*6.3. The CIII2+CIV SC*

Isolated CIII<sup>2</sup> and CIV coexists with the CIII2+CIV SC, which has not been structurally characterized in mammalian tissues, likely due to its low relative abundance. Cryo-EM studies on yeast, which lacks CI, detected a CIII dimer at the core of the SC flanked by a CIV monomer on either side. The CIII–CIV interface revealed protein-protein interactions on either side of the membrane and with lipids within the membrane. The majority of interactions occur on the matrix side between Cor1/UQCRC1 and the N-terminus of COX5A, whereas the C-terminal domain of COX5A interacts with both Qcr6/UQCRH and cytochrome *c*1 on the intermembrane space. Within the membrane, COX5A contacts the N-terminal helix of Rip1/UQCRH and Qcr8/UQCRQ via a cardiolipin molecule and another lipid modelled as phosphocholine. Two other cardiolipins indirectly support the CIII–CIV interface highlighting again their crucial role in SC formation [75].

#### **7. SCs Assembly Factors**

Several molecules able to directly or indirectly promote mitochondrial SCs assembly and stability have been identified, such as cardiolipins, stomatin-like protein 2, prohibitin 1 and 2, and others (for a recent review, see [76]). Here we consider the protein factors involved in SCs assembly, the most studied being HIGD2A (hypoxia inducible domain family member 2A) [66,77–80] and COX7A2L [81]. HIGD2A mainly functions in CIV assembly [77,82] and mediates CIV integration within the respirasome, as well [66,80]. The mammalian homologous HIGD1A has been reported to assist the assembly of CIII and CIII-containing SCs biogenesis, having overlapping functions with HIGD2A [82]. These findings suggest the involvement of multiple pathways to assemble the respiratory complexes and to gather the SCs [82–84].

COX7A2L (also called SCAF1) has been the subject of various studies with contradictory results, due to the presence of two variants (long and short forms) differently expressed in mouse strains and tissues [81]. There is an agreement that long COX7A2L can bind to both CIII<sup>2</sup> and CIV and is required for formation of the CIII2+IV SC [79,83,84], as well as for the assembly of the megacomplexes [73], but is dispensable for the respirasome [66,80].

Assembly factors specifically interacting with CIII were identified in yeast and their orthologs validated in human cells, namely, Cbp3 (UQCC1) and Cbp6 (UQCC2) [85] and Cbp4 (UQCC3) [86]. The best-characterized is UQCC3 (ubiquinol-cytochrome *c* reductase complex assembly factor 3; also known as C11orf83), which is involved in the early phase of CIII assembly and in the stabilization of CIII-containing SCs [86,87]. Interestingly, UQCC3 was reported to be indispensable for simultaneously maintaining both OXPHOS and glycolysis during hepatocarcinoma cells hypoxia adaption, suggesting a role in energetic reprogramming [88].

Table 1 summarizes the genetic alterations affecting CIII structural subunits and assembly factors, which will be considered in the following paragraphs, also indicating the redox activities and the assembly state of complexes/SCs.

#### **8. Models of SC Organization**

The question of the organization of the respiratory chain has been debated since the pioneering studies aimed at identifying the molecular components and catalytic mechanisms [89,90]. Hackenbrock et al. (1986) described the "fluid" or "random collision model" of electron transfer, where each complex acts as an individual entity, CoQ and cytochrome *c* freely diffuse within the lipid bilayer, and electron transfer occurs during random and transient collision events [91]. The alternative solid-state model derives from the early observations reporting that CI and CIII preferentially associated in the native membranes [92]. Since 2000, the solid-state model has received strong support from isolation of SCs by BN-PAGE and then by development of protein crystallography and cryo-EM approaches. According to the solid model, a unique multicomplex unit is able to execute all the steps of respiration. More recently, the two models have been merged in the so-called "plasticity" model [64,93], based on the coexistence of individual CIII and CIV complexes, with a variable combination of SCs. Dynamic SC association/dissociation can be triggered under physiological conditions by availability of different substrates (NADH and FADH2), determining a variety of different structural options that allow to adapt the efficiency of the respiratory chain to metabolic demands. One consequence of the possibility to preferentially use NADH- or FADH2-dependent substrates through isolated complexes and SCs is the presence of partially dedicated pools of the mobile electron carrier CoQ [81,94]. However, demonstrating the presence of two CoQ pools is experimentally difficult and therefore this issue has been disputed [95,96]. According to the working model proposed by the Enriquez group, under normal circumstances, the superassembly in the respirasome generates a CoQ fraction within the SCs functionally dedicated to NADH oxidation. Given that also individual CIII<sup>2</sup> co-exist with the SC, CoQH<sup>2</sup> generated by CII or by other FADH2 dependent enzymes can be oxidized by the free CIII out of respirasome. Under conditions of block/lack of CIII or CIV, all CIII is associated with CI and CoQH<sup>2</sup> can diffuse out and be oxidized outside the SCs, and, on the other hand, the ubiquinol generated by CII can diffuse in and be oxidized by CI+CIII<sup>2</sup> SC [81,94]. Additionally, for cytochrome *c* the existence of a unique pool is also unlikely, given that this protein, besides transferring electrons between CIII and IV, can interact with many mitochondrial and non-mitochondrial components, exerting a variety of roles, among them the trigger of apoptotic cell death. SCAF1, which has been shown to be required for CIII and CIV interaction, plays an important role in the cytochrome *c* pool functional segmentation and likely in the efficient use of respiratory substrates. For a detailed discussion on CoQ and cytochrome *c* segmentation in SCs, we refer to [97].

#### **9. Functional Roles of the SCs and CIII Involvement**

Even if the evidence in support of SCs in detergent extracts is beyond question and the SCs in situ arrangement in the mitochondrial inner membrane has been defined, their physiological functional significance is still debated. The hypothesis of a catalytic advantage provided by SCs [67,81,98] has been questioned [65,72,95,96,99]. Catalytic advantage would imply substrate channelling, i.e., a defined conduit for the hydrophobic quinol from its reduction site in CI to CIII, where it is oxidized. However, no structural evidence for such a protein-defined conduit between CI and CIII<sup>2</sup> was obtained [100].

As alternative hypotheses, it has been proposed that SC organization could support CI assembly and stability [101,102], preventing dangerous casual protein aggregations within the membrane [100,103]. Finally, it has to be recalled that CI and CIII<sup>2</sup> are the main sources of reactive oxygen species (ROS) in the mitochondrial matrix and in the inner membrane [104–107], and that experimental dissociation of the SCs results in increased ROS production from CI [108]. It was, therefore, proposed that SCs organization could reduce ROS generation and subsequent oxidative damage to membrane components [108,109], although the molecular mechanism was not clearly defined. Analysis of recent structural data allowed to propose that in the respirasome, the interactions of CIII<sup>2</sup> with CI and CIV break the symmetry of CIII2, ensuring efficient oxidation of QH<sup>2</sup> and allowing CI to operate at full rate, thus helping to reduce ROS formation at CI [100].

The relationship between SCs and ROS is intriguing, because it is difficult to dissect how important SCs are to prevent ROS production and how ROS production depends on SCs disassembly. There is evidence that the production of ROS due to CIII dysfunction plays a role in the stability of SCs. It has been reported that some pathogenic mutations in *BCS1L*, the chaperone protein needed to incorporate the UQCFRS1 protein in CIII, affected CIII activity, inducing ROS production and secondary CI and CIV activity/stability alteration [110]. Furthermore, in a mouse cellular model lacking UQCFRS1 protein, the increase of ROS production due to CIII deficit induced a general reduction in the assembly and stability of CI, CIV, and in turn of SCs; however, this latter defect was partially rescued in presence of antioxidants treatment or hypoxia [111].

In addition to a direct ROS production by CIII due to its defective activity, it has been shown that the lack of CIII, by increasing the CoQH2/CoQ ratio, could promote the backflow of electrons from CoQH<sup>2</sup> to CI by the reverse electron transfer (RET) reaction through CI. RET produced the oxidation of cysteine residues of CI, triggering its degradation and in turn hampering SC formation [94]. Although in this study the occurrence of RET and ROS production was not directly demonstrated, the ROS involvement was suggested by partial rescue of CI assembly after treatment with the CI inhibitor rotenone and under hypoxic conditions, all interventions that prevent RET. The overexpression of SOD2, however, was not effective [94]. An interesting finding has been recently reported, showing that Na<sup>+</sup> modulates ROS production during acute hypoxia through the regulation of inner membrane fluidity [112]. Noticeably, this increased ROS production was associated with reduction in combined CII+CIII enzymatic activity and respiratory capacity, whereas combined CI+CIII activity and respiration remained unchanged. This finding reinforces the general idea of that formation of the CI+CIII<sup>2</sup> SC limits the production of ROS.

Finally, in a human cellular model carrying an 18-bp frame deletion in *MTCYB* associated with a severe impairment of CIII, CI and CIV assembly and ROS production, treatment with the antioxidant N-acetyl cysteine partially rescued respirasome formation [113]. Taken together, these data indicate that ROS production may affect respiratory complexes and SCs assembly, thus, it is reasonable to speculate that the correct assembly of SCs could be useful to reduce ROS production in a healthy respiratory chain.

The fact that the mutual arrangement of CI and CIII<sup>2</sup> is essentially conserved from obligate aerobic yeast to mammals, and plants seems to favour the hypothesis that specifically these two complexes have a functional role in maintaining the respiratory chain stability reducing ROS production [51]. Further structural and biochemical work is needed, also considering that in some cases a slight increase of ROS was detected, but the assembly of

respiratory complexes and SCs was normal, suggesting that ROS levels could modulate the response to this supermolecular organization [114–116]. The influence of tissue-specific subunit isoforms is also to be taken into account [83].

#### **10. SCs Biogenesis and Role of CIII**

An important issue to be elucidated concerns the SCs biogenesis, i.e., how CI, CIII and CIV interact in the respirasome formation. Currently, two models have been presented. The first model proposes that the close interplay among the three complexes favours better structural stability, implying that the isolated forms of respiratory complexes, in particular CI, are more prone to degradation. Accordingly, the complexes are supposed to follow separate assembly pathways to build mature individual complexes and to form SCs in a second step [93,117]. The second model suggests a central role of CI acting as a scaffold for the sequential incorporation of CIII<sup>2</sup> and CIV<sup>n</sup> subassemblies to form mature SCs [102]. However, this latter model does not rule out the occurrence of a dynamic exchange of CIII<sup>2</sup> and CIV once the respirasome assembly has been completed, allowing the formation of the other SCs and the co-existence with free complexes [66]. Both models have been developed from data obtained in experiments in which SCs biogenesis was studied following the time-course of respiratory complexes assembly after mitochondrial translation inhibition, by using diverse antibiotics and different experimental conditions, and this may explain some conflicting results. Recently, interesting information obtained by combining the Stable Isotope Labelling by Amino acids in Cell culture (SILAC) and complexome profiling techniques, suggested a new main role of CIII in SCs formation as a structural and functional platform for the overall respiratory chain biogenesis [118,119]. In particular, Protasoni et al., 2020 [118] analyzed human cells bearing a 4-bp *MTCYB* deletion that induced a frameshift with the loss of the encoded protein and of CIII2, associated with hampered CI biogenesis due to the stall in N-module incorporation and decreased CI stability. Furthermore, the CIV assembly was also defective because some CIV subunits were recruited within the accumulated CIII subassemblies. A similar analysis was carried out by Páleníková et al. (2021) [119], in cells bearing a 18-bp frame deletion in *MTCYB* gene that produced a protein shortened of six amino acids and induced a strong defect of CIII<sup>2</sup> activity/assembly as well as of CI and CIV [113,120,121]. These cells, differently from those with the 4-bp deletion, exhibited some amount of isolated CIII<sup>2</sup> and CI, that were associated in the enzymatically active CI+III<sup>2</sup> SC [113,121], in agreement with structural data demonstrating the existence of multiple interactions between the two complexes [74]. However, Páleníková et al. (2021) showed that several sub-complexes with mixed CIII, CI and CIV subunits were also present, indicating that the decreased CIII<sup>2</sup> assembly leads the formation of intermediates that trap other respiratory complex subunits impairing their assembly [119]. Taken together, these data support a cooperative assembly model in the respiratory chain structural and functional maturation and in SCs biogenesis, highlighting the central role of CIII<sup>2</sup> as scaffold for the ordered association with mixed CI and CIV subunits. Exhaustive studies focused on the mechanisms of SCs biogenesis in presence of CIII deficiency are still lacking; however, after reviewing literature it appears that mutations that disrupt CIII assembly also induce an impairment of CI and CIV and of SCs formation as well. Indeed, severe mutations in both CIII structural subunits or in early assembly factors strongly affect CIII structure and reduce CI and CIV stability [85,86,111,112,118,122–128] (Table 1). Unfortunately, several reported mutations that could be relevant for CIII structure, such as truncating mutations in cytochrome *b*, were poorly or not at all investigated for CIII assembly and related influence on CI, CIV and SCs structure (Table 1). However, it appears that most *MTCYB* missense mutations affect neither the assembly of CIII nor CI and CIV (Table 1), except for the p.Y278C mutation that was associated with a slight reduction of CI activity only and a reduced amount of the CIII<sup>2</sup> +CIV SC [129], and the p.E373K that disassembled CIII and CI, but not CIV [130]. Taken together, these findings indicate that the assembly of CI and CIV depends on the physical presence of assembled CIII species and not on their catalytic activity, although ROS production could play a role in specific cases [111].

However, the role of CIII structure in SCs biogenesis is far from being clarified; for instance, some authors suggested that the complete formation of the CIII2+CIV SC is important to safely activate CI only when the respiratory chain is fully assembled [102], but some data show that CI could interact with a pre-CIII to form enough SC to ensure sufficient respiratory chain activity [115,131]. These data suggest that regardless of the mutation, there is a tendency to maintain respiratory complex stability and SC assembly to mitigate CIII dysfunction. This latter piece of evidence is supported by recent papers showing that missense mutations in *MTCYB* that induce defective CIII enzymatic activity when detected in the isolated complex, are mitigated when CIII activity is measured under conditions in which the respiratory complex is organized into SCs [129,132]. Although further work is clearly needed, recent structural details obtained from cryo-EM analysis of active SC particles from sheep mitochondria highlighted the specific involvement of cytochrome *b* in the crosstalk between CI and CIII2, confirming its role in structural/functional interactions between the two complexes [74].

#### **11. Metabolic Disturbances and Treatment Options**

Defective CIII and associated perturbation of the supramolecular organization of CIII-containing SCs result in several significant metabolic alterations, which are briefly summarized in Figure 3 and discussed in detail below, providing also some hints on experimental treatments specifically aimed at improving these metabolic disturbances. For an extensive recent review of the therapeutic strategies to treat mitochondrial disease, we refer to [133].

**Figure 3.** Major metabolic alterations associated with perturbed supramolecular organization of CIII-containing SCs and proposed treatments.

#### *11.1. Unbalanced Intracellular Redox Homeostasis*

It is widely accepted that even under physiological conditions the electron flow through the mitochondrial respiratory chain results in mild ROS production [134,135], with CI and CIII being the main redox components responsible for molecular oxygen reduction to superoxide anions [104–107,136]. As mentioned above (point 9), one of the proposed roles of the supramolecular organization into SCs is to limit ROS formation, avoiding the diffusion of free radicals with damaging effects at protein and lipid levels [100,137,138]. Accordingly, different experimental conditions causing disruption or prevention of the association between CI and CIII were shown to increase ROS production, supporting the view that dissociation of SCs may strictly link oxidative stress and energy failure [108]. Furthermore, pathological conditions leading to dismantling the SCs organization, such those described in cells with 4- and 18-bp *MTCYB* deletions, were also shown to enhance ROS generation [94], and, due to up-regulation of intracellular efficient antioxidant defences in both cytosol and mitochondrial compartments, to cause a significant unbalance in the redox homeostasis [113,121,139]. Of note, mild to moderate increase in oxidative stress associated with parallel SCs depletion was detected in different brain areas of neuronspecific mice KO for the UQCRFS1 gene, demonstrating that ROS can modulate the SCs architecture to cope with a high level of ROS [140]. This conclusion is further supported by previous finding that a superoxide dismutase mimetic compound and SOD2 overexpression induced a partial increase in SCs in the UQCRFS1 KO cells [111,140]. Moreover, the prolonged treatment of cells bearing the 18-bp *MTCYB* deletion with N-acetyl cysteine (NAC) significantly increased the rate of ATP synthesis driven by CI substrates as well as the amount of free CI, CIII and CIV and of the respirasome. It is likely that NAC may provide optimal redox conditions for respiratory complexes interactions and SCs re-organization [113], also considering that both CI subunit NDUFB7 and CIII subunit UQCRH contain disulphide bonds [69].
