The Role of Mitochondrial Dysfunction in Preeclampsia: Causative Factor or Collateral Damage?

Abstract

Preeclampsia, new onset hypertension in pregnancy, affects ~5%–10% of the world’s population. Preeclampsia is the leading cause of morbidity and mortality for both the mother and fetus. As of today, there is no cure for this disease except for delivery of the fetal–placental unit. The exact causation and onset of the disease are unknown. However, recent studies have shown a strong correlation between mitochondrial dysfunction and preeclampsia. Circulating mitochondrial DNA, elevated reactive oxygen species, angiotensin II type 1 receptor agonistic autoantibodies (AT1-AA), activated natural killer cells, and upregulated inflammatory responses all contribute to mitochondrial dysfunction and the pathophysiology of preeclampsia. This review summarizes the current literature of both experimental and clinical observations that support the hypothesis that mitochondrial dysfunction contributes to the pathophysiology of preeclampsia and may be a precursor to the disease onset. This review will also address the use of therapies to improve mitochondrial dysfunction in preeclampsia.

OVERVIEW

Normal pregnancy

There are several unique maternal adaptions that must occur in order to have a successful normal pregnancy. These adaptions include an increase in cardiac output, glomerular filtration rate and metabolic demand, and a decrease in blood pressure.¹ The increase in cardiac output is caused by an elevation in heart rate and stroke volume.¹ The stroke volume is elevated due to plasma volume expansion via rises in sodium and water reabsorption in the kidney.¹ Despite the increase in cardiac output, there is a decrease in blood pressure because total peripheral resistance is reduced by elevated nitric oxide production that causes systemic vasodilation.^2,3 A lack of changes in these maternal adaptions during pregnancy is evident in complicated pregnancies, such as preeclampsia.^1–3

Normal pregnancy is a state of elevated inflammation and oxidative stress, in which women have elevated oxidized low-density lipoprotein (oxLDL), urinary isoprostanes, urinary 8-hydroxydeoxyguanosine, and decreased total antioxidant capacity, along with an increase in proinflammatory cytokines and cells compared with nonpregnant women.^4–11 Although there is a surge in inflammation and oxidative stress in normal pregnant women, there is a greater increase in women with preeclampsia and other gestational diseases.¹¹ Even though there is a dramatic increase in metabolic demand during pregnancy, little is reported in the literature about changes in mitochondrial function and morphology in a normal pregnancy.

Throughout pregnancy the placenta grows and changes structurally to meet the changing metabolic demands of transporting nutrients and blood between the maternal and fetal systems. To meet these increasing energy demands, placental mitochondria increase respiration (mitochondrial activity), and content throughout gestation.¹² No changes in cardiac mitochondrial structure, amount, or function are recorded during pregnancy, along with mitochondria in other cells and tissues of the body.¹³

Preeclampsia

Approximately 5%–10% of pregnancies worldwide are affected by preeclampsia, which is the leading cause of maternal and perinatal death.^14,15 Preeclampsia is defined as new onset hypertension during pregnancy that usually occurs late in pregnancy.¹⁵ The common clinical symptoms of preeclampsia are hypertension (blood pressure ≥140/90 mm Hg), proteinuria (24-hour protein urine concentration of ≥300 mg/dl), and/or end-organ damage, especially to the liver and the kidneys.^15–17 Although proteinuria and end-organ damage are a hallmark of preeclampsia and are often associated with preeclampsia, they are not necessary to be present for the diagnosis of preeclampsia.^15–17

In 2017, the American College of Cardiology and American Heart Association (ACC/AHA) changed the guidelines to diagnose all hypertension as a systolic blood pressure >130 mm Hg and diastolic blood pressure >80 mm Hg.¹⁸ While this change in the blood pressure threshold will increase the number of women diagnosed with preeclampsia, and perhaps begin care and treatment of women with preeclampsia earlier and at lower blood pressures to improve maternal and fetal outcomes, there is no evidence that care and treatment at lower blood pressures in preeclamptic women is beneficial.¹⁸ The American College of Obstetricians and Gynecologists (ACOG) recommends treatment only for severe hypertension in pregnancy, as there is no evidence that treatment of mild hypertension is beneficial to the mother and fetus.¹⁸ Thus, the current diagnosis set by the ACOG for preeclampsia is still a blood pressure ≥140/90 mm Hg.

If hypertension is presented before 34 weeks of gestation it is considered early-onset preeclampsia, which is normally more severe, as there is a greater increase in blood pressure, risk of end-organ damage, death, and eclampsia (which is the onset of seizures that can occur in women with preeclampsia).^15–17,19 On the other hand, if hypertension occurs after 34 weeks of gestation, it is referred to as late-onset preeclampsia, which consists of the majority of preeclampsia cases.^15–17,19 The rate of disease progression and array of other symptoms that are associated with preeclampsia can vary from patient to patient.²⁰

The exact cause of preeclampsia is unknown. However, the etiology of preeclampsia is hypothesized to be caused by apoptosis of trophoblast cells and inability of trophoblast cells to invade into the maternal decidua, leading to deficient uterine spiral artery remodeling for proper placentation.¹⁵ Deficient spiral artery remodeling will cause insufficient blood supply through the placenta to the fetus resulting in placental ischemia and dysfunction. The results from placental dysfunction are fetal growth restriction, as well as the increased risk of mortality and morbidity to the mother and fetus.²¹ Along with placental ischemia, reactive oxygen species (ROS) formation, inflammation via natural killer (NK) cells activation, endothelial dysfunction, angiotensin II type 1 receptor agonistic autoantibody (AT1-AA) production, mitochondria ROS (mtROS), and mitochondrial dysfunction are present in women with preeclampsia and play a role in its pathophysiology and possible pathogenesis.^15,22–27

The goal of this review is to provide a summary of the evidence linking mitochondrial dysfunction to the pathogenesis and pathology of preeclampsia (Figure 1). This review will discuss (i) mitochondrial dysfunction, (ii) clinical observations of mitochondrial dysfunction in preeclampsia, (iii) animal models of preeclampsia with mitochondrial dysfunction, and (iv) the role of oxidative stress, inflammation, and NK cell activation to cause mitochondrial dysfunction in pregnancy. Furthermore, we will examine the possibility of targeting mitochondrial dysfunction and oxidative stress as a therapeutic approach to preeclampsia.

Figure 1.

Scheme showing the role of several circulating factors (AT1-AA, homocysteine, sFlt-1, endothelin 1, and inflammatory cytokines) associated with preeclampsia and mitochondrial dysfunction, ROS, and DAMPs (N-formyl peptides and mtDNA) in preeclampsia via activation of TLRs and N-formyl peptide receptors to damage trophoblast cells, increase NK cell cytolytic activity, and cause endothelial dysfunction. See text for additional details. Some images are adapted from Servier Medical Art (https://smart.servier.com/). Abbreviations: DAMPs, damage-associated molecular patterns; ETC, electron transport chain; mtDNA, mitochondrial DNA; NK cells, natural killer cells; ROS, reactive oxygen species; sFlt-1, soluble fms-like tyrosine kinase-1; Th1/2, T-helper cell 1/2; TLR, toll-like receptor.

MITOCHONDRIAL DYSFUNCTION

Mitochondria are a double membrane organelles located within cells that are chiefly responsible for adenosine-5′-triphosphate (ATP) formation and fatty acid oxidation for cellular energy.^28–30 Mitochondria aid in steroid synthesis, cell signaling, apoptosis, oxidative stress production, and regulating cellular responses to environmental stress.^28–30 ATP synthesis is coupled with oxidative phosphorylation (OXPHOS) via the transfer of electrons through multimeric complexes, namely complexes I, II, III, and IV, which form the electron transport chain (ETC) on the inner mitochondrial membrane.²⁹ Mitochondrial dysfunction occurs when there is a loss and/or reduction of mitochondria amount, cellular respiration, efficiency of the ETC, electrical and chemical transmembrane potential of the inner mitochondrial membrane, and transport of metabolites to the mitochondria for ATP production.^28,30 Dysfunctional mitochondria are a major source for ROS production, which in normal physiological settings plays a role in cellular signaling. However, elevated levels of mtROS will cause damage and death to cells, endothelial cell dysfunction, and damage to tissues leading to end-organ damage. Mitochondrial dysfunction and mtROS are correlated and a part of the pathogenesis of aging and several diseases, such as atherosclerosis, hypertension, inflammation, heart disease, and preeclampsia.^19,29,30

CLINICAL OBSERVATIONS OF MITOCHONDRIAL DYSFUNCTION IN PREECLAMPSIA

In the early 1990s, researchers began noting that mitochondria of the uterine wall or placenta from preeclamptic patients displayed morphological and physiological changes in their structure and function compared with normal pregnant patients.^31–33 These alterations included increases in mitochondria size, cristae spacing, ETC efficiency, and oxygen consumption rate. Within the mitochondria ETC efficiency, cytochrome c oxidase activity is decreased in the preeclamptic placenta.^34–36 This is the last enzyme in the ETC cascade that passes electrons to oxygen to produce water and aids in the generation of the proton gradient for cellular respiration and energy supply. In preeclampsia, small, unidentifiable microvesicles were located within the proximity of the mitochondria, perhaps indicative of a metabolic disorder or stress.³² Moreover, increases in circulating serum levels of mitochondrial DNA (mtDNA) and cytochrome c were noted from blood samples obtained from mothers who developed preeclampsia.³³

In one study, cytotrophoblast cells of the placenta from preeclamptic patients displayed vacuolated mitochondria, which are mitochondria that appear spherical and fluid filled under the electron microscope, with loss of cristae.³⁷ Mitochondria vacuolization is a sign of mitochondrial swelling and cell injury. Often mitochondrial swelling is the result of increased oxidative stress damage to lipids in the mitochondrial membrane that causes further generation of ROS, damage to the ETC, and uncoupling of the mitochondria to decrease its activity and function.^38–40 Dysfunctional mitochondria decrease the cell’s bioenergetic capacity and lead to cell death.⁴¹ To counteract this decline in bioenergetic capacity, mitochondria go through the processes of fusion (to rescue nonfunctional mitochondria) and/or fission (to eliminate damaged mitochondria) to maintain mitochondrial quality.⁴¹

Impaired mitochondrial fusion, along with autophagy and lipid metabolism deficiencies have been associated with preeclampsia.^42,43 Zhou et al. examined several metabolic factors that influence mitochondrial function in normal and preeclamptic placentas after birth.⁴² ATP production and fatty acid catabolism were decreased, along with an elevation of fatty acids in preeclamptic placentas vs. normal pregnant placentas.⁴² These 2 observations indicate that beta oxidative pathways within mitochondria of preeclamptic placentas are compromised due to mitochondrial damage and/or insufficiency.

Zhou et al. also demonstrated that levels of mitochondrial fusion proteins MFN1, MFN2, and OPA1 were downregulated in the placenta of preeclamptic women, indicating that mitochondria in these placentas are unable to fuse to accommodate stress and energy demands.⁴² A decrease in placental OPA1 expression was also observed in another study with increased levels and activation of DRP1.⁴³ DRP1 is an important mitochondrial fission protein that plays a central role in discarding and targeting unhealthy and nonfunctional mitochondria for degradation.⁴³

Interestingly, in the previous study by Zhou et al., BNIP3, a regulator of mitochondrial autophagy, was decreased in preeclamptic placentas.⁴² Mitochondrial autophagy is a natural process that assures mitochondria are functioning at their highest capacity.^44,45 A decrease in mitochondrial autophagy with large amounts of nonfunctioning mitochondria could exacerbate the damage from placental ischemia in preeclampsia.⁴⁵

A small study from Moscow also examined the bioenergetic role that mitochondria play in the pathophysiology of preeclampsia involved the collection of human myometrium samples from early- and late-onset preeclamptic pregnancies, and normal pregnancies immediately after delivery.⁴⁶ Their results where different from the others, in which levels of OPA1, a mitochondrial fusion protein, were slightly elevated in the late-onset preeclampsia group and significantly elevated in the early-onset preeclampsia group compared with normal pregnancies, while there were no changes in MFN1 and MFN2 expression.⁴⁶ In addition, another study that separated preeclamptic women into 2 subgroups, preterm and term delivery, displayed an increase in OPA1 and MFN1 protein amounts at term, along with a decrease in FIS 1 (fission protein), with no changes in fission/fusion proteins preterm.⁴⁷ These changes are indicative of an increase in mitochondrial fusion as opposed to fission.

Note that in the literature contradictory findings are reported on placental expression of mitochondrial fusion and fission proteins in preeclampsia. Perhaps the difference in expression of these proteins may be due to the severity of preeclampsia and the populations chosen for each study. For example, early-onset preeclampsia, which is the more severe form of preeclampsia, may display an upregulation of OPA1, a fusion protein, as a protective mechanism to preserve mitochondrial structure and function, while late-onset preeclampsia mitochondria may favor the process of fission in creating new mitochondria, as evident by a decrease in OPA1 and elevation in DRP1. Both mitochondrial fission and fusion play an important role in maintaining mitochondria function in response to cellular stress.^48,49 Specifically, fusion helps to preserve mitochondria function by mixing contents of damaged mitochondria, while fission is the process required to create new mitochondria.^48,49 Disruptions in the balance of fission and fusion of mitochondria are implicated in several neurodegenerative and cardiovascular diseases, and as discussed in preeclampsia.⁴⁸

Levels of voltage-dependent anion selective channel 1 (VDAC1), a regulator of mitochondria permeability and contributor to apoptosis, were reported to be doubled in the late-onset preeclampsia groups compared with the control group, and increased in early-onset preeclampsia.⁴⁶ These results may reflect a higher susceptibility of cells in the preeclamptic tissue to undergo apoptosis. There was a 2.5-fold increase in mitochondrial transcription factor A in the late-onset preeclamptic samples; however, mtDNA quantity and copy number were unaltered with preeclampsia.⁴⁶ Thus, evidence from this study suggests that mitochondria turnover is occurring rapidly and is increased in preeclampsia.

Mitochondrial dysfunction in trophoblast cells

During placentation, the inability of trophoblast cells to invade the myometrium is a key factor in the pathogenesis of preeclampsia (although trophoblast invasion abnormalities do not occur in every case of preeclampsia). Syncytiotrophoblast cells are trophoblast cells predominantly on the maternal side that aid in the formation of the placental bridge from the maternal blood space to the placental villa. In preeclampsia, the ultrastructure of the syncytiotrophoblast cells undergo adverse changes in cellular structure and function.⁵⁰ In preeclampsia, there is an upregulation in fibrinoid deposition, microvilli loss, and swelling/dilation of the endoplasmic reticulum, Golgi bodies, and the mitochondria in syncytiotrophoblast cells.

Homocysteine, a homologue of cysteine, has been shown to induce mitochondrial dysfunction and elicit mtROS production.⁵¹ It is a sulfur-containing amino acid that is produced from the conversion of methionine (an essential amino acid) to cysteine. Elevated levels of circulating homocysteine are correlated with decreased body levels of vitamins B6, B9, and B12, which are involved in homocysteine metabolism.⁵² Homocysteines are elevated in women throughout gestation and in both severe and nonsevere forms of preeclampsia.^53,54 Vitamin B12 is reduced in women with preeclampsia.^53,54 Homocysteine can cause apoptosis and damage to the mitochondria in trophoblasts.^55,56 Furthermore, homocysteine attenuates the production of human chorionic gonadotropin,⁵⁴ a hormone necessary for placental remodeling, in trophoblasts.⁵⁵

Increased levels of placental synctiotrophoblast extracellular vesicles and fetal hemoglobin have been observed in the circulation of women with preeclampsia.^57,58 The extracellular vesicles originating from the trophoblast cells induce endothelial dysfunction in the placenta.⁵⁹ Under stressful conditions, such as placental ischemia, the extracellular vesicles may contain mtDNA, RNA, soluble fms-like tyrosine kinase-1 (s-Flt1), and antiangiogenic enzymes.^59,60 The content in the vesicles can facilitate inflammatory responses and endothelial cell dysfunction that can hinder spiral artery remodeling, and halt the process of placentation.⁶¹

In preeclampsia, mitochondrial damage and oxidative stress will affect the mitochondria’s ability to utilize beta oxidation to catabolize fatty acids for ATP production.⁶² Genetic defects of fatty acid oxidation within the mitochondria are associated with maternal, placental, and fetal complications, such as maternal fatty liver, preterm delivery, and preeclampsia.^62,63 Thus impaired beta oxidation in trophoblasts cells could result in increased production of vesicles that contain mtDNA, which could further exacerbate the damage to the mitochondria, trophoblast cells, and inflammatory responses during preeclampsia. Experiments correlating preeclamptic environments to vesicle formation and content would potentially provide insight into the increase in circulating mtDNA and proinflammatory responses that occurs in preeclampsia.

Mitochondrial dysfunction in endothelial cells

Endothelial cells play a crucial role in angiogenesis and placentation.^64,65 The spiral arteries, with assistance of invading trophoblasts and NK cells, establish the placental–maternal barrier and blood vessels for exchange of maternal and fetal blood supply and nutrients. The formation of these new blood vessels will adapt from high resistance and moderate blood flow vessels to low resistance and high blood flow vessels to assure proper blood flow to the fetus.^66,67 Preeclampsia is known to alter the genetic expression and morphology of endothelial cells causing endothelial cell dysfunction.⁶⁸ Endothelial cells in preeclampsia express the antiangiogenic protein, sFlt-1, which serves as a decoy receptor for vascular endothelial growth factor, a necessary component in the development and remodeling of vascular tissue.^69,70 Endothelial cells also secret endothelin 1, a potent vasoconstrictor that is elevated in preeclampsia.

Oxidative stress causes endothelial cell dysfunction, which in preeclampsia leads to insufficient angiogenesis and inadequate oxygen supply and nitric oxide bioavailability.⁷¹ It has been hypothesized that preexisting mitochondrial factors could play a role in the development of endothelial cell dysfunction. Thus, experiments have been conducted to target mtROS in attempts to improve endothelial dysfunction.^72,73 One group utilized human endothelial cells, exposing them to plasma from women exhibiting preeclampsia, as well as plasma from women with a normal pregnancy.⁷³ Cells exposed to the plasma from preeclamptic pregnancies exhibited a decrease in mitochondrial function and increase in ROS production. The preeclamptic plasma group also exhibited increased levels of TNF-α, toll-like receptor 9 (TLR-9), and intercellular adhesion molecule-1 (ICAM-1), all known markers of inflammation and preeclampsia. MitoTempo, an antioxidant that localizes to mitochondria, reduced levels of mtROS and decreased the amount of inflammatory cytokines produced in response to oxidative stress, while preserving mitochondrial functionality.⁷⁴ Furthermore, experiments performed by our group also showed that human endothelial cells incubated with serum from preclinical rat model of preeclampsia have increased endothelin 1 secretion and mtROS production.^75,76

Endothelial cells are the mediators of the entry of NK cells and other leukocytes into infected and inflamed tissue through selectin- and integrin-mediated diapedesis.⁷⁷ Soluble vascular cell adhesion molecule-1 (sVCAM-1), ICAM-1, and E-selectin levels are all elevated in mothers with preeclampsia and associated with an increased in NK cells^78,79 The influx of the immune cells will cause tissue degradation and increase ROS production.⁸⁰

MITOCHONDRIAL DYSFUNCTION IN PRECLINICAL ANIMAL MODELS OF PREECLAMPSIA

There are several animal models of preeclampsia that involve genetic manipulation, pharmacological intervention, and surgical intervention.^81–86 However, only a few models have examined mitochondria and mitochondrial function, which include the sFlt-1-induced model, storkhead box 1 (STOX1) overexpression model, reduced uterine perfusion pressure (RUPP) model, and AT1-AA infused model.

Mitochondrial dysfunction in the sFlt-1 and STOX1 overexpression models of preeclampsia

The sFlt-1 inducible mouse model of preeclampsia, where sFlt-1 (an antiangiogenic factor upregulated in preeclampsia) is injected into pregnant mice for 8 consecutive days, starting at gestational day 9, showed swollen and enlarged mitochondria along with extensive DNA damage and increased apoptotic protein expression in trophoblast cells.⁸³ In the STOX1 overexpression mouse model of preeclampsia, placental ROS generation and peroxynitrite are increased along with mitochondrial mass and alterations in mitochondrial function regulator genes.⁸⁴ STOX1 is a transcription factor belonging to the Forkhead Box gene family that is upregulated in trophoblast cells of women with preeclampsia. Mice overexpressing the human STOX1 gene, display several characteristics of preeclampsia, which are gestational hypertension, proteinuria, and elevated circulating sFlt-1.⁸⁷ STOX1 overexpression in cell culture experiments under hypoxic conditions, increased mitochondrial membrane potential and mtDNA replication and transcription.⁸⁴ This increase in mitochondria activity was observed despite a decrease in mitochondria mass, which was not displayed in the STOX1 overexpression mice.⁸⁴ Perhaps the difference between the in vivo and in vitro experiments can be explained by the influence of other environmental factors present in vivo. Thus far in summary, these data together suggest that preeclampsia in animal models displays alterations in mitochondria function and morphology.

Mitochondrial dysfunction in the RUPP model of preeclampsia

Mitochondrial dysfunction has been studied extensively in the RUPP model, which has been used to replicate the pathogenesis and onset of preeclampsia.⁸⁸ This model has hypertension, along with placental ischemia, oxidative stress, and mitochondrial dysfunction. The RUPP model is surgically induced in pregnant rats at day 14 of gestation by placing silver clips around the abdominal aorta before the iliac bifurcation and on the left and right uterine arteries to reduce blood flow by ~40%.^88,89 Studies have shown that both placental and renal mitochondrial dysfunctions in the RUPP model are evident by a reduction in mitochondria respiration (a marker of mitochondrial function), protein abundance and activity of complexes I, II/IV (components of the ETC), and mitochondria amount.^75,90 Note that these alterations in mitochondrial function are also observed in women with preeclampsia.^38–40 Studies in which antioxidant treatment was administered to RUPP rats demonstrated that a major source of oxidative stress in the RUPP rat, and possible women with preeclampsia, is the mitochondria.⁷⁵

Briefly, NK cells, which will be discussed in more detail later, play an important role in pathophysiology of preeclampsia. Numerous studies in both clinic and animal models of preeclampsia display robust elevations in cytotoxic NK cells.^91–93 NK cell depletion in the RUPP model of preeclampsia improves hypertension, fetal growth restriction, and inflammation during pregnancy, along with placental and renal mitochondrial function and mtROS.^90,93

Mitochondrial dysfunction in the AT1-AA model of preeclampsia

AT1-AAs are elevated early in women with preeclampsia and in the RUPP model.^{24,85,86,94–96} AT1-AAs bind to a specific 7 amino acid region on the secondary extracellular loop of the AT1 receptor to activate the receptor and downstream singling cascades that contribute to the pathophysiology of preeclampsia.^24,94 AT1-AAs, isolated from preeclampsia patients and RUPP rats, administered to normal pregnant rats caused an increase in blood pressure along with other factors, such as oxidative stress and inflammation, which are associated with preeclampsia.^24,85,97 Studies show that the AT1-AA-induced rat model of preeclampsia has elevated placental and renal NK cell activity and mtROS, which are inhibited by a modified and unique AT1-AA inhibitory peptide.⁸⁵ Furthermore, administration of the AT1-A7A inhibitory peptide to RUPP rats improved hypertension along with mitochondrial respiration and oxidative stress.⁸⁶ These data taken together suggest that the AT1-AA elevated in preeclampsia causes mitochondrial dysfunction, mtROS, and NK cell activation in the placenta and kidney, which may contribute to the hypertension and pathophysiology of preeclampsia. Whether the actions of AT1-AA on mitochondria in preeclampsia are direct or indirect is unknown. Experiments examining the direct role of AT1-AA on mitochondria are warranted along with determining whether AT1-AA or mitochondrial dysfunction/oxidative stress occurs first in preeclampsia.

OXIDATIVE STRESS AND INFLAMMATION MEDIATING MITOCHONDRIAL DYSFUNCTION

Oxidative stress and mitochondrial dysfunction

Oxidative stress is defined as an increase in ROS production and/or reduced antioxidant capacity. Women with preeclampsia have an increase in oxidative stress (as evident by increased levels of isoprostane) via an increase in ROS production (e.g., hydrogen peroxide), as well as a decrease in antioxidant capacity.^11,98–103 Although, mitochondria are a major source of ROS, there are several other sources in preeclampsia.^11,100,104 Others sources are xanthine oxidase, NADPH oxidase (NOX), cyclooxygenase, lipoxygenase, and cytochrome P450.¹⁰⁴

Multiple studies have examined the role of NOX abundance, activity, and isoforms in AT1-AA, RUPP, and antiangiogenic factor (sFlt-1)-mediated animal models of preeclampsia.¹⁰⁴ Specifically, NOX 1 and 5 have been extensively investigated and shown to be the major source of ROS generation during preeclampsia, with the potential to disrupt the mitochondrial ETC to generate more ROS (circulating/placental ROS-induced placental mtROS release).^105,106

Mitochondria can serve as a mediator of inflammatory responses, with the formation of damage-associated molecular patterns (DAMPs) and the production of ROS.^102,107 Both DAMP production and ROS generation contribute to the pathophysiology of preeclampsia.^108,109 DAMPs produced by mitochondria include mtDNA and N-formyl peptides.

Mitochondria have their own separate set of DNA (mtDNA) that is inherited from the maternal side. Elevated levels of circulating mtDNA serve as a biomarker indicating mitochondrial damage and dysfunction.^110,111 Elevated levels of circulating mtDNA have been reported in women diagnosed with preeclampsia.¹¹² This damage is associated with an increase in DAMPs and ROS that increase oxidative stress and inflammation, both of which contribute to the progression of preeclampsia.

Inflammation and mitochondrial dysfunction

Mitochondria are bacterial in origin and express molecular patterns that can invoke the inflammatory response.^113,114 The mtDNA and N-formyl peptides released by damaged mitochondria activate pattern-recognition receptors that activate the innate immune system.¹¹⁴ DNA of both bacteria and mitochondria have unmethylated CpG DNA repeats that can activate the body’s innate immune system via toll-like receptors (TLRs), such as TLR-9.^107,115 Elegant studies by Goulopoulou et al. demonstrated that administration of high doses of synthetic CpG oligonucleotides to pregnant mice cause placental cell death, mtDNA formation and release, and activation of TLR-9 receptor to generate a model of preeclampsia in mice, which consists of inflammation, maternal hypertension, fetal growth restriction, and vascular dysfunction.^107,116 Therefore, activation of TLR-9 by mtDNA suggests that mtDNA plays a significant role in the development of preeclampsia.^107,116

Differences in TLR expression account for hypertension and the differences in blood pressure between males and females, where males have higher pressures than females after puberty and before menapause.¹⁰⁸ Placentas and trophoblast cells from preeclamptic women overexpress TLR-2, TLR-4, TLR-3, TLR-7, TLR-8, and TLR-9 compared with women with healthy pregnancies. When TLR-9 and TLR-3 are overactivated during pregnancy in animal models, preeclamptic-like symptoms are observed.^117,118 As mentioned above, circulating mtDNA in preeclampsia may activate TLR-9, which would result in an increase in acute inflammation, hypertension, and trophoblast cell cycle arrest.^107,116 Trophoblast cell cycle arrest causes apoptosis of trophoblast cells leading to placental dysfunction and placental ischemia. mtDNA is also reported to activate the TLR-3 receptor.¹⁰⁸ TLR-3 binds viral double stranded RNA and induces mitochondrial dysfunction ultimately resulting in apoptosis.¹¹⁹ Furthermore, TLR-4 can be activated by LPS, hyaluronic acid, and fibrinogen.¹²⁰ Fibrinogen levels are elevated in preeclampsia and may induce an inflammatory response through TLR-4 activation during pregnancy.^121,122

N-Formyl peptides are also a major source of mitochondrial DAMPs. In the mitochondria, 13 proteins have the N-formyl groups attached to their amino terminus. All 13 proteins are important for mitochondrial function because they play a role in oxidative phosphorylation.¹²³ Upon damage to mitochondria, the N-formyl peptides are released and can bind to formyl peptide receptors 1 and 2, and poorly to receptor 3.¹²⁴ These receptors are G-protein-coupled receptors that are expressed on innate immune system cells, such as neutrophils and monocytes, which trigger an inflammatory response.^111,124,125 Little is known about the function of formyl peptide receptor 3. However, formyl peptide receptors 1 and 2 have been studied in inflammatory diseases.¹²⁴ Remodeling of the uterine spiral arteries is disrupted in preeclampsia, and leukocytes could serve as a means of attack on the angiogenic remodeling that is occurring during placentation.^98,125,126 Oxidative stress produced from placental ischemia as well as the release of mtDNA and N-formyl peptides, could elicit an immunological response, causing more damage to the placenta to further decrease blood flow to the fetus leading to fetal growth restriction.

One key leukocyte in the process of placentation during pregnancy is regulatory NK cells. The subsets of regulatory NK cells that play a functional role in pregnancy are circulating, uterine, decidual, and endometrial NK cells.¹²⁷ During pregnancy uterine and decidual regulatory NK cells help to clear the space at the sight of implantation for trophoblast cell invasion and spiral artery remodeling for placentation.^{92,93,127–129} These regulatory NK cells are elevated in the blood, uterus, and decidua during pregnancy and play a critical role in tolerance against fetal antigens, trophoblast invasion, spiral artery remodeling, and angiogenesis.⁹³ Circulating, uterine, and decidual NK cells are similar in morphology and function during pregnancy. In fact, it is hypothesized that NK cells in the uterus and decidua are derived from circulating NK cells that have infiltrated into placental tissues.¹²⁷ Not much is known about the role of endometrial NK cells in the pathophysiology of preeclampsia.

During preeclampsia, there is a shift toward an increase in circulating, uterine, and decidual NK cells that are cytolytic and not regulatory.^91–93,130 This shift is due primarily to the upregulation of cytokines, such as IL-17, IL-2, TNF-α, and IFN-γ. The upregulation of these cytokines promotes alteration of a T-helper 2 cell immunosuppressive response, to a T-helper 1 cell immunoactive response.^{92,93,131,132} Release of these cytokines coupled with deactivation of immunosuppressive Th2 cells drive NK cells to shift from regulatory/noncytotoxic NK cells to cytolytic NK cells, which can secrete granenzymes and perforin that damage tissues, cells, and mitochondria.^91,130,133

Increasing circulating mtDNA will activate TLR-9 to cause an immunological cascade that releases cytokines to trigger T-helper 1 cells as well as cytotoxic NK cells from the periphery. The influx of cytotoxic peripheral NK cells, or the switching of noncytotoxic NK cells to cytolytic NK cells at the site of spiral artery remodeling can cause damage to the placenta and further contribute to placental ischemia and dysfunction. Other cytokines elevated in preeclampsia, such as TNF-α and IL-17, can also enhance NK cell cytolytic activity.^131,132 The exact mechanism of how these cytokines influence NK cell cytolytic activity will require further experimentation.

It is important to note that overactivation of cytolytic NK cells can cause damage to mitochondria. Studies have shown that blockade of NK cell activity in the RUPP rat, rat model of preeclampisa, will improve not only the pathophysiology of preeclampsia, but also mitochondrial respiration, mtROS, and the enzyme activities of complexes I and IV.⁹⁰ A decrease in complex I and IV activity has been shown to decrease ECT efficiency in mitochondria causing a decrease in ATP production and creating an energy crisis for the cell.¹³⁴ As mentioned, complex IV or cytochrome c oxidase activity is decreased in the preeclamptic placenta, aging, and other CV diseases.^34–36,134 Furthermore, a decrease in complex I and IV activity may generate mtROS.¹³⁴

PHARMACOLOGICAL TREATMENTS AND THERAPIES

Due to prevalence and significant amount of maternal and perinatal mortality and morbidity associated with preeclampsia, researchers have explored various treatment options. The only known treatment of preeclampsia is birth⁴⁷ and thus novel treatments to reduce mtROS and improve mitochondrial dysfunction are likely to be beneficial.

One therapeutic agent shown to attenuate the adverse effects associated with preeclampsia is 1,25-dihydroxyvitamin D (1,25-(OH)2D).¹³⁵ Vitamin D deficiency early in pregnancy is associated with preeclampsia.¹³⁵ Lower levels of plasma vitamin D are observed in both preeclamptic women and RUPP rodents.^135,136 Vitamin D supplementation, which has been proven to be safe during pregnancy, decreases the risk of preterm labor in women with preeclampsia.¹³⁵ 1,25-(OH)2D administered to RUPP rats prevents the increased blood pressure, proteinuria, and fetal mortality associated with preeclampsia.^135–137 In addition, 1,25-(OH)2D averts trophoblast cell damage and attenuates apoptosis of placental cells in RUPP rats, while simultaneously inducing autophagy of damaged placental cells and mitochondria to promote cell survival.¹³⁵ Vitamin D supplementation also reduces AT1-AAs, which are elevated in the rat models of preeclampsia and women.¹³⁷ Additionally, high-dose vitamin D inhibits placental mitochondria cytochrome P450scc, an enzyme associated with increased levels of oxidative stress through production of isoprostanes and lipid peroxidation.¹³⁸ Isoprostanes are prostaglandin-like compounds formed in vivo from the free radical-catalyzed peroxidation, via superoxide, of essential fatty acids.¹³⁹ An increase in isoprostanes correlates with an increase ROS generation.¹³⁹

Therapeutic treatments utilizing classical antioxidants have not been successful in preeclampsia.¹⁴⁰ One explanation is their inability to target-specific forms and sources of ROS, such as mtROS. Recent preclinical studies have shown that mitochondria-targeted antioxidants alleviate many of the pathophysiology associated with preeclampsia. Pretreatment of human umbilical vein endothelial cells exposed to serum from preeclamptic women, with MitoTempo, a superoxide dismutase mimetic that localizes to the inner mitochondrial membrane, resulted in a decrease in mtROS, mRNA levels of TNF-α, mitochondrial uncoupling protein 1, and TLRs.⁷³ MitoTempo stabilizes mitochondrial function and reduces hydrogen peroxide (a major ROS)-induced cell death.

MitoQ, which is composed of the natural antioxidant ubiquinol attached to a lipophilic cation, is another mitochondria targeting antioxidant that reduces mitochondrial oxidative stress, hypertension, and other markers of preeclampsia. In a rat model of hypoxic pregnancy, MitoQ treatment increased placental maternal blood space surface area and volume, and blocked activation of mitochondrial and endoplasmic reticulum stress, which can activate signaling cascades that destroy cells and organelles.¹⁴¹ Renal and placental respiratory complex (I, II/IV) activities and expression levels, and mitochondrial respiration rates were reduced in RUPP vs. normal pregnant rats, while mtROS were elevated.⁷⁵ These changes were prevented by treating the animals with MitoQ.⁷⁵

Although mitochondria-specific antioxidant therapy seems promising, this approach may be limited in that it only treats the symptoms, and not the etiology of preeclampsia. These mitochondria-specific antioxidant therapy treatments clearly show that mitochondrial dysfunction is a major contributor to preeclampsia, and that using these therapies will reduce and/or alleviate the oxidative stress and inflammation in preeclampsia. However, the question still remains, what is the initial factor that causes preeclampsia? Is it mitochondria dysfunction? If not, then why is mitochondrial dysfunction a recurring theme observed in women with preeclampsia?

Evidence from the literature and this review suggests that mitochondrial dysfunction is a byproduct of preeclampsia and an initial factor. We hypothesize that early in preeclampsia, placental ischemia occurs, which will elicit an immune response and several other factors, such as AT1-AAs and inflammatory cytokines, that can cause damage to the mitochondria. The damage to the mitochondria can be the result of indirect effects of the factors just mentioned or a direct effect of placental ischemia. The damage and dysfunction of the mitochondria will release DAMPS and generate mtROS that can damage trophoblast cells, endothelial cells, and tissues directly or through TLRs and N-formyl peptide receptors (Figure 1).

CONCLUSIONS AND FUTURE PERSPECTIVES

Various physiological observations utilizing human and animal models of preeclampsia show a strong correlation between preeclampsia and mitochondrial dysfunction. Increased mtDNA, mtROS, AT1-AAs, NK cell activity, and overall systemic oxidative stress are present in women with preeclampsia, and can cause further mitochondrial damage and dysfunction. In this review, we make the case that mitochondrial dysfunction plays role in preeclampsia development and progression. Elevated circulating mtDNA, DAMPS, mtROS, and mitochondrial dysfunction can facilitate oxidative stress and inflammation, which is present in preeclampsia and plays a role in its pathogenesis.

Moving forward, timeline studies utilizing animal models and human observation experiments could provide a better understanding of when mitochondrial dysfunction begins in preeclampsia and how this affects NK cells, AT1-AAs, oxidative stress, and inflammation. Some of our studies administering mitochondria-specific antioxidants at the same time as our RUPP surgical protocol to elicit the preeclamptic phenotype, suggest that mitochondria dysfunction and mtROS are early and important players in the pathophysiology of preeclampsia. Circulating mtDNA and DAMPs levels could be monitored throughout the stages of the preeclamptic pregnancy compared with normal pregnancy to evaluate (i) when these factors first appear, (ii) how these factors change with the course and severity of preeclampsia, and (iii) how these factors overall affect pregnancy. Screening for mitochondrial dysfunction, injury, damage, ROS, and abnormalities could be performed to predict preeclampsia and its outcomes.