An AAA-ATPase links mitochondrial division with DNA nucleoids

Mitochondria are unique organelles with their own circular genome in the matrix (1, 2). In humans, the 16.5 kb of mitochondrial DNA (mtDNA) encodes only 37 genes, including 22 transfer RNAs (tRNAs), two ribosomal RNAs (rRNAs), and 13 proteins. These proteins are subunits of the electron transport chain complexes, essential for cellular energy production, and account for approximately 1% of the total mitochondrial proteome. Depending on cell type, the copy number of mtDNA varies, ranging from several hundreds to thousands per human cell. Each mtDNA molecule is packaged into DNA–protein complexes called nucleoids and is distributed throughout mitochondria. Mitochondrial nucleoids play important roles in DNA replication and transcription. A major protein of mitochondrial nucleoids is mitochondrial transcription factor A (TFAM), which directly binds mtDNA and compacts it into roughly 100-nm structures. The depletion of TFAM affects nucleoid structure and impairs mitochondrial respiratory activity. Along with TFAM, nucleoids also contain many other proteins involved in DNA replication and gene expression. Nucleoids move within mitochondria to ensure their even distribution; however, the underlying mechanisms are largely unknown.

Mitochondria are highly dynamic organelles, continuously dividing and fusing (3, 4). In addition to intramitochondrial movements of nucleoids, these membrane dynamics play a vital role in nucleoid distribution and activity. In particular, mitochondrial division is crucial for the segregation of mitochondrial nucleoids. In the absence of mitochondrial division, for example, in cells lacking an essential mechanochemical GTPase for mitochondrial division (Drp1 for humans and Dnm1 for yeasts), nucleoids become prone to clustering in extensively connected mitochondria (5, 6). Interestingly, mitochondrial division occurs in the vicinity of mitochondrial nucleoids, and mtDNA replication is spatially coupled to mitochondrial division occurring at ER–mitochondria contact sites (5, 7). This regulation of mitochondrial division facilitates equal inheritance of mtDNA into two daughter mitochondria after DNA synthesis. Since mitochondrial division machinery is located on the outer membrane, a link connecting the outer membrane–located division machinery to matrix-located nucleoids is predicted. Yet, its molecular identity remains elusive.

In an article in PNAS, the excellent work by Ishihara et al. (8) discovered this missing link bridging the mitochondrial division machinery and nucleoids and its role in nucleoid movement (Fig. 1). These authors tested whether ATPase family AAA-domain containing protein 3A (ATAD3A), an inner membrane ATPase carrying an AAA (ATPases Associated with diverse cellular Activities) domain with a single transmembrane stretch, has a role in nucleoid dynamics; previous studies have reported that ATAD3A is associated with nucleoids and is vital for nucleoid morphology and mitochondrial division (9–11). Using coimmunoprecipitation, Ishihara et al. (8) found that ATAD3A coprecipitates with TFAM and an outer membrane–anchored Drp1 receptor, Mff (mitochondrial fission factor), in HeLa cells. The authors demonstrated that the matrix-exposed ATPase domain of ATAD3A directly interacts with TFAM. It remains to be determined whether the two coiled-coil domains of ATAD3A in the intermembrane space directly associate with Mff. These findings identify a crucial physical connection between the matrix and outer membrane, traversing mitochondria’s inner and outer membranes.

Fig. 1.

The Mff–ATAD3A–TFAM complex spatially couples mitochondrial division to mitochondrial nucleoids.

“Using co-immunoprecipitation, Ishihara et al. (8) found that ATAD3A co-precipitates with TFAM and an outer membrane-anchored Drp1 receptor, Mff (mitochondrial fission factor), in HeLa cells.”

Using gene knockdown, Ishihara et al. (8) showed that ATAD3A is critical for the size and number of mitochondrial nucleoids in Hela cells. Strikingly, live cell imaging further revealed that the intramitochondrial movement of nucleoids is dramatically decreased in ATAD3A-knockdown cells. Interestingly, the clustering of mtDNA in highly merged mitochondria in Drp1-knockdown cells was rescued by additional depletion of ATAD3A without affecting mitochondrial elongation. These data strongly indicate that ATAD3A plays a critical role in intramitochondrial nucleoid movements and that nucleoid clustering in excessively fused mitochondria in the absence of Drp1 is driven by ATAD3A-dependent nucleoid displacement. To define the underlying molecular mechanisms, Ishihara et al. (8) performed mutational analyses of ATAD3A. They demonstrated that two coiled-coil domains in the intermembrane space and the ATPase domain in the matrix are both necessary for nucleoid dynamics. Since ATP hydrolysis is not required for interactions of ATAD3A with TFAM, ATP hydrolysis could be specifically required for the movement of nucleoids.

What is the cellular significance of this nucleoid translocation in mitochondria? Drp1-knockout Hela cells contain decreased levels of multiple subunits of the electron transport chain. Remarkably, additional depletion of ATAD3A improved these decreased subunit levels in the knockout cells. Similarly, the depletion of ATAD3A in wild-type cells also increased the amounts of subunits of electron transport chain complexes. These data strongly suggest that ATAD3A controls the stability of oxidative phosphorylation proteins and potentially their assembly into the complexes.

In summary, thanks to Ishihara et al. (8), these discoveries reveal a role of ATAD3A in moving nucleoids in the matrix and connecting nucleoids to mitochondrial division. Furthermore, the identification of the Mff–ATAD3A–TFAM complex opens new fascinating research directions to further understand the dynamic nature of the mitochondrial organization and raises many thrilling questions (Fig. 1). First, understanding how ATAD3A binds to Mff at the contact site between the inner membrane and outer membrane would be exciting. Since Mff has only a very short stretch of a few amino acids in the intermembrane space (12), there are likely other proteins yet to be found to establish the intramitochondrial connectors spanning the two membranes. Related to this point, since the loss of ATAD3A does not disrupt the close localization of nucleoids to Drp1 or Mff, there are likely redundant mechanisms that connect mitochondrial division machinery and nucleoids. Second, Drp1 has additional receptors, such as MiD49, MiD51, and Fis1. Testing whether these receptor proteins also mediate the dynamic interaction of nucleoids and mitochondrial division would be interesting. Given that distinct types of mitochondrial division mechanisms have been shown to operate for biogenesis and mitophagy via specific Drp1 receptors, it would be critical to elucidate how ATAD3A joins nucleoids to a specific type of division (13). ATAD3A may couple nucleoids to biogenic division but not autophagic division. Third, another critical question is how ATAD3A moves nucleotides. Because ATAD3A is an ATPase, it may act as a molecular motor (like myosin, kinesin, and dynein) that travels along the actin cytoskeleton and microtubules. If so, mitochondria may contain a mitoskeleton that provides a track for the ATAD3A motor. Alternatively, ATAD3A oligomerization potentially generates force to push nucleoids forward, like actin comets for vesicular and bacterial movements in the cytoplasm (14). Fourth, how does the loss of ATAD3A increase the level of inner membrane proteins? It would be interesting to determine whether this role of ATAD3A is coupled to nucleoid movement or separable as ATAD3A has multiple functions in lipid homeostasis, assembly of oxidative phosphorylation machinery, and inner membrane cristae. Finally, in humans, deletions of ATAD3 genes cause cerebellar dysfunction with altered mtDNA (15). It would be important to understand whether defects in mitochondrial nucleoid movements cause this brain disease.