Huntington’s disease (HD) is an inherited illness caused by a dominant mutation in the hungtintin (HTT) gene, located on chromosome 4 (4p16.3). HTT codes for a 348 kDa scaffold protein which participates in a series of different processes: vesicle trafficking, cell division, ciliogenesis, endocytosis, autophagy and transcription regulation ^[1][45]. The mutation involves an abnormal repetition of a CAG trinucleotide sequence in exon 1 of HTT ^[2][64]. The CAG repeats are responsible for the production of an abnormal form of the huntingtin protein that contains an expanded polyglutamine (polyQ) tract named mutant huntingtin (mHTT). The length of the CAG repeat is inversely correlated with the age of onset of HD, with longer CAG repeats associated with earlier onset and more severe symptoms ^[3][65]. The exact mechanisms by which mHTT causes neurodegeneration in HD are complex and not yet fully understood. However, it is known to mediate toxicity through the formation of protein aggregates whithin cells. The toxic effects of mHTT aggregates result in impaired cellular functions, disrupted intracellular signaling pathways and altered gene expression. These effects lead to mitochondrial dysfunction, oxidative stress, and excitotoxicity, and other pathological processes ^[4][66].

The relationship between HD and mitochondrial dysfunction has been an area of active research. Subsequently, studies have shown that mHTT can directly interact with mitochondrial components (such as protein transporter TIM23), disrupting mitochondrial dynamics and impairing mitochondrial function ^[5][6][67,68]. Dynamic events allow mitochondria to adapt to changing cellular needs, maintain their integrity, and coordinate energy production. Mitochondrial fission and fusion are crucial processes that ensure mitochondrial dynamic and regulate the morphology, function, and homeostasis of mitochondria. In HD patients, the expression of pro-fusion proteins MFN1, MFN2 and OPA1 was downregulated, while DRP1 and FIS1 were upregulated ^[7][69]. These data sustain an imbalance between mitochondrial fusion and fission in HD. The increase of mitochondrial fission in HD causes more fragmented and less motile mitochondria. This excessive mitochondrial fragmentation can be due in part to the interaction of mHTT with DRP1, ^[8][44] which significantly increases the activity of the latter. Consequently, expanded polyQ tracts induce mitochondrial morphology alteration and fragmentation ^[9][70]. These alterations were associated with a decrease in PGC-1α levels, a protein implicated in mitochondrial biogenesis and function ^[8][44]. PGC-1α regulates the expression of antioxidant enzymes that protect cells against an excessive production of ROS emanating from oxidative phosphorylation activity ^[10][71]. It has been shown that mHTT affect mitochondrial biogenesis by interacting and thus interfering with cAMP response element-binding protein (CREB) and TATA-binding protein-associated factor 4 (TAF4) ^[11][72]. This leads to a reduction in PGC-1α activity that affects the antioxidant enzyme superoxide dismutase (SOD2) gene expression ^[11][72].

While some studies reporting mitochondrial respiration impairment did not observe evidence supporting changes in the activity of the oxidative phosphorylation (OXPHOS) complexes ^[12][13][50,73], others have found a decrease in their activity, particularly affecting complex II and III ^[14][15][74,75]. Despite all of that, ROS production in HD samples was seen augmented, even before the clinical manifestations appear ^[16][53]. Moreover, studies focusing on TCA suggest a decrease in aconitase, α-ketoglutarate dehydrogenase (α-KGDH) and succinate dehydrogenase (SDH) activities in mice and human brain HD samples ^[17][18][76,77]. However, other research has also found an increase in SDH activity in human and mice HD cortex samples, with a rise of aconitase found only on mice ^[19][78]. Overall, the TCA cycle is not spared in HD, it being the primary source of ATP generation within cells ^[20][25]. Therefore, the alterations of enzymes involved in TCA cycle may contribute to energy deficits and metabolic disturbances in HD.

All of these studies support an alteration in mitochondrial function. Consequently, the downregulation of mitochondrial biogenesis in HD may be the trigger for excessive mitochondrial fission. This mechanism does not appear to be compensatory during mHTT-mediated toxicity, as a partial inhibition of mitochondrial fission ^[8][44] was enough to restore a protective balance in mitochondrial dynamics.

Alhtough mitochondrial biogenesis was reduced, an increase in mitochondrial mass was observed ^[21][43]. Of note, HTT participates in the formation of the autophagy initiation complex by binding ULK1 and promoting the interaction between ubiquitin receptor p62, ubiquitinated substrates and the autophagic vesicle marker LC3-II ^[1][45]. In addition to its role in the initiation of autofagosome formation, HTT is involved in the organelle and vesicle trafficking through a direct interaction with dynein or through the hungtintin-associated protein 1 (HAP-1) ^[22][57]. The association of mHTT with HAP-1 and the sequestration of other components of the trafficking process, such as microtubules or P150, has been demonstrated to impede mitochondrial trafficking in mice and in human brain samples ^[23][24][79,80]. Indeed, mHTT impairs the initiation complex, leading to a decrease in autophagy and subsequent impairment of mitophagy ^[21][43]. This is in consistence with what was observed in striatal cells, where despite the ubiquitination of depolorized mitochondria, mHTT lead to the accumulation of damaged mitochondria by precluding the recruitment of mitophagy receptors, such as OPTN and CALCOCO2, to LC3-II, which is associated with the autophagosome membrane ^[21][43]. The specific mechanism by which mHTT interferes with the recruitment of mitophagy/autophagy components was not established, However, it will be interesting to investigate whether mitophagy receptors were also located in ubiquitinated damaged mitochondria. An interesting fact is that mHTT does not seem to interact with the ubiquitin system, which provide an alternative pathway independent to the Parkin protein. However, it still requires PINK1, the kinase protein implicated in ubiquitin-dependent mitophagy, which was essential for ensuring a basal mitophagy process in neuronal HD. Accordingly, overexpression of PINK1 lead to a partial removal of mitochondria in striatal cells derived from HD mice ^[25][81].

Together, all of these data do not distinguish whether mitochondria are a direct target of mHTT, or if mitochondrial dysfunction is a collateral damage resulting from the impairment of signaling pathways, such as Ca²⁺ signaling. In HD, aberrant calcium signaling has been implicated in the pathogenesis of the disease ^[26][61]. The mHTT protein can alters NMDA receptor (NMDAR) activity or impair the receptor signaling ^[27][82]. NMDAR is an ion-channel glutamate receptor and when activated, it allows the entry of calcium into neurons (81). In parallell, a scaffold protein named postsynaptic density-95 protein (PSD-95) interacts with NMDAR. This interaction enables the clustering of NMDAR at the postsynaptic membrane and inhibits their negative regulator STEP61, thereby modulating their activation and signaling ^[28][29][83,84]. In postmortem studies and mouse model of HD, the association between NMDAR and PSD-95 is enhanced, causing NMDA-mediated exitotoxicity and neuronal death ^[27][30][82,85]. Moreover, mHTT can bind to inositol-3-phosphate receptors (I3PR) via HAP-1, leading to the release of Ca²⁺ from the ER ^[31][86].

This increase in calcium influx compromises the capacity of Ca²⁺ uptake into mitochondria ^[26][32][61,87], resulting in mitochondrial membrane depolarization. Indeed, mitochondria from HD are more vulnerable to calcium overload, and mitochondrial depolarisation is preceded by a change in mitochondrial membrane potential (MMP) until the calcium burden causes the opening of the mPTP ^[33][56]. This opening leads to the release of sequestered calcium from mitochondria and subsequently triggers apoptosis in various HD models. These events are supported by the increase in cytosolic levels of Smac/DIABLO, which inhibits the anti-apoptotic properties of other proteins, such as X chromosome-linked inhibitor of apoptosis (XIAP) and inhibitor of apoptosis protein-1 (IAP1), in striatal cell lines expressing mHTT ^[34][88]. The binding of mHTT to p53 facilitates its nuclear translocation and, therefore, the gene expression of some pro-apoptotic factors, such as Bax ^[35][89]. Both mRNA and Bax protein levels were increased in R6/1 mice brain samples ^[36][37][90,91]. A study conducted in HD mice demonstrated that caspases 1, 3, 8 and 9 were progressively more active, with caspase 1 being the first to become active at 7 weeks of disease progression ^[37][91]. Also, caspase-2 can cleave mHTT, producing toxic N-terminal fragments, and caspase-7 interaction with mHTT is known to activate other caspases ^[11][72]. These events emphasize the detrimental effects of mitochondrial dysfunction, impaired autophagy/mitophagy, and the interaction with calcium signaling and apoptotic pathways in the development of HD and its neurodegenerative processes. Based on the consequences of mitochondrial dysfunction in HD, potential therapeutic strategies could target enhancing mitochondrial function and restoring energy production through effective mitophagy/autophagy activiation.

The chromosome 9 open reading frame 72 or C9orf72 is a gene located in the short arm of chromosome 9 (9p21.2) ^[38][92]. C9orf72 was first described in 2011 following studies in familiar frontotemporal dementia (FTD) and amyotrophic lateral sclerosis (ALS) cases that described a pathogenic GGGGCC hexanucleotide repeat expansion (HRE) in the first intron of the gene. This pathogenic HRE is the primary cause of familiar forms ^[38][39][92,93] and is also responsible for approximatively 6% of the sporadic cases ^[40][94]. C9orf72 codes for a protein of the same name, mainly expressed in neurons, that forms part of a multimeric complex with Smith–Magenis chromosome region 8 (SMCR8) and WD40-repeat containing protein 41 (WDR41) proteins ^[41][95]. The C9orf72-WDR41-SMCR8 complex was initially proposed to have GEF activity on Rab GTPases, such as Rab8A and Rab39B, resulting in their activation ^[42][43][44][49,96,97]. However, recent in vitro studies have suggested opposing GAP activity ^[41][45][46][95,98,99]. Nevertheless, C9orf72-WDR41-SMCR8 participates in the regulation of a wide variety of processes, for example autophagy, actin dynamics, endo-exocytosis and inflammation ^[47][100]. There are three pathogenic molecular mechanisms associated to the pathogenic HRE (1) the large HRE is transcribed into sense and antisense repeat RNA which form aggregates termed RNA foci. The HRE-containing RNA transcrips may adopt different secondary structures, being one of the most characteristic RNA G-quadruplexes ^[48][101], which can interact with several RNA-binding proteins, interfering with their functions. Examples of such proteins are the heterogeneous nuclear riboprotein or hnRNP family (being the most common hnRNP H) ^[49][48], nucleolin ^[48][101], zinc finger protein 106 (Zfp106) ^[50][102] and Pur-α ^[51][103]. (2) The expanded RNA molecules can be translated in both direction into five different dipeptide-repeat proteins (DPRs), which form protein aggregates in C9orf72-ALS/FTD patient brain and spinal cord. Some of them have shown toxic properties in animal models ^[52][53][16,104]. (3) Finally, the HRE affects the expression of the gene provoking a reduction in C9orf72 protein and so alterations in the pathways in which the protein takes part. This loss of function mechanism has been also proposed to synergize the gain of toxicity mechanisms described above ^[54][55][105,106].

Despite C9orf72 recent discovery 12 years ago, there are many evidences of mitochondrial deregulation in C9orf72-mediated ALS/FTD. While in prefrontal cortex post-mortem samples of C9orf72 patients the number of mtDNA showed a 50% reduction in contrast to control samples ^[56][107], mtDNA and mitochondrial mass were increased in C9orf72 patient-derived fibroblasts, due to the upregulation of PGC1-α protein levels ^[57][108]. DPRs are a crucial element of C9orf72 pathogenesis. As such, it is expected that they could interact with mitochondria. Indeed, there are many evidences that mutant C9orf72-derived DPRs, particularly those with arginine, can cause mitochondrial damage. For example, a study in a mouse model expressing polyGR₈₀ found shorter, less motile mitochondria in cortical neurons. These changes could be explained by an increase in DRP1 and a decrease in OPA1 ^[58][52]. In contrast, C9orf72 patient-derived fibroblasts displayed elevated MFN1 protein levels ^[57][108].

Nevertheless, these data shows that mitochondrial dynamics in C9orf72 are altered in cortical neurons towards an increase in mitochondrial fission. Regarding this, the increase in mitochondrial biogenesis could be explained as a compensatory mechanism to replace damaged mitochondria ^[5][13][67,73].

C9orf72 HRE also affects mitocohondrial metabolism. OXPHOS complex I and ATPase activities in mice expressing polyGR₈₀ were reduced. Additionally, a reduction in complexes I and IV gene expression has been reported in C9orf72-human induced pluripotent stem cells (iPSC)-derived MNs ^[59][60][109,110]. PolyGR can bind to ATP synthase F1 subunit alpha (ATP5A1), one component of ATP synthase complex, favoring its proteasomal degradation ^[58][52]. The reduction of complex I activity can be explained by a reduction in C9orf72 protein levels, which prevents the degradation of the translocase of the inner mitochondrial membrane domain containing 1 protein (TIMMDC1), needed for the assembly of OXPHOS complex I ^[61][111]. Dysfunctions in OXPHOS can provoke an increase in ROS generation, as observed in C9orf72 patient fibroblasts, iPSC-derived MNs transduced with a polyGR80 lentivirus and NSC34 MNs transfected with GFP-polyGR₅₀ or GFP-polyPR₅₀ ^[57][62][63][47,108,112]. In the latter cell model the impairment in NRF2 synthesis, an enzyme with antioxidant activity, was also reported ^[62][47]. This data confirms the hindrance of mitochondrial metabolism in C9orf72 pathogenesis, which will affect ATP production and energy-dependent celullar processes.

A decrease in autophagic clearance of mitochondria may also contribute to the increase in mitochondrial mass observed in C9orf72-related disorders. It is important to remark the widely accepted role of C9orf72 protein in autophagy and so in the degradation of damaged/old mitochondria via mitophagy. A growing body of evidence has revealed multiples roles for the C9orf72 protein in autophagy; while some studies suggest a role in the activation of autophagy via the interaction of the C9orf72/SMCR8/WDR41 complex with different proteins involved in the ULK complex activation of autophagy ^[55][64][106,113]. On the contrary, others suggest a negative regulation of autophagy via upstream modulation of mTORC1 signalling ^[65][66][67][114,115,116]. Nevertheless, loss of C9orf72 protein has a known impact on autophagy ^[68][59]. The lysosomal localization of C9orf72 also argues for a role of the protein in autophagy degradation ^[69][70][71][117,118,119]. Recently, a study linked disruption of nucleocytoplasmic transport of the autophagy-lysosomal transcription factor TFEB caused by C9orf72-HRE with impaired autophagy. Interestingly, in this case, the effect on autophagy was not linked to haploinsufficiency of the C9orf72 protein but to the expanded transcripts ^[72][120], demonstrating that pathogenic HRE in C9orf72 alters autophagy. Consequently, C9orf72 HRE might alter mitophagy, impairing their degradation and thus promoting an increase in their numbers. However, mitochondria spared from autophagic degradation may already be damaged. Thus, their accumulation will have negative effects in ROS generation and ATP production.

One of the main regulators of mitochondrial activity is Ca²⁺. Deregulation of mitochondrial Ca²⁺ levels disrupts correct mitochondrial metabolism and is also responsible of increased cell death ^[73][38]. Indeed, Ca²⁺ dyshomeostais have been reported in iPSC-derived MNs from patients carrying the C9orf72 mutation. These neurons showed higher Ca²⁺ cytosolic concentration than control cells, paired with a lower Ca²⁺ buffering capacity of their mitochondria ^[22][57]. These defects were linked to a decreased expression of the mitochondrial Ca²⁺ uniporter (MCU) and its regulatory protein mitochondrial Ca²⁺ uptake 2 (MICU2) ^[22][57]. Likewise, mitochondria contacts with the endoplasmic reticulum (ER) regulate key cellular processes such as Ca²⁺ homeostasis or autophagy ^[74][121]. Reduced ER-mitochondria contacts have been reported in iPSC-derived cortical neurons from patients carrying pathogenic C9orf72 expansions and in affected neurons of a mutant C9orf72 transgenic mice. These studies involved the toxic DPRs poly-GA, poly-GR and poly-PR in the mechanism ^[75][122]. In addition to metabolism alterations, mitochondrial Ca²⁺ levels imbalances can affect MMP, as seen in C9orf72 patient-derived fibroblasts and iPSC-derived MNs expressing polyGR80, where an increase in MMP was found ^[57][63][108,112] and in GFP-polyGR50/polyPR50 expressing NSC34 MNs-like cells, where a decrease in MMP was observed ^[62][47].

However, while a reduction in mitochondrial Ca²⁺ could represent a decrease in mitochondrial-mediated apoptotic cell death, this is not what happens in C9orf72 FTD/ALS. Instead, an increase in apoptotic cell death has been widely described in different C9orf72 models, although neuronal lineage cells tend to be more susceptible to this process ^{[76][77][78][79]}[62,123,124,125]. Upregulation of Bax mRNA, higher Bax protein levels and cleaved caspase 3 as a result of increased p53 phosphorylation was found in iPSC-derived C9orf72 MNs ^[80][126]. Moreover, iPSC-MNs carrying the HRE exhibit reduced anti-apoptotic Bcl-2 levels and Bcl-XL mRNA, while there was an upregulation of proapoptotic BAK mRNA and elevated cytochrome c release in comparison to control iPSC-MNs ^[59][109]. Additionally, SH-SY5Y cells and mice primary cortical neurons treated with synthetic PR20 peptides showed higher p53 and Bax protein levels ^[81][127]. The upregulation of apoptosis in absence of increased mitochondrial Ca²⁺ levels could mean that, at least in C9orf72 FTD/ALS, other sources of mitochondrial damage would contribute more to the apoptotic process.

Myotonic dystrophy type 1 (DM1) is an autosomal dominant neuromuscular disease with mutlisystemic features caused by the expansion of a CTG trinucleotide found in the 3’ untranslated region (UTR) of the DMPK gene, located on chromosome 19q13.3 ^[82][128]. DMPK codes for a serine/threonine protein kinase involved in the correct skeletal muscle structure and function, involved in processes such as Ca²⁺ homeostasis, muscle-related gene regulation, nuclear envelope organization and myotube differentiation, while also playing a role in cardiac muscle activity and synaptic plasticity ^[83][129].

The three common molecular mechanisms reported for other repeat expansion diseases have also been described for DM1. Firstly, the expansion affects the levels of DMPK leading to a decrease in the protein, a serine threonine kinase involved in maintaining the skeletal muscle structure and function ^[84][130]. Secondly, mRNA molecules containing the expanded tracts can exist as aggregated intranuclear RNA foci ^[85][46]. They interact with CUG RNA binding proteins, preferentially muscleblind-like (MBNL) family members and CUG binding protein Elav-like family member 1 (CELF1). These proteins are RNA metabolism regulators, implicated mainly in alternative splicing. The union of CUG-containing RNAs to them alters their function, causing abnormal splicing pattern leading to many of the biochemical alterations observed in DM1 ^[85][46]. Thirdly, the expanded transcripts can generate up to five different homopolypeptides via RAN translation ^[86][6].

Mitochondrial alterations in DM1 patient samples have been supported by the presence of increased mtDNA deletions ^[87][131]. The mitochondrial genes affected by these deletions are CO3 gene and ND5 gene, which encode for cytochrome c oxidase (COX) (known as complex IV) and NADH dehydrogenease subunit 5 of the complex I, respectively ^[87][131]. These two complexes are members of the OXPHOS proteins involved in the electron transport chain of mitochondrial respiration, enabling oxidative phosphorylation and ATP synthesis ^[88][27]. Therefore, mutations in one of the OXPHOS proteins could compromise mitochondrial metabolism, as previously reported in fibroblasts from DM1 patients ^[89][54].

Further evidence support an impairment of muscle oxidative metabolism in DM1 patients, which result in reduced ATP levels and compromise muscle performance ^[90][132]. Although there are no consistent data establishing a direct relation between mitochondrial mutations and metabolism impairment in DM1, gene and protein expression levels of complexes I to IV were downregulated in DM1 patients and mice muscle samples ^[91][55]. Rewsearchers think that the decrease in inner mitochondrial membrane proteins, including Coenzyme Q10 (CoQ10), could be partly responsible for ROS generation in DM1. Interestingly, defects in mitochondrial metabolism and the increase in ROS generation observed in DM1 patient-derived fibroblasts was mitigated by metformin, an activator of AMPK ^[89][54]. Moreover, all these changes could be partially reversed by aerobic exercise ^[92][51] through the activation of AMPK pathway, which restores the level of OXPHOS proteins to that of control in DM1 ^[91][55].

Additionally, the increase of mitochondrial fission (phosphor-DRP1Ser616), autophagic and mitophagic markers (BNIP3 and Parkin) in DM1 models allow people to think that basal autophagy might be triggered, but it is not enough to conduct the clearance of dysfunctional mitochondria and needs to be enhanced. It would be interesting to study whether mitochondria are being sequestered for degradation or if the degradation process is temporary delayed, as reported in DM1-derived fibroblasts ^[93][60].

However, efficient degradation is necessary to maintain cellular homeostasis and prevent the accumulation of potentially harmful material. A deficit in mitochondrial turnover may accelerate DM1 muscle wasting when calcium levels are disrupted. This increase in calcium is due to defects in the alternative splicing of the voltage-dependent T-tubule Ca²⁺ channel CaV.1.1 ^[94][133], the sarcoplasmic reticulum Ca²⁺ ATPase 1 (SERCA-1), the dihydropyridine receptor (DHPR) present in T-tubules and the ER ryanodine receptor (RyR1) ^[85][46]. Importantly, SERCA-1 and RyR1 are critical for calcium release and uptake in the sarcoplasmic reticulum (SR) of muscle cells. Therefore, if SR fails to pump Ca²⁺, dysfunctional mitochondrial will be unable to buffer increased cytoslic calcium levels. This can subsquently lead to damage organelles and cell death. However, further researchs are needed to establish a clear link between mitochondrial turnover and calcium homeostasis in DM1.

Another molecular mechanism potentially related to mitochondrial alterations in DM1 could involve the loss of function of DMPK, which is caused by the presence of the pathogenic expansion ^[49][48]. Through alternative splicing, up to seven DMPK isoforms (A to G) may be synthesized, each with a common N-terminus leucine-rich domain, a kinase domain and a coiled coil domain. Variations in the C-terminus tail and the presence or absence of a VSGGG motif are responsible of the A to F isoforms of DMPK, with the exception of the human-only isoform DMPK-G, produced via the rare splicing event of a sixteenth exon ^[95][134]. Human DMPK-A is found primarily on skeletal muscle fibers, among other tissues, where it is known to bound to the OMM ^[96][97][135,136]. One of its roles is to protect the cells from ROS damage and apoptosis by forming a multimeric complex with the Src kinase and hexokinase II (HK-II) in the OMM ^[96][135]. HK-II is the first enzyme of the glycolysis pathway and a negative regulator of the permeability transition pore (PTP) via its OMM localization ^[98][137], while Src is a tyrosine kinase able to sense ROS thanks to cysteine residues ^[99][138]. Thus, the reduction of DMPK protein levels associated to the CTG expansion may lead to an increase in ROS generation and apoptotic cell death. Additionally, a study shows that human DMPK-A, in a C-terminal tail dependent but kinase independent manner, causes mitochondrial morphology alterations and perinuclear clustering, which ultimately leads to apoptotic cell death ^[97][136].

MBNL1 is one the most studied proteins associated to DM1 pathogenesis. It is a nuclear protein that participates in alternative splicing regulation during skeletal muscle differentiation, affecting genes such as cardiac troponin-T (c-TNT), insulin receptor (IR) and SERCA-1, among others ^[100][101][139,140]. It recognizes and binds to YGCY sequences in mRNAs, allowing or repressing their processing by competition with other splicing factors ^[102][103][141,142]. MBNL1 is also able to interact with CUG(exp) tracts found in DM1, involving the RNA gain-of-function mechanism of the expansion. In consequence, the RNA expansions sequester MBNL1 in nuclear foci, reducing the available levels of this protein and therefore preventing its functions ^[85][46]. A recent study by Yokoyama and colleagues suggests that MBNL1 knockdown in C2C12 myotubes decreases mRNA and protein levels of mitochondrial biogenesis marker PGC-1α and increases Bax/Bcl-2 ratio, leading to an increase in apoptosis of these cells ^[104][63]. However, another previous research showed no changes in PGC-1α levels in DM1 patient fibroblasts ^[89][54]. These results highlight the requirement of a more extensive research.