1. MCU Complex Structure and Function
The MCU complex is composed of pore forming subunits, i.e. MCU and its dominant-negative isoform, MCUb, the essential MCU regulator (EMRE), which allows interaction with the MICU mitochondrial calcium uptake regulatory subunits (MICU1, MICU2 and MICU3), and possibly other modulators, such as the mitochondrial calcium uniporter regulator 1 (MCUR1) (
Figure 1).
The molecular identity of MCU was discovered in 2011 by two different studies. MCU is a 40 KDa protein, highly conserved and ubiquitously expressed, located in the IMM. These reports show that the downregulation of MCU strongly reduces mtCa
2+ uptake in cells after Ca
2+ release from the ER after treatment with IP3-generating agonist. Notably, these changes occur without impinging on mitochondrial morphology and membrane potential. Coherently, MCU overexpression strongly enhances mtCa
2+ uptake after agonist-induced stimulation [
1,
2]. Structure analyses revealed that both the N- and C-termini of the protein face the mitochondrial matrix and that MCU contains two transmembrane domains linked by a highly conserved loop located in the intermembrane space, containing a “DIME” motif with negatively charged amino acids critical for Ca
2+ permeation [
3].
Raffaello et al., in 2013 described an alternative isoform of MCU, MCUb, a 33 KDa protein which shares high structural similarity with MCU. However, critical amino acids substitutions in the DIME motif explains why it exerts a dominant-negative effect, reducing the [Ca
2+]mt rise evoked by agonist stimulation. The opposite effect on mtCa
2+ uptake was obtained by MCUb silencing, confirming its role as a negative regulator of the channel [
4]. The MCU/MCUb ratio varies greatly between different mammalian tissues, and this impinges on the intrinsic capacity of mitochondria of a tissue to rapidly accumulate Ca
2+. Not surprisingly, the heart, that undergoes repetitive Ca
2+ spiking, at the risk of mitochondrial Ca
2+ overload, displays a high MCUb/MCU ratio, while skeletal muscle ensures maximal and sustained metabolic upregulation in phasic responses though a very low expression of MCUb. In general, the great variability in expression and stoichiometry of MCU and MCUb accounts, at least in part, for the wide differences in the MCU currents measured in different mammalian tissues [
5].
The Essential MCU Regulator (EMRE) is a 10 KDa metazoan-specific protein, identified by quantitative mass spectrometry of affinity-purified MCU complex components [
6]. It is inserted in the IMM by a single transmembrane domain and is required for the interaction between MCU and MICU1, acting as a bridge between the pore-forming and the regulatory subunits of the complex. Experiments performed on EMRE knockout cells clearly demonstrate that mtCa
2+ uptake is impaired, phenocopying the effect of MCU silencing. This evidence enlightens the critical role of EMRE as an essential component of the MCU complex, required for its activity [
6]. Coherently, reconstitution of the human MCU protein alone in yeast cells is not sufficient for uniporter activity, because the MCU channel is active only if MCU and EMRE are co-expressed [
7]. It is not surprising that the proteolytic regulation of EMRE is a finely-tuned process, essential for the formation of a functional MCU complex [
8].
MICU1 was the first member of the MCU complex to be identified in 2010. It is a 54 KDa protein located in the IMS where it regulates the activity of the MCU channel in Ca
2+-dependent way [
9]. Indeed, it contains two EF-hand Ca
2+ binding domains at its N-terminal sequence. This regulation ensured by MICU1 explains the sigmoidal response to cytCa
2+ levels of mtCa
2+ uptake. On one hand, at low cytCa
2+ concentrations, MICU1 keeps the channels closed and mtCa
2+ uptake is negligible, thus avoiding a constant entry of Ca
2+ into the mitochondrial matrix. On the other hand, when cytCa
2+ concentration increases, MICU1 acts as cooperative activator of MCU, leading to an exponential increase of mtCa
2+ uptake [
10]. An alternative splicing isoform of MICU1, MICU1.1, characterized by the addiction of a short exon (4 amino acids), was shown to be expressed predominantly in skeletal muscle, and at a lower level in the brain. Similar to MICU1, MICU1.1 act as a positive modulator of the channel, leading to even higher increases of mtCa
2+ uptake after stimulation. Interestingly, MICU1.1 when co-expressed with MICU2 does not show any reduction of [Ca
2+] peaks, implying that the MICU1.1/MICU2 heterodimer is less sensitive to negative regulation of MICU2 (see below) [
11].
Two paralogs of MICU1 residing in the IMS contributing to the regulation of mtCa
2+ handling: MICU2 and MICU3 [
12]. MICU2 has an expression pattern similar to MICU1, it contains two EF-hand domains and interacts with MICU1 and MCU, forming obligate heterodimers with MICU1 [
13]. Interestingly, MICU2 protein stability seems to be dependent on the presence of MICU1. Indeed, MICU1 silencing leads to the loss also of MICU2 [
12]. Patron et al., proposed an inhibitory effect of MICU2 on MCU activity at low cytCa
2+ concentration [
13], while Kamer et al., proposed that both MICU1 and MICU2 act as gatekeeping of the channel [
14]. Notably, another study indicates that MICU2 can regulate the Ca
2+ threshold of MCU activation [
15].
MICU3 is another MICU1 paralog mainly expressed in the nervous system, and to a lower extent in skeletal muscle. Similarly to MICU1 and MICU2, also MICU3 contains two EF-hand domains. Patron et al. demonstrated that MICU3 forms heterodimers exclusively with MICU1, and not with MICU2. Notably, MICU3 has a reduced gatekeeping activity compared to MICU1, mediating more rapid responses to elevated cytCa
2+ levels. Thus, MICU3 is proposed to be a positive modulator of MCU channel ensuring mtCa
2+ uptake in response to fast cytCa
2+ rises, typical of neuronal stimulation [
16].
MCUR1 is a 35KDa protein located in the IMM that interacts with MCU [
17]. Its downregulation leads to a decrease in mtCa
2+ uptake and ATP production [
18]. In contrast to this view, Paupe et al., showed that MCUR1 is not a regulator of the MCU complex, but it is a cytochrome c oxidase assembly factor [
19]. Thus, the precise role of MCUR1 related to mtCa
2+ homeostasis still remains to be fully elucidated.
2. Cardiovascular Diseases
The dysregulation of mitochondrial calcium (mtCa
2+) homeostasis occurring in acute myocardial ischaemia reperfusion (I/R) injury points to a critical role of MCU in cardiovascular diseases. Upon reperfusion, mtCa
2+ overload leads to excessive reactive oxygen species (ROS) production, promoting the opening of the mitochondrial permeability transition pore (PTP), which triggers cardiomyocyte death [
20]. Thus, MCU was conceptually expected to be involved in ischemia-dependent sensitization of cardiac muscle to reperfusion damage.
The analysis of hearts from MCU knockout mice showed not only unaffected basal cardiac parameters, but, surprisingly, were not protected from ischemia/reperfusion damage, as expected of a role of mitochondrial Ca
2+ loading upstream of permeability transition pore (PTP) opening. In addition, the hearts of MCU
-/- mice were not protected by treatment with the pore desensitizer cyclosporin A (CsA) [
21]. The same lack of protection was also observed in a transgenic mouse model in which MCU downregulation was achieved by the overexpression of a dominant-negative form of MCU (DN-MCU) [
22]. Overall, these results suggest that constitutive ablation of MCU (and hence loss of mtCa
2+ signals) from the embryonic phase could lead to mitochondrial compensatory mechanisms (e.g. in sensitivity to Ca
2+ or modulators of PTP or prevalence of other cell death pathways).
To overcome this problem, a mouse model with acute deletion of MCU in adult cardiomyocytes was generated [
23,
24]. In contrast to constitutive MCU deletion, the conditional knockout model showed protection from I/R injury and cell death, thus confirming the view of MCU-dependent dysregulated Ca
2+ signals upstream of PTP opening and myocardial cell death.
Interestingly, MCUb and mNCLX gene expression increases after ischemic damage to the heart, and their overexpression, that leads to a decrease in mtCa
2+ by limiting the uptake or enhancing the efflux respectively, is protective against I/R injury [
25,
26].
In this contest the role of MICU1 is still controversial. Indeed, while on one hand it is protective in early stages after reperfusion since its knockdown worsens the I/R damage, on the other hand MICU1 is also found to be upregulated during the late stages after reperfusion. However, the mechanism behind this increase is still not clear [
25,
27].
Together, these findings highlight the relevance of MCU modulation as potential therapeutic approach in the treatment of cardiovascular diseases. However, further studies are needed to translate these findings in a clinical approach.
3. Metabolic Diseases
Normal blood glucose concentrations are ensured by glucose-induced insulin secretion from pancreatic β-cells [
28]. In this process mitochondrial oxidative metabolism plays a key role. Indeed, an increase in cytosolic ATP level [
29] results in the closure of ATP-sensitive K
+ channels (K
ATP) [
30], leading to plasma membrane depolarization. This in turn promotes the opening of voltage-gated Ca
2+ channels, allowing Ca
2+ entry into the cell, ultimately leading to insulin release [
31].
Considering the critical role of mitochondrial oxidative metabolism in glucose-induced insulin secretion, a properly functional MCU complex is required in pancreatic β-cells for the correct functioning of this process [
32]. ATP rise upon glucose stimulation is impaired when MCU is downregulated [
33], resulting in a decrease in insulin secretion [
34]. Interestingly, not only the pore-forming subunit MCU, but also the regulatory subunit MICU1 is important for insulin secretion. Similar to MCU silencing, MICU1 downregulation leads to a decrease in ATP levels and glucose-induced insulin secretion in pancreatic β-cells [
34]. Surprisingly, the strongest reduction in insulin secretion in β-cells is observed when the regulatory subunit MICU2 is downregulated [
35]. Although the precise mechanism by which different subunits of the MCU complex influence insulin release still needs to be fully elucidated, the critical role of the MCU complex in the secretory function of pancreatic β-cells is undoubted.
MCU complex components are upregulated in insulin-resistant adipocytes and in mouse and human visceral adipose tissue (VAT) in conditions of obesity and diabetes. Interestingly, normal levels of MCU expression are restored in VAT of patients after bariatric surgery-induced weight loss. As for insulin secretion in pancreatic β-cells, also in this scenario not all the MCU complex components behave similarly, but a critical role emerges for MICU1, the only MCU complex component strongly upregulated during the transition from obesity to diabetes [
36]. These data suggest a key role of mitochondrial Ca
2+ dysregulation in obesity and diabetes, highlighting the relevance of MCU as a putative therapeutic target for the treatment of these metabolic diseases.
5. Skeletal Muscle Diseases
The molecular identification of MCU was followed by intensive studies on skeletal muscle aimed at characterizing the role of mtCa
2+ homeostasis in this tissue characterized with a specific physiology and Ca2+ signaling repertoire [
65]. The study of the first global MCU knockout mouse model exhibited the most prominent alterations in the skeletal muscle. As expected, in this model both resting mtCa
2+ concentrations and stimulated mtCa
2+ uptake were reduced. These alterations caused an impairment in mitochondrial oxidative metabolism with an increase in the phosphorylation level of pyruvate dehydrogenase, leading to a reduction in TCA cycle activity. The defective mitochondrial energetic control is responsible for the reduction in exercise performance and muscle force [
21]. Mammucari et al., studied the role of MCU in adult skeletal muscle, avoiding compensatory effects that can be present in the global knockout model. MCU expression was modulated through silencing and overexpression in vivo: overexpression of MCU caused muscle hypertrophy, while the silencing of MCU led to muscle atrophy. Interestingly, the control of muscle size and trophism by MCU, observed in both developing and adult muscles, did not depend on the effect on aerobic metabolism, but on the regulation of two major pathways of skeletal muscle hypertrophy, IGF1-Akt and PGC-1α4 [
66]. To further characterize the role of MCU in skeletal muscle physiology, Gherardi et al., generated a skeletal muscle-specific MCU knockout, characterized by myofiber-specific impairment of mtCa
2+ uptake. This triggered a decrease in muscle exercise performance and force, and a fiber-type switch, from slow to fast MHC expression. Notably, loss of MCU rewired skeletal muscle metabolism toward fatty acid oxidation [
67] (
Figure 3).
Interestingly, another MCU complex component, MCUb, the dominant-negative form of MCU, plays a key role during skeletal muscle regeneration, by modulating macrophage-driven stimulation and differentiation of satellite cells after muscle damage. In particular, MCUb was shown to drive macrophage polarization from the pro-inflammatory phenotype to the anti-inflammatory phenotype, with secretion of cytokines that promote satellite cells differentiation and fusion [
68].
The role of the MCU complex in skeletal muscle physiology is critical, and mutations in MICU1 gene were reported in human patients with a disease phenotype characterized by learning difficulties, a progressive extrapyramidal movement disorder and learning difficulties. Clinically, the disease was characterized by early onset proximal muscle weakness, intellectual impairment and elevated levels of serum creatine kinases. At the genetic level, different loss-of-function mutations were found, resulting in the loss of MICU1 protein. This leads to in increased mtCa
2+ load, increasing sensitivity to cell death stimuli but also resulting in lower cytoplasmic Ca
2+ level, potentially impinging on muscle contraction and synaptic transmission [
69]. To further understand the mechanisms behind this neuromuscular disease, Debattisti et al., characterized patient cells and skeletal muscle-specific MICU1 knockout mice. Lack of MICU1 was associated with a low threshold for MCU-mediated Ca
2+ uptake. Notably, MICU1 loss causes muscle atrophy and a decrease in force. The alterations in mtCa
2+ uptake during sarcolemma injury, leads to an ineffective muscle repair [
70]. Recently, the potential other side of the coin, the neural pathogenesis, was characterized by Singh et al. They generated a neuron-specific MICU1-KO mouse model showing progressive motor and cognitive degeneration. MICU1-KO neurons are more susceptible to mtCa
2+ overload and cell death, and this is reverted by the inhibition of the mPTP [
71].
MICU1 was shown to be critical also in another skeletal muscle disorder, the Barth syndrome, characterized by cardiolipin deficiency. Ghosh et al., utilized several Barth syndrome models including yeast, mouse model, and patient cells, and showed that cardiolipin is required for the stability of MICU1, which is reduced in Barth syndrome patient-derived cells, together with MCU and MICU2. The reduction in mtCa
2+ uptake results in reduced mitochondrial respiration [
72].
Finally, mtCa
2+ uptake was found to be critical in embryonal rhabdomyosarcoma (ERMS). Indeed, MCU expression is upregulated in ERMS and its downregulation causes mROS level decrease, and an increased propensity to differentiate, inhibiting the oncogenic phenotype [
73].