**4. Discussion**

The ANT has an important role in maintaining mitochondrial bioenergetics [21] and recently, it has been proposed to play a role in RCS formation [8]. Therefore, in this study, we sought to clarify whether genetic downregulation of ANT1, the main isoform of ANT found in the heart and skeletal muscle cells [1], affects RCS assembly in H9c2 cardioblasts. Our results demonstrate that ANT1 downregulation by 37% does not affect cell viability with no remarkable changes in mitochondria bioenergetics. Furthermore, the activity of all ETC complexes and the mitochondrial OCR was not dependent on ANT1; however, ANT1 appears to be important in the assembly (structural integrity) of the RCS, particularly the respirasome. Additionally, we demonstrate that hyperacetylation of mitochondrial proteins due to SIRT3 ablation stimulates RCS disassembly. The novel role of acetylation on RCS stability may provide additional information as to the mechanism of how acetylation of mitochondrial proteins is involved in the pathogenesis of cardiovascular diseases such as hypertrophy [22–24], IR [20,25,26] and heart failure [27,28]. The current study was performed in H9c2 cardiomyoblasts, but not in primary cardiomyocytes because the latter are quite sensitive to genetic manipulations. It should be noted that H9c2 cardiomyoblasts are more energetically similar (at least, in comparison with atrial HL-1 cells) to primary cardiomyocytes and can be successfully used to simulate an in vitro model of cardiac diseases [29].

Apparently, the role of ANT in the regulation of RCS assembly is not associated with its acetylation as *SIRT3* KO did not increase ANT acetylation in liver mitochondria. Interestingly, we are the first to demonstrate that acetylation per se affects RCS assembly, which could contribute to the mitochondrial dysfunction observed in *SIRT3* KO hearts [20,25]. Disruption of the ANT has been

linked to various cardiac diseases. In a mouse model of IR, ANT1 expression was significantly reduced, and cardiac-specific ANT1 overexpression prevented the detrimental e ffects associated with IR injury [7]. *ANT1* KO mice develop cardiac hypertrophy and lactic acidosis [2], similar to that observed in patients. Therefore, ANT1 has an important role in maintaining cardiac function and potentially mediating the detrimental e ffects associated with heart IR injury [30].

In our studies, *ANT1* KD increased cell number without a ffecting cell viability (Figure 1B–D). Previous studies have observed an increase in mitochondrial number, size [2,31], and upregulation of mitochondrial genes, including OXPHOS components [31] in *ANT1* KD hearts and skeletal muscle. It is tempting to speculate that *ANT1* KD cells display an increase in cellular proliferation as a reflection of upregulated mitochondrial genes and an increase in mitochondrial number and size as an adaptive response. The lack of any e ffects of ANT deficiency on cell viability might be explained, at least in part, by (i) insu fficient (37%) downregulation of ANT1 to induce mitochondrial dysfunction, or (ii) upregulation of other ANT isoforms, such as ANT2, as a compensatory mechanism, and their functional redundancy. Indeed, ANT2 has been shown to have opposite properties to ANT1 as it has been found capable of importing cytosolic ATP into the mitochondrial matrix [32], possibly maintaining normal mitochondrial function, although these findings are somewhat controversial [33]. In addition, ANT2 is regarded as a proliferative marker and correlated to loss of cell cycle control, which could partially explain why *ANT1* KD cells have an increase in cell number [32].

Interestingly, *ANT1* KD increased the number of total cells by 32% and alive cells by 22% (Figure 1) associated with a 36% increase of ATP levels (Figure 2B). The increase of ATP levels in *ANT1* KD cells might be due to the increase in cell number; however, this suggestion was excluded after normalization of ATP to total cellular protein (Figure 2B). Since aerobic (non-glycolytic) ATP production is coupled to ΔΨ m, we sought to examine the possibility of having disturbances in mitochondrial ATP production that could hint towards a glycolytic compensation. Previous studies have reported an increase in anaerobic metabolism and lactic acidosis [2,3] in ANT1 deficiency. Our results demonstrated that *ANT1* KD cells display a decrease in ΔΨ m (Figure 2B), which could be due to an impaired ETC activity and OXPHOS. However, neither we (Figure 2E–H) nor other groups using HEK293 cells [8] reported di fferences in the activity of individual ETC complexes due to ANT1 downregulation or ablation. In addition, we were unable to detect di fferences in basal and maximal mitochondrial respiration (Figure 3C,D) and ATP production (Figure 3E). Interestingly, although beyond the scope of our experiments, *ANT1* KD cells displayed a significant increase in cellular ROS levels (Figure 2C) and non-mitochondrial oxygen consumption (*data not shown)*. The production of ROS can occur outside the mitochondria, such as in the cytosol (xanthine oxidase, nitric oxide synthase), peroxisomes, and plasma membrane (NADPH oxidases) [34], possibly suggesting a cross-talk between ANT1-deficient mitochondria and other cellular compartments.

The physiological significance of the RCS is still under debate [35]. The mitochondrial RCS have been suggested to increase the e ffectiveness of electron transport through the ETC complexes, optimize ATP production, and reduce mtROS production by reducing electron leakage [11]. Disassembly of the RCSs, particularly the respirasome, was observed in cardiovascular diseases such as IR [16] and heart failure [36]. However, the mechanisms underlying the assembly of the RCSs, as well as their physiological role in the heart, are not fully understood. Our previous studies demonstrated that high Ca2+ and pharmacological/genetic inhibition of complex I (Figure 4C,D) stimulate disruption of the RCS in H9c2 cells and isolated mitochondria [12,37]. These studies suggested crosstalk between RCS assembly and permeability transition pore (PTP) opening as Ca2+ is the strong inducer of pore opening and complex I is the PTP regulator. This point is further supported by the current study that demonstrates that genetic downregulation of ANT, a PTP regulator, induces disorganization of the RCS. However, the cause–e ffect relationship between RCS and PTP seems to be more complex. Despite RCS disassembly, inhibition of complex I by rotenone prevented Ca2+-induced PTP opening in cardiac mitochondria [12], and *ANT1* KD did not increase mtROS, a PTP inducer in H9c2 cells (Figure 2D). Finally, we demonstrate that acetylation of mitochondrial proteins due to SIRT3 deficiency induces

RCS disassembly in an ANT-independent manner because there was no difference in ANT acetylation between WT and SIRT3−/− mitochondria (Figure 5). Disruption of the RCS could be a result of direct mechanisms involving disruption of protein–protein interactions due to changes in lysine residue charges, or indirect mechanisms through inactivation of RCS regulatory proteins (e.g., RCS assembly factors) due to their hyperacetylation.

In conclusion, this study suggests that ANT is involved in RCS assembly, although RCS may not be solely dependent on ANT. ANT may physically interact with ETC complexes I, III, and IV [8] and thus, be involved in the respirasome structure or play a regulatory role in the formation/maintenance of the RCS assembly. Further studies are required to elucidate the role of ANT in the structural integrity and regulation of RCS and other mitochondrial supercomplexes (e.g., ATP synthasome) in cardiac cells.

#### **5. Limitations of the Study**

We elucidated the contribution of only *ANT1* downregulation to mitochondrial bioenergetics and RCS assembly. ANT family proteins contain four isoforms (ANT1-4) that play a differential role and perform distinctly opposite functions in cell life and death. We were not able to verify protein expression of other ANT isoforms in *ANT1* KD H9c2 cells due to lack of reliable ANT2, ANT3, and ANT4 antibodies. Functional redundancy of other ANT isoforms could compensate for the effects induced by *ANT1* deficiency.

**Author Contributions:** Conceptualization: S.J. (Sabzali Javadov); Methodology: R.M.P.-R., S.J. (Sehwan Jang), C.A.T.-R., S.A.-P.; Validation: All authors; Formal analysis: R.M.P.-R., S.J. (Sehwan Jang), X.C.-D.; Investigation: All authors; Writing—original draft: R.M.P.-R.; Writing—review & editing: S.J. (Sabzali Javadov); Supervision: S.J. (Sabzali Javadov); Project administration: S.J. (Sabzali Javadov); Funding acquisition: S.J. (Sabzali Javadov).

**Funding:** This research was supported by the National Institute of General Medical Sciences (Grants SC1GM128210 and R25GM061838) of the National Institutes of Health.

**Conflicts of Interest:** The authors declare no conflict of interest.
