Matrix metalloproteinases comprise a family of zinc-dependent proteases that degrade extracellular matrices. Specifically, matrix metalloproteinase-9 (MMP-9) degrades type IV, V, and IX collagens, gelatin, and elastin and is involved in numerous processes, including implantation, placentation, and embryogenesis (Vu et al., 1998; Silvia and Serakides, 2016; Espino et al., 2017; Quintero-Fabián et al., 2019; Timokhina et al., 2020). During pregnancy, MMP-9 has been shown to be involved in endometrial remodeling and the invasion of placental extravillous trophoblasts (Zhang et al., 2020). Conversely, decreased levels of MMP-9 are associated with impaired trophoblast invasion (Staun-Ram et al., 2004; Chen and Khalil, 2017).

The data demonstrating an association of MMP-9 in human preeclampsia are variable. Some reports have shown reduced levels of circulating MMP-9 in maternal plasma from preeclamptic pregnancies; however, other studies have found elevated levels or no significant differences (Montagnana et al., 2009; Plaks et al., 2013; Eleuterio et al., 2015; Laskowska, 2017; Timokhina et al., 2020). In a study looking at the association of MMP-9 in severe preeclampsia, mild preeclampsia, and normal pregnancies; MMP-9 expression was reduced in severely preeclamptic patients, but was not different between mild preeclampsia and normal pregnancies (Wang et al., 2015; Zhang et al., 2019). Some of these discrepancies may be attributable to MMP-9 levels in pregnancy, as MMP-9 have been shown to be positively correlated with gestational age (Montagnana et al., 2009).

Dubois et al. initially reported impaired reproduction in MMP-9 null mice. In these mice, MMP-9 deficiency was associated with a decrease in the number of pregnancies as well as a reduction in litter size (Dubois et al., 1999; Dubois et al., 2000). Plaks and Rinkenburger also investigated the role of MMP-9 in pregnancy using MMP-9 deficient mice and found these mice are sub-fertile with decreased implantation and increased fetal demise (Plaks et al., 2013). In pure or mixed backgrounds with homozygous MMP-9 null crosses (129SV/J on CD1 or Swiss black backgrounds), there was up to a 20% reduction in litter size; however, on a C57BL/6J mouse background, MMP-9 null homozygous crosses exhibited a 50% reduction in litter size (Plaks et al., 2013). The variation in litter size among different strains of mice suggests that genetic background may be a contributing factor in the varying effects observed.

When embryos from MMP-9 null homozygous crosses were analyzed at E10.5, 18.4% exhibited fetal growth restriction when compared to heterozygous control. Additionally, of the embryos that were growth restricted, it was reported that they were in a “twisted” and “constrained” state. Homozygous MMP-9 knockout mice had reduced and malformed ectoplacental cones, surrounded by blood pools at E7.5, impaired trophoblast differentiation, and reduced invasion (Plaks et al., 2013). The MMP-9 null placentas exhibited altered morphology with an increased number of trophoblast giant cells and diminished spongiotrophoblast and labyrinth layers (Plaks et al., 2013). Interestingly, it was reported that normal placental development required both maternal and fetal MMP-9 expression (Plaks et al., 2013).

When compared to non-pregnant controls, non-pregnant homozygous MMP-9 knockout mice were reported to have an elevated mean systolic blood pressure (Plaks et al., 2013). Importantly, pregnant homozygous MMP-9 knockout mice with viable fetuses exhibited a prolonged decrease in blood pressure as gestation progressed, compared to controls (Plaks et al., 2013). Kidneys from pregnant and non-pregnant MMP-9 null homozygous mice were also assessed. During pregnancy, the percentage of “open” glomerular capillaries in MMP-9 null mice was significantly reduced and proteinuria was present, which may indicate a pathology similar to glomerular endotheliosis (Plaks et al., 2013). While MMP-9 knockout mice have preexisting hypertension, the use of these mice as a comorbid model that develops preeclampsia-like symptoms requires consideration due to the uncharacteristic lack of elevated blood pressure during pregnancy.

The renin-angiotensin-aldosterone system (RAAS) maintains blood pressure by regulating sodium, potassium, and fluid volume. Dysregulation of the renin-angiotensin-aldosterone system leads to the development of chronic hypertension (Santos et al., 2012; Carey, 2015; Te Riet et al., 2015). Activation of renin-angiotensin-aldosterone system occurs when the precursor, prorenin, is converted to renin by the juxtaglomerular cells of the kidney (Hsueh & Baxter, 1991; von Lutterotti et al., 1994). Renin proteolytically cleaves angiotensinogen and subsequent processing results in the generation of angiotensin I and II. Angiotensin II, the predominant end-product of the renin-angiotensin-aldosterone system, binds to the angiotensin II type 1 receptor 1 (AT1) and results in the activation of downstream signaling pathways that ultimately lead to vasoconstriction and increased retention of sodium and water by the kidney (Spaan & Brown, 2012). Furthermore, angiotensin II stimulates release of aldosterone from the adrenal glands to promote blood volume expansion. The physiological response to vasoconstriction, sodium reabsorption, and volume expansion is increased blood pressure (Spaan & Brown, 2012).

In a normal healthy pregnancy, the levels of renin, angiotensinogen, angiotensin I and II, and aldosterone are elevated compared to non-pregnant women, but a vasodilatory state is present such that hypertension does not typically occur (Brown et al., 1963; Brown et al., 1997). The increased level of progesterone during pregnancy decreases the sensitivity to angiotensin II and changes the state of AT1 receptor binding so that twice as much angiotensin II is required to elicit an elevation in blood pressure that would typically occur while not pregnant (Assali & Westersten, 1961; Gant et al., 1973; AbdAlla et al., 2001). Thus, increased levels of angiotensin II are required during pregnancy to maintain a normotensive blood pressure (Irani & Xia, 2008). In patients with preeclampsia; however, the maternal levels of plasma renin, angiotensin II, and aldosterone have been shown to be lower than normotensive pregnancies (Weir et al., 1973; Gant et al., 1980; Brown et al., 1997; Irani and Xia, 2008; Leaños-Miranda et al., 2018).

Falcao et al., sought to investigate the occurrence of preeclampsia-like features in a hypertensive mouse model during pregnancy using double transgenic mice that overexpress the genes for human renin (REN, R+) and human angiotensinogen (AGT, A+) (R+A+ mice) (Sigmund et al., 1992; Falcao et al., 2009). Non-pregnant R+A+ mice have significantly elevated angiotensin II and mean arterial pressure compared to non-transgenic, non-pregnant control mice (Falcao et al., 2009). Pregnant R+A+ mice had significantly elevated blood pressure above mean arterial pressure levels on gestational days 5 and 17, compared to non-pregnant R+A+ mice (Falcao et al., 2009). The mean arterial pressure at gestational day 18 for R+A+ mice was significantly higher than non-transgenic mice, indicating that in R+A+ mice, pregnancy further exacerbates their hypertensive state (Falcao et al., 2009). The R+A+ mice MAP decreased 24 h after birth but did not completely return to prepregnant levels.

Pregnant R+A+ mice developed proteinuria by the end of gestation but did not have signs of glomeruloendotheliosis or other renal pathology (Falcao et al., 2009). In addition, pregnant and non-pregnant R+A+ mice exhibited cardiac hypertrophy. Placental pathology included increased necrosis and loss of labyrinthine trophoblast structure. Although no differences in litter size were observed, fetal and placental weights were both significantly reduced, compared to non-transgenic mice.

The Falcao et al. model is similar to one generated by Takimoto et al., who analyzed AGT overexpressing female mice crossed with REN overexpressing male mice (Takimoto et al., 1996). Takimoto reported that female AGT mice became hypertensive after day 14 of gestation and had glomerular enlargement and increased urinary protein, coinciding with the development of the placental renin overexpression. This indicates that these transgenic mice exhibit a pregnancy-specific increase in blood pressure, a hallmark of preeclampsia. These pregnant mice also developed myocardial concentric hypertrophy with only 38% surviving pregnancy (Takimoto et al., 1996). In addition, 15% of these transgenic hypertensive mice had generalized convulsions late in pregnancy. Placental analysis indicated necrotic cell death in the spongiotrophoblasts and decidual cells and chorionic congestion (Takimoto et al., 1996).

Both models show that the excess secretion of renin and other renin-angiotensin-aldosterone system proteins or angiotensin-like peptides by the placenta can lead to the development of preeclampsia-like symptoms during mouse pregnancy (Takimoto et al., 1996; Shah et al., 2000; Falcao et al., 2009; Denney et al., 2017). The paradox between overexpression of renin and angiotensinogen creating preeclampsia-like symptoms in mice and the reduced levels observed in humans with preeclampsia suggests that renin-angiotensin-aldosterone system may be a downstream effect of other dysregulated systems. Additionally, most currently available therapeutic inhibitors of renin-angiotensin-aldosterone system are teratogenic and fetotoxic, precluding their use to treat preeclampsia (Alwan et al., 2005; Alwasel et al., 2010; Ferreira et al., 2010).

Nitric oxide synthases [NOS I (nNOS), NOS II (iNOS), and NOS III (eNOS)] are a family of enzymes that create nitric oxide (NO) via the reduction of L-arginine to L-citrulline (Moncada and Higgs, 1993; Hefler et al., 2001; Förstermann and Sessa, 2012; Qian and Fulton, 2013). In the endothelium, the primary function of NO is to relax vascular smooth muscle tissue and it serves as an important regulator of arterial blood pressure (Shesely et al., 2001; Endres et al., 2004; Chen and Zheng, 2014; Guerby et al., 2021). Nitric oxide and the NOS enzymes play an important role in cardiovascular remodeling during development and are involved in the synthesis of vascular endothelial growth factor (VEGF), stimulation of endothelial progenitor cell activity, and angiogenesis via hypoxia inducible factor 1 alpha (HIF-1α) (Shesely et al., 2001; Endres et al., 2004; Chen and Zheng, 2014; Guerby et al., 2021). In a healthy normotensive pregnancy, nitric oxide levels spike early in gestation and gradually increase through the third trimester (Owusu Darkwa et al., 2018). In contrast, the levels of nitric oxide reported in preeclamptic women are conflicting; with reports showing increased, decreased, or no differences, compared to control (Seligman et al., 1994; Lyall et al., 1995; Smárason et al., 1997; Choi et al., 2002; Marshall et al., 2018).

Several groups have investigated the effect of eNOS knockout on blood pressure and the generation of preeclampsia-like symptoms in mice, following reports that chronic non-specific pharmacologic inhibition of NOS recapitulates some symptoms of preeclampsia (Baylis & Engels, 1992; Yallampalli & Garfield, 1993; Shesely et al., 1996; Hefler et al., 2001; Shesely et al., 2001; Kusinski et al., 2012). Three independent studies have reported that non-pregnant eNOS knockout mice have elevated blood pressure compared to controls (Huang et al., 1995; Shesely et al., 1996; Kusinski et al., 2012).

Hefler et al. examined pregnant eNOS homozygous knockout mice and reported reduced fetal weights; however, no placental abnormalities were noted. In that study, blood pressure and proteinuria were not reported, but severe limb abnormalities were identified (Hefler et al., 2001). Shesely et al., found that non-pregnant homozygous eNOS knockout mice had significantly increased blood pressure compared to control mice; however, blood pressure did not significantly increase during pregnancy (Shesely et al., 1996; Shesely et al., 2001). Similarly, Kusinski et al., observed that eNOS knockout mice had elevated blood pressure before and during pregnancy at gestational day 17.5, when compared to wild type mice (Kusinski et al., 2012). However, there was no significant increase in blood pressure between pregnant eNOS knockout mice and non-pregnant eNOS knockout mice (Kusinski et al., 2012). Investigations into placental structure determined that spiral artery remodeling was dysregulated, the labyrinth zone was reduced, uteroplacental hypoxia was present, and placental nutrient transport was reduced in eNOS knockout mice (Kulandavelu et al., 2012; Kusinski et al., 2012; Kulandavelu et al., 2013). Levels of VEGF were reduced and HIF-1α protein was increased in eNOS knockout placentas; whereas, levels of souble fms-like tyrosine kinase-1 (sFLT-1) in maternal plasma were not different between eNOS knockout and control (Kulandavelu et al., 2012; Kusinski et al., 2012; Kulandavelu et al., 2013).

Litter size was also reduced in eNOS knockout compared to control mice (Kulandavelu et al., 2013). Additionally, maternal and fetal weights were reduced in eNOS knockout mice compared to control at gestational day 17.5, but placental weights were not different (Kulandavelu et al., 2012; Kusinski et al., 2012; Kulandavelu et al., 2013). eNOS null mice have consistently been reported to be small and could serve as a model of fetal growth restriction; however, these mice have not been shown to exhibit a significant elevation in blood pressure during gestation that is characteristic of the preeclamptic condition (Medica et al., 2007; McCarthy et al., 2011; Kulandavelu et al., 2012; Kusinski et al., 2012; Qi et al., 2013; Alpoim et al., 2014).

Non-obese diabetic (NOD) mice are an inbred strain characterized by the spontaneous development of autoimmunity and Type 1 diabetes mellitus. These mice have a mutation that results in depletion of regulatory T-cells and leads to the death of the pancreatic islet beta cells (D’Alise et al., 2008). NOD mice are widely used to study the pathophysiology of Type 1 diabetes mellitus, as they demonstrate autoimmune cell infiltration into pancreatic islet beta cells characteristic of the disease (Chen et al., 2018). Notably, not all non-obese diabetic mice will develop hyperglycemia in their lifetime, as the degree of pancreatic immune cell infiltration determines whether NOD mice will develop diabetes (Chen et al., 2018). Non-obese diabetic mice with mild pancreatic autoimmune cell infiltration can maintain normal blood glucose levels and never progress to a Type 1 diabetic phenotype; whereas, non-obese diabetic mice that have extensive insulitis develop a phenotype more severe than what occurs in the human condition. Nevertheless, the NOD mouse model most closely mirrors the spontaneous onset of Type 1 diabetes mellitus (Burke et al., 2011; Chen et al., 2018).

Pregnant non-obese diabetic mice are reported to have significantly increased proteinuria as gestation progresses with accompanying renal histopathology indicative of acute kidney injury (Burke et al., 2011). Pregnant non-obese diabetic mice also exhibit progressive bradycardia and reduced blood pressure from gestational day 10 through gestational day 18, compared to pregnant non-obese nondiabetic mice (Burke et al., 2011). The placentas in pregnant non-obese diabetic mice demonstrated impaired spiral artery remodeling, as a reduction in the number of spiral arteries and decreased luminal diameters (Burke et al., 2007; Burke et al., 2011). Placental weights from non-obese diabetic pregnancies were significantly increased; whereas, pup weights at birth were significantly lower compared to non-diabetic controls (Burke et al., 2007). Pregnant non-obese diabetic mice do exhibit kidney dysfunction and proteinuria in the diabetic animals, compared to the non-diabetic non-obese diabetic mice; however, the variability in the generation of a diabetic phenotype and unexpected cardiac features indicate further work is needed to model this comorbidity. Additionally, the decrease in blood pressure in pregnant non-obese diabetic mice is not representative of the preeclamptic condition in humans with pregestational Type 1 diabetes mellitus (Burke et al., 2011; Shub, 2020; Shub and Lappas, 2020).

Obesity is typically studied by feeding animals a diet high in fat to increase body weight (Christians et al., 2019). Masuyama and Hiramatsu studied the effect of obesity on pregnancy in adult, eight-week-old ICR mice after being fed a high-fat diet (HFD) consisting of 62% of calories from fat for 4 weeks (Masuyama and Hiramatsu, 2012). The prepregnancy weights of female ICR mice fed the high-fat diet were not reported; however, at the end of gestation, pregnant female high-fat diet mice were on average 16 g heavier than those on the control diet (Masuyama and Hiramatsu, 2012; Masuyama et al., 2016). High-fat diet pregnant mice exhibited increased insulin resistance with poor glucose tolerance, as well as increased serum levels of triglycerides and leptin, but had decreased levels of adiponectin (Masuyama and Hiramatsu, 2012; Masuyama et al., 2016).

Pregnant high-fat diet mice developed significantly elevated blood pressure at E18.5 accompanied by proteinuria (Masuyama and Hiramatsu, 2012). Also present was a significant increase in fetal weight, but no change in placental weight or litter size was observed (Masuyama and Hiramatsu, 2012; Masuyama et al., 2016). Placental morphology, lineage, immune, or angiogenic markers were not assessed in these studies (Masuyama and Hiramatsu, 2012). However, placental morphology was reported to be altered in a different study using pregnant obese mice (45% high-fat diet), where high-fat diet was associated with decreased labyrinth thickness, compared to the control diet (Kim et al., 2014).

A standard way to induce obesity in mice is to use a diet high in fat content (i.e., ∼60% of calories from fat). It should be noted; however, that this percentage of calories to induce obesity is not generally representative of typical Western diets, which is closer to ∼35–45% fat (Hintze et al., 2018). Also, not all strains of mice develop obesity on a high-fat diet, which suggests that an underlying genetic predisposition for obesity may be present in certain strains (Nishikawa et al., 2007; Huang et al., 2020; Li et al., 2020). Overall, the high-fat diet mouse model exhibits some of the classical hallmarks of preeclampsia, but further studies are required to better define inflammatory and angiogenic factors in these obese mice that develop preeclampsia-like symptoms.

Work by Davisson et al., led to the identification of a mouse line, Blood Pressure High 5 (BPH/5), that is borderline hypertensive throughout adult life and spontaneously develops the hallmarks of preclampsia during pregnancy (Davisson et al., 2002). BPH/5 mice are a substantially inbred subline, derived from >20 generations of brother-sister matings of the more well-known, hypertensive BPH/2 strain (Schlager and Sides, 1997; Davisson et al., 2002). BPH/5 mice have significantly increased blood pressure prior to pregnancy, compared to C57BL/6 mice, as well as a significantly increased mean arterial pressure during pregnancy, beginning on gestational day 14 through parturition. The blood pressure of BPH/5 returned to the baseline “borderline hypertensive” levels after birth (Davisson et al., 2002). Endothelial dysfunction was also noted in pregnant BPH/5 mice and proteinuria as well as glomerulosclerosis were observed. In addition, fetal weights were significantly reduced and litter sizes were notably smaller (Davisson et al., 2002).

Analysis of BPH/5 pregnancies revealed that placental weights of BPH/5 placentas were similar to controls in late gestation (Dokras et al., 2006). All placental lineages were present; however, disruption of placental structure was noted, as trophoblast layers were disorganized and the expression of placental lineage markers were significantly reduced at E14.5 (Dokras et al., 2006). Vascular pathologies were also present in BPH/5 placentas, as blood spaces and branching morphogenesis were reduced and decidual vessels were characterized by thickened vessel walls, narrowed lumens, and increased uterine artery resistance (Dokras et al., 2006). Decreased VEGF and placental growth factor (PlGF) mRNA were observed in BPH/5 placentas at E10.5, while sFLT-1 mRNA was significantly increased, consistent with preeclampsia symptoms (Sones et al., 2018).

The BPH/5 mouse model presents with borderline hypertension and recapitulates several hallmarks of preeclampsia during pregnancy (Sones and Davisson, 2016). Disruption of normal placental development and smaller fetal weight at birth suggests that this model may be similar to early-onset preeclampsia with fetal growth restriction (Waker et al., 2021). Recent studies; however, have changed the context of the BPH/5 model and expanded the comorbidity beyond spontaneous chronic borderline hypertension and to now include preexisting metabolic disease, obesity, and fatty liver disease (Sutton et al., 2017; Reijnders et al., 2019; Johnston et al., 2021; Sones et al., 2021).

In addition, the reproductive axis in these mice is disrupted as the BPH/5 mice have reduced serum 17β-estradiol during estrous (Sutton et al., 2017). During pregnancy these mice continue to exhibit increased body weight and white adipose tissue with elevated cholesterol and triglyceride levels compared to C57BL/6 control mice, indicative of obesity (Reijnders et al., 2019). Caloric restriction ameliorated some of these pathologies and suggests that the BPH/5 phenotype might be linked to the metabolic disease, in addition to being chronically borderline hypertensive (Reijnders et al., 2019; Olson et al., 2020). The conjunction of these multiple comorbidities may make it challenging to determine the individual effects of chronic hypertension or obesity on the generation of preeclampsia. However, women that are both obese and hypertensive are a segment of the population at high risk of developing the condition and a mouse model that exhibits these preexisting conditions may be useful to study the interaction and impact of both comorbidities on the development of preeclampsia.