Friday, 22 May 2015

Mechanisms linking mtDNA damage and aging

Milena Pinto and Carlos T. Moraes

A nice review on mitochondria and ageing, summarizing the links between ageing tissues and mtDNA turnover, damage and repair. A discussion is given on the free radical theory of ageing and its controversies, clonal expansion of mtDNA mutations, and the role of ROS damage/signalling in ageing. Finally, they comment on how mutations in mtDNA can lead to tissue failure.

Tuesday, 19 May 2015

Frequent somatic transfer of mitochondrial DNA into the nuclear genome of human cancer cells

Young Seok Ju, Jose M.C. Tubio, Peter J. Campbell, Michael R. Stratton et al.

Mitochondrial DNA (mtDNA) is physically separated from nuclear DNA, with each being contained within separate double membranes. However, in this study, the authors show that mitochondrial-nuclear genome fusion events occur in cancer cells at a rate similar to nuclear interchromosomal rearrangements.

Across the samples investigated, the authors found 10 primary cancers (1.8%, 10/559) and 2 cancer cell lines (7.1%, 2/28) with somatic mtDNA integrations into their nuclear genomes. Of the 12 cancers, 2 had more than one mitochondrial-nuclear DNA translocation event (aside: suggesting that this is not a phenomenon one must worry about if trying to count mtDNA copy number in cancer, with techniques such as qPCR? But interesting nonetheless!). By counting the total amount of mtDNA in a cancer cell, the authors calculate that the average frequency of mitochondrial-nuclear DNA fusion is ~5.1e-3 junctions per million base pairs of mtDNA, which is about half the intranuclear interchromosomal translocation rate (1.2e-2 junctions per million bp of nDNA).

Further studies are required to determine the mechanism by which mtDNA and nDNA can overcome the physical barriers separating them. However, the authors note that nonhomologous end joining and/or replication-dependent DNA double-strand break repair are the dominant mechanisms involved, suggesting that these fusion events coincide with nuclear genomic rearrangements.

Sunday, 17 May 2015

On being the right (cell) size

Miriam B. Ginzberg, Ran Kafri, and Marc Kirschner

Across tissues, cell size is highly variable: pancreatic beta cells have a characteristic length of ~10um whereas adipocytes are ~80um. Despite this inter-tissue diversity, within individual tissues, cell size is much more homogeneous. This review discusses mechanisms to explain this homogeneity.

In order for a population of cells to control their size, cells can either vary the amount of time they spend growing, or the rate at which they grow. The authors discuss evidence supporting the former hypothesis, that cells vary the amount of time they spend in G1 phase, such that cells only progress to S phase of the cell cycle once they reach a particular target size. One possibility for how this may be achieved mechanistically is the diffusion of intracellular 'ruler' proteins, whose concentration gradient can be used to measure distance.

The reason why cells have a tight control over their size is not completely clear. A 1945 study by Fankhauser in polyploid salamander larvae showed that normal structures could be formed even with large alterations in cell size, across many (but not all) cell types. A notable exception to this is the brain, where cell size affects its morphological complexity.

Perhaps a mitochondrial perspective may aid in the explanation of some of these observations? It is known that inheritance of mitochondrial content correlates with global transcription rate (see here and here). Furthermore, there is evidence to suggest that mitochondrial replication occurs in G1 phase, coinciding with the cell size checkpoint. Perhaps there exists some mitochondrial set point before progression is allowed to S-phase, with a corresponding global transcription rate and characteristic cell size?

Wednesday, 13 May 2015

Mosaic Deficiency in Mitochondrial Oxidative Metabolism Promotes Cardiac Arrhythmia during Aging

Olivier R. Baris, Stefan Ederer, Johannes F.G. Neuhaus, Jürgen-Christoph von Kleist-Retzow, Claudia M. Wunderlich, Martin Pal, F. Thomas Wunderlich, Viktoriya Peeva, Gabor Zsurka, Wolfram S. Kunz, Tilman Hickethier, Alexander C. Bunck, Florian Stöckigt, Jan W. Schrickel, Rudolf J. Wiesner

It has previously been observed that, during aging, tissues can accumulate a small number of cells with high levels of mutated mitochondrial DNA (mtDNA). It was previously unknown whether mosaic mitochondrial deficiency was able to cause disease. Here, the authors investigate whether mosaicisim causes an age-related degeneration of heart tissue, in vivo.

To do this, the authors generated a mouse model which has a dominant-negative mutant of the mitochondrial replicative helicase Twinkle. The mice are engineered such that the mutant form of Twinkle is only active in myocardium. The result is an accelerated accumulation of mtDNA deletion mutations in the heart (10.92%+/-9.5% mutant load at 18 months), which replicates a decades-long process during human aging. By staining heart muscle for Complex IV (COX), the authors found a mosaic pattern of COX-negative cells by 18 months (0.56%+/-0.34% of cells by 18-months), whereas control mice showed none. COX-negative cells also showed a compensatory increase in mtDNA copy number, by around a factor of ~4.

As a result, the mutant mice were ~x8 more susceptible to spontaneous ventricular premature contractions at rest, and ~x3 more susceptible to atrioventricular blocks (loss of conductivity between the atria and ventricles of the heart) during stress. By analysing heart homogenate, the authors show that there is little difference between the enzyme activity between the control and mutant, across the whole tissue. This demonstrates that mosaicism causes a heart defect in these mice, not a general mitochondrial defect.

Wednesday, 6 May 2015

The complete structure of the 55S mammalian mitochondrial ribosome

Basil J. Greber, Philipp Bieri, Marc Leibundgut, Alexander Leitner, Ruedi Aebersold, Daniel Boehringer, Nenad Ban

Mitochondria contain their own genomes, as well as their own transcription and translation machinery. In this study, the authors solve the structure of the mammalian mitochondrial ribosome using electron microscopy.  It highlights many differences that exist between ribosomes which translate nuclear DNA in the cytosol, and mitoribosomes which exist in the mitochondrial matrix, tethered to the inner mitochondrial membrane. For instance, the mitoribosome contains a structurally-incorporated valine tRNA (perhaps explaining the pathogenicity of mitochondrial tRNA mutations?). Furthermore, mitoribosomes contain substantially less rRNA than their cytosolic counterparts - which is restricted to the catalytic core of the particle. Instead, mitoribosomes contain a substantially higher protein content, perhaps to protect itself from oxidative damage inside the matrix. It is currently unknown how the mitoribosome recognises mRNAs, but this structural insight suggests that protein, rather than rRNA, may have a larger role to play.

(see also
Roland Beckmann, Johannes M. Herrmann)

The critical glucose concentration for respiration-independent proliferation of fission yeast, Schizosaccharomyces pombe

Kojiro Takeda, Caroline Starzynski, Ayaka Mori and Mitsuhiro Yanagida

The authors use fission yeast (S. Pombe) as a model organism to explore the relationship between cell proliferation and energy availability. S. Pombe is a more natural model system to consider than budding yeast (S. Cerevisiae), for a number of reasons. It divides symmetrically into two equally-sized daughters upon division; loses the ability to proliferate when depleted of mtDNA; and is unable to use ethanol as a respiratory substrate, all of which are held in common with higher animals and plants, unlike S. Cerevisiae.

By sweeping the glucose concentration, the authors find that the proliferation rate of S. Pombe abruptly halts at ~0.04% glucose, presumably since an energy deficiency prevents further division. They inhibit OXPHOS (using antimycin A), to show that the proliferation rate diminished by ~25% at high glucose concentration, and the halt in proliferation occurs sooner at ~ 0.1% glucose. By studying the oxygen consumption rate, the authors show that reducing glucose concentration shifted S. Pombe from anaerobic to aerobic respiration. Combined, these data suggest that by inhibiting OXPHOS, S. Pombe is unable to upregulate aerobic respiration in response to diminishing glucose supply, and thus ceases to proliferate at a lower glucose concentration.

Friday, 1 May 2015

Convergence of parkin, PINK1 and α-synuclein on stress-induced mitochondrial morphological remodelling

Kristi L. Norris et al.

Parkin is a ubiquitin ligase proposed to promote the removal of damaged mitochondria. In this paper they observe that under moderate stress conditions in MEFs, parkin does not stimulate the degradation of mitochondria (mitophagy) but rather increases mitochondrial fusion.

In previous studies, treating cells with the mitochondrial uncoupler CCCP causes parkin to translocate to depolarized mitochondria which are then removed by mitophagy. However, as observed in this paper, when lowering concentrations of CCCP, no translocation of parkin to mitochondria or mitophagy is observed. Instead, mitochondria elongate.  Parkin negative MEFs underwent fission upon CCCP treatment. Various kinds of mitochondrial stress (rotenone and hypdrogen peroxide treatments) showed similar results.

Both the mitochondrial fusion proteins Mfn1,2 and PINK1 are required for Parkin-mediated mitochondrial fusion, as well as ubiquitin E3 ligase activity of Parkin. Under stress conditions, Parkin seems to form a complex with alpha-synuclein (alpha-synuclein has been shown to induce mitochondrial fission) and Parkin catalyzes ubiquitination of species. The interaction between Parkin and alpha-synuclein during stress negatively regulates the activity of alpha-synuclein which leads to the observed mitochondrial fusion.

Selection against Heteroplasmy Explains the Evolution of Uniparental Inheritance of Mitochondria

Joshua R. Christie, Timothy M. Schaerf and Madeleine Beekman

Mitochondria are maternally inherited, but the selective advantage of this is debated. Nuclear DNA obeys Mendelian inheritance, which increases genetic variance in offspring, and allows the selection of advantageous genotypes. Why is this actively avoided in mitochondrial DNA (mtDNA)?

This study hinges on the following observation: the mixture (heteroplasmy, h) of two normal mtDNA haplotypes in an individual can cause physiological and behavioural abnormalities. Thus an individual is most fit when it is homoplasmic (h = 0 or h = 1) in its mtDNA. The authors explore this mathematically by generating a model of single-celled eukaryotes, obeying biparental inheritance, undergoing a cell cycle of four stages:  random mating of cells; mutation of mtDNA; selection based on a fitness function of h; and finally meiosis. The fitness functions considered by the authors all had a minimum at h = 1/2.

The authors initialise their model of biparental-inheritance, and allow it to equilibriate. They then introduce a small population of cells (1%) which instead have a nuclear allele which encodes mitochondrial uniparental inheritance. The authors find that, despite this small initial population of uniparental cells, they eventually come to dominate, such that the system purely obeys uniparental inheritance. Thus, selection against heteroplasmy is sufficient to explain a fitness advantage of uniparental inheritance. 

Mitofusin 2 ablation increases endoplasmic reticulum–mitochondria coupling

Riccardo Filadi, Elisa Greotti, Gabriele Turacchio, Alberto Luini, Tullio Pozzan and Paola Pizzo

Interactions between mitochondria and the endoplasmic reticulum (ER) have been studied a lot and have important physiological functions (e.g. lipid metabolism and Ca²⁺ signalling). Several proteins are involved in regulating the interactions between the ER  and mitochondria. The mitochondrial fusion protein Mfn2 has been proposed to be involved in the physical tethering. Quantitative EM analysis, however, challenges this view. In this paper they find that Mfn2 actually prevents an excessive proximity between mitochondria and ER and that cells with reduced Mfn2 show more ER-mitochondria contact points which causes a greater sensitivity to death stimuli.

EM analysis was done in mouse embryonic fibroblasts  with knocked down (KD) or knocked out (KO) Mfn2. An increase in the number of close (<15 nm) mitochondria-ER contacts in Mfn2 KO cells and Mfn2 KD cells. When using confocal miscroscopy, however, a decrease in the overlapping area between ER and mitochondria was seen in mutant cells compared to wild-type cells. 

The authors explain this discrepancy by arguing that Mfn2 ablation causes mitochondria to swell. EM measures the number and size of contacts between perimeters of the organelles, whereas fluorescence microscopy usually uses luminal fluorescent proteins and measures overlapping areas/volumes. Mitochondrial swelling makes its area increase more than the perimeter which could results in an apparent decrease in ER-mitochondria tethering when using fluorescent microscopy.

More ER-mitochondria contact points should lead to more calcium uptake by mitochondria. However, mitochondria took up less calcium in Mfn2 KO cells compared to wildtype cells. The authors explain this by arguing that expression levels of the mitochondrial calcium uniporter (MCU) were reduced by about 50% in Mfn2 KO cells. The reduced calcium uptake by mitochondria is thus suggested to be a cause of a reduction in MCU expression level, rather than a change in the distance between ER and mitochondria.

A comparison between wildtype and Mfn2 KD cells showed that Mfn2 KD cells show similar ATP levels and respiration levels but lowered membrane potential compared to wildtype. The increase in contact points in Mfn2 KD cells led to a greater sensitivity to apoptotic stimuli.