Resolving mitochondrial cristae: introducing a new model into the fold

The many functions of mitochondria—the powerhouses of our cells—are intimately linked with their ultrastructure and network morphology. In this issue, Stephan et al (2020) apply a tour de force of microscopic techniques to examine the contributions of specific mitochondrial proteins to crista architecture.

A recent study uncovers the hierarchy of MICOS components in the biogenesis and maintenance of mitochondrial cristae in human cells.

Since Palade's first electron micrographs of mitochondria showing an intricately folded inner membrane encapsulated by an outer membrane (Palade, 1952), mitochondrial ultrastructures have been the subject of intense research and, coincidingly, intense biological imaging efforts. The mitochondrial inner membrane can be divided into two subcompartments, the tubular and bag‐like folds termed cristae and the inner boundary membrane (IBM) that sits adjacent to the outer membrane. The crista membrane and IBM are joined at sites termed crista junctions (CJs). The shapes, sizes and arrangements of cristae can differ greatly between tissues, developmental stages and cellular states (Brandt et al, 2017). Changes in crista morphologies are also observed in a multitude of diseases including neurodegenerative disorders, obesity and cardiomyopathy (Eramo et al, 2020). Furthermore, even mutations in the tiny mitochondrial genome can cause remarkable ultrastructural changes ranging from striking geometric angles to concentric onion‐like circles (Vincent et al, 2016). So, why is subcompartmentation of the inner mitochondrial membrane necessary? The answer lies in the recent knowledge that cristae are the hubs for ATP generation (Gilkerson et al, 2003) with the respiratory chain complexes predominantly found along the length of the crista membrane and F₁F_o‐ATPase complexes located towards the apex (Strauss et al, 2008; Davies et al, 2011).

The mitochondrial contact site and crista organising system (MICOS) is the primary architect of crista morphology. MICOS is a multi‐subunit protein complex located at CJs and makes contacts with protein complexes embedded in the outer membrane (Harner et al, 2011). MICOS is composed of seven main subunits and is organised into Mic60 and Mic10 subcomplexes (Fig 1). While the MICOS is widely understood to be the main controller of crista morphology, the exact contributions of each of its subunits are unclear.

Figure 1. Loss of MICOS subunits alters cristae morphology.

(A) Schematic of cristae architecture in wild‐type (WT) and MICOS subunit‐deficient cells summarised from Stephan et al. Inset shows schematic of the seven known MICOS protein subunits (MicXX, numbered based on their molecular weight), and organised into two subcomplexes (Mic60 subcomplex shown in white, Mic10 subcomplex shaded). (B) Model of cristae remodelling during Mic10 rescue, adapted from Stephan et al. OMM—outer mitochondrial membrane; IMM—inner mitochondrial membrane; IBM—inner boundary membrane; CM—crista membrane.

In the current work, Stephan et al (2020) address the contributions played by these proteins by utilising CRISPR‐engineered cell lines deficient in each of the seven MICOS subunits and then imaging the mitochondrial cristae that ensue. Until relatively recently, electron microscopy (EM) was the only way cristae could be examined. Here, the authors undertake elegant reconstructions of mitochondria and demonstrate how powerful super‐resolution light microscopy (or nanoscopy) techniques have become, particularly when used in concert with EM. For many years, light microscopy was bounded by what is known as the diffraction barrier—that is, due to the intrinsic properties of light, the inability to focus a beam of light to a point small enough that allows the differentiation of single objects any smaller than approx. 200 nm. No sooner were super‐resolution nanoscopy techniques such as PALM, STORM and STED described, than were they quickly aimed at mitochondrial cristae. Stefan Jakobs and co‐workers are pioneers in this, with landmark reports of the first crista images by fixed‐ and live‐STED. They now also introduce the latest advance, MINFLUX, into the fold. MINFLUX is a super‐resolution technique that combines the strengths of both STED and PALM techniques (Balzarotti et al, 2017). Using stochastically excited fluorophores (like PALM/STORM) and a “doughnut”‐shaped beam of light (like STED), MINFLUX cleverly locates molecules based on minimal emission fluxes (MINFLUX) and boasts resolving capabilities of molecules < 10 nm apart.

Here, the authors document the dependencies of expression levels between individual subunits and establish that the Mic60 subcomplex, which is stable in the absence of the Mic10 subcomplex, is essential for the maintenance of CJs and the stability of the holo‐MICOS complex. Furthermore, they found that the Mic10 subcomplex appears important for lamellar crista formation. Using MINFLUX, the authors found that in the absence of Mic10, Mic60 localised to two bands along the mitochondria much like the arc‐like cristae seen by EM. Through knockdown studies, the authors also established that the fusion GTPase Opa1 on the mitochondrial inner membrane, which is important for maintaining the diameters of CJs (Frezza et al, 2006), is also required for stabilising tubular CJs in mitochondria lacking Mic10. This suggests that both Mic10 subcomplexes and Opa1 play complementary roles in CJ stability. Finally, the presence of F₁F_o‐ATPase dimers also influences an ordered distribution of the MICOS complex and CJs, potentially by interacting with the Mic10 subcomplex.

Stephan et al were able to comprehensively characterise the resultant aberrant cristae into six distinct categories. Using inducible expression vectors to re‐introduce specific MICOS subunits over time, the authors assessed the ability of mitochondria to reform cristae. They found a lack of any short cristae, suggesting that there was no cristae biogenesis, but rather the remodelling of existing cristae. Furthermore, as CJ numbers increased, many CJs were apparent on abnormal cristae, giving the impression that secondary CJs were formed to drive remodelling. It should be noted that the authors were unable to completely knockout Mic60 and mitochondria from these cells showed onion ring‐shaped cristae with 1–2 CJs still present. It may be that residual Mic60 forms these CJs and this connection is essential for mammalian cell viability.

The work complements new findings into the organisation and dynamics of cristae. Recently, it has been shown that mitochondrial cristae constantly undergo cycles of membrane remodelling events (Kondadi et al, 2020). Strikingly, when imaged on the timescales of seconds, crista membranes and junctions appear to merge and separate quite continuously within mitochondria (Kondadi et al, 2020). In this sense, the static images from the rescue experiments presented by Stephan et al provide tantalising suggestions as to how mitochondria may repair aberrant cristae. However, much higher time resolution will be required before a significant understanding of the sequences of events that drive crista remodelling—and importantly, also crista biogenesis—can be attained.

It is clear that the true potential of MINFLUX nanoscopy is yet to be realised and its application in the current study pales in comparison with the types of biological insight it could be capable of. However, the introduction of this exciting new technique to the field of mitochondrial biology is a leap forward to get excited about. Perhaps in the near future, we will be watching the incorporation of new MICOS complex subunits in the curves and bends of newly synthesised cristae in developing tissues. Only time will tell how useful MINFLUX nanoscopy and its offspring will become.

The EMBO Journal (2020) 39: e105714