahl S rats and SD rats after the RUPP procedure.

INTRODUCTION

A growing body of literature suggests that immune system dysfunction has a role in both the development of preeclampsia and the downstream manifestations of the disease (1–3). The maternal immune system undergoes many changes over the course of pregnancy to prevent a maternal immune response to paternal antigens and establish a cooperative status with semiallogeneic trophoblast cells, all while maintaining defenses against pathogens (4). It was proposed by Sargent et al. (5) that an aberrant maternal immune response against the trophoblast, including abnormal cytokine production and altered decidual natural killer (NK) cell activity, is responsible for the initiation of preeclampsia. In addition, cytokines and cells of both the innate and adaptive immune system have been implicated in the downstream maternal inflammatory syndrome (2).

Multiple experimental models of preeclampsia have been developed based on causes and mediators of the disease, however, none fully recapitulate the human condition (6, 7). The pregnant Dahl salt-sensitive (Dahl S) rat has been described as a spontaneous model of superimposed preeclampsia (8–10). The Dahl S pregnancy is consistent with many of the characteristics observed in superimposed preeclampsia, including exacerbated hypertension, increased proteinuria, placental hypoxia, increased plasma soluble fms-like tyrosine kinase-1 (sFlt-1), and intrauterine growth restriction (8). Because the immune system has been implicated in the development of hypertension and renal injury in the Dahl S rat (11, 12), and clinical evidence indicates that immune activation contributes to the maternal syndrome of preeclampsia, we hypothesize that activation of immune system is prevalent during the pregnancy of the Dahl S rat. An additional well-established rodent model of uteroplacental ischemia is the reduced uterine perfusion pressure (RUPP) model (13). The RUPP model is generated by inserting silver clips to cause mechanical constriction of the uterine arteries and aorta on GD14, restricting blood flow to the placenta. Previous studies report that the RUPP model exhibits increases in mean arterial pressure, impaired renal function (14), endothelial dysfunction (15), and immune system activation and systemic inflammation that are characteristic of the maternal syndrome in preeclampsia (16). The goal of the present study was to assess and compare immunological parameters in female virgin and pregnant Dahl S rats to determine the potential change in immune cell types during pregnancy and to compare those to immunological changes observed in Sprague-Dawley (SD) rats that have undergone the RUPP procedure.

MATERIALS AND METHODS

Animals

Dahl S rats.

Dahl salt-sensitive rats (SS/Jr, n = 14) were obtained from the colony maintained by Dr. Michael Garrett at the University of Mississippi Medical Center. Timed breeding was performed in 7 Dahl S rats starting at 16 wk of age, and presence of sperm in vaginal smears was indicative of gestational day (GD) 0. All rats were maintained on normal rodent chow (TD7034, 0.3% NaCl, Harlan Teklad, Madison, WI) and water ad libitum on a 12-h light-dark cycle.

Sprague-Dawley rats.

Timed pregnant (n = 13) and virgin (n = 9) SD rats were obtained from Charles River Laboratories (SAS SD, strain code: 400). For RUPP rats, pregnant animals were received on GD11 and subjected to aortic and bilateral uterine artery constriction on GD14. Briefly, rats were anesthetized and maintained on 3% isoflurane and a midline abdominal incision was made. After externalization of both uterine horns, uterine arteries and abdominal aortas were dissected from surrounding tissue, and one single 0.203-mm silver surgical clip was placed on the abdominal aorta above the iliac bifurcation. One 0.100-mm silver surgical clip was placed on both the left and right uterine arteries that supply the uterus to prevent compensatory flow; with placement caudal to the ovarian artery to prevent restricted blood flow to the ovary. Animals were excluded from the study if there was either total reabsorption of pups on GD19 or if there was less than 25% fetal demise, indicating insufficient reduction in blood flow. Rats undergoing sham procedures were anesthetized on GD14, maintained on 3% isoflurane, a midline abdominal incision was made, and both uterine horns were externalized. Both the uterine arteries and abdominal aorta were isolated as aforementioned, but no surgical clips were placed. Virgin SD, sham SD, and RUPP SD animals were anesthetized on GD18 and implanted with an indwelling carotid artery catheter for the assessment of mean arterial pressure.

An additional subset of SD rats (n = 7) from the colony maintained by Dr. Jennifer Sasser at the University of Mississippi Medical Center (denoted as SD-UMMC in Table 2, original breeders from Harlan, Indianapolis, IN) were also used for analysis of circulating cytokines. Timed breeding with SD males was performed, and the presence of sperm in vaginal smears was indicative of GD0 as described in the in Dahl S rats.

All experiments were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved and monitored by the University of Mississippi Medical Center Institutional Animal Care and Use Committee.

Tissue Harvest

Both Dahl S and SD rats were euthanized on GD19 while under isoflurane anesthesia. A blood sample was taken from the abdominal aorta for plasma and peripheral blood leukocyte isolation, and placentae (2 or 3 per animal) and spleens were harvested from each animal. Single cell suspensions were prepared from each placenta and spleen (details in Cell Isolation), and whole placenta and a spleen sections were snap frozen in liquid nitrogen and stored at −80°C until use.

Plasma Immunoglobulins

IgM and IgG concentrations were measured in plasma using the Rat IgM total enzyme-linked immunosorbent assay (ELISA) and Rat IgG total ELISA (eBioscience, San Diego, CA) according to the manufacturer’s instructions.

Cytokine Analyses

Plasma.

Plasma cytokines were analyzed using the MILLIPLEX MAP Rat Cytokine/Chemokine Magnetic Bead Panel (Millipore Sigma, Burlington, MA) according to the manufacturer’s instructions. Samples were run on a Milliplex Analyzer with Luminex Technology, and the mean fluorescence intensity data were analyzed using a 5-parameter logistic method to analyze cytokine concentrations in each sample.

Placenta.

One-hundred-milligram segments of frozen placenta were combined with 1 mL radio immunoprecipitation assay (RIPA) lysis buffer containing 10 mM Tris·HCl (pH 8.0), 140 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% SDS, and complete mini EDTA-free protease inhibitor cocktail (Roche, Basel, Switzerland) and agitated for 2 min in a Bullet Blender tissue homogenizer (Next Advance Laboratory Instruments, Troy, NY). Samples were then centrifuged at 11,000 g for 10 min to remove debris. The supernatant was removed, and total protein was quantified in the supernatant using the bicinchoninic acid assay (BCA) method. Lysates were analyzed using the Rat tumor necrosis factor-α (TNF-α) and Mouse/Rat IL-17F ELISAs (R and D systems, Minneapolis, MN).

Cell Isolation

Peripheral blood leukocytes.

Erythrocytes were lysed by adding 10× volume of 1× PharmLyse (BD Biosciences, San Jose, CA). After incubation for 6 min at room temperature, the blood was centrifuged at 200 g for 5 min. The pelleted cells were washed 1× phosphate-buffered saline solution (PBS), 2% fetal calf serum (FCS) and centrifuged at 350 g for 5 min. The purified peripheral blood leukocytes (PBLs) were resuspended in 90% FCS and 10% dimethyl sulfoxide (DMSO) and stored at −80°C until use.

Spleen.

One-third of the spleen was homogenized using the Spleen Dissociation Kit (Miltenyi Biotec, Bergisch Gladbach, Germany) with the GentleMACS Octo Dissociator (Miltenyi Biotec) according to the manufacturer’s instructions. The resulting homogenate was filtered through a 70 μM cell strainer and centrifuged at 300 g for 10 min. The cell pellet was then resuspended in 3 mL of 1× PharmLyse and incubated at room temperature for 5 min. Twenty-seven millilters of 1× PBS, 2% FCS were added and the samples were centrifuged at 300 g for 10 min. Purified splenocytes were then subjected to in vitro culture or resuspended in 90% FCS, 10% DMSO, and stored at −80°C until use.

Placenta.

The decidua of each placenta was removed, and the placenta was homogenized in 5 mL Roswell Park Memorial Institute (growth medium) (RPMI) media containing 200 U/mL DNase and 10 mg/mL collagenase IV using the GentleMACS Octo Dissociator and a user-defined protocol for placenta. The resulting homogenate was filtered through a 70 μM cell strainer and washed with 1× PBS containing 2% FCS and 2 mM EDTA. The single cell suspension was centrifuged at 300 g for 10 min. The resulting cell pellet was then resuspended in 3 mL of 1× PharmLyse and incubated at room temperature for 5 min. Twenty-seven milliliters of 1× PBS containing 2% FCS were added, and the samples were centrifuged at 300 g for 10 min. Cells were then resuspended and immediately analyzed using flow cytometry.

Flow Cytometry

Surface immunophenotyping.

For all flow cytometric analyses of cell surface markers of placenta, spleen, and PBLs, cells (2 × 10⁷ cells/mL) were washed and resuspended in PBS. Cells were then stained using Zombie Green Live/Dead stain (BioLegend) for 15 min at room temperature. The cells were subsequently washed in PBS containing 2% FCS and 0.9% sodium azide and nonspecific binding was blocked using anti-rat CD32 (BD Biosciences) according to the manufacturer’s instructions. Cells were stained with antibodies diluted in PBS containing 2% FCS and 0.09% sodium azide on ice for 30 min. The antibodies used are listed in Table 1, and the representative gating strategy is shown in Figs. 1 and 3A, 5A, and 7A. After incubation, the samples were washed two times in 2 mL of PBS containing 2% FCS and 0.09% sodium azide. All samples were analyzed on a Gallios (Becton Dickinson, Franklin Lakes, NJ) flow cytometer at the UMMC Flow Cytometry core facility. A total of 100,000 events were acquired for each sample. Data were analyzed using Kaluza software.

Table 1.

Monoclonal antibodies used for flow cytometry

Antibody	Supplier	Clone
CD32	BD Biosciences	D34-485
CD45	BD Biosciences	OX-1
CD3	BD Biosciences	G4.18
CD4	BD Biosciences	OX-35
CD8α	BD Biosciences	OX-8
CD45RA	BD Biosciences	OX-33
CD43	Biolegend	W3/13
Anti-rat granulocytes	BD Biosciences	His48
CD161	BD Biosciences	10/78
TNF-α	BD Biosciences	TN3-19.12
IFN-γ	BD Biosciences	DB-1
IL-4	BD Biosciences	OX-81
IL-17a	eBioscience	eBio17B7

IL-4, interleukin-4; IFN-γ, interferon-γ; TNF-α, tumor necrosis factor-α.

Figure 1.

Representative gating strategy used to identify immune cells in peripheral blood, spleen, and placenta.

Figure 3.

Effect of preeclamptic pregnancy on circulating immune cell populations in Sprague-Dawley (SD) and Dahl S rats. A: representative gating strategy for peripheral blood leukocytes. Cells were purified from blood isolated from the abdominal aorta. B: percentage of circulating CD3⁺CD4⁺ T cells in virgin, sham, and RUPP SD rats and virgin and pregnant Dahl S rats. C: percentage of circulating CD3⁺CD8⁺ T cells. D: percentage of circulating CD45RA⁺ B cells. E: percentage of CD3⁻CD161⁺ NK cells. F: percentage of neutrophils. G: percentage of monocytes, including CD43Lo/His48Hi and CD43Hi/His48Int-Lo cells. n = 9 SD-virgin, n = 6 SD-sham, n = 7 SD-RUPP, n = 7 Dahl-virgin, n = 7 Dahl-pregnant. *P < 0.05 using one-way ANOVA for SD and unpaired t test for Dahl. Dahl S, Dahl salt-sensitive; NK, natural killer; RUPP, reduced uterine perfusion pressure.

Figure 5.

Effect of preeclamptic pregnancy on splenic immune cell populations in Sprague-Dawley (SD) and Dahl S rats. A: representative gating strategy for splenocytes. Single cell suspensions were prepared from spleens isolated from virgin animals or at GD19 of pregnancy. B: percentage of splenic CD3⁺CD4⁺ T cells in virgin, sham, and RUPP SD rats and virgin and pregnant Dahl S rats. C: percentage of CD3⁺CD8⁺ T cells. D: percentage of CD45RA⁺ B cells. E: percentage of CD3-CD161⁺ NK cells. F: percentage of neutrophils. G: percentage of monocytes, including CD43Lo/His48Hi and CD43Hi/His48Int-Lo cells. n = 9 SD-virgin, n = 6 SD-sham, n = 7 SD-RUPP, n = 7 Dahl-virgin, n = 7 Dahl-pregnant. *P < 0.05 using one-way ANOVA for SD and unpaired t test for Dahl. Dahl S, Dahl salt sensitive; RUPP, reduced uterine perfusion pressure.

Figure 7.

Analysis of placental T and B lymphocytes and cytokines. A: representative gating strategy for placental cells. B: percentage of CD3⁺CD4⁺ T cells. C: percentage of CD3⁺CD8⁺ T cells. D: percentage of CD45RA⁺ B cells. E: placental TNF-α levels, as measured by ELISA. F: placental IL-17F levels, as measured by ELISA. n = 6 SD-sham, n = 7 SD-RUPP, n = 7 Dahl-pregnant *P < 0.05 using t test for SD. ELISA, enzyme-linked immunosorbent assay; RUPP, reduced uterine perfusion pressure; SD, Sprague-Dawley; TNF-α, tumor necrosis factor-α.

Intracellular cytokine staining.

For flow cytometric analyses of intracellular cytokines, splenocyte cell pellets were resuspended in complete RPMI media and the cells were counted using a hemocytometer. After adjusting the cell concentration to 2 × 10⁶/mL, 6 mL (12 × 10⁶ cells) were transferred to a six-well culture plate. Ten microliters of Leukocyte Activation Cocktail (BD Biosciences) containing phorbol 12-myristate 13-acetate, ionomycin, and BD GolgiPlug were added to each well. The cells were then incubated at 37°C/5% CO₂ for 4 h. After incubation, the cells were harvested, transferred to a 50 mL tube, and centrifuged at 400 g for 5 min at 8°C. The cells were washed once and resuspended in PBS containing 2% FCS and 0.09% sodium azide. Cells were then aliquoted (1 × 10⁶ cells/tube) into flow cytometry tubes blocked using anti-rat CD32 (BD Biosciences) as stated in Surface Immunophenotyping. Cells were stained with anti-rat CD3 and anti-rat CD4 for 30 min on ice. After washing was completed, the cells were resuspended in 250 µL of BD Cytofix containing 4% paraformaldehyde and incubated at 4°C for 20 min. The cells were then resuspended in 1 mL of 1× Perm/Wash Buffer (BD Biosciences) and incubated for 15 min at 4°C and pelleted by centrifugation. The cells were resuspended in 50 µL Perm/Wash buffer and then stained with anti-TNF, anti-IL-17, anti-IL-2, or anti-IL-4 for 30 min. Cells were then washed twice in perm/wash buffer and resuspended in in PBS containing 2% FCS and 0.09% sodium azide. All samples were analyzed on a Gallios (Becton Dickinson) flow cytometer at the UMMC Flow Cytometry core facility. A total of 100,000 events were acquired for each sample, and data were analyzed using Kaluza software.

Statistical Analysis

Data are presented as means ± SE. Statistical analyses were performed using GraphPad Prism 7. One-way ANOVA followed Tukey’s post hoc test for multiple comparisons was used to analyze differences among the SD-virgin, SD-sham, and SD-RUPP. An unpaired t test was used to analyze differences between virgin and pregnant Dahl S rats. An unpaired t test was used to analyze differences between SD-sham and SD-RUPP placentas. A P value of less than 0.05 was considered statistically significant.

RESULTS

Plasma Immunoglobulin Levels

To analyze the impact of pregnancy on circulating immunoglobulins, we measured total IgG and IgM levels in plasma of virgin SD and Dahl S rats as well as pregnant rats subjected to either a sham or RUPP procedure (Fig. 2). Pregnancy had no effect on circulating IgG levels in either strain, and the RUPP rats did not differ from sham-operated pregnant SD rats. Circulating IgM levels were increased in SD-sham and SD-RUPP animals as compared with SD-virgin.

Figure 2.

Effect of preeclamptic pregnancy on plasma IgG and IgM levels in SD and Dahl S rats. A: circulating IgG. B: circulating IgM. n = 9 SD-virgin, n = 6 SD-sham, n = 7 SD-RUPP, n = 7 Dahl-virgin, n = 7 Dahl-pregnant. SD, Sprague-Dawley; Dahl S, Dahl salt-sensitive; IgG, immunoglobulin G; IgM, immunoglobulin M; RUPP, reduced uterine perfusion pressure. *P < 0.05 using one-way ANOVA for SD.

Circulating Cytokine Levels

We next analyzed circulating cytokines and chemokines using a magnetic bead-based assay (Table 2). Notably, SD-virgin, SD-sham, and SD-RUPP animals all had elevated levels of several inflammatory cytokines. There were, however, some differences among the three SD groups. Pregnant SD rats receiving either the sham or RUPP procedure had significantly different plasma concentrations of CXCL5 (decreased), IL-1α (decreased), IL-4 (increased), and IL-5 (increased) as compared with the virgin animals. In addition, SD-sham animals had significantly increased levels of IL-12p70 and IL-17a as compared with virgin animals. SD-RUPP rats had significantly higher levels of IL-18 and TNF-α as compared with SD-virgin. Plasma levels of IL-10 in SD-RUPP rats were also greater than both SD-virgin and SD-sham rats. We hypothesized that the elevated circulating cytokines observed in the SD rats could result from transportation during pregnancy, the RUPP surgical procedure, or the implantation of carotid catheters for blood pressure measurements in these groups. Therefore, we also assessed cytokine levels in virgin SD rats that were bred on-site without indwelling catheters. In this SD-UMMC group, many of the cytokines were below the detection limit of this kit and were also significantly lower compared with the virgin rats from Charles River that had undergone surgical manipulations. In comparison, the majority of cytokines were at undetectable levels in both virgin and pregnant Dahl rats and were similar to those in the SD-UMMC group. Of the few cytokines that were detected, no differences were identified when comparing pregnant and virgin Dahl S rats.

Table 2.

Circulating cytokines/chemokines from virgin and pregnant SD and Dahl S rats

	SD-Charles River–Implanted with carotid catheter			SD-UMMC	Dahl S
Cytokine/ Chemokine, pg/mL	Virgin	Sham	RUPP	Virgin	Virgin	Pregnant
CXCL5	5,214 ± 372	2,315 ± 254a	3,054 ± 388a	283 ± 52	772 ± 408	618 ± 204
CXCL10	490 ± 161	701 ± 230	737 ± 346	52 ± 5	106 ± 10	127 ± 13
Eotaxin	121 ± 105	20.5 ± 0.2	20.3 ± 0.4	16.9 ± 0.03	17.3 ± 0.1	17.3 ± 0.03
G-CSF	27.6 ± 6.2	13.3 ± 3.6	8.6 ± 3.6a	ND	ND	ND
GM-CSF	135 ± 48	218 ± 16	215 ± 35	ND	ND	ND
Fractalkine	36.7 ± 3.2	31.2 ± 3.0	42.1 ± 7.3	13.9 ± 2.5	17.7 ± 0.9	16.3 ± 1.5
GRO/KC/CINC-1	197 ± 61	237 ± 103	383 ± 182	ND	ND	ND
IL-1 α	319 ± 26	175 ± 21a	217 ± 37a	ND	ND	ND
IL-1 β	28.5 ± 9.1	39.3 ± 4.1	35.3 ± 4.8	ND	ND	ND
IL-2	108.7 ± 60.9	57.2 ± 7.9	107.6 ± 37.3	ND	ND	ND
IL-4	16.4 ± 5.6	68.2 ± 7.4a	56.6 ± 12.0a	ND	ND	ND
IL-5	40.3 ± 3.4	82.7 ± 3.9a	74.3 ± 5.6a	20.3 ± 3.3	30.0 ± 4.6	28.0 ± 4.7
IL-6	258 ± 104	474 ± 163	571 ± 103	ND	ND	ND
IL-10	16.0 ± 3.9	17.9 ± 6.6	63.6 ± 21.5a,b	ND	ND	ND
IL-12p70	55.2 ± 35.4	260 ± 49a	172 ± 37	ND	ND	ND
IL-13	ND	ND	ND	ND	ND	ND
IL-17a	7.67 ± 0.76	18.10 ± 1.41a	14.80 ± 3.63	ND	ND	ND
IL-18	22.0 ± 5.7	87.3 ± 10.5	195.1 ± 72.0a	ND	70.5 ± 29.0	84.2 ± 31.4
MCP-1	566 ± 257	598 ± 98	674 ± 130	170 ± 27	267 ± 41	230 ± 41
MIP-1a	25.6 ± 3.4	23.4 ± 4.0	53.2 ± 17.4	7.1 ± 0.1	9.9 ± 1.0	9.5 ± 0.9
MIP-2	ND	128 ± 13	123 ± 15	ND	ND	ND
RANTES	1,913 ± 736	563 ± 70	1,014 ± 169	168 ± 22	404 ± 120	494 ± 230
TNF-α	3.0 ± 1.8	22.4 ± 10.3	49.5 ± 21.6a	ND	ND	ND

^a P < 0.05 versus SD-virgin (CR);

^bP < 0.05 versus SD-sham. Dahl S, Dahl salt-sensitive; ND, not detected; SD-UMMC, SD-University of Mississippi Medical Center; TNF-α, tumor necrosis factor-α.

Circulating Immune Cells

Because both pregnancy and preeclampsia cause alterations in circulating immune cell populations, we examined circulating immune cells isolated at GD19 from pregnant dams and from virgin animals using flow cytometric analyses. The antibodies used are listed in Table 1, and the representative gating strategy is shown in Figs. 1 and 3A. CD4⁺ and CD8⁺ T-cell populations were analyzed, and no significant differences in circulating CD3⁺CD4⁺ T cells were found among SD-virgin, SD-sham, or SD-RUPP rats (Fig. 3B). Similarly, there were also no differences between virgin and pregnant Dahl rats. There was a significant decrease in CD3⁺CD8⁺ T cells in SD-sham rats as compared with SD-virgin (Fig. 3C), but pregnancy had no effect on circulating CD8⁺ T cells in Dahl S rats. CD45RA⁺ B cells were also analyzed in the periphery (Fig. 3D). There were no significant differences in circulating B cells among virgin, sham, or SD-RUPP rats. Virgin Dahl rats had a much higher percentage of CD45RA⁺ B cells (Dahl-virgin: 48.5% ± 3.1%) than what is reported in the literature for other rat strains (17), and pregnant animals had decreased circulating CD45RA⁺ B cells (Fig. 3D). There were no significant changes in NK cells, as defined by a CD3^-CD161^high surface phenotype (18), in either the SD or Dahl S rats (Fig. 3E).

Circulating neutrophil and monocyte cell populations were analyzed according to the method of Barnett-Vanes et al. (19), which uses differential staining of monocytes and neutrophils using anti-His48 and anti-CD43 antibodies. There were no differences in the percentages of neutrophils among SD-virgin, SD-sham, and SD-RUPP animals (Fig. 3F), and the neutrophil levels were elevated in all three groups compared with normal reported values for rats (14%–20%). In a small subset of animals, we evaluated the effect of carotid catheter surgery on neutrophil percentages in SD rats. PBLs were isolated from SD-sham and SD-RUPP animals who did not receive carotid catheter surgery, and compared with SD-sham and SD-RUPP that did receive carotid catheter surgery. Similar percentages of neutrophils were seen in all groups (Fig. 4). Notably, Regal et al. (20) reported similar neutrophil percentages in pregnant SD rats who received either a sham or RUPP procedure at GD14 followed by implantation of a carotid catheter on GD18. Virgin Dahl S rats had neutrophil percentages in agreement with other studies in which no surgical procedures had been performed (19), and pregnant Dahl S rats had increased neutrophils as compared with Dahl-virgin. SD-RUPP rats had elevated percentages of monocytes as compared with SD-virgin rats (Fig. 3G), and although not statistically significant, there was a trend for an increase in monocytes in SD-sham as compared with SD-virgin animals.

Figure 4.

Percentages of neutrophils identified in the peripheral blood of Sprague-Dawley (SD)-sham and SD-RUPP rats with and without catheters implanted on GD18. RUPP, reduced uterine perfusion pressure.

Splenic Immune Cells

We also analyzed percentages of splenocytes, using the same surface marker antibodies as in the peripheral blood. The representative gating is shown in Fig. 5A. No changes were detected in the percentages of CD3⁺CD4⁺ T cells in SD rats, but there were significantly more splenic CD4⁺ T cells in pregnant Dahl rats as compared with Dahl-virgin (Fig. 5B). SD-RUPP had a higher percentage of CD8⁺ T cells as compared with SD-virgin (Fig. 5C). SD-sham and SD-RUPP rats had a significant decrease in CD45RA⁺ B cells as compared with the SD-virgin animals, but pregnancy had no effect on splenic CD45RA⁺ B-cell percentages in the Dahl S rats (Fig. 5D). Similar to the results in the peripheral blood, pregnancy had no impact on splenic NK cells (Fig. 5E). Myeloid cell populations were also examined (Fig. 5, F–G), and SD-sham and SD-RUPP rats had significantly lower neutrophils as compared with SD-virgin rats, but pregnancy had no impact on splenic neutrophil percentages in Dahl S rats as compared with Dahl-virgin. SD-sham had a lower percentage of splenic monocytes as compared with SD-virgin (Fig. 5G).

Because there are several T_H subsets, we further characterized CD4⁺ T lymphocytes in the spleen. Briefly, freshly isolated splenocytes were stimulated for 4 h in culture using phorbol 12-myristate 13-acetate (PMA) and calcium ionophore and we performed surface staining with CD3 and CD4 followed by intracellular cytokine staining (Fig. 6). There were no significant differences in percentages of TNF-α⁺, IL-4⁺, or IL-17⁺ CD4⁺ T cells among SD-virgin, SD-sham, and SD-rats. SD-sham had significantly increased IFN-ɣ⁺ cells compared with SD-virgin. In Dahl rats, pregnancy resulted in a significant increase in TNF-α⁺ CD4⁺ T lymphocytes but did not alter IFN-ɣ⁺, IL-4⁺, or IL-17⁺ splenic CD4⁺ T cells.

Figure 6.

Effect of preeclamptic pregnancy on splenic T cell cytokine production. Splenocytes were stimulated for 4 h using leukocyte activation cocktail (BD Biosciences) containing phorbol 12-myristate 13-acetate, ionomycin, and BD GolgiPlug. A: percentage of CD3⁺CD4⁺ T cells that produce TNF-α. B: percentage of CD3⁺CD4⁺ T cells that produce IL-4. C: percentage of CD3⁺CD4⁺ T cells that produce IFN-γ. D: percentage of CD3⁺CD4⁺ T cells that produce IL-17a. n = 7 SD-virgin, n = 6 SD-sham, n = 6 SD-RUPP, n = 6 Dahl-virgin, n = 5 Dahl-pregnant. *P < 0.05 using one-way ANOVA for SD and unpaired t test for Dahl. IFN-γ, interferon; IL-4, interleukin-4; IL-17a, interleukin-17a; RUPP, reduced uterine perfusion pressure; SD, Sprague-Dawley; TNF-α, tumor necrosis factor-α.

Placental lymphocyte and Cytokine Profiles

Placental cytokine production and changes in immune cell homeostasis may also contribute to the risk for developing preeclampsia. Therefore, we examined placental T and B lymphocytes using flow cytometry and placental cytokine levels using ELISA (Fig. 7). SD-RUPP rats had significantly higher percentages of both placental CD4⁺ T cells and CD8⁺ T cells as compared with SD-sham animals. There were no differences in placental CD45RA⁺ B cells between SD-sham and SD-RUPP groups. In the SD rats, the percentages of T cells and B cells were lower in the placenta than in the circulation; in contrast, placentas from Dahl S rats had slightly higher percentages of T cells compared with the peripheral blood, but much lower percentages of B cells. TNF-α and IL-17 levels were measured from placental total cell lysates using ELISAs, and no significant differences were found in TNF-α or IL-17 levels between SD-sham and SD-RUPP.

DISCUSSION

A multitude of adaptations occur in the immune system during pregnancy to allow for the optimal growth and delivery of the fetus; however, dysregulated immunological processes may be a risk for the development of preeclampsia. The major goal of this study was to compare alterations in immune status in pregnant Dahl S rats (a model of superimposed preeclampsia) with pregnant SD rats either with or without a RUPP procedure that induces characteristics of preeclampsia. Notably, the immunological changes that occurred in each model were distinct and different in both the circulation and the spleen. The major findings in the Dahl model were as follows: the pregnant Dahl rat has an increase in neutrophils and a decrease in B cells in the circulation, and an increase in CD4⁺ T cells, and specifically TNFα⁺CD4⁺ T cells in the spleen compared with virgin rats. In the pregnant SD rats, SD-sham rats had a decrease in CD8⁺ T cells in the circulation and SD-RUPP rats had increased circulating monocytes. Several other changes in T cells, B cells, monocytes, and neutrophils in the spleen were also present in the pregnant SD rats. Finally, SD-RUPP rats had increased placental T cells. An additional important finding was that circulating cytokines were greatly elevated in the SD rats that received carotid catheter surgeries the day before blood collection. These results emphasize the need for caution when interpreting results from data collected using measurements in the systemic circulation to assess immune system function in rodents immediately following surgical procedures.

Changes in Circulating Immune Cells

In the present study, we analyzed the predominant immune cell populations in SD-virgin rats and pregnant SD rats at GD19. The dynamics of circulating immune cells in pregnant women are quite complex, although there is no appreciable evidence to support that the maternal immune system is globally suppressed during pregnancy. Two significant changes were observed in the SD rats: 1) SD-sham rats have significantly lower CD8⁺ T cells as compared with virgin, and 2) SD-RUPP rats have an increase in circulating monocytes compared with SD-virgin animals. Similarly, in healthy human pregnancies, there is a decrease in CD8⁺ cytotoxic T cells in the circulation, so this may be indicative of normal pregnancy-associated changes in immune cells (21). Changes in circulating monocytes have been reported in pregnancy, with most studies reporting increases in intermediate monocytes that are both inflammatory and phagocytic. The percentages of intermediate monocytes are even higher in preeclamptic women and rats subjected to adenosine triphosphate (ATP) infusion to induce a preeclamptic phenotype (22). Notably, we did not see an increase in B cells in the periphery of SD-RUPP. This agrees with studies by Regal’s group (23), but contrasts with work by LaMarca et al. (24).

In the pregnant Dahl S rats, there was a decrease in circulating B cells and an increase in neutrophils. A recent clinical study examined B cells in the third trimester of pregnancy and determined that B-cell counts in the peripheral blood are lower in late third trimester than in nonpregnant women (25); thus, it may be that the decreases in B cells observed in the pregnant Dahl S are reflective of normal late pregnancy changes. Jensen et al. reported that preeclamptic women had an expansion of CD19⁺CD5⁺ B1a B cells that are responsible for the production of natural and polyreactive antibodies of the IgM isotype. In addition, the authors showed that these cells produced autoantibodies to the angiotensin II type 1 receptor (AT1R-AA) (26), which have been identified in the plasma of preeclamptic women (27). A similar population of CD5⁺ B cells has been identified in mice, but not in rats (28).

Neutrophils are short-lived cells that are a first-line defense during infections. It is increasingly well-appreciated that neutrophils play a role in the regulation of adaptive immune responses, including antibody production (29), T-cell function (30), and antigen presentation (31). Investigators have reported a neutrophilia in pregnancy that increases with gestational age (32), and also evidence of increased neutrophil activation and delayed neutrophil apoptosis as compared with nonpregnant values (33). In addition, it has been reported that the increases in neutrophil activation and delayed apoptosis are exacerbated in preeclampsia (33, 34). The importance of neutrophils in pregnancy was demonstrated in studies by Nadkarni et al. (35), in which neutrophils were shown to induce proangiogenic T cells with a regulatory phenotype. In this study, we found an increase in circulating neutrophils in the pregnant Dahl S rat as compared with the virgin animals, in line with studies in humans. It is unknown, however, whether these neutrophils have an activated phenotype or contributed to the preeclamptic phenotype. Regal et al. (20) demonstrated that neutrophil depletion lowers blood pressure in SD-RUPP rats, but does not affect blood pressure (BP) significantly in sham animals, who also have neutrophilia. These data suggest that neutrophils in the RUPP animals are potentially phenotypically different than neutrophils in the sham animal. Another possibility is that neutrophils infiltrate resistance vessels and therefore could contribute to the endothelial dysfunction and hypertension associated with preeclampsia (36).

Changes in Splenic Immune Cells

There was a decrease in splenic B cells in the SD-sham and RUPP animals as compared with virgin, and the percentages are similar to those reported for B2 B cells on GD19 in SD rats by Laule et al. (23), which were defined using IgM or CD45RA (used in this study). However, it is unknown why the percentages are significantly different from the virgin animals. A comprehensive analysis of B-cell development in different stages of pregnancy in mice found that there are lower numbers of B cells in the spleen due to decreased B-cell production in the bone marrow (37). Although speculative, this decrease in B-cell generation in the bone marrow could be responsible for the decrease in circulating B cells in the pregnant Dahls in this study. We also observed decreased neutrophils and monocytes in the spleen of pregnant SD rats, but there is a paucity of data on the potential implications of these changes.

Finally, there were several changes in T-cell populations in pregnant SD and Dahl rats. There was an increase in CD8⁺ cytotoxic T lymphocytes in the spleen of SD-RUPP animals. It is known that large numbers of CD8⁺ T cells accumulate in the decidua by late pregnancy, and these cells have an activated effector memory phenotype (38). Elevated levels of CD8⁺ T cells are associated with preeclampsia in women (39); however, less is known about this cell type as compared with the CD4⁺ T cells and investigation into the role of these cells and cytokines they produce is needed. In addition, the increase in CD4⁺IFN-ɣ⁺ T cells in the spleen suggests an increase in T_H1 activity in the RUPP model. An increase in T_H1 cells has been documented in women with preeclampsia (40), and studies in rodents showed that the adoptive transfer of activated T_H1 cells into pregnant mice leads to the development of preeclamptic symptoms, including increased blood pressure, altered kidney function, and decidual inflammation (41). Splenic CD4⁺ T cells were higher in pregnant Dahl S rats and compared with virgin rats, and we also identified significantly higher percentages of splenic CD4⁺ T cells that stained positive for TNF-α (Fig. 6) in pregnant Dahl S rats. T_H1 cells can produce TNF-ɑ, which also suggests increased T_H1 activity in the pregnant Dahl. The increase in TNF-α-producing CD4⁺ T cells, coupled with the increased levels of TNF-α previously reported (8), suggests that TNF-α may play a pathogenic role in the maternal syndrome in Dahl S rats. These previous results were obtained using a more sensitive ELISA than the current Luminex bead assay to measure TNF-α levels in the plasma. Future studies are needed to fully examine the role that TNF-α-producing CD4⁺ T cells play in the development of the maternal syndrome of superimposed preeclampsia and how this cytokine affects the different organ systems affected by the disease. It is well established that TNF-ɑ blockade can reduce blood pressure and renal injury in several animal models of hypertension (42), and was also reported to improve the preeclamptic phenotype in the RUPP model (43).

Circulating Immunoglobulins and Cytokines/Chemokines

Circulating antibodies can have both beneficial and pathogenic roles in pregnancy: antibodies against paternal antigens have a protective effect, whereas pathogenic autoantibodies can be detrimental to pregnancy outcome (44). Although there have been limited studies regarding plasma IgG concentrations throughout pregnancy in humans, recent reports have shown than mean IgG levels drop throughout pregnancy, with the concentration of IgG at term being ∼90% of that in the first trimester (45). An earlier study by Benster and Wood revealed that women in the third trimester have lower concentrations of IgG as compared with nonpregnant women, and that women experiencing hypertension during pregnancy had significantly lower IgG levels as compared with healthy pregnant women (46). In this study, we were unable to detect any changes in circulating IgG levels in either the SD or Dahl S rats, although there was a trend for lower IgG levels in the pregnant Dahl S as compared with virgin Dahl S animals (Fig. 2). Interestingly, both virgin and pregnant Dahl S rats had increased concentrations of IgG in their plasma as compared with what has been reported in the literature for other rat strains (47–49). Whether this has functional significance is unknown. The IgM concentrations in SD-sham and SD-RUPP are similar to the values obtained by Regal et al. (50), which also reported no difference between these two groups. They did, however, see increased IgM deposition in both the kidney and placenta of RUPP animals as compared with sham. Most studies show that circulating IgM levels are similar throughout pregnancy (46, 51), or may slightly increase by the end of gestation (52).

Cytokines and chemokines are important to maintain homeostasis during pregnancy, with each cytokine following a specific pattern of expression over the gestational period in both humans and rodents (53). In humans, early and late pregnancy are characterized by more proinflammatory/T_H1 type cytokine expression, including IL-2, TNF-α, and IFN-ɣ, whereas midpregnancy is marked by a more T_H2-dominated immune response that favors the production of cytokines such as IL-4 and IL-10 (54). We detected elevated levels of most cytokines in the virgin SD rats as well as the pregnant rats who underwent either a sham or RUPP surgery. As stated earlier, these rats were also implanted with carotid artery catheters one day before the harvest of blood and tissues, which is the standard approach in many studies using the RUPP model. A recent study by Morton et al. (55) also showed elevated levels of cytokines in rats that underwent a sham operation, a standard RUPP procedure (used in this study), or a selective RUPP which does not restrict blood flow in the aorta. These animals were also implanted with telemetry probes for BP measurement. Thus, many of the elevations of cytokines are likely due to a systemic inflammatory response to surgery (56, 57).

Despite the effects of the surgical procedures, some differences among the three groups of SD rats were detected. Both sham and RUPP SD rats had significantly lower levels of the cytokine CXCL5 as compared with the virgin SD rats. In addition, there was a trend for decreased CXCL5 in the pregnant Dahl S as compared with the virgins. CXC chemokines play diverse roles in angiogenesis, cell trafficking, and have been extensively studied for their role in placental trophoblast invasion (58). CXCL5 binds to chemokine receptor CXCR2 and plays a role in neutrophil chemotaxis, and has also been shown to play a role in angiogenesis in nonsmall cell lung cancer (59, 60). To our knowledge, no studies published to date have examined CXCL5 levels in pregnancy or preeclampsia.

IL-12p70, which is the bioactive form of IL-12, was also elevated in SD-sham and SD-RUPP rats as compared with SD-virgin rats. The primary function of IL-12 is to increase IFN-ɣ production by promoting the differentiation of naïve T lymphocytes into T_H1 cells, which secrete IFN-ɣ (61). This agrees with the increases in IFN-ɣ⁺ T cells that were detected in the spleen of pregnant SD rats. In addition, IL-18 levels were higher in SD-sham and SD-RUPP rats. IL-18 also mediates T_H1-type immunity by inducing IFN-ɣ production with IL-12 (62), but in the absence of IL-12, IL-18 can promote T_H2-type responses (63, 64). Studies in pregnant women show an increase in IL-18 over the course of pregnancy, with the highest levels being during labor or in complicated pregnancies (65). IL-12 levels are generally lower in pregnant woman compared with nonpregnant controls and increased in women with preeclampsia (66). Sakai et al. (66) studied both IL-18 and IL-12 in pregnant women and determined that the ratio of IL-18 to IL-12 is higher in pregnant women than in nonpregnant women, and the IL-18/IL-12 ratios are significantly lower in women with preeclampsia as compared with women experiencing a healthy pregnancy. We did not detect a similar IL-18/IL-12 ratio pattern in the SD-sham and SD-RUPP animals; and in fact, the SD-sham animals had a lower IL-18/IL-12 ratio than the SD-RUPP animals.

Finally, both TNF-α and IL-17α were elevated in both SD-sham and SD-RUPP animals as compared with the SD-virgin group. Preeclamptic women exhibit increased levels of circulating (67, 68) and placental (69, 70) TNF-α, and these results have been echoed in previous studies using the RUPP model of placental ischemia (43, 71–73). Other rodent hypertensive models of pregnancy, including the stroke-prone spontaneously hypertensive rat, have elevated circulating TNF-α (74). Although the levels of TNF-α in the Dahl S rats were below the detection limit of the Milliplex assay in this study, previous studies in our laboratory showed a small but significant increase in circulating TNF-α levels in pregnant Dahl rats compared with virgin Dahl rats and pregnant SD rats when measured by a more sensitive ELISA (10). IL-17 is predominately secreted by T cells, ɣδ T cells, NK cells, and neutrophils (75). IL-17 increases during late pregnancy and plays a role in the initiation of labor (76). IL-17a levels were higher in both the SD-sham and SD-RUPP rats compared with virgin rats, but were not different between sham and RUPP animals. Previous studies in the RUPP model report that administration of a soluble IL-17 receptor to block IL-17 activity lowered blood pressure (71); however, the controls used were pregnant animals that did not have sham surgical procedures performed. Thus, it may be that the increases in IL-17 in late pregnancy in these animals are due to surgical procedures at GD14 of pregnancy.

SD rats also exhibited alterations in several T_H2 or anti-inflammatory cytokines, including IL-4, IL-5, and IL-10. IL-4 induces naïve T cells to differentiate into T_H2 cells, and it also decreases the production of T_H1 cells. IL-4 levels increase throughout normal pregnancy due to the high levels of progesterone, which is a known inducer of IL-4 production (77). Loss of IL-4 in mice leads to the development of preeclampsia-like symptoms, including proteinuria and increases in blood pressure (78). IL-4 levels were significantly higher in SD-sham and SD-RUPP animals as compared with the virgin animals, likely due to the increases during normal pregnancy. IL-10, which is primarily produced by monocytes and T_H2 cells, increases significantly during early pregnancy and decreases in the third trimester until the onset of labor (79). Data are conflicting regarding IL-10 levels in late pregnancy in preeclampsia, and our current results showing significantly increased IL-10 levels in the SD-RUPP animals as compared with the SD-virgin and SD-sham groups, which conflicts with a previous study using this model. Cornelius et al. (80) reported decreased IL-10 in the circulation on GD19 in RUPP rats as compared with normal pregnant animals, and a study by Davidge’s group showed similar IL-10 levels between sham and RUPP animals (55). A recent meta-analysis showed that more studies reported higher circulating IL-10 in the third trimester in preeclamptic women as compared with normal pregnant controls (81); thus, it may be that high IL-10 in late pregnancy has immunostimulatory activity. The proinflammatory potential of IL-10 has been documented in studies of other inflammatory conditions, including human experimental endotoxemia (82), Crohn’s disease (83), and psoriasis (84). Together, these findings suggest that high IL-10 levels in late pregnancy have a detrimental effect and could contribute to the preeclamptic phenotype.

A limitation of this study is that we did not confirm the preeclamptic-like phenotype in either model as our focus was on the immunology of the two strains. The intent of the study was to ultimately have a better understanding of the fundamental immunological changes that are occurring in the Dahl S rat during pregnancy, and then compare that with the well-characterized RUPP model in the SD rat. One advantage of the Dahl S rat is that it develops characteristics of preeclampsia spontaneously, so we could examine the immunological effects of pregnancy in the absence of any surgical manipulations or stress. Therefore, we did not implant telemetry transmitters/catheters (risking altered immune status due to surgery) or use tail cuff (thereby inducing restraint stress)/metabolic cage studies (unusual housing stress) in these rats. Although the lack of blood pressure data is certainly a limitation of the current study, we have observed a very consistent phenotype in this inbred rat strain (8, 9). We have also published consistent results using the RUPP model in the SD rat (85, 86). Ultimately, this deeper understanding of the immunological changes during pregnancy in these strains will inform new questions that will help identify mechanisms contributing to preeclampsia.

Perspectives and Significance

Although the ability of the immune system to distinguish self from nonself is the basis of immunity to pathogens, pregnancy success is dependent on a favorable interaction between maternal and semiallogeneic fetal cells. There is significant evidence that preeclampsia likely initiates from a defective maternal-fetal immune response that is perpetuated by alterations in maternal cytokines and leukocytes, which lead to an exaggerated maternal inflammatory state. The current study provides important new insight to the immunological changes that occur during pregnancy in two rat strains that are commonly used as models of preeclampsia. The immunological changes described herein can serve as the foundation for asking new questions about the mechanistic underpinnings of preeclampsia. Therefore, future work is needed to directly test how and when these changes occur during implantation and placental development, and whether they ultimately contribute to the initial phase of preeclampsia. The findings of this study also extend on previous work in both the Dahl S model of superimposed preeclampsia and RUPP model of placental ischemia to compare and contrast immunological changes that occur in these two different strains. Understanding these changes could help us to understand the pathogenesis of the systemic maternal disease to develop strategies to manage the disease more effectively in the clinic.

GRANTS

E. B. Taylor was supported by F32HL137393 and K99HL146888. This work was supported by R01HL134711, R01HL136684, R01HL137791, K01DK095018, R00HL116774, P20GM104357, Veteran’s Administration Merit award (BX002604-01A2), and the Carl W. Gottschalk Research Scholar Grant.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

E.B.T. and J.M.S. conceived and designed research; E.B.T., E.M.G., and J.M.S. performed experiments; E.B.T. and J.M.S. analyzed data; E.B.T. and J.M.S. interpreted results of experiments; E.B.T. prepared figures; E.B.T. drafted manuscript; E.B.T., E.M.G., J.M.S., M.R.G., and M.J.R. edited and revised manuscript; E.B.T., E.M.G., J.M.S., M.R.G., and M.J.R. approved the final version of manuscript.