Animal Models of Preeclampsia: Investigating Pathophysiology and Therapeutic Targets

Abstract

Animal models have been critical in investigating pathogenesis, mediators and even therapeutic options for a number of diseases, including preeclampsia. Preeclampsia is the leading cause of maternal and fetal morbidity and mortality, worldwide. The placenta is thought to play a central role in the pathogenesis of this disease as it releases anti-angiogenic and pro-inflammatory factors into the maternal circulation, resulting in the maternal syndrome. Despite the deleterious effects preeclampsia has been shown to have on mother and baby during pregnancy and postpartum, there is still no effective treatment for this disease. While clinical studies in patients are crucial to identify the involvement of pathogenic factors in preeclampsia, there are obvious limitations that prevent detailed investigation of the quantitative importance of time dependent mechanisms involved in this syndrome. Animal models allow investigators to perform proof of concept studies and examine whether certain factors found in preeclamptic women mediate hypertension and other manifestations of this disease. In this brief review, we summarize some of the more widely studied models used to investigate pathophysiological mechanisms that are thought to be involved in preeclampsia. These include models of placental ischemia, angiogenic imbalance and maternal immune activation. Infusion of preeclampsia-related factors into animals has been widely studied to understand the specific mediators of this disease. These models have been included, in addition to a number of genetic models involved in overexpression of the renin-angiotensin system, complement activation and trophoblast differentiation. Together these models cover multiple mechanisms of preeclampsia from trophoblast dysfunction and impaired placental vascularlization to the excess circulating placental factors and clinical manifestation of this disease. The majority of animal studies have been performed in rats and mice, however we have also incorporated non-human primate models in this review. Preclinical animal models have been instrumental in not only understanding the pathophysiology of preeclampsia but they continue to be important tools in the search for novel therapeutic options for the treatment of this disease.

Introduction

Despite its position as the leading cause of maternal death and major contributor to maternal and perinatal morbidity, there is no effective drug treatment to delay the progression of preeclampsia, and current management therapies have significant limitations.¹ At present, the only effective treatment for preeclampsia is early delivery (removal of the placenta). Thus, the discovery of novel approaches for the treatment of preeclampsia is a major unmet need in the field. Identification of therapeutic targets for PE treatment can only result from the interplay between basic research involving animal models and clinical research in humans. The study of preeclampsia in humans is of critical importance to identify potential biomarkers and pathogenic factors that associate with the progression of preeclampsia.^2-4 The data obtained from human studies, however, are often correlative and unable to establish cause and effect relationships. Moreover, clinical studies in humans have obvious limitations that prevent detailed investigation of the quantitative importance of time dependent mechanisms involved in this syndrome. In contrast, experimental studies in animal models, despite their limitations, allow investigators to perform proof of concept studies. In addition, studies in animal models allow investigators to examine whether certain factors found in preeclamptic women can indeed lead to hypertension and other manifestations of this disease. In this brief review, we attempt to summarize some of the more widely studied and/or recently developed animal models used to investigate pathophysiological mechanisms that are thought to be involved in preeclampsia. These mechanisms and resulting models include placental ischemia (reduced uterine perfusion pressure models), impaired angiogenesis (sFlt-1 chronic excess model), models used to study maternal immune activation in preeclampsia and genetic mouse models that examine specific pathogenic pathways.

Animal Models of Preeclampsia

Since the spontaneous development of preeclampsia is essentially limited to the human species, the study of preeclampsia in patients is of critical importance to identify biomarkers and potential pathogenic factors that correlate with the progression of the syndrome. However, experimental studies in pregnant women have obvious limitations that prevent complete investigation of many of the mechanisms involved. In contrast, studies in animal models, despite their limitations, allow investigators to isolate and study the many pathways involved in preeclampsia. Preclinical studies in rodent animal models have been critical in understanding the pathogenesis of this disease. In investigating therapeutic options, targets identified in small animal models must be tested in non-human primates for efficacy and safety before transitioning to human trials. Limited studies have been performed in non-human primates to study preeclampsia, most likely due to expense and other limitations in resources. Still, these studies have contributed to better understanding of this disease and have been included in this review.

Models used to study the relationship between placental ischemia and maternal syndrome

Experimental induction of chronic uteroplacental ischemia appears to be a promising animal model to study potential mechanisms of preeclampsia. Uterine artery resistance is markedly increased in preeclampsia as a result of impaired spiral artery remodeling and placental ischemia.⁵ Indeed, the most accurate predictive measurement of early-onset preeclampsia is measurement of angiogenic balance and uterine artery Doppler assessment.⁶ Reductions in uteroplacental blood flow in a variety of species lead to a maternal cardiovascular phenotype that closely resembles preeclampsia in women.^7-13 The chronic reduced uteroplacental perfusion pressure (RUPP) rat model was developed and characterized by our laboratory to examine potential pathophysiological mechanisms that mediate cardiovascular and endothelial dysfunction in response to placental ischemia.⁷

The RUPP surgical procedure is typically performed in timed-pregnant Sprague-Dawley rats (gestation day 14),¹² but the procedure has also been successfully performed in other rat strains, including Wistar rats.¹⁴ Uterine perfusion pressure in the gravid rat is reduced by slipping a silver constriction clip around the aorta below the renal arteries, right above the iliac bifurcation where the uterine arteries lie.^12,15 We have found this procedure to reduce uterine perfusion pressure by ~40%.¹⁶ Because compensation of blood flow to the placenta occurs in pregnant rats through an adaptive increase in ovarian blood flow, branches of both the right and left ovarian arteries that supply the uterus are also clipped.¹² The timing of the vessel constriction is an important consideration. During a series of pilot studies to determine the appropriate clip size and the ideal gestational time for reducing uterine perfusion pressure, it was found that placement of the clips prior to day 14 of gestation in the rat resulted in a significant increase in fetal death. However, placement of clips on gestational day 14 to reduce placental perfusion produces a consistent blood pressure effect and minimizes total fetal reabsorption.^12,17,18

In response to RUPP, both uterine and placental blood flow are decreased by approximately 40%.¹⁶ Arterial pressure increases by approximately 20-25 mmHg by day 19 of pregnancy. In contrast, RUPP in virgin rats results in no significant effect on arterial pressure relative to control virgin rats.¹⁷ The hypertension in RUPP rats is associated with increases in total peripheral resistance and decreased cardiac output, systemic hemodynamic characteristics consistent with preeclampsia in women.¹⁶ Glomerular filtration rate and to a lesser extent renal blood flow decreases in the RUPP model relative to normal pregnancy at day 19.¹⁷ Although quite variable, an increase in urinary protein excretion has also been observed in the RUPP model when compared with normal pregnancy.¹² The reason for the variability is unknown but may be due to the short time frame of exposure to placental ischemia.

Angiogenic imbalance, endothelial dysfunction, reduced production of nitric oxide in vascular tissue, and increases in vascular endothelin-1 and reactive oxygen species production are all characteristics of preeclampsia also present in the RUPP rat.^1,19-24 Placental ischemia in rats also increases placental expression of hypoxia-inducible factor (HIF)-1α and immunoreactive sFlt1 in the placenta and plasma.²³ sEndoglin levels are also elevated in the RUPP model.²⁴ Moreover, RUPP is associated with decreases in plasma concentrations of vascular endothelial growth factor (VEGF) and placental growth factor (PlGF).²³

Fetal growth restriction (FGR), another feature of human preeclampsia, is also evident as pup weights are decreased in RUPP rats relative to controls.²⁵ Alexander and colleagues have performed a series of studies utilizing the FGR rat offspring from the RUPP model to examine the mechanisms underlying the relationship between birthweight and cardiovascular diseases later in life.²⁶ Thus, offspring from RUPP rats are also an important model to examine mechanisms of fetal programming of cardiovascular disease.

The RUPP model has been proven to be a useful tool to examine the mechanisms whereby ischemia initiates a cascade of events resulting in placental inflammation and subsequent systemic inflammation and hypertension in the mother.^27-30 RUPP rats have increased placental expression and circulating levels of pro-hypertensive cytokines (TNF-α, IL-6, IL-17, etc.) and immune cell counts (T and B lymphocytes), as well as increases in the production of the agonistic autoantibody to the angiotensin type 1 receptor (AT1-AA), and imbalance in T helper and T regulatory cells.^27-30 Activation of the complement system also occurs in RUPP rats, as detected in the circulation by decreased complement component 3 and increased complement activation product C3a.³¹

Cerebrovascular disturbances are now part of the diagnostic criteria for preeclampsia when accompanied by new onset hypertension following the 20^th week of gestation.¹ Blood-brain barrier (BBB) disruption, cerebral edema, and impaired cerebral blood flow (CBF) regulation are common cerebrovascular findings in preeclamptic women.^32-35 The mechanisms contributing to these cerebrovascular abnormalities during pregnancy, however, are not completely understood. In the last few years, the RUPP model of placental ischemia has been used to determine whether placental ischemia causes similar characteristics as the clinical syndrome and to begin to dissect the mechanisms that contribute to RUPP-induced cerebrovascular abnormalities. Placental ischemia in the RUPP model leads to increased BBB permeability and cerebral edema.^36,37 In these studies, impaired cerebral blood flow and vascular myogenic reactivity was particularly evident at higher perfusion pressures.^36,38 The RUPP model also has a shorter latency to the onset of drug-induced seizures, and is associated with increased concentrations of pro-inflammatory cytokines in the cerebrospinal fluid.³⁹ These findings were consistent with a study utilizing the RUPP rat coupled with high cholesterol in which the threshold for seizures was reduced compared to normal pregnant rats.³⁹ Importantly, magnesium sulfate has the capability of reducing placental ischemia-induced increases in pro-inflammatory cytokines in the cerebrospinal fluid.⁴⁰ While these studies demonstrated that the RUPP model can be utilized to study mechanisms of cerebrovascular abnormalities, it is still not known what factors are responsible for BBB disruption, cerebral edema, and impaired cerebral blood flow following placental ischemia.

Preeclampsia is associated with peripartum and postpartum myocardial fibrosis and heart failure.^41-43 Despite high incidence of serious morbidity and mortality due to postpartum heart failure after a preeclamptic pregnancy, there is a substantial gap in our knowledge regarding the cardiac dysfunction during preeclampsia, and how to prevent or reverse it. Recent studies from our laboratory indicate that the RUPP model has many features of cardiac dysfunction seen in preeclamptic women. Importantly, the RUPP model of preeclampsia also demonstrates reduced myocardial function (ejection fraction and global longitudinal strain) seen in preeclamptic women. Furthermore, biomarkers consistent with cardiac fibrosis including cardiac troponin, increased collagen mRNA expression, markers of pathological hypertrophy including increased brain natriuretic peptide (BNP), myosin heavy chain (MHC) α/β, and atrial natriuretic peptide (ANP) levels are present in RUPP rats.^44-46 We also have data showing that reduced EF and fractional shortening persists up to 8 weeks postpartum in the RUPP model despite blood pressure returning to normal by this time.⁴⁵ Thus, the RUPP model may also be useful tool to provide a better understanding of the mechanisms underlying cardiac dysfunction that occurs in during preeclampsia and postpartum.

Thus, RUPP-induced hypertension in pregnant rats is associated with endothelial dysfunction, angiogenic imbalance, immune activation, and cardiac and cerebrovascular dysfunction; all features seen in preeclampsia women (Table 1). A major strength of the RUPP model is that investigators have been able to assess the functional and quantitative role for each of the systems in mediating the hypertension in the RUPP model by utilizing pharmacological agents to disrupt their actions. The RUPP model has also been used by numerous laboratories as a valuable tool to investigate potential therapeutic targets for the treatment of preeclampsia. Importantly, the RUPP model involves mechanical constriction of vessels that feed the uteroplacental unit as a means to reduce blood flow to the placenta. As a result, the RUPP rat displays a similar phenotype to preeclampsia, where placental ischemia is also thought to play a central role. A limitation of this model, however, is that it is not useful for studying the mechanisms involved in the abnormal spiral artery remodeling proposed to underlie placental ischemia in women with preeclampsia.

Table 1:

Features of reduced uterine perfusion models compared to preeclampsia.

Features	Preeclampsia	RUPP Rat
Mean arterial pressure	↑	↑
Total peripheral resistance	↑	↑
Circulating sFlt-1	↑	↑
Circulating sEng	↑	↑
Circulating free PlGF	↓	↓
Circulating free VEGF	↓	↓
Oxidative stress	↑	↑
AT1-AA	↑	↑
Pro-hypertensive cytokines	↑	↑
T and B lymphocytes	↑	↑
ET-1	↑	↑
NO	↓	↓
Renal plasma flow	↓	↓
GFR	↓	↓
GLS	↓	↓
Ejection fraction	↓	↓
Cardiac output	↓	↓
Cerebral blood flow regulation	↓	↓
Cerebral edema	↑	↑
BBB permeability	↑	↑
FGR	↑	↑

RUPP, reduced uterine perfusion pressure; UPI, uteroplacental ischemia; sEng, soluble endoglin; PlGF, placental growth factor; VEGF, vascular endothelial growth factor; ROS, reactive oxygen species; AT1-AA, angiotensin II type I receptor autoantibody; TNF-α, tumor necrosis factor-α; ET-1, endothelin-1; NO, nitric oxide, GFR, glomerular filtration rate; GLS, global longitudinal strain; BBB, blood brain barrier; FGR, fetal growth restriction, ↑, increased; ↓, decreased.

While numerous laboratories have utilized the rat RUPP model, the placental ischemic model of preeclampsia has also been applied in mice and in non-human primate models. The RUPP technique has been performed in mice, which is especially beneficial in using genetic engineered mice to enhance our understanding of the pathogenesis of placental ischemia-induced hypertension.^11,47 Makris et al. also characterized an uteroplacental (UPI) model in radiotelemetered pregnant baboons by selective ligation of one uterine artery resulting in a 40% decrease in uteroplacental blood flow.⁴⁸ Hypertension, proteinuria, and increased production of antiangiogenic markers were reported in the pregnant baboons compared to control animals who underwent a sham procedure.⁴⁸ Makris and colleagues also recently reported that RNAi modulation of placental sFLT1 significantly attenuated the blood pressure response to placental ischemia.⁴⁹

Models used to study role of angiogenic imbalance

Angiogenic imbalance is thought to play a central role in the development of preeclampsia. Pro-angiogenic factors, VEGF and PlGF, are produced during pregnancy and are critical in maintaining endothelial integrity. These angiogenic factors bind to vascular endothelial growth factor receptors (VEGFR), including VEGFR-1 or Flt-1.^50,51 sFlt-1 is an alternately spliced variant of the full-length receptor in which the transmembrane and cytosolic domains have been removed, leaving only the extracellular recognition domain.⁵² This recognition domain acts as a VEGF antagonist by binding free VEGF, and making it unavailable for proper signaling. Increased circulating levels of sFlt-1 and reduced PlGF levels have been documented in women during preeclampsia, as well as prior to the onset of clinical symptoms.⁵³ Excess sFlt-1 and soluble endoglin (sEng), which is also significantly increased in preeclampsia⁵⁴ cause depletion in the bioavailablity of VEGF and PlGF at the detriment of the endothelium and blood pressure regulation.

While there is a strong association between angiogenic imbalance and severity of preeclampsia, the relative importance of angiogenic factors in preeclampsia remains an open area of investigation. A number of experimental models have been used to uncover the mechanisms that link angiogenic imbalance and preeclampsia. Overexpression and infusion of sFlt-1 have both been shown to induce a preeclampsia-like phenotype ^53,55,56. Maynard et al. showed that infusion of an adenovirus (Ad) expressing sFlt in Sprague-Dawley rats on day 8 or 9 of pregnancy develops into hypertension, proteinuria and glomerular endotheliosis by day 16 or 17 of pregnancy (Figure 1).⁵³ Similarly infusion of sFlt-1 into pregnant rats induces preeclamptic-like symptoms, including hypertension, reduced fetal weight, and increases in vascular ROS.⁵⁵ Interestingly, excess sFlt-1 itself is not thought to cause hypertension in preeclampsia, but rather the deficiency of free pro-angiogenic factors such as VEGF and PlGF. In one study, mice infused with Ad-sFlt-1, animals develop hypertension, proteinuria, endotheliosis and vascular damage. However, when Ad-VEGF is administered in conjunction, these features are alleviated⁵⁷ These data indicate that free excess sFlt-1 and depletions of free VEGF and PlGF play a major role in preeclampsia. Importantly, studies have also investigated full-length human sFlt-1-e15a isoform, which is most abundant in placentas of preeclamptic women. Studies by Szalai et al show that infusion of full-length human sFlt-1 in mice also results in hypertension, proteinuria, glomerular injury and vascular damage.^58,59

Figure 1:

Mean arterial pressure (MAP; A) and proteinuria (B) is shown for rats following administration of an sFlt-1 adenovirus (Ad-sFlt-1) on during pregnancy.⁵³ Sprague-Dawley rats were injected on gestational day 8/9 of pregnancy and data was collected on gestational day 16/17. Data are presented as mean+SEM. *P<0.05.

As reported in the RUPP and UPI models, placental ischemia induces significant increases in circulating sFlt-1, and administration of PlGF in these models has been shown to improve the preeclampsia-like phenotype. In the RUPP rat, administration of 180ug/kg/day recombinant human (rh) PlGF for 5 days reduced blood pressure and fetal reabsorption in the in the RUPP rat, in addition to significant increases in placental and fetal weight.⁶⁰ Subsequently, another study showed that 100ug/kg/day rhPlGF administration in the UPI non-human primate model results in significant improvements in hypertension and proteinuria.¹⁰ Most recently, Logue et al. found that administration of an elastin-like polypeptide linked VEGF construct (ELP-VEGF) in the RUPP rat significantly reduces blood pressure.⁶¹ These and other data have led to the hypothesis that the ratio between sFlt-1 and PlGF/VEGF is critical to maintain a healthy endothelium and normal vascular activity. Furthermore, these studies suggest that improving the angiogenic balance in preeclampsia could be a therapeutic avenue for treatment. Current therapeutic efforts have also focused on reducing circulating sFlt-1 via apheresis, and in doing so improving the angiogenic balance, with promising results.⁶² These clinical data highlight the importance of angiogenic factors in the development of preeclampsia.^62,63

Animal models of angiogenic imbalance have also been useful to study the closely associated pregnancy disorder, HELLP syndrome (hemolysis, elevated liver enzymes, low platelet count). While HELLP syndrome only occurs in 0.2-0.8% of pregnancies, it is diagnosed alongside preeclampsia in 70-80% of cases.⁶⁴ Therefore, it is not surprising that factors induced in animal models of preeclampsia also produce a HELLP-like phenotype. Infusion of sFlt-1 and sEng on gestational day 12 results in increased mean arterial pressure, elevated liver enzymes, reduced platelets and pro-inflammatory factors such as TNF-α, IL-17, IL-6, CD4⁺ T cells and CD8⁺ T cells^65,66. Interestingly, this model is one of the few that has been studied postpartum to show blood-brain barrier permeability, persistent hypertension and indicators of anxiety.⁶⁷ The profound effect on immune factors in this model and others of angiogenic imbalance highlight in interactions between these two pathways.

Models used to examine role of Agnostic Angiotensin II Type 1 Receptor Autoantibodies

The renin-angiotensin system (RAS) plays an important role in normal pregnancy and in preeclampsia. Normal pregnancy is associated with activation of RAS components with diminished vascular responsiveness to angiotensin II In contrast, women with preeclampsia, have reduced angiotensinogen, plasma renin activity and angiotensin II, but typically exhibit increased vascular responsiveness to angiotensin II.⁶⁸ Dechend and colleagues previously reported that sera from preeclamptic women contain an IgG (type 3) autoantibody reacts with the AT1 receptor. A number of studies have now indicated that women with preeclampsia produce this novel agonistic autoantibody to the angiotensin II type I receptor (AT1-AA) during pregnancy and up to 8 years postpartum. ^69,70 One animal model utilized to examine the role of AT1-AA is the AT1-AA chronic excess model where purified AT1-AA is infused into normal pregnant rats. A number of investigators have shown that AT1-AA signaling, via the AT1 receptor, results in a variety of physiological effects, including TNF-α and ROS generation, both of which have been implicated in preeclampsia.⁷¹

Zhou et al demonstrated that immunoglobulin isolated from preeclamptic women increases systolic pressure in pregnant mice.⁷² Similar studies from LaMarca and colleagues have reported that infusion of purified rat AT1-AA, isolated from serum collected from a pregnant transgenic rat overproducing components of the renin angiotensin system, into pregnant rats from day 12 to day 19 of gestation, increased serum AT1-AA and blood pressure.⁷¹ While these data suggest AT1-AA causes hypertension by direct activation of AT1 receptor, additional pathways involving endothelin and anti-angiogenic factors may also be involved during pregnancy. More recently, Cunningham et al. showed that renal natural killer cells are activated in renal tissue from AT1-AA infused rats, and that mitochondrial dysfunction is also present in this model.⁷³ Blood pressure response in AT1-AA-infused animals have been blocked by AT1 receptor antagonists or co-injection of an AT1 receptor antagonist or the 7 amino acid peptide (n7AAc) that selectively blocks the actions of the AT1-AA.⁷⁴ Administration of n7AAc in the RUPP rat has also been shown to reduce blood pressure (Figure 2).⁷⁵ Another approach to block the effects of endogenous AT1-AA in the RUPP rat is B cell depletion, which results in a blunted blood pressure response to placental ischemia.⁷⁶

Figure 2:

Mean arterial pressure (MAP) is shown for normal pregnant rats, the reduced uterine perfusion pressure (RUPP) rat model and inhibition of the angiotensin type 1 receptor autoantibody (AT1-AA) in the RUPP rat (with ‘n7AAc’). These data suggest AT1-AA play a role in placental-ischemia induced hypertension. Data are mean+SEM and was retrieved from Cunningham et al. ⁷⁵. *P<0.05.

AT1-AA infusion in rats is also associated with angiogenic imbalance and elevated tissue levels of preproendothelin-1 (prepro-ET-1).⁷¹ Interestingly, blood pressure response in AT1-AA-infused animals can be blocked by ET type A receptor antagonist.⁷⁷ These data suggest that AT1-AA induced hypertension during pregnancy is in part due to activation of the endothelin system.⁷⁷ sFlt-1 and sEng are significantly elevated in this model.⁷⁷ sFlt-1 levels measured in media from placental explants from AT1-AA-infused rats are also significantly increased.⁷⁸ Interestingly sEng was not increased in media from placental explant, suggesting another source for its production.⁷⁸ These studies demonstrate an important interaction between inflammatory and angiogenic markers found to be produced excessively in response to placental ischemia.

Immune activation and cytokines

One of the earliest and most persistent theories about the origins of preeclampsia is that preeclampsia is a disorder of immunity and inflammation. This maternal immune tolerance involves crucial interactions between regulatory T cells (T regs) and uterine natural killer cells which recognize and accept the fetal antigens and facilitating placental growth. Complete failure leads to spontaneous miscarriage while partial failure of this crucial step leads to poor placentation and dysfunctional placental perfusion and chronic immune activation. Preeclamptic women have an increase in pro-inflammatory cytokines and decreases in uterine natural killer cells and Tregs.^29,79 In a mouse model, acute Treg-depletion from gestational day 3.5 results in fetal loss, increases in pro-inflammatory markers and uterine artery resistance. Interestingly, Treg-depletion alone does not increase blood pressure. However, in the presence of L-NAME (L-N^G-Nitro arginine methyl ester), which blocks nitric oxide production for proper endothelial function, blood pressure increases by approximately 25% in Treg-depleted mice, but only 15% in normal pregnant mice.⁸⁰ Treg cell populations have also been shown to be important in late gestation. Partial or total Treg depletion from gestational day 14.5 in a murine model of susceptibility to pre-term birth, resulted in greater fetal loss, and reduced survival of pups up to 3 weeks of age.⁸¹ The RUPP rat is another model that is induced late in pregnancy that also exhibit a reduction in Tregs⁸² and increases in T-helper (Th)-17 cells. Cornelius et al. showed that Th-17 cells isolated from the spleen of RUPP rats, cultured and injected interperitoneally into normal pregnant rats, induced a preeclampsia-like phenotype.^82,83 Meanwhile Tregs administration to the RUPP rat reduce hypertension and pro-inflammatory factors. ^84,85. These data suggest that the increased pro-inflammatory immune cells populations play a deleterious role in preeclampsia. These factors all act to cause pathological activation of the maternal endothelium as well as directly impacting multiple organ systems.

Pro-inflammatory cytokines are produced in abundance in preeclampsia coupled with deficiencies of anti-inflammatory factors, such as IL-10. IL-10 plays an important role in normal pregnancy in regulating the polarization of T cells to a Th-2 phenotype over the pro-inflammatory Th-1 phenotype.⁸⁶ Preeclamptic women have a twofold elevation in placental and plasma TNF-α protein levels. It is becoming increasingly evident that pro-inflammatory cytokines such as IL-6 and TNF-α interact with important blood pressure regulatory systems such as the RAS system, sympathetic nervous system, and endothelial factors.²⁹ TNF-α-infusion (50ng/day) in normal pregnant rats from gestational day 14 to 19 results in elevated blood pressure and increased expression of renal, placental, and aortic prepro-ET-1.⁸⁷ Moreover, the increase in mean arterial pressure in response to TNF-α is completely abolished in pregnant rats treated with an ET_A receptor antagonist.⁸⁸ Sera from pregnant rats exposed to chronic RUPP increases ET-1 production by cultured endothelial cells, and these levels of ET-1 production are mimicked when cells are exposed to TNF-α.⁸⁹ Further studies in this model suggest that TNF-α-induced hypertension could also be caused by decreases in renal nitric oxide synthase expression,⁸⁷ and increases in AT1-AA production.⁹⁰ Most recently, TNF-α-infusion in pregnant rats has been shown to exhibit impaired cerebrovascular regulation due to reduced β-epithelial Na+ channel expression.⁹¹

Systemic activation of the immune cells during pregnancy has been studied as a model of preeclampsia. Faas et al showed that injection of low-dose endotoxin to induce an pro-inflammatory state on gestational day 14 results in hypertension and proteinuria.⁹² These animals were also found to have platelet coagulopathy and glomerular injury. An alternative approach to induce systemic inflammation is via administration of lipopolysaccharide (LPS). A low-dose of LPS is adequate to induce fetal growth restriction, placental nitrosative stress, reduced spiral artery area, hypertension and proteinuria.⁹³ Although these models are not induced via specific preeclampsia-related factors, they have been useful in understanding how widespread activation of leukocytes in this disease contribute to its pathogenesis.

A novel model of preeclampsia was recently published where mice were treated with placental-derived extracellular vesicles (pcEV) on gestational day 17-18 from preeclamptic women, which resulted in hypertension, proteinuria and kidney injury. This was associated with increased vascular wall tension and reduced cerebral blood flow. Interestingly in virgin mice, hypertension developed within 30 minutes of pcEV infusion.⁹⁴ pcEV shedding from the placenta is considered normal throughout pregnancy. However, excess levels are seen due to the dysfunction placental in preeclampsia.⁹⁵

Models used to study role of endothelin and nitric oxide in preeclampsia

Endothelial dysfunction is another known stimulus for ET-1 synthesis. Plasma concentration of ET-1 has been measured in a number of studies involving normal pregnant and preeclamptic women. Most investigators have found higher ET-1 plasma concentrations of approximately two-to threefold in women with preeclampsia.^96,97 Previous studies have reported that 2-3 fold elevation in plasma levels of ET-1 in animals is adequate to impart significant long-term effects on systemic hemodynamics and arterial pressure regulation. Thus, long-term elevations in plasma levels of ET-1 comparable to those measured in women with preeclampsia could play a role in mediating the reductions in renal function and elevations in arterial pressure observed in women with preeclampsia.

There are robust increases in ET-1 coupled with marked deficiencies in vasodilatory mediators, including nitric oxide (NO), in preeclampsia.^98,99 Substantial production of ROS during preeclampsia leads to oxidation of NO and soluble guanylate cyclase, a critical molecule involved in NO-mediated vasodilation. NO production is reduced in preeclamptic women and in the RUPP and sFlt-1 excess models.²¹ An important model used to examine the role of NO deficiency in preeclampsia is the L-NAME model. In normal pregnant rats, L-NAME administration to block NO production from gestational day 1-19 leads to hypertension, proteinuria, elevated sFlt-1 and reduced fetal and placental weight.¹⁰⁰

Genetic models used to study preeclampsia mechanisms

BPH/5

Perhaps the best characterized genetically linked model for the study of preeclampsia is the BPH/5 mouse. The BPH/5 is a strain derivation of the BPH/2 “borderline hypertensive” mouse, and exhibits mildly elevated blood pressure throughout the adult lifespan of the animal. However, Davisson et al demonstrated that the BPH/5 strain also demonstrates late gestational acute elevations in blood pressure (up to ~25mmHg), which resolve immediately following parturition, mimicking the effects on blood pressure seen in the typical preeclampsia patient. This was accompanied by proteinuria, glomerulosclerosis, intrauterine growth restriction, maternal endothelial dysfunction, and increased fetal mortality.¹⁰¹ Work from the Davisson and Sones laboratories in subsequent years has uncovered a number of additional characteristics similar to human preeclampsia, such as altered extravillous trophoblast invasion and placental abnormalities which coincide with increased uterine artery vascular resistance. Importantly, these effects are seen prior to onset of hypertension, as is postulated in the human disorder.¹⁰² It should be noted, however, that while total circulating angiogenic potential is decreased in the BPH/5, this seems to be associated with decreased VEGF and PlGF rather than increased sFlt-1, an interesting difference from the human syndrome. Recent studies from Sones et al. shows that angiogenic imbalance precedes complement activation, which has been implicated in the pathogenesis of preeclampsia.¹⁰³

Angiotensinogen/Renin Overexpression

Though its role in long-term maintenance of blood pressure is well established, the exact pathophysiological role of the RAS in the development of preeclampsia is less clear. To investigate the role of the RAS in pregnancy-induced hypertension, several groups have utilized transgenic mouse and rodent strains in which females overexpressing human angiotensinogen (hAGN) are crossed with males overexpressing human renin hREN.^104-106 This model has been shown to exhibit late gestational hypertension and proteinuria. Recent studies have also showed the development of postpartum cardiac dysfunction in this model.¹⁰⁷ Importantly, when gender of the transgenic animals is swapped (i.e. male angiotensinogen and female renin), no phenotypic effect is noted, likely due to species specificity of renin activity and lack of secretion of AGN from the placenta itself. A similar model has been reported in mice overexpressing both REN and AGN, which are chronically hypertensive, as a model of superimposed preeclampsia; which also develops very late gestational hypertension and fetal growth restriction.¹⁰⁸

STOX1 Overexpression

Another new potential model that has recently been reported is transgenic overexpression of the transcription factor storkhead box 1 (STOX1) in a murine model. A significant, though often contradictory, body evidence has previously implicated STOX1 dysregulation in the etiology of preeclampsia.^109,110 Interestingly, Doridot et al. showed that STOX1 overexpression lead to increases in systolic blood pressure from very early gestation, proteinuria, renal capillary swelling, fibrin deposition, and elevated levels of sFlt-1 and sEng.^111,112 Early onset of hypertension suggests abnormal placentation may not be the cause of elevated blood pressure in this model. Nevertheless, the STOX1 overexpression model has been used to study many organ systems related to preeclampsia. Recently, studies have shown that STOX1 overexpression in mice results in increased renal artery resistance, cardiac hypertrophy, FGR and a trend towards increased umbilical resistance.^113,114 In a 2019 study, Miralles et al. showed that STOX1 overexpression in mice with a preeclampsia-like phenotype during pregnancy exhibited left ventricular hypertrophy, cardiac fibrosis and markers of inflammation and cellular stress up to eight months postpartum.¹¹⁵

Asb4 deletion

Another model that may be useful in studying vascularization of the placenta in the early stages of preeclampsia is the ubiquitin ligase Ankyrin repeat, SOCS box-containing 4 (ASB4), which promotes differentiation of vascular lineages in trophoblasts. Townley-Tilson et al showed that in placentas of Asb4^−/− mice, trophoblast-to-endothelial cell differentiation is impaired resulting in fewer mature endothelial cells and reduced placental vascularization.¹¹⁶ Consequently, these mice produced smaller litter sizes and developed hypertension and proteinuria later in gestation.^116,117

ELABELA deficiency

ELABELA, an endogenous ligand of the apelin receptor, has recently garnered much attention as a biomarker and therapeutic target of preeclampsia. Serum levels of ELABELA have been during early- and late-onset in a number of studies with varying results.^118,119 While some of these studies indicate that ELABELA may be increased in some cohorts, it is generally accepted that ELABELA plays a role in proper extravillous trophoblast migration and is deficient at least in the placentas of preeclamptic women.^120,121 Animal studies of ELABELA knockout in mice results in hypertension, proteinuria, impaired placental vascularization and fetal growth restriction.¹²² Interestingly, the investigators did not see the same results in apelin knockout mice. Furthermore, exogenous ELABELA attenuated these outcomes.¹²²

Complement C1q-deficiency

Complement activation has been shown to play a role in not only the clinical manifestation of preeclampsia but also spiral artery remodeling. Women with both early and late preeclampsia have decreased serum levels of complement component 1q (C1q) compared to those with uncomplicated pregnancies.¹²³ In a mouse model of C1q deficiency (C1q^−/−), animals develop hypertension, proteinuria, endotheliosis, decreased circulating VEGF and elevated sFlt-1.¹²⁴ These characteristics are accompanied by increased fetal death.¹²⁴

Dahl salt-sensitive rat

Clinically, preeclampsia can be classed as ‘superimposed’ when a patient is hypertensive prior to pregnancy. Similarly, the Dahl salt-sensitive (Dahl S) rat is hypertensive and develops a preeclampsia-like phenotype during pregnancy, including further elevations in blood pressure, proteinuria, glomerulomegly, increased uterine artery resistance, FGR, and elevated circulating levels of TNF-α and sFlt-1.¹²⁵ Further studies show postpartum renal injury in these animals despite no long-term blood pressure differences.¹²⁶ Interestingly, not all strains of hypertensive rats develop a preeclampsia-like phenotype during pregnancy. In contrast to the Dahl S rat, blood pressure in the spontaneously hypertensive rat (SHR) decreases towards late pregnancy. Utilizing this model Gillis et al. also demonstrated that sildenafil treatment ameliorates the maternal syndrome of preeclampsia and rescues fetal growth in the dahl salt-sensitive rat.¹²⁷

African green monkey

The African green monkey (Chlorocebus aethiops sabaeus) has been shown to develop hypertension during adulthood.¹²⁸ More recently Osbourne and colleagues showed that in non-hypertensive adults, some animals develop hypertension and fetal growth restriction during pregnancy similar to preeclampsia.¹²⁹

Many pathways in this disease have been targeted to produce animal models of preeclampsia (Table 2). A summary of the characteristics of each of the models discussed in this review are detailed in Table 3.

Table 2:

Summary of pathways involved in preeclampsia that are studied using animal models

Pathway	Animal model	References
Placental ischemia	Reduced uterine perfusion pressure	11,12
	Uteroplacental ischemia	48
Angiogenic imbalance	Adenovirus-sFlt-1 excess	53,57
	sFlt-1 infusion	55,59
Immune activation	T regulatory cell depletion	80
	TNF-α infusion	87
	Low-dose endotoxin infusion	92
	Low-dose lipopolysaccharide infusion	93
	Placenta-derived extracellular vesicle infusion	94
Renin-angiotensin system activation	Agnostic angiotensin II Type 1 receptor autoantibody infusion	71,72
	Angiotensinogen/renin overexpression	106,108
Endothelial dysfunction	L-NAME infusion	100
Abnormal placentation	STOX1 overexpression	111,112
	Asb4 deletion	116
	ELABELA deficiency	122
Complement system activation	C1q−/− deficiency	124
Superimposed or spontaneous models of preeclampsia	BPH/5 model	101
	Dahl Salt-sensitive rat	125
	African green monkey	129

Table 3:

Summary of features in animal models of preeclampsia during pregnancy

Model	Animal	Features	References
RUPP	Rat	↑MAP, ↑proteinuria, ↑sFlt-1, ↑sEng, ↓PlGF, ↓VEGF, ↑ROS, ↑AT1-AA, ↑TNF-α, ↑Tregs, ↑NK cells, ↑ET-1, ↓NO, ↓RPF, ↓GFR, ↑cardiac dysfunction, ↑cardaic hypertrophy, ↓CBF, ↑cerebral edema, ↑BBB permeability, ↓placental weight, ↑FGR	1,12,17-24,27-31,36,37,44,46
Uteroplacental ischemia	Non-human primate	↑MAP, ↑proteinuria, ↑sFlt-1,	48,49
Ad-sFlt-1-infusion	Rat	↑MAP, ↑proteinuria, ↑glomerular endotheliosis	53
sFlt-1-infusion	Rat	↑MAP, ↑proteinuria, ↑prepro-ET-1, ↑ROS, ↓placental weight, ↑FGR	55
T regulatory cell depletion	Mouse	↔MAP, ↑pro-inflammatory markers, ↑UARI, ↑fetal loss	80
AT-AA-infusion	Rat	↑MAP, ↑NK cells, ↑sFlt-1, ↑sEng, ↑prepro-ET-1, ↑AT1-AA	71,73,77
TNF-α-infusion	Rat	↑MAP, ↑prepro-ET-1, ↓NOS, ↑AT1-AA, ↓CBF	87,90,91
L-NAME-infusion	Rat	↑MAP, ↑proteinuria, ↑sFlt-1, ↓placental weight, ↑FGR	100
pcEV infusion	Mouse	↑MAP, ↑proteinuria, ↑kidney injury, ↓CBF	94
BPH/5	Mouse	↑MAP, ↑proteinuria, ↑endothelial dysfunction, ↑glomerulosclerosis, ↑fetal mortality, ↑UARI, ↔Flt-1, ↓VEGF, ↓PlGF, ↓ impaired cytotrophoblast invasion, ↓placental weight, ↑FGR	101-103
hAGTxhRen overexpression	Rat	↑MAP, ↑proteinuria, ↑cardiac hypertrophy	104,105
STOX1 overexpression	Mouse	↑systolic pressure, ↑proteinuria, ↑renal capillary swelling, ↑sFlt-1, ↑sEng, ↑cardiac hypertrophy, ↑FGR, ↑renal artery resistance	111-114
Complement C1q deficiency	Mouse	↑MAP, ↑proteinuria, ↑endotheliosis, ↑sFlt-1, ↓VEGF, ↑fetal death	124
Asb4 deletion	Mouse	↓placental vascularization, ↑MAP, ↑proteinuria, ↓litter size	116,117
ELABELA deficiency	Mouse	↓placental vascularization, ↑MAP, ↑proteinuria, ↑FGR	122
Dahl S rat	Rat	↑MAP, ↑proteinuria, αsFlt-1, glomerularomegly, ↑UARI, ↑FGR	125,127
African green monkey	Non-human primate	↑MAP, ↑FGR	128,129

RUPP, reduced uterine perfusion pressure; sEng, soluble endoglin; PlGF, placental growth factor; VEGF, vascular endothelial growth factor; ROS, reactive oxygen species; AT1-AA, angiotensin II type I receptor autoantibody; TNF-α, tumor necrosis factor-α; Tregs, T regulatory cells; NK, natural killer cells; ET-1, endothelin-1; NO, nitric oxide; RPF, renal plasma flow; GFR, glomerular filtration rate; CBF, cerebral blood flow; BBB, blood brain permeability; FGR, fetal growth restriction, hAGT, human angiotensinogen; hREN, human renin; ↑, increased; ↓, decreased; ↔, unchanged.

Use of animal models to study long-term consequences of preeclampsia

Since more recent studies have revealed the profound impact of preeclampsia later in life for both mother and baby, long term consequences have become an important consideration in developing animal models. While studies in this area are limited, some models have been shown to have persistent features beyond pregnancy. In some cases, such as in the RUPP, postpartum cardiac and renal dysfunction are present despite return to normal blood pressure in these animals.⁴⁵ These data suggest that placental ischemia and the factors involved cause irreversible damage and therefore do substantiate further studies in this area. Long-term studies where animals undergo multiple pregnancies may also be interesting, as cardiovascular risk in women later in life increases with each pregnancy complicated by preeclampsia.

Investigating therapeutic options for preeclampsia using animal models

Currently, treatment for preeclampsia is limited to managing symptoms and in severe cases premature delivery of the fetus, which poses significant risk to both mother and baby. Since initial studies in preeclamptic women are obviously impossible, animal models represent a critical tool in this area of study. While many studies have been performed in animal models, unfortunately limited therapies have reached clinical trials. Still, investigations in rodent and non-human primates continue in the search for a treatment for preeclampsia.

Angiogenic imbalance is an early predictor and central player in the development of preeclampsia, therefore sFlt-1, PlGF and VEGF remain key therapeutic targets. Administration of PlGF in RUPP rats and UPI non-human primates have showed reductions in sFlt-1, blood pressure and proteinuria.^10,60 Another approach that has been studied in non-human primates is administration of short interfering RNAs (siRNAs) that silence the three sFlt-1 mRNA isoforms that are responsible for sFlt-1 overexpression in the placenta. In pregnant baboons, UPI was induced and the human siRNA mixture (hsiRNA^sFLT1-2283/2519) was administered on gestational day 133.⁴⁹ Data was collected at 4-6 week intervals and showed significant decreases in circulating sFLt-1 levels, as well as reductions in systolic blood pressure and proteinuria (Figure 4). Thadhani et al are currently investigating clearance of sFlt-1 (via apheresis) to improve angiogenic balance in women with preeclampsia. These studies show transient decrease in mean arterial pressure, following apheresis and prolonged gestation by up to 15 days ^62,63

Figure 4:

Circulating levels of sFlt-1 (A), systolic blood pressure (B) and proteinuria (C) in the uteroplacental ischemia (UPI) non-human primate model of preeclampsia, following a single dose of human siRNA that silence 3 sFlt-1 isoforms (hsiRNA^sFLT1-2283/2519). These 3 isoforms are responsible for placental overexpression of sFlt-1 but do not reduce full-length Flt-1 mRNA.⁴⁹ Data are mean+SEM. ***P<0.001.

Preeclampsia in women is associated with increases in vascular expression of ET-1 and endothelial activation²¹ In addition, a number of experimental models of preeclampsia are also associated with elevated tissue levels of prepro-ET-1 mRNA. These models have been used to determine whether blockade of the endothelin system could improve hypertension. Interestingly, in the RUPP model, sFlt-1 infusion, TNF-α infusion, and AT1-AA infusion, administration of the ET type A receptor antagonist reduces mean arterial pressure^13,55,71,88 (Figure 3). Results from these studies suggest that ET-1 may be a final common pathway whereby placental factors act on the maternal vasculature to cause vasoconstriction and hypertension. Another avenue to improve endothelial function is by stimulation of the vasodilatory nitric oxide (NO)-soluble guanylate cyclase (sGC)-cyclic guanosine monophosphate (cGMP) pathway. Compounds that promote NO production or blockade of cGMP degradation to increase activity of this pathway are two methods that have been tested in animal models ^{100,127,130-134} and reached clinical trials. ^135-141 In animal models, administration of a phosphodiesterase type 5 inhibitor (Sildenafil) can result in reduced blood pressure, increased fetal weight, decreased uterine artery resistance and improved angiogenic balance.^100,127,133. However, clinical studies showed little to no improvement in preeclamptic women.^140,141

Figure 3:

Mean arterial pressure (MAP) in animal models of preeclampsia, with and without administration of an endothelin Type A antagonist. Data is shown for studies in the reduced uterine perfusion pressure (RUPP)¹³, sFlt-1 infusion⁵⁵, angiotensin type 1 receptor autoantibody (AT1-AA) infusion⁷¹ and tumor necrosis factor-α (TNF-α) infusion⁸⁸ models, suggesting endothelial activation may be a common final pathway in the preeclampsia-related hypertesion. Data are mean+SEM. *P<0.05.

In addition to these studies to directly block culprit pathways in preeclampsia, a number of vitamins and drug have been tested in animal models, including Vitamin D, Vitamin B and statins. In the RUPP model, Vitamin D administration reduced blood pressure, ET-1, sFlt-1 and AT1-AA, but did not improve fetal outcomes.^142-144 In the L-NAME model, Vitamin D reduced sFlt-1 and TNF-α.¹⁴⁵ Administration of the cholesterol-lowering drug, pravastatin has been shown to improve placental bloodflow and weight and longtern cardiovascular outcomes in the C1q^−/− mouse.¹⁴⁶ In the RUPP rat, pravastatin treatment results in reduced blood pressure, improved angiogenic balance and reduced reactive oxygen species.¹⁴⁷ While clinical studies have not shown comparable results, no adverse effects have been reported in women. ^148-150 These avenues continue to be researched.

Conclusions

Preeclampsia is a complex, multiorgan disease during pregnancy. The discovery of pathophysiological processes involved in preeclampsia has resulted from the interplay between basic research involving animal models and clinical research in humans. A number of animal models have been developed to address the many pathways involved in preeclampsia (Figure 4). A major consideration for any model of preeclampsia is that the animal is manipulated, whether it be surgically, pharmacologically or genetically, to express these features. Moreover, some of the animal models discussed above focus on a single characteristic or mediator in the development of preeclampsia. While these preclinical models have been crucial in understanding the pathophysiological importance of individual factors, it is important to remember that preeclampsia is a multi-organ, multi-faceted disease that first develops as a result of impaired spiral artery remodeling and placental development. Some of the models discussed above such as the Asb4 deletion model and the Dahl S rat could be useful in studying these very early stages of preeclampsia. Regardless of the mechanism, each of these animal models has been critical in understanding the two phases of preeclampsia; 1) impaired extravillous trophoblast migration and invasion of the endometrium causing placenta ischemia, and 2) the release of factors from the ischemic placenta into the maternal circulation leading to endothelial dysfunction and the clinical syndrome. Preclinical studies in animal models have been instrumental in not only understanding the pathophysiology of preeclampsia but also in the search for novel therapeutic options for the treatment of this disease.

Figure 5:

Pathways in the pathogenesis of preeclampsia that have been shown and studied in animal models. sFlt-1, solube Flt-1; sEng, soluble endoglin; VEGF, vascular endothelial growth factor; PlGF, placental growth factor; ROS, reactive oxygen species; AT1-AA, angiotensin II Type 1 receptor autoantibody; Th, T helper cell; Tregs, T regulatory cells; IL, interleukin; TNF-α, tumor necrosis factor-α; ET, endothelin; NO, nitric oxide; FGR, fetal growth restriction.