Short summary: A supporter asks if “backing up” copies of the mitochondrially-encoded genes in the nucleus is really viable, granted free radical damage in the nucleus. We emphasize the many additional ways that the nuclear copies will be safer than the mitochondrial originals, that the “backup copies” can be backed up again, and how they and additional strategies will buy us time for even better solutions.
SENSible Question: A big part of the reason for making “backup copies” of the mitochondrially-encoded genes in the nucleus is that they will be safe from damage by mitochondrial free radical production. So I was surprised to learn that the type of free radical damage to DNA that leads to mitochondrial mutations also impacts the DNA of the nucleus. Doesn’t this undermine nuclear “backup copies” as a solution?
Engineering “backup copies” of the mutation-prone mitochondrial genes into the safe harbor of the nucleus is our strongest strategy for protecting the body from large mitochondrial deletion mutations. It’s these deletions — particularly one that is found in so many cells that it is literally called “the common deletion” — that are most tightly linked to aging and diseases and disabilities of aging like Alzheimer’s and Parkinson’s diseases and the loss of functioning muscle fibers and strength with age. The concept of these engineered “backup copies” (technically, allotopic expression (AE)) was in fact the first of the “Strategies for Engineered Negligible Senescence” (SENS) rejuvenation biotechnologies proposed by founding CSO Dr. Aubrey de Grey. It was several years later that he realized that a comprehensive panel of similar damage-repair-based longevity therapeutics was the best way to conquer all of aging. And once SENS Research Foundation established our own Research Center, our very first in-house research project was on allotopic expression as a MitoSENS strategy.
There are several things about the nature of the problem and the tools we have at our disposal that make AE not just viable but an enduring solution for mitochondrial mutations, even though nuclear DNA is also susceptible to free radical damage just like mitochondrial DNA is.
The first is the sheer amount of oxidative damage[1] in mitochondrial versus nuclear DNA. The mitochondrial DNA’s exceptional vulnerability to free radical damage results, first and foremost, from its being located so close to the mitochondrial energy-production machinery, which is one of the major sources of free radical production in our bodies.[2] Most of these free radicals are generated in the innermost chamber of the mitochondria (the matrix), which is also where the mitochondrial DNA is located. Free radicals are, by definition, unstable; they therefore don’t get the chance to travel very far, but instead react with and damage whatever they bump into first.
Moreover, the specific free radical that mitochondria most directly generate (superoxide) gets trapped in the matrix once formed because of its negative electrical charge and the high pH of the matrix. This further increases the odds that mitochondrial superoxide (or radicals formed from it) will careen into the mitochondrial DNA. And any superoxide that does escape the matrix (Do you know what I’m talking about?) into the more acidic middle chamber of the mitochondria can be converted into perhydroxyl radicals, which also damage the mito DNA indirectly. They do this by oxidizing the matrix’s membrane, which then propagates the free radical damage like a wave rippling across a calm pond struck by a stone, until the wave of oxidized fats reaches the mito DNA, to which the membrane is attached.
Despite the target-rich environment of the mitochondria, you might still suspect that mitochondrial free radicals must still be responsible for a lot of the oxidative DNA damage in the nucleus. Researchers have long had reason not to think so, and a recent elegant study put the issue to bed. We can now say with great confidence that the impact of mitochondrial free radicals on the nuclear DNA is negligible.
So while it’s true that there is oxidative damage to nuclear as well as mitochondrial DNA, there’s about fifteen times as much of it in the former as there is in the latter. This is true even though OGG1 — the main enzyme responsible for keeping initial oxidative “hits” to the DNA from developing into full-on mutations — is more than three times as active in the mitochondria as in the nucleus.
But the susceptibility of mitochondrial DNA to oxidative damage is only one of the factors that would make nuclear copies of mitochondrially-encoded genes far more secure than the mitochondrial originals. For one thing, the nuclear DNA is usually wrapped up tightly around protein spools called histones, and this wrapped configuration protects it against some kinds of damage. Mitochondria are packed into similar complexes with proteins, but they have no protective histones. For another thing, the nucleus has more sophisticated DNA repair systems than the mitochondria.
And even when mutations do occur in the nuclear DNA, the consequences are less likely to be harmful than those in the mitochondrial DNA. This is true in a couple of different senses. First, most mutations in the nuclear DNA are self-contained, causing defects in the protein(s) produced by one or a small number of genes that are directly damaged. By contrast, the “common deletion” mutation in mitochondrial DNA not only wipes out several genes that code directly for proteins, but also the genes that encode some of the machinery that the mitochondria need to assemble the proteins coded for by any gene in the mitochondrial genome. Without this machinery, the mitochondria can’t produce any of the proteins encoded in the mitochondrial genome, including proteins whose genes are completely intact.
There is no parallel catastrophic failure in garden-variety nuclear mutations. So while the AE copy of an individual mitochondrial gene might be mutated, the other AE mitochondrial genes would still be produced normally. And if the mutation in a single AE gene were not too severe, it might be able to send mitochondria a good enough protein to keep carrying on more or less normal metabolism, rather than requiring the metabolic rewiring that Dr. de Grey postulates may spread oxidative stress across the body when deletion mutations take over a cell.
A second way that mutations in the nuclear DNA are less likely to cause problems than mitochondrial DNA mutations derives from how much less of its nuclear DNA a given cell type needs to carry out its function. The mitochondrial DNA is a very “lean” operating system: almost every “letter” in its code carries essential instructions for producing some machinery that the mitochondria require for their function throughout life.
By contrast, each cell houses lots of DNA in its nucleus that it can do without, or that can suffer a significant amount of mutation without harming the cell. This includes genes that are epigenetically deactivated because they are inappropriate for a given kind of cell (genes for liver enzymes in brain neurons, for example); DNA that codes for temporary regions of the “working copy” of a gene a protein that the cell’s machinery will ultimately cut out once the “working copy” or protein is put together, like the tearaway stabilizers used in embroidery; and the shrinking pool of “junk DNA” — genetic detritus from our evolutionary history that doesn’t code for anything functional.
Another kind of nuclear genetic material whose mutation is unlikely to harm us is transposons, the remnants of viruses that inserted their genetic material into our ancestors’ genomes that are still capable of “waking up” and wreaking havoc in the cell. When we’re young, our cells suppress transposons, but our transposon-repressing machinery tends to fail with age, allowing them to rise up and trigger inflammation, senescence, and mutations. Additional mutations in that genetic material are more likely to do us good than harm.
The Least of Our Worries
Despite the much lower rate of both oxidative damage and full-on mutation in the nuclear than in the mitochondrial DNA, and the fact that large amounts of a cell’s genetic material can suffer mutations without it having any effect at all on the cell’s function, many people assume that large numbers of our cells accumulate mutations during aging that make them dysfunctional and drive age-related disease. This idea was given renewed plausibility by a recent study that found that the rate of nuclear mutations is strongly correlated with lifespan across species, such that mice that only live for two years wind up with a similar number of mutations in their cells at the end of their “natural” lives as do humans, who live to be 85.
SRF founding CSO Aubrey de Grey has argued that it is unlikely that a meaningful fraction of our cells would suffer such crippling mutations because of what he called “protagonistic pleiotropy.” This starts from the fact that while many individual mutations in cells of a given type may make the cell less functional or even completely non-functional, individual cells losing their function will rarely be fatal, because there are many other cells of the same type in the same tissue to pick up the slack.
There is a critical exception to this general principle, however. Mutations (even, initially, a single mutation) in just one cell in the body can, in principle, evolve into cancer, which will then threaten a person’s life in a way that neither non-cancer-causing nuclear mutations nor mitochondrial mutations can, since those mutations are restricted to one or a small number of cells. Thus, he argued, the fact that evolution has to hold cancer-causing mutations in check well enough to give us the multi-decade cancer-free lives we already enjoy implies that it has equipped us with more than sufficient DNA repair machinery to suppress nearly all functional mutations in almost all of our cells, including mutations that would cause “general cellular malaise.”
So the tight correlation between the rate of nuclear mutation and lifespan across species doesn’t mean that the specific mutations present in the cells of animals at the end of their lives cause their aging and death. Instead, the fact that humans have better defenses against mutation accumulation can be explained by the fact that we need sufficiently strong protections against nuclear mutation to keep ourselves cancer-free for decades to live as long as we already do, whereas mice can get away with holding back cancer for just one year — just as “protagonistic pleiotropy” predicts.
There are certainly limitations to this argument, such as the existence of ways to hold cancer in check other than through DNA repair. But the best examples of such ways are apoptosis (cellular suicide), senescence (which turns off out-of-control cell growth), and immune senescence (since the immune system plays a critical role in destroying cancer cells). And the first two of these mechanisms either are themselves forms of aging damage that can also be caused by things other than mutations, and which are in any case the prime targets of RepleniSENS and ApoptoSENS, respectively. The third (immune senescence) is a more complicated part of aging, but there are many clear SENSible strategies to rejuvenate it, including engineering new thymic tissue (more RepleniSENS), destroying abnormal immune cells (more ApoptoSENS), and clearing out other damage that causes inflammation, including senescent cells, oxidized cholesterol in arterial foam cells (LysoSENS), and beta-amyloid (AmyloSENS).
Buying Time
Similar to other non-cancer-driving mutations in the nucleus, mutations in the AE copies of mitochondrial genes in any one cell would not be life-threatening. To impair our health, such mutations would have to be disabling to the cell in which they occur, and afflict a large enough fraction of the cells in a tissue for the tissue to be rendered non-functional. But barring catastrophic exposure to some mutagenic driver in our environment (like ionizing radiation), it would likely take many hundreds of years for this to occur. There are three reasons why this is so.
First, it already takes us many decades of life before we accumulate enough cells overtaken by deletion-bearing mitochondria for it to become a health problem. And it would take many times as long for a similar number of dysfunctional mutations to occur in our AE copies, since (again) AE copies would suffer mutations at a much lower rate than their mitochondrial originals because of the lower rate of mutations in the nucleus.
Second, some of the cells that suffer mutations in their AE copies will still have intact mitochondrial originals, which would cancel out the effect of the AE gene mutations, just as wiping out the cloud backup of a file is harmless if you still have the original on your hard drive.
And third, the consequences of single mutations to AE copies of mitochondrial genes are generally restricted to the protein produced by that gene, rather than leading to the full-scale protein-production paralysis that results from the “common deletion.”
Moreover, SENS Research Foundation is in the early stages of researching a drug-based approach that — if successful — would slow the hostile takeover of cells by mitochondria bearing deletion mutations. If it works, this new class of drugs would lengthen the amount of time that a cell with intact copies of mitochondrial genes could keep functioning after suffering its first deletion mutation, buying it time before it would have to draw on its AE copies. (Remember that even in people at the extremes of current lifespans, only a small fraction of cells actually accumulate large deletions — though those cells likely have an outsize impact on whole-body aging beyond their tissue-specific effects on sarcopenia, Parkinson’s, and other particular diseases of aging). Again, the mitochondria in such cells would keep on powering cellular activity even if they suffered mutations in their AE backups, delaying the need for additional rounds of AE.
Another way we might be able to slow the spread of mitochondrial deletions is by using TAL Effector Nucleases (TALENs). This CRISPR-like gene editing platform can be delivered into mitochondria and edit mitochondrial DNA., whereas CRISPR-Cas9 itself probably can’t, because it requires a sequence of genetic material to match up against its target sequence of DNA in order to edit it, and it’s generally accepted that mitochondria can’t adequately take in genetic material. Way back in 2013, researchers showed that they could use mitochondrially-targeted TALENS (mitoTALENS) to identify and destroy mitochondrial genomes carrying the “common deletion,” reducing the number of them in cells with a mixture of intact and deletion-bearing mitochondria.
This wouldn’t be a practical solution for mitochondrial deletions today, both because of the difficulty of getting mitoTALENs broadly into cells and because deletion-bearing mitochondria appear to take over long-lived cells so rapidly. But it could become viable in the future if we had better ways to deliver it and if we could combine it with a drug that slowed the deletions’ blitzkrieg across the cell enough for mitoTALENs to catch up.
Put Another Bite Back in the Apple
But still, our AE copies of mitochondrial genes are likely to suffer disabling mutations at some point in the future, even if that point is many times further away than anyone alive today has yet lived. What can we anticipate our future options to fix the problem to be?
The most obvious approach is a “simple” do-over: put another set of AE mitochondrial genes in the nucleus of our cells. If we were to again deliver all 13 AE genes together for this second round of therapy, we would face the same challenges of delivering gene therapy for mitochondrial genes as we did the first time around, with two differences.
First, as we’ve just noted, a very long time would likely have to pass after a person’s first round of AE gene therapy before they would be due for a round of maintenance treatment. This would give scientists decades — perhaps even centuries — to either develop improved versions of the AE genes themselves, or new gene therapy technologies that do a better job of delivering them, or entirely new MitoSENS strategies to supersede AE. Indeed, SENS Research Foundation scientists are already in the early stages of developing a mitochondrial “gene drive” rejuvenation biotechnology that would replace all mutated mitochondrial genomes directly, without making any modifications to the nuclear genome. Long before anyone needs a second round of AE gene therapy, this “gene drive” approach might render AE obsolete.
Another difference between a person’s first round of AE gene therapy and subsequent rounds is that while a person has no existing AE genes in any of their cells when they undergo their first round of AE, in later rounds they would still have some intact AE genes left over from their first round of therapy, even if other AE genes had mutated. In the great majority of cases, any such mutations would occur in one or a few AE genes rather than as large deletions. And those localized mutations would not impact the production of the proteins encoded in non-mutated AE genes as happens in mitochondria with the “common deletion.”
Thus, while it wouldn’t make sense for a person to undergo a new round of AE gene therapy until he or she had lost a sufficient number of AE genes from a sufficient number of cells, the fact that most people would likely still have several of the AE genes from previous rounds still intact might make it unnecessary to repeatedly deliver the entire set of 13 mitochondrial genes at once, as would be the case for the first round.
Instead, people’s pre-existing AE copies and the slow rate of nuclear mutation might allow us to perform an “AE topup” of just those AE genes that had mutated. More likely, it would give us room to set up a protocol in which people receive successive rounds of partial AE gene therapy. Each round would deliver a single AE gene or a small number of them; the regimen would work through all 13 AE genes in turn over the course of years or decades and then start over again at the beginning. One feature of this approach is that delivering these smaller packets of genetic material would not require the very large payload capacity of phage integrases, which would allow us to use gene delivery systems that we can’t use to deliver the full 13 for the first round of AE.
A second, entirely different way to address the long-term risk of mutations in AE genes is to replace the cells entirely. We will need to replace cells anyway on an ongoing basis to keep up with losses due to aging, trauma, and other causes (RepleniSENS). Additionally, systematically replacing each tissue’s adult stem cells and/or its mature cells is at the core of the WILT defense against cancer. We can both screen cells used for cell therapy or for WILT to make sure that their mitochondrial genomes are pristine, and also pre-engineer AE genes into such cells, which is a lot easier than delivering AE genes into the existing cells in a person’s body.
Bottom line for this section: no single cell in the body has to last indefinitely for us to last indefinitely, so the AE copies of mitochondrial genes in any given cell don’t have to last indefinitely either.
When you combine the dispensability of individual cells with the ability to perform maintenance rounds of AE, the much lower rate of both oxidative damage to and actual mutations in mitochondrial versus nuclear DNA, the additional defenses against mutations and their consequences in the nucleus versus the mitochondria, and incipient rejuvenation biotechnologies to slow the spread of mitochondrial DNA deletion mutations, we can be confident that “backing up” our mitochondrial genes via allotopic expression from the nucleus will secure our cells’ ability to fuel themselves and maintain normal metabolism beyond our current horizon.
[1] When we refer to ‘damage’ in the context of free radicals and DNA in this post, we are using the term somewhat loosely. For the purposes of this post, we mean the initial ‘hit’ to DNA bases by free radicals. However, because the cell routinely repairs these initial lesions, they aren’t true ‘damage’ in the sense denoted by the word in SENS. In SENS, ‘damage’ refers to stable, structural damage to cells and biomolecules that the cell can’t reverse on its own.
[2] This machinery — known as the electron transport chain (ETC) — creates cellular energy by sending high-energy particles harvested from our food through a controlled fall. Like the devices in a Rube Goldberg machine, parts of the ETC harness the energy from the electrons’ stepwise “downhill” tumble, ultimately leading to cellular energy production. But as the ETC machinery passes the “hot potato” from one working unit to the next, it occasionally “fumbles” an electron. Having escaped the tight path that the mitochondria intend for them, the electrons are free to interact with whatever comes their way, leading to free radical production and (if not rapidly quenched) free radical damage.
