Supporting Evidence for the Importance of Mitochondrial DNA Deletions in Aging

Mitochondrial DNA damage is thought to be important in aging, but not all such damage is similarly relevant to aging. For example, researchers have produced mice that generate excessive numbers of point mutations in mitochondrial DNA, and these mice appear to suffer little harm as a result (with the caveat that different groups have found different degrees of outcome in this sort of investigation). Deletion mutations, however, are a different story. Some deletions result in mitochondria that are both dysfunctional and privileged in some way, better able to replicate or evade quality control mechanisms than their peers, even while they fail to properly perform their assigned tasks. These broken mitochondria quickly take over the mitochondrial population of a cell, turning that cell into a malfunctioning exporter of damaging oxidative molecules.

Unfortunately comprehensive proof of this picture, as opposed to the existing strong indirect evidence, has yet to be assembled. That proof may or may not arrive before the development of some form of rejuvenation therapy based on prevention or repair or working around deletion mutations, such as the allotopic expression of mitochondrial genes. Such a therapy would in and of itself provide strong evidence for or against mitochondrial mutations as a cause of aging, based on whether or not it works in animal studies. For now, more indirect evidence is what we have, however, and here researchers here provide a new set of supporting evidence for the importance of mitochondrial DNA deletions in degenerative aging by comparing samples from people with and without Alzheimer's disease. On average, comparing people of the same chronological age, those suffering from later stages of age-related disease should have a higher load of the forms of cell and tissue damage that cause aging.

Research suggests that mitochondrial changes are a driving force, rather than a consequence, of the aging process and Alzheimer's disease pathogenesis. Although point mutations of mitochondrial DNA have been hypothesized as being a critical cause of aging, there is evidence that they may not be fully explanatory. Mitochondria are dynamic organelles with very short half-lives. Continuous replication of mitochondrial DNA (mtDNA) is required for assignment to new mitochondria, resulting in a significant error rate and accumulation of mutated in mtDNA genome over time and space. We hypothesized that, beyond point mutations, different types of mtDNA rearrangements should be extensively distributed in aging cells. As these rearrangements are often not detected by routine methods such as polymerase chain reaction, we applied the approach of directly sequencing mtDNA from isolated mitochondria derived from fresh frozen brain samples.

Our data show that different types of mitochondrial rearrangements are very common in both the aging brain and Alzheimer's disease (AD) brain. Three types of mitochondrial DNA (mtDNA) rearrangements have been seen in post mortem human brain tissue from patients with AD and age matched controls. These observed rearrangements include deletion, F-type rearrangement, and R-type rearrangement. F-type rearrangement is defined as fragments with two different sections of mtDNA joined together in the same direction. R-type rearrangement is defined as rearrangement of mtDNA originating from two different orientations of mtDNA fragments. We detected a high level of mtDNA rearrangement in brain tissue from cognitively normal subjects, as well as the patients with Alzheimer's disease (AD). The rate of rearrangements was calculated by dividing the number of positive rearrangements by the coverage depth. The rearrangement rate was significantly higher in AD brain tissue than in control brain tissue (17.9% versus 6.7%). Of specific types of rearrangement, deletions were markedly increased in AD (9.2% versus 2.3%).

Evidence indicates that mitochondrial dysfunction has an early and preponderant role in Alzheimer's disease. Our data supports this, as the AD brain samples had more than 2.7 times the recombinant rate of similarly-aged controls. Significantly, the rate for deletion in AD was 4 times that of the control samples. The position of deletion joining points was not evenly distributed across the entire genome and instead was concentrated between regions 6kb and 15kb of the mitochondrial genome, which happens to be the area containing the DNA sequences for synthesizing all three cytochrome oxidases necessary for correct electron transport chain function. This makes it is reasonable to advance the concept that increased deletions in this area may affect the ability of mtDNA to synthesize cytochrome oxidase. Our results are consistent with reports of decreased cytochrome oxidase activity in AD brain samples.



"The position of deletion joining points was not evenly distributed across the entire genome and instead was concentrated between regions 6kb and 15kb of the mitochondrial genome, which happens to be the area containing the DNA sequences for synthesizing all three cytochrome oxidases necessary for correct electron transport chain function."

So maybe SRF should prioritize research on those genes.

Posted by: Antonio at July 5th, 2017 8:51 AM

Great paper.

" 'The "tip of the iceberg' hypothesis, which
assumes that mitochondrial DNA rearrangements are much more frequent than indicated by
the frequency of common deletions. These data suggest
that using "common" deletions as markers of the overall deletion frequency is inappropriate "

It appears re-arrangements are not just bugs but often features. Standard ones even.

" It is probable that the occurrence of mtDNA rearrangement
events is a semi-random process that is relatively unique to each cell "

When differention is veiwed as a largely mitochondria-driven events, as I believe the current evidence clearly shows, (for example specification of gonad progenitors and the dynamics of targeting paternal mitochondria to sperm in blastula 4 stage in mussels to resolve the apparent violation paradox of uniparental inheritance), one can surmise that the various iufoldings of the nucleoid in fact represents cell specific optimizations and defacto differention of mitochondria within them.

Posted by: john hewitt at July 5th, 2017 9:07 AM

Both ATP6, ATP8 and ND4 are a part of that region.
But there's still some big genes left, ND5 and COX1. The rest are about as big as the ATPs so they could be engineered on shoestring budget. ND5 and COX1 will require serious funding.

The company doing ND4 did say they want to do ND5 as well didn't they?

Posted by: Anonymoose at July 5th, 2017 9:23 AM

Therefore, disease situations like Alzheimers or Parkinsons can not be viewed simply as faults of mitochondrial rearrangements within single cells, but rather as alluded to previously, they are more likely varous disruptions, bloackages, backups, shortfalls, or excesses of the larger scale vectors of mitochondrial transexudations and transmissions across neuron and astrocytic circuits.
For example, has anyone ever generated a legit theory as to why mitochondria in dopaminergic cells in the parkinsons cicuit are so vulernabgle to MPP+ ? No. Hint: Would one suppose that these particular deep brain cells are net generators or net consumers of mitochondria relative to the larger cortical regions they ennervate? I'll leave that as an exercise for the moment.

One further consideration in this vein is that aging while aging folks are quite concerned about their brains, they actually have invested (as a whole) little effort to understand the nervous system as the showcase for mitochondrial disruption (point mutations and these semi-desireable deletions and duplications with various types of inversions (R and F type on heavy and light strands), and most glaringly, the fact that what ties this all together is that neurons are the exemplar senescent cell in the body.

Posted by: john hewitt at July 5th, 2017 9:24 AM

@antonio, @anonymoose
If we must insist on the allotopic Holy Grail lests at least take a moment to discuss how we might actual regulate the thing shall we?
Perhaps the most reliable mechanism the nucleus has for tailoring its supply of mito-targeted genes to mito demand is by sensing mito uptake of these proteins.
One way, for example, the nucleus senses and responds to cytosolic concentration of a protein that fails to be taken up by mitochondria because of insufficient or otherwise inappropriate membrane potential is to attach a nuclear localization sequence to the back end of the protein so that if the mitos don't get it, the nucleus eventually gets will, as is the case for ATFS1, as in the UPRmt (unfolded protein response)
Pretty tricky.
If we therefore imagine doing this NLS sequence addition (on a shoestring budget of course) for allotopic COX or whatever, what would be a good way to turn of the COX fawcett once we sense excess COX?

Posted by: john hewitt at July 5th, 2017 10:13 AM

chirp chirp,.... chirp chirp

Posted by: john hewitt at July 5th, 2017 10:36 AM

Okay, we have another data point now for the assembly sequence and regulation of complex I, now for drosophila;
Waiting for Leo Nijtman's (whodid the human complex i sequence recently) response on this because, voila, drosophila does not code for complex I subunits in their mtDNA ( ) so it will clearly be useful for anyone trying to get clean complex 1's with allotopic ND subunits, or alternatively, it maybe at least be useful for modern day recreations of Jeff Golblum's transformations in the 1986 thriller 'The Fly'

Posted by: john hewitt at July 5th, 2017 1:59 PM

@john hewitt
Any idea how we could enforce quality control on those 'priviliged' mitochondria ?
Like increasing NAD+/NADH ratio, prolonged fasting, exercise, hypoxia, heat exposure etc

Posted by: AndeyR at July 6th, 2017 3:25 AM


A couple thoughts. In his recent 'On being the right Q: Shaping Eukaryotic evolution' ( ) David Speijer makes a good case for mitochondrial operation at different FADH2/NADH ratios as an explanation of the evolution of peroxisomes, carnitine shuttles, uncoupling proteins, and even decreased fatty acid utilization in neurons (which must operate at low F/N ratios).

Similarly Michael Murphy makes a successful appeal in 'How mitochondria produce reactive oxygen species' ( ) for a simplified interpretation of mitochondrial operation into 3 general modes based on key parameters including Δp (protonmotive force), matrix NADH/NAD, reduced CoQ, and whether much ATP is being made.

Posted by: john hewitt at July 6th, 2017 6:59 AM

@john hewitt
Okay, now we have a battle on our hands. Just got back the comments from Leo, and it seems none of us here have bothered to check the literature to even see if complex I subunits have yet been identified in drosophila mtDNA since 1984 so lets get on it. Here is you chance allotopists:

Hi John,
Thank you for het paper of CI assembly in Drosophyla. I did not see it yet.
It seems to me that the CI assembly in Drosophyla resembles the scheme in mamalia very much. The second paper you are referring to is from 1984. I think that the unidentified reading frames are the CI subunits. To me it seems impossible that there is a CI without any mitochondrially encoded subunits. So I think there is no difference of this system regarding allotopic expression.


Posted by: john hewitt at July 6th, 2017 8:34 AM

ah well, seems drosophila has them too after all, quite similar to ours, very conserved for some very important reason no doubt :

Posted by: john hewitt at July 6th, 2017 8:45 AM

If DNA deletions are the cause of all of this mess, it seems to me that some sort of CRISPR therapy is in order to fix them. Allotopic expression would be unnecessary.

Posted by: Abelard Lindsey at July 6th, 2017 9:56 AM

agreed, but CRISPR in its present form ultimately needs varios ligated function including stand non-homologous end joing processes which are not part of the mitochondrial reportoire. furthermore streamlined mitochondria are not always bugs that 'take over cells' they can act purposefully with their own built-in self destruct timers while at the same time optimally perform various extranumerary functions independant of respiration and the specific membrane potentials and radical load it entails, in tight spaces like dendritic spines of purkinje and pyramidal cells

Posted by: john hewitt at July 6th, 2017 11:09 AM
Comment Submission

Post a comment; thoughtful, considered opinions are valued. New comments can be edited for a few minutes following submission. Comments incorporating ad hominem attacks, advertising, and other forms of inappropriate behavior are likely to be deleted.

Note that there is a comment feed for those who like to keep up with conversations.