Mitochondrial Dysfunction in Obesity and Reproduction

Abstract

Mounting evidence suggests a role for mitochondrial dysfunction in the pathogenesis of many diseases, including type 2 diabetes, aging, and ovarian failure. Because of the central role of mitochondria in energy production, heme biosynthesis, calcium buffering, steroidogenesis, and apoptosis signaling within cells, understanding the molecular mechanisms behind mitochondrial dysregulation and its potential implications in disease is critical. This review will take a journey through the past and summarize what is known about mitochondrial dysfunction in various disorders, focusing on metabolic alterations and reproductive abnormalities. Evidence is presented from studies in different human populations, and rodents with genetic manipulations of pathways known to affect mitochondrial function.

Mitochondria are essential organelles for many aspects of cellular homeostasis, including cellular energy production through oxidative phosphorylation, acting as a critical component of apoptosis, and maintaining calcium homeostasis. Because of these myriad roles within cells, mitochondrial function plays an important role in cellular and metabolic health. Alteration of mitochondrial function may lead to a wide range of seemingly unrelated disorders, such as dementia, Alzheimer disease, epilepsy, strokes, Parkinson disease, ataxia, cardiomyopathy, coronary artery disease, chronic fatigue syndrome, hepatitis C, primary biliary cirrhosis, diabetes, and ovarian dysfunction (1). Therefore, understanding mitochondrial dysfunction and its implication in pathological states is of prime importance.

Though research on mitochondria started in the late 19th century, their function and relevance to health and disease is not yet fully understood. For instance, although mitochondrial dysfunction has been associated with insulin resistance (IR), a large variety of association and intervention studies, as well as genetic manipulations in rodents, have reported contradicting results (2). Indeed, even 39 years after the first publication describing a relationship between IR and diminished mitochondrial function, it is still unclear whether a direct relationship exists, and more importantly if changes in mitochondrial capacity are a cause or consequence of IR. Other studies have suggested that mitochondrial dysfunction may be a key mechanism for inter-generational transmission of metabolic diseases through reprogramming of IR in offspring (3, 4). If true, this provides an important link between the role of mitochondrial dysfunction in metabolism and also in reproduction. Like IR, the age-related decline in fertility is also associated with mitochondrial dysfunction that may cause ovarian aging (5). This idea is supported by the association of mutations in mitochondrial genes with rare cases of primary ovarian insufficiency (POI); however, data are scant as to whether these genes in particular, and mitochondrial dysfunction in general, contribute to the majority of POI cases that lack a known etiology (6).

In light of 4 recent mini-reviews on mitochondrial dysfunction in insulin resistance and ovarian insufficiency (5-8), this review will take a broad look at the mechanisms underlying mitochondrial dysfunction and summarize the proposed role of mitochondria in both metabolic alterations and reproductive abnormalities.

Mitochondria structure and function

Mitochondria are maternally-inherited, double-membrane organelles that are essential for producing the cellular energy source 5′-adenosine triphosphate (ATP) from the oxidation of fatty acid, glucose, and amino acid metabolites through the combined action of the tricarboxylic acid (TCA) cycle in the mitochondrial matrix and oxidative phosphorylation (OXPHOS) in the electron transport chain (ETC) in the mitochondrial inner membrane. Mitochondria are semiautonomous organelles that have a genome spanning 16.7 kilobase and containing 37 genes that encode 13 proteins functioning within the OXPHOS pathway, as well as 22 transfer RNAs and 2 ribosomal RNAs. However, human mitochondrial proteome studies have detected at least 1500 proteins, and indeed the large majority of proteins localized to mitochondria are encoded by the nuclear genome (9, 10). Unlike the nuclear genome, the mitochondrial genome lacks protective histones and an effective DNA repair mechanism, and is, thus, more vulnerable to DNA mutations (11). Normal respiratory chain activity requires an intact and functional mitochondrial genome (12), and coordinated interaction between nuclear and mitochondrial DNA (mtDNA) is essential for proper mitochondrial function to support many critical cellular functions such as protein and nucleic acid biogenesis, lipid synthesis, regulation of cell cycle and programmed cell death, calcium signaling, the citric acid cycle, generation of reactive oxygen species (ROS), and antioxidant protection (13). To protect their functional integrity, mitochondria deploy several quality control (QC) pathways to survey mitochondrial stress to maintain their healthy metabolic functions. Mitochondria are extremely dynamic organelles that can undergo fission and fusion to facilitate adaptation to various cues and stresses by remodeling the mitochondrial network. This QC process also aids with the tagging of dysfunctional mitochondria for destruction by mitophagy (14), a specific mitochondrial form of autophagy (15). Another essential QC pathway is the mitochondrial unfolded protein response (mtUPR), which senses matrix protein misfolding and induces an adaptive transcriptional program to maintain mitochondrial proteostasis (16, 17), and also activates a cytosolic response by inducing components of the heat shock response (17). Finally, when cellular damage is too great, mitochondria play an essential role in apoptosis, a well-defined mechanism of programmed cell death (18).

Mitochondrial dysfunction in disease

The activation of stress response pathways in individual mitochondrion can lead to deterioration of its function. The origin of global mitochondrial dysfunction, however, can be from many causes, including a decrease in mitochondrial biogenesis or increased mitochondrial loss, reduced mitochondrial protein content, or reduced activity of enzymes in the TCA cycle or ETC (19). Accordingly, some groups refer to mitochondrial dysfunction as a decrease in mitochondrial activity and oxidative phosphorylation, some as a diminished mitochondrial number, and yet others focus on the generation of ROS. Importantly, the deterioration of mitochondrial function, whatever its cause, has become a prominent signature of metabolic, cardiovascular, renal, inflammatory, reproductive, muscular, and neurodegenerative diseases as well as cancer (18, 20). For instance, several metabolic studies in humans suggested the existence of mitochondrial dysfunction in obese and insulin-resistant individuals and in patients with type 2 diabetes mellitus (T2DM), with these individuals exhibiting lower oxidative enzyme activities and decreased lipid metabolism in muscle compared with lean control individuals (2, 21). Mitochondrial dysfunction has also been documented in women with gestational diabetes (GDM) who are at risk of developing T2DM later in life. Specifically, patients with GDM have reduced placental mitochondrial content (22), which could affect fetal development. Furthermore, optimal mitochondrial function is crucial for embryonic development (5) and the preservation of oocyte quality and ovarian reserve (6). In this review, we will thus focus on the relationship of mitochondrial function with metabolic disease and reproductive dysfunction.

Type 2 diabetes mellitus

T2DM is a serious health concern worldwide, with more than 380 million people suffering from the disease. As mentioned earlier, mitochondrial dysfunction has been reported in patients with T2DM and also in nondiabetic individuals with IR (2, 21). Many studies have shown that IR is fundamental to the development of T2DM and is present in most prediabetic individuals. It is also now generally accepted that T2DM requires both IR and insulin secretion defects. Metabolic analysis of normal glucose-tolerant individuals with a family history of T2DM indicated that IR in the face of normal insulin secretion predicts the development of T2DM over a mean of 25 years (23). In contrast, longitudinal analysis over 13 years demonstrated that defects in insulin secretion emerge relatively late in those who ultimately develop T2DM (24). IR is defined as the reduced ability of insulin to promote storage of nutrients in skeletal muscle and adipose tissue, and to restrain lipolysis and hepatic glucose production. It is thus important to understand how mitochondrial dysfunction could contribute to insulin resistance.

Mechanistic link of mitochondrial dysfunction and insulin resistance

Mitochondrial dysfunction is primarily associated with a disruption in cellular homeostasis that is exacerbated in the state of disease (8) (Fig. 1). Multiple factors such as mutation in mtDNA, decrease in mitochondrial biogenesis, impaired mitochondrial dynamics, reduced activity of enzymes in the TCA cycle or ETC, reduced mitochondrial clearance by mitophagy, impaired bioenergetics, and imbalance of calcium homeostasis all contribute to mitochondrial dysfunction. Firstly, this impairment in mitochondrial function leads to a reduction in mitochondrial oxidation of substrates including fatty acids resulting in ectopic lipid accumulation and increased levels of the diacylglycerols (DAG) and ceramides (CER) (2). DAGs and CERs have both been shown to inhibit insulin signaling. Numerous studies suggest that DAGs elicit skeletal muscle and liver IR through activation of novel protein kinase Cs (PKCθ, PKCε) (25-30) that phosphorylate the insulin receptor and inhibit its tyrosine kinase activity (31); but studies have also reported that CERs cause IR by activating protein phosphatases that inhibit insulin-dependent signaling at the level of the protein kinase AKT downstream of the INSR (32, 33). Second, a decrease in electron flow through the mitochondrial ETC increases electron leakage toward oxygen and the formation of superoxide (34, 35). ROS production damages mitochondrial and cellular components, which normally results in mitophagy to maintain mitochondrial homeostasis (19). If unchecked, high levels of ROS and metabolic stress can also impair mitophagy resulting in the failure to clear damaged and dysfunctional mitochondria, further exacerbating the oxidative stress and causing apoptosis (36, 37). In vitro, mitochondrial ROS production attenuates insulin action in adipocytes, myotubes, and mice (38), and abolishes insulin-stimulated GLUT4 translocation in 3T3L1 cells by interfering with insulin activation of IRS-1 and PI3-kinase (39). More recent studies indicate that dysregulation of intracellular Ca²⁺ homeostasis may also play a role in the pathogenesis of insulin insensitivity and T2DM (17, 40) as Ca²⁺ uptake into mitochondria regulates Ca²⁺-dependent enzymes that participate in fatty acid metabolism via oxidative phosphorylation in the TCA cycle (41).

Figure 1.

Mitochondrial dysfunction can involve mutation of mitochondrial DNA, reduction in mitochondrial content and/or biogenesis, impaired dynamics (fission/fusion), impaired mitophagy, failure in bioenergetics, reduced enzyme activity, augmented oxidative stress, or an imbalance in calcium homeostasis that cumulatively lead to decreased glucose and lipid oxidation. This disruption in lipid oxidation leads to lipid accumulation, which is manifested as an increase in active lipid intermediates, such as diacylglycerols (DAG) and ceramide (CER) that can inhibit insulin action. The decrease in substrate oxidation and reduced oxidative phosphorylation leads to diminished electron flow through the electron transport chain (ETC), which subsequently causes electron leakage and superoxide generation, followed by oxidative stress and mitochondria damage. Mitophagy maintains mitochondrial homeostasis by removing damaged mitochondria. Impaired mitophagy exacerbates the problem by failing to remove damaged mitochondria leading to further increases in oxidative stress and mitochondrial damage.

Debate on mitochondrial dysfunction and insulin resistance in humans

Recent insights from studies of individuals with insulin resistance, either with established T2DM or in prediabetics, have confirmed some of these defects in mitochondrial function (42-44).

Indeed, reduced mitochondrial function has been demonstrated in healthy, young, lean, IR offspring of patients with T2DM, which is accompanied both by a reduction in mitochondrial density/content and a decrease in OXPHOS in muscle, and an increase in the intramyocellular lipid concentration (45-47). Similarly, obese IR humans fed a high-fat diet exhibit an enhancement of skeletal muscle H₂O₂ production (48). However, it remains unclear whether mitochondrial dysfunction is primary or secondary to the derangements in glucose and lipid metabolism. Indeed, this controversy has been the source of much debate in the scientific community (49-51). In the section that follows, we present the evidence for and against a causative link between mitochondrial dysfunction and insulin action in different tissues to dissect this key question in T2DM pathophysiology (Table 1).

Table 1.

Selected studies exploring links between mitochondrial dysfunction and insulin resistance

Sample group	Observation	Reference
Young, IR offspring of parents with T2DM	Decreased mitochondrial activity, increased intramyocellular FA, and lower ratio of type 1 to type 2 fibers in muscle.	(45)
	Reduced mitochondrial density in muscle of IR offspring of T2DM parents.	(46, 47)
Overweight/obese individuals	Improved IR without increasing mitochondrial capacity.	(63)
Elderly, lean individuals	Severe IR in skeletal muscle, with decreased mitochondrial oxidative capacity, and decreased ATP synthesis.	(52)
T2DM patients/IR individuals	Mitochondrial fatty acid oxidation normal.	(64)
	Mitochondrial dysfunction in IR, 30%-40% lower mitochondrial enzyme activity.	(44, 53, 65, 66)
	PGC 1α-dependent genes involved in oxidative metabolism are downregulated in skeletal muscle of individuals with family history of T2DM and T2DM patients compared to healthy controls.	(67)
	Increased daily physical activity improves lipid oxidation independent of change in mitochondrial activity in people with T2DM.	(68)
	IR Asian Indians with T2DM showed muscle mitochondrial capacity similar to nondiabetic Indians, but higher than Northern European Americans. Asian Indians have greater IR than Northern Europeans.	(69)
	Obese diabetic and normal males showed altered activities of enzymes involved in carbohydrate breakdown, and aerobic metabolism in muscle.	(70)
	T2DM patients have normal muscle mitochondrial function.	(71)
Rodent studies	High-fat diet reduces mitochondrial gene expression in muscle.	(72)
	Decreased mitochondrial function does not cause IR in mice.	(73)
	UCP-mediated energy depletion increases glucose uptake despite reduced mitochondrial function.	(74)
	High-fat diet causes IR but increases mitochondrial capacity.	(75-77)
	Lowering plasma FFA reduces mitochondrial gene expression in muscle.	(78)

Abbreviations: ATP, adenosine 5′-triphosphate; FA, fatty acid; FFA, free fatty acid; IR, insulin resistance; PGC1α, peroxisome proliferator-activated receptor gamma coactivator 1-alpha; T2DM, type 2 diabetes mellitus; UCP, uncoupling protein.

Decreased mitochondrial function with insulin resistance.

Some studies have reported abnormalities in mitochondrial function in skeletal muscle biopsies from patient with IR and T2DM with resting ATP synthesis being reduced in IR vs insulin-sensitive individuals (52). Other studies reported a decrement in mitochondrial number and ETC activity in T2DM and obese individuals compared with lean volunteers (21). At the molecular level, citrate synthase, carnitine-palmitoyl transferase 1, and cyclooxygenase are reduced in obese individuals (44). Studies in muscle biopsy samples evaluated ex vivo have corroborated these defects. Lower muscle OXPHOS capacity was found both in muscle biopsy samples and isolated mitochondria from T2DM individuals (53-56). Mitochondrial biogenesis may also contribute to observed alterations in mitochondrial number and function. Transcriptomic studies revealed a coordinated reduction in expression of genes regulated by peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC1α), a key transcription factor driving mitochondrial biogenesis, in patients with T2DM both in the fasting (57) and insulin-stimulated states (58). Mitochondrial dynamics are also important. Obese IR individuals have reduced expression of the mitochondrial fusion gene mitofusin 2 in skeletal muscle (59, 60), while mutations in the fission protein optic atrophy 1 reduce muscle ATP synthesis (61, 62). Most of these studies are correlative, however, and it is unclearwhether a causative relationship exists. Arguing against a causative role, muscle-specific knockout of mitochondrial transcription factor A (Tfam), an important gene controlling mitochondrial biogenesis and protein expression, causes mitochondrial abnormalities and deteriorated respiratory chain function but does not cause IR (73). Loss of the estrogen receptor in skeletal muscle or removal of the ovaries in female mice causes IR and decreased mitochondrial mass and function that may underlie the relative protection against metabolic disease in females (7, 79). Conversely, deletion of Ant2 in liver or Ant1 in muscle increases mitochondrial number and enhances uncoupled respiration, and prevents hepatic steatosis, insulin resistance and diet-induced obesity (80, 81).

Unchanged mitochondrial function despite insulin resistance

In disagreement with studies showing reduced mitochondrial function with IR, a number of publications have reported normal mitochondrial activity both in T2DM (71, 82, 83) and in obese, IR individuals. Consistent with this, muscle mitochondrial ATP production is increased by post-prandial insulin levels in normal individuals, but T2DM patients are unresponsive despite having normal basal ATP production (84, 85). The authors suggested that reduced insulin-stimulated ATP production was secondary to reduced glucose disposal in these individuals rather than causative (44). Another study showed increased ROS production, along with a lower abundance of complex I subunits and carnitine palmitoyltransferase 1B protein, despite normal mitochondrial respiration in IR human skeletal muscle (86). Experimentally, nonobese sedentary humans who were overfed for 28 days exhibited peripheral IR without changes in several markers of mitochondrial content in muscle (87). Similarly, rats fed a high-fat diet displayed unchanged messenger RNA levels of several energy and glucose metabolism markers in muscle as well as comparable hepatic mitochondrial and peroxisomal fatty acid oxidation capacity compared to low-fat diet controls (88, 89).

Increased mitochondrial function in insulin resistance

Several research groups have indicated a compensatory increase in mitochondrial oxidative capacity associated with IR (75, 77, 90, 91). Mice and rats fed high-fat diets showed impairments in glucose tolerance and insulin sensitivity, with a concurrent increase in fatty acid oxidative capacity, as well as activity of mitochondrial oxidative proteins and their content in muscle (75, 77, 90-93). Consistent with these animal results, studies in South Asian Indians demonstrated increased OXPHOS activity in IR-participants (69).

Decreased mitochondrial function in improved insulin sensitivity

It is important to note that some studies have observed improved insulin sensitivity with reduced mitochondrial function (94). Increased mitochondrial damage in the polymerase gamma mitochondrial DNA (D257A) mutator (POLG) mice causes oxidative stress and activation of the mtUPR in skeletal muscle leading to elevated expression of growth differentiation factor 15 (GDF15) and FGF21, which may protect against IR (95, 96). Additionally, the POLG mutation can overcome the diabetes phenotype in Akita mice (97). Improved whole-body insulin sensitivity is also observed in mice with deficiency of mitochondrial fusion protein optic atrophy 1 in muscle, again possibly due to an increase in FGF21 production, in spite of progressive mitochondrial dysfunction (61).

Collectively, the 4 previously mentioned possible scenarios describing decreased, unchanged, or a compensatory increase in mitochondrial function in the context of IR suggest mitochondrial dysfunction is not a prerequisite for IR in all circumstances and is highly dependent on the methodological approach, the model system examined (eg, human vs rodent models) and the population studied.

Gestational diabetes and mitochondrial dysfunction

A number of studies suggest that the mitochondrial dysfunction in T2DM and IR also has effects on reproductive health and metabolic reprogramming in the offspring (22, 98). The theory of developmental origins of health and disease proposes that the risk of developing disease at a later time point in life is increased by exposure to environmental factors before birth (99, 100). Mitochondria are inherited from the mother so maternal mitochondrial dysfunction could be passed to the fetus. GDM is a form of diabetes that can occur during pregnancy and confers an increased lifetime risk for the development of T2DM both in mother and child (101). Children born to mothers with and without GDM were categorized at birth for being large-for-gestational-age and appropriate-for-gestational age, and followed at ages 6, 7, 9, and 11 years. The results show an increased prevalence of IR in the large-for-gestational-age children born to a mother with GDM (102). Women with GDM have mitochondrial dysfunction with reduced placental mitochondrial content (22), which could potentially be transmitted to the child and affect fetal development. These observations are supported by studies in mice. One study showed that an imbalance in metabolic homeostasis led to oxidative stress and mitochondrial dysfunction in oocytes from IR mice both at the germinal vesicle and metaphase II stages and predicted poor oocyte quality (103). Maternal obesity prior to conception is coupled with altered mitochondria function in oocytes and zygotes that causes developmental deficiency as measured by blastocyst formation (104). Another study showed that mitochondrial dysfunction in muscle was passed to offspring from obese IR females potentially through mitochondrial dysfunction in the oocytes (4). Although there is evidence pointing toward the effects of the mother’s metabolism on the offspring’s predisposition to disease, it remains unclear whether poor-quality mitochondria can be passed down through the oocyte and cause metabolic abnormalities directly, or whether they create a stressful environment for fetal development and indirectly cause mitochondrial dysfunction in offspring.

Mitochondrial Dysfunction in Ovarian Function

Polycystic ovary syndrome (PCOS) is typically characterized by hyperandrogenism and oligo-ovulation, and occurs in 6% to 10% of women of reproductive age worldwide (11). POI, also referred to as premature ovarian failure, is distinguished by the loss or irregular function of ovarian follicles before age 40 years and is typically characterized by hot flashes, sleep disturbances, mood liability, and decreased energy with patients typically presenting with oligomenorrhea or amenorrhea (6, 105). Both disorders carry long-term risks. Individuals with PCOS have IR and are at elevated risk for the development of T2DM, cardiovascular disease, and endometrial cancer (106), whereas individuals with POI are at increased risk of suffering from T2DM (107), ischemic heart disease, and decreased bone density (6, 108, 109). Interestingly, the propensity to increased adiposity and risk of T2DM do not correlate with diminished ovarian reserve (110). Both of these reproductive abnormalities have been linked to mitochondrial dysfunction as a key modulator in disease progression. Whether impaired mitochondrial function has a causal effect, however, or simply serves to aggravate the disease, is not yet understood (Table 2). Age is a factor because female fertility decreases after peaking in the early 20s and severely decreases by age 35 years (5, 111). Importantly, both diseases play a pivotal role in the ability of women worldwide to reproduce, and finding new alternatives to natural reproduction has led many investigators to tackle the question of what mechanisms lead to impaired follicular development and the overall onset of both of these diseases.

Table 2.

Selected studies exploring links between mitochondrial dysfunction and ovarian function

Sample group	Observation	Reference
Human	mtDNA mutations in diabetic women with PCOS.	(117)
	mtDNA mutations and low mtDNA copy number in women with PCOS.	(11, 126)
	Higher levels of ROS in GCs from women with PCOS, reduced IVF-ET pregnancies.	(127)
	Decreased GSH and increased DNA damage in blood cells from women with PCOS.	(128)
	Reviews on role of mitochondria in reproduction and use of mitochondrial replacement therapies.	(113, 134)
	Ooplasmic transfer from fertile donor oocytes led to successful births but heteroplasmy was observed.	(140)
	Monogenic mitochondrial disorders associated with POI.	(6, 141-149)
	Cytoplasm transfer into oocytes from IVF recipients with history of poor embryo quality allowed successful fertilization.	(150)
Rodent	POLG mitochondrial mutator mice accumulate mtDNA mutations and have a premature infertility phenotype.	(121, 122)
	Resveratrol supplementation improved number and quality of oocytes in mice.	(135)
	Melatonin mROS and reduces ovarian aging.	(138)
	Fragile X premutation reduces mitochondrial number and quality in oocytes and GCs, and reduces fertility.	(151)
	Low expression of mitofusin-1 or -2 causes mitochondrial damage and POI.	(152, 153)
	Mitochondrial fission regulating GTPase DRP1 deficiency reduces oocyte quality.	(154)
	Single-cell RNA-seq identifies mitochondrial stress in oocyte aging.	(155)
	Mitochondrial UPR gene CLPP is required to maintain ovarian follicular reserve.	(156)
Bovine	Resveratrol improved ATP production from bovine oocyte and GC complexes from early antral follicles, and improved overall quality of oocytes.	(136)
	Resveratrol supplementation enhanced SIRT1 protein expression and improved oocyte quality and fertilization.	(137)

Abbreviations: ATP, adenosine 5′-triphosphate; ERα, estrogen receptor alpha; ESR1, estrogen receptor 1 gene; ET, estrogen therapy; GCs, granulosa cells; IR, insulin resistance; IVF, in vitro fertilization; mROS, mitochondrial reactive oxygen species; mtDNA, mitochondrial DNA; PCOS, polycystic ovary syndrome; POI, primary ovarian insufficiency; POLG, polymerase gamma mitochondrial DNA (D257A) mutator; ROS, reactive oxygen species; RNA-seq, RNA sequencing; SIRT1, sirtuin-1; UPR, unfolded protein response.

Oocyte quality and mitochondrial function

Both in PCOS and POI, impaired oocyte quality has been linked to disease onset and progression ultimately leading to disrupted ovulation (11, 104, 112, 113) (Fig. 2). Mitochondrial content increases dramatically during oogenesis with a primordial germ cell containing roughly 10 mitochondria per cell and ultimately having more than 100 000 total mitochondria per mature preovulatory oocyte (114). The mitochondria are small and spherical, and contain few cristae and a dense matrix (115). With this increased number of mitochondria, the probability of mtDNA mutations increases, which can also lead to heteroplasmy and could ultimately affect offspring (114); however, most mitochondria are inactive in the oocyte and become replicative only in the blastocyst stage (116). Consistent with this increase in mitochondria, mutations in genes controlling mitochondrial homeostasis, including those involved in fusion, fission, stress responses, and energy production, are found in individuals with PCOS or POI (6, 112, 117, 118). The increased mutational burden leads to an increase in ROS production by the mitochondria, ultimately driving premature aging (112, 119). Forty percent of infertile women undergoing in vitro fertilization (IVF) harbored mtDNA mutations in oocytes and blastocysts, and the oocytes with mtDNA mutations showed low fertilization rates (120). This has been borne out by studies in mice. The POLG mitochondrial mutator mice accumulate mtDNA mutations and have a premature aging phenotype leading to infertility by age 20 weeks (121, 122); however, a mouse heterozygous for this mutation does not show accelerated aging in spite of increased mutational burden (123). Although this is an attractive model, no increase in mtDNA mutational burden has been found in women older than 40 years (124). Mitochondrial copy number is now an established marker for embryo assessment during IVF, with a high mtDNA copy number correlating with implantation failure. Consistent with this finding, blastocysts from older women have higher mtDNA content (115). In contrast, mtDNA copy number has been negatively correlated to ROS levels, and a lower mtDNA copy number and quality have been seen in women with PCOS and POI (11, 125, 126).

Figure 2.

A schematic of the mature follicle in the ovary depicting disrupted mitochondrial homeostasis due to increased mitochondrial content in the mature preovulatory oocytes and granulosa cells, which increases the probability of mitochondrial DNA (mtDNA) mutations, activation of unfolded protein response, and increased production of reactive oxygen species (ROS) that ultimately attenuate oocyte quality and reproduction.

Mitochondrial dysfunction in granulosa cells

Oocyte quality depends on the ability of granulosa cells (GCs) to generate the energy required during oogenesis. As such, GCs are key players in oocyte development and ovulation. Accumulation of ROS in follicular fluid and GCs could be the cause of impaired oocyte quality and poor embryonic development in women with PCOS and POI (6, 11, 127). Indeed, a clinical study by Lai et al showed that 22 women with PCOS had significantly higher ROS levels in GCs than 25 women with tubal factor infertility, and these increased ROS levels induced apoptosis in GCs (127). Thus, tightly regulating ROS levels within the mitochondria is critical because ROS levels dictate the stress response, inducing the mtUPR, lipid uptake, steroid production, and cell death (112). Additionally, higher levels of testosterone seen in women with PCOS may increase the susceptibility to DNA damage (128). The mitochondrion is the site for cholesterol metabolism and sterol production. Cholesterol is transported to the inner mitochondrial membrane in the thecal cell by the steroid acute regulatory protein, where it is converted into progesterone and androgens, which are then absorbed by the GCs and converted to estrogens (6, 129). Thus, mitochondrial function is critical for steroid production. Women with PCOS have elevated androgen to estrogen ratios due to impaired aromatization (11). It is interesting to speculate that the increased ROS in GCs is due to the imbalance in steroid hormones as estrogen protects against mitochondrial dysfunction. Obesity has also been linked to increased androgen expression/secretion and low estrogen in patients with POI (11, 130).

Targeting mitochondria to improve ovarian function

In the ETC, electron carrier coenzyme Q10 serves the important task of transporting electrons from complexes I and II to complex III and also serves as an antioxidant (112, 131). For this reason, coenzyme Q10 has become a target for therapeutic development and improves oocyte quality in animal studies (132, 133). The NAD-dependent protein deacetylase sirtuin-1 (SIRT1) promotes mitochondrial biogenesis through effects on PGC1α acetylation. Resveratrol, a SIRT1 activator, increases mtDNA levels and membrane potentials, thus increasing ATP levels in oocytes, and decreases ROS levels as well as improving fertilization in animal studies (134-137). Similarly, melatonin can reduce mitochondrial oxidative stress and prevent fertility decline (138). It remains to be seen, however, whether these approaches will work in women. The idea that increasing mitochondrial energy production can restore oocyte quality and function during IVF was tested in studies by adding mitochondria or cytoplasm from donor oocytes that have functional mitochondria, to heterologous or dysfunctional oocytes (139, 140). Donor cytoplasm transfer injections have been shown to restore mitochondrial function and led to the successful delivery of 25 babies; however, this process has become a point of controversy because it carries an increased risk of heteroplasmy (139, 140). Thus, oocyte rejuvenation by enhancing mitochondrial function remains an attractive avenue of treatment to improve the rate of successful pregnancies, but more studies will be needed because of concerns over the detrimental effects of heteroplasmy.

Conclusions and Future Perspectives

During the last 25 years, many studies have reported changes in mitochondrial function in the pathogenesis of diseases, including T2DM and ovarian dysfunction. The finding that mitochondrial dysfunction and IR can be passed from mother to offspring suggests that mitochondrial dysfunction may be the common underlying defect that links metabolic imbalance with reproductive problems. Collectively, the studies highlighted in this review indicate that the relationship between mitochondria and insulin action or ovarian function is highly complex and interdependent. Given the transgenerational relationship of diet with mitochondrial function, IR, and reproductive function in animal studies (50), it would be informative to investigate whether mitochondrial dysfunction is responsible for the observed disease susceptibility in children of obese, IR mothers (157). Future studies should address whether such a causative relationship exists and whether correcting mitochondrial dysfunction can prevent the increased disease susceptibility in the offspring. How this inter-relationship works mechanistically also remains to be uncovered. The description of mitochondrial dysfunction is often inadequate and in-depth metabolic flux studies are needed to better understand the dynamic changes in mitochondrial function in response to environmental stresses in vivo and ex vivo. Finally, if we are to intervene to prevent transgenerational transmission, we need to know at what point in prenatal development mitochondrial dysfunction is detrimental (100).

Glossary

Abbreviations

ATP

adenosine 5′-triphosphate

CERs

ceramides

DAGs

diacylglycerols

ETC

electron transport chain

GDM

gestational diabetes

IR

insulin resistance

IVF

in vitro fertilization

mtDNA

mitochondrial DNA

mtUPR

mitochondrial unfolded protein response

OXPHOS

oxidative phosphorylation

PCOS

polycystic ovary syndrome

PGC1α

peroxisome proliferator-activated receptor gamma coactivator 1-alpha

POI

primary ovarian insufficiency

POLG

polymerase gamma mitochondrial DNA (D257A) mutator

QC

quality control

ROS

reactive oxygen species

T2DM

type 2 diabetes mellitus

TCA

tricarboxylic acid

Additional Information

Disclosure Summary: The authors have nothing to disclose. The authors declares no conflict of interest.

Data Availability

Data sharing is not applicable to this article because no data sets were generated or analyzed during the present study.