Inhibition of Mitochondrial Complex I Extends Life in Killifish and Zebrafish

In the open access paper linked below, researchers demonstrate modest life extension in the short-lived killifish and zebrafish species by inhibiting a specific portion of the protein machinery inside mitochondria, the power plants of the cell responsible for - among many other things - producing a supply of the chemical energy store molecule adenosine triphosphate (ATP). Mitochondria swarm in animal cells by the hundred. They are the evolved remnants of symbiotic bacteria, contain their own mitochondrial DNA, separate from the chromosomal DNA in the cell nucleus, and still replicate like bacteria even though they are tightly integrated into the cellular processes of monitoring and damage control. The cell culls the herd on a continual basis, destroying mitochondria that show signs of damage.

Mitochondria are known to be important in aging, but there are a number of different mechanisms involved. For one, there is a robust association between the details of mitochondrial biochemistry and longevity across species. Species with more resilient mitochondria, made up of a mix of lipids that is on average more resistant to oxidative damage, tend to be longer lived. Secondly, if mitochondria become dysfunctional or limited in number due to any sort of damage or change in environment - such as the sweeping changes of aging - then tissues with high energy requirements begin to suffer. The brain is particularly vulnerable from this perspective, and loss of mitochondrial function over time is associated with the progression of neurodegenerative conditions. Thirdly, mitochondrial signaling is involved in all sorts of processes known to be associated with aging and longevity, such as programmed cell death and triggering of cellular recycling and maintenance mechanisms. Many of the long-lived mutant lineages created over the past two decades in the lab are characterized by altered mitochondrial function and greater cellular repair activity. Lastly, and probably most importantly, rare forms of mitochondrial damage, such as large deletions in mitochondrial DNA, can evade quality control mechanisms, causing cells to be taken over by mutant mitochondria and fall into a harmful state. These cells grow in number with age, and export large quantities of reactive molecules out into tissues, contributing to many forms of age-related damage. For example, this increases the presence of the oxidized lipids that are the seed for the development of atherosclerosis in blood vessel walls.

The SENS rejuvenation research approach to mitochondrial damage is genetic engineering to create a backup copy of mitochondrial DNA in the cell nucleus. Thus there is always a supply of the necessary proteins, and mitochondria can't fall into a state in which they are malfunctioning due to DNA damage. Nuclear DNA is much more robustly protected and repaired than mitochondrial DNA. The challenge lies in the changes and additions needed to route the generated proteins from the nucleus back to the mitochondria. So far this has been achieved for only a few of the necessary genes, and it is a time-consuming process. Gensight is trialing this technology for a gene involved in an inherited mitochondrial disorder, for example, but everything they come up with as a technology platform is applicable to the end goal of carrying out this backup gene therapy for all mitochondrial genes, so as to remove this contribution to aging.

Longitudinal RNA-Seq Analysis of Vertebrate Aging Identifies Mitochondrial Complex I as a Small-Molecule-Sensitive Modifier of Lifespan

Here, we have used the short-lived killifish N. furzeri to perform a longitudinal study of gene expression during adult life. N. furzeri is the shortest-lived vertebrate that can be cultured in captivity and replicates many of the typical hallmarks of aging. The recent sequencing of its genome, and the establishment of genome-editing techniques makes it a convenient model species for experimental investigations on aging in vertebrates. Here, we report the observation that individual N. furzeri of different lifespans differ in their transcript levels at an early adult age. Further, we observed that genome-wide the rate of age-dependent gene modulation was lowest in the longest-lived individuals, suggesting that they are characterized globally by a slower aging rate.

Intuitively, differences in gene expression between individuals that differ in their aging rate should become larger as age progresses. However, we do not observe this consistently as differences between the longevity groups were larger at 10 weeks than at 20 weeks, and numbers of differentially expressed genes between adjacent age steps showed a U-shape. Our observations in N. furzeri are rather consistent with the results of a large-scale study of human aging in the prefrontal cortex: rates of age-dependent changes in gene expression are high during childhood, decline until age 20 years, rise again after 40 years, and, by the age of 60, exceed those observed during teenage years. The main result of this paper is that conditions favoring longevity are laid out during early adult life when inter-individual differences in gene expression are larger, and this result is consistent with observations in C. elegans where knock down of complex I genes or mitochondrial ribosomal proteins during development is necessary and sufficient for life extension.

Reduced mitochondrial mass and function is among the most conserved hallmarks of aging and is specifically observed also in N. furzeri at the levels of gene expression, mitochondrial mass, and mitochondrial functional parameters. Mitochondrial biogenesis is intimately connected to conserved longevity pathways such as the mTOR- and IGF1-pathways. Improved mitochondrial function is currently considered as a crucial component for the health-promoting action of physical exercise and calorie restriction. However, knock down of complex I genes expression induces life-extension in worms and flies. This contradiction between physiological age-dependent regulation and effects observed after genetic manipulations is also observed for another major longevity pathway: the IGF-I pathway. Genetic dampening of IGF-I signaling is life-extending in several models, yet growth hormone and IGF-I concentrations in blood decline during aging. Also, expression of mitochondrial ribosomal proteins declines during aging, but knock down of these proteins induces life-extension.

Complex I of the respiratory chain can be potently inhibited by small molecules, such as rotenone (ROT). The effects of ROT may also be explained by the mitohormesis hypothesis postulating that life-extending interventions act via a transient burst of free radical oxygen species that induce adaptive stress responses. In C. elegans, life-extending effects of calorie restriction or RNAi of the insulin signaling pathway are blocked by antioxidants, and partial inhibition of complex I by ROT prolongs lifespan, generates a burst of ROS, and antioxidants block the life-extending effects of ROT. Increasing the dosage of ROT, however, is life-shortening in N. furzeri, as it is expected by a hormetic effect. Life-extending effects of metformin on mice may also be mediated by mitohormesis, since this drug can inhibit complex I, and effects of metformin in C. elegans were directly linked to mitohormesis via induction of peroxiredoxin.

We observed that treatment with a dose of ROT three orders of magnitude below the median lethal concentration can revert the transcriptional profile of brain, liver, and skin to patterns characteristic of younger animals. This effect was seen not only in N. furzeri, but was replicated in the zebrafish D. rerio, showing that ROT effects are not linked to the peculiar physiology of this short-lived species. In D. rerio, effects of ROT were dependent of the length of treatment: treatment for 3 weeks had a smaller effect than a treatment of 8 weeks. The median lifespan of D. rerio is in the order of 3 years, therefore 8 weeks represent ∼5% of median lifespan indicating that a relatively short treatment can cause rejuvenation of the transcriptome. In summary, our data suggest complex I as a new potential target for prevention of age-related dysfunctions.


Hey! Hello there,

''Accordingly, partial pharmacological inhibition of complex I by the small molecule rotenone reversed aging-related regulation of gene expression and extended lifespan in N. furzeri by 15%''.

This is very interesting and shows that mitochondrias, once again, are the heart of all of this aging process. Complex I inhibition is simply an evolutionary form of pseudo hypoxia; pseudo because mitochondrias need respiration (OXPHOS Oxidative Phosphorylation) to produce ATP (state 3 and state 4 respiration, electron forward and back flow through the ETC and proton ATPase pump) but evolution has found ways to rig the game :
UCP1/2 (uncoupling proteins) that uncouples électrons transport/Complex OXPHOS usage for ATP production (the proton-gradient is lost like a valve that is opened to let the steam out, thus reducing the throughput but improving the effiency (per ATP unit/per electron/per O2/per ROS)). Certain animals, such as humans, use UCPs and thermogenesis strongly. Thermogenesis is activated by UCPs, dépendent on HSPs (Heat Shock Proteins ) and HIF-1 (Hypoxia-Inducible Factor 1) in the brown adipose tissue and white adipose tissue (BAT/WAT). It's a good metabolic increase switch to burn fat and stay thinner. It's not surprising, really, that HIF and HSPs; and UCPs, work in tandem to create a sort pseudo hypoxia around mitochondrial ETC Complex I; HIF is a master regulator of O2 sensing (and in a sense guage/communicator of energy by O2 entry) and hypoxia onset. It is also a double-edged sword, it helps out for cancers (who use HSPs (HSP90) and HIF-1) who survive/thrive in O2-deprived hypoxic environment.
Glucose that feeds through the pentose phosphate shunt pathway (G6PD glucose-6-phosphade dehydrogenase) also does the same thing by creating an 'hypoxic' environment but, at the same time, increase HIF-1 availability (normally in normoxia HIF-1 is kept shut and degraded by ubiquitin proteasome; when hypoxia sets in, HIF-1 is master survival switch that must be shunted away from ubiquitination and allowed entry to make cells adapt to hypoxic environment).

Evolution also used another trick, mitochondrial membrane potential modulation to alter mitochodnrial ROS release (mV millivoltage reduced potential). ATP production is very closely tied to membrane potential; though it fluctuates depending on state (3/4 respiration, i.e for example, membrane potential rises dramatically when the cell is about to die, like a huge ROS burst and then collapses as the cytochrome c is lost through mitochondrial permeability pore opening). Complex inhibition is just one of them. Membrane bilayer phospholipid fatty acid peroxidizable-to-less peroxidizable reordering is another one.

One study in 5 types of bivalves found that the longest lived one had the lowest amount of complex I activity and also, surprisingly, they found complex II also was lower.
For humans, Complex IV (Cytochrome C Oxidase) is the most important one as it is the one that finalized the ATP energy creation availability with ATPases; it's also called the 'respiration' complex; wherease COmplex I is really the ATP creating engine when fed electrons. It is also the one creates the most ROS; hence the zebrafish is living longer because of a reduced COmplex I activivity means a reduced ROS Hydrogen perodixe and electron leakage at Complex I; meaning more efficiency and less ROS damaging the phospholipids of the surrounding bilayer membrane (IMM/OMM).

The effect is still very mild though, just 15% increase in median lifespan, it is akin to CR. Longterm CR has been shown to increase ETC coupling, rather than decrease (or put another way, inhibits proton-gradient uncoupling because this shunt-mechanism is costly (you lose proton/electron to try to mitigate ROS production, this can affect ATP production).

The mitohormesis effect, visible immediately on the membrane potential (main decider of ROS production), is when ROS increase to a 'tolerable' level - and increase uncoupling proteins, HPS, HIF-1 and other sensors of ROS; plus, this also, increases antioxidant
Nrf2-ARE/EpRE detoxyfying enzymes of the Redox (Nuclear-Responnse Factor 2 is activated upon increased ROS and oxidative stress, it is transported in the nucleus and activâtes DAF-16/DAF-2 which alter IGF-1 via SIR1/HIStones). ROS are damaging at a catastrophic point, but 'small-to-moderate' oxidative stress strenghtns the system by this 'compensation' hormesis response. Telomerase is also shunted away from the nucleuse and towards the mitochondrion to try to maintain mitochondrial OXPHOS function upon increased ROS (hTERT mitigates mild ox.stress ROS).

THis is very interesting in very fast living animal, the zebrafish. It also makes me understand, that mitochondrial mutations are one the key element of intrinsic aging.
The reason is because, mitochondrial mutations accumulate exponentially with age and have myriad of effects - but all of them point to one direction - mitochondrial OXPHOS Complex assembly/energy creation dysfunction. Any of these mutations, wether it be Progeria (accelerated aging, like a zebrafish, in 'fast forward' basically), Down Syndrome, Werner Sydrome, Diabètes, etc...
scientist can detect mitochondrial deletions, lesions, nucleotide combinational errors and mutations that have a final effect = making the mitochondrias Complex totally dysfunctional = ROS production rise exponential/membrane potential lost = ATP energy deficity = death.
It's no Wonder, mitochondrial H2O2 release correlates and causates to MLSP in mammals of exponential aging rate. But, we don't even need to look elsewhere, just look at fast aging people (progeria) and we see 'aging' in the speed of light. And it points to mitochondrias the culprits causing human death. For without energy (by mitochondrias) there is no cell life, and no organ life, and no human (body) life.

Posted by: CANanonymity at May 5th, 2016 1:09 AM

The theory that the 13 genes kept in the mitochondria because they are needed there to 'regulate' the mitochondria's activity gets a bit of a boost in this 'simulation thought experiment' paper:

On the other hand, if this is true, why has Gensight's expression of the ND4 gene outside the mitochondria rescued retinal cell function in a stage 2 trial? I'm a bit surprised that this data point isn't mentioned in The Scientist article.

Posted by: Jim at May 5th, 2016 7:11 AM


Hi Jim! THank you for the link, very intriguing.

''In the end, we identified three features that together predict whether a gene is likely to be retained in the mitochondrion, rather than transferred to the nucleus: 1) it encodes a protein that forms the center of a complex, 2) it encodes hydrophobic (water-repelling) proteins, and 3) it contains many Gs and Cs in the DNA sequence.''

I'm not surprised by their findings, the fact that mitochondrias would have lost thousands of genes from their once-filled genome and are now left with a measly 13 genes (13, if you're superstitious, is an ironic number here; could be a 'bad sign/bad omen' for our mitochondrias)...well, yes and no. These 13 genes are Crucial (and though I believe mitochondrias could (find a way) to without them; if they kept them they are essential at this point).

Why these 13 genes in the mitochondria and not in the nucleus:
''1) it encodes a protein that forms the center of a complex, 2) it encodes hydrophobic (water-repelling) proteins, and 3) it contains many Gs and Cs in the DNA sequence''

Because these 13 ones all point to Complex assembly, Complex activity, Complex ATP creation, *Energy Creation*, cristae formation (the folds Inside the mitochondrion), hydrophobic proteins because DHA and EPA phospholipids fatty acids are water hydrophilic fattya acids that are Major responsible ones for peroxidizingt the membrane and the mtDNA; they release hydroperoxide* which catalyze the creation of aldehyde products (MDA, TBARS) from the
peroxidation of DHA/EPA fatty acids in the membrane - which happens to be - right next to the mitochondrial innermembrane full of Complexes who produce the ROS which Attack the peroxidizable phospholipids. Basically, mtiochondrias 'covered their a**es', by making sure they avoid water-environment rich or water-loving proteins, since that's where major damage happens - and thus, would make catastrophic mitochondrial energy failure (Something evolution éliminâtes immediately, for it is incompatible with specie survival (survival is the whole point)). Same thing for Gs and Cs, Guanine and Cytosine, they found more of them, Guanines form Quadruplexs structures (G-Quads) which 'solidify' integrity and safety of the genome and DNA. Cytosine with Guanine has been found in nuclear DNA and CpG islands (Cytosine-Guanine islands) and is a marker of DNA methylation aging (5-methylcytosine loss with aging), and of télomères status (télomères that are hypermethylated are strong and stable; telomere shortening and demethylation is accompanied by 5-methylcytosine loss, and thus cytosine itself, loss. DNA is lost progressively, Guanine and Cytosine are the nucleotide 'blocks' of DNA. Evidently, mitochondrias kept them specifically for integrity and stability in their genome. G-Quads and oligonucléotides increase overall immune system too).

ND4 gene, ND4 NADH dehydrogenase, subunit 4 (complex I)

I believe you are refering to this study :

Nuclear expression of mitochondrial ND4 leads to the protein assembling in complex I and prevents optic atrophy and visual loss

The title is the answer to your question. The reason why the retinal function is rescued - is by complex I reassembly in the mitochondrion ETC and thus,
cell ATP energy recreation (need)/maintenance.

This proves, yet again, that mitochondrion/mtDNA (mutations/deletions/alteration to them), are the main deciders of actual organismal 'aging'. The Nucleus is also in tandem with that, but mitochondrias are Critically important (they are far more naked-like right close in front of Complex ROS and vulnerable (10-folds more oxidizable than nucleus and producing the bad ROS - just to 'help' themselves during their energy manufacturing), for they themselves being inside the cell, produce the cell's energy to sustain its life. Without them, cell death is inevitable (except rare cells, like RBCs who have no mitochondria and still are capable of getting ATP...proof that evolution even found a way to work it (make energy) without our bacterial mitochondrias). The paper you give is very thoughtful and make us understand that mitochondrias and the cell is a evolutionary bargain bacterial life win-win mutual arrangement (Symbiosis/ Osmosis/Host-Hosted/Yin-Yan) billions of years in the making when Earth had its transfer from ambient volcanic sulfurous gases to ambient oxygen(ation).

Posted by: CANanonymity at May 5th, 2016 11:21 AM

Filippo Scialò et al., show that mtROS production increases with age and that un-detoxified ROS can be detrimental to Drosophila lifespan, while increasing ROS production specifically from reduced CoQ, possibly via reverse electron transport through respiratory complex I, acts as a signal to maintain mitochondrial function (notably complex I) and extend lifespan. These results are similar to those reported in worms where small doses of rotenone, paraquat, and mutations in complex I that increase ROS extend lifespan. However, they did not observe lifespan extension in flies fed with varying doses of rotenone or paraquat. Why?
Manipulation of the redox state of CoQ may be a strategy for the extension of both mean and maximum lifespan and the road to new therapeutic interventions for aging and age-related diseases. See

Posted by: Dmitry Dzhagarov at May 5th, 2016 2:09 PM

Speaking of mitochondrial genes that are subject to damage, does anyone know how Matt O'Connor is doing with the MitoSENS project since the crowdfunder last October? I believe that they said that they would post updates.

Posted by: Morpheus at May 5th, 2016 10:09 PM

@Dmitry: Mitogen of New Zealand claim the reason the "standard" anti-oxidants don't work is that they don't make it to the mitochondria. They claim that their product does make it.

Posted by: xd at May 5th, 2016 10:53 PM


Hi Dmitry! Thank you for that!

''These results are similar to those reported in worms where small doses of rotenone, paraquat, and mutations in complex I that increase ROS extend lifespan. However, they did not observe lifespan extension in flies fed with varying doses of rotenone or paraquat. Why?''

Because of specie differences that are non-translatable. These studies in flies and worms nematode you have to take them with a grain of salt. They are not mammals.

It's far more complicated in mammals, and it's also why fly results or nematode results don'T translate 1:1 into a mouse or a naked-mole rat.

As the animal increases in complexity (organs etc...) most of the results are lost and dampened. Simply because longer-lived species have already been optimized and adapted with longevity genes - it's how they live such long lives.

I mean I take these fly and worm studies as important but they are not 100% reliable, many contradict themselves and especially, when they try to apply their results in another specie.
It's not surprising, these species have different evolutionnary survival stratégies and pathways. IGF-1 might be bad in a dauer larva, but in us it's good to a certain extent, we need it - different specie, different survival strategy, different longevity, different usage of the same gene 'in another specie survival context'. Most of these short-live animals is the 'procreate, live fast, spend energy, burn the candle, high metabolism, die Young after accomplishing 'duty' of sexual procreation for specie survival'...we are not made that way, we are the total inverse, as such have different genetic stratégies that are incompatible/non-translatable...

PLus on this subject of Complex I ROS production acting as a protection signal to increase lifespan as mitohormetic compensation effect, by keeping the Complexes/ETC/CoQ in reduced redox state hits a ceiling. Hormesis is compensation mechanism, the rising ROS signals the protective response to improve mitochondrial OXPHOS function and respiration/leading to higher or maintained ATP production. But, it's not enough...this is just a tiny calorie restriction like effect to promote better efficiency.

Much much longer lived animals and mammals, show exactly what I told, they show a decrease of Complexes activity as a necessary mechanism to limit ROS release at Complex I. What I'm getting at is that hormetic effect is weak and only a 'safeguard' mechanism to allow a longer life in small short-lived animals. Long-lived animals also benefit from hormesis but it's not what makes them live long in the first place. In long-lived animals, it's different, ROS production must be Carefully controlled to mitigate Attack of the membranes, because if you increase ROS you have to increase Antioxidant Response/Enzymes (SOD, Catalase, GSH, GSR, GPx, etc). Why ? Because membranes peroxidation create Deadly aldéhydes that reach the mitochondrial DNA (creating mitochondrial lesions, mutations and deletions), fractures it and in doing so, disassembles the OXPHOS Complexes of the mitochondrial innmembrane - increasing ROS production at Complex I exponentially (Orders of Magnitude).

How did evolution solve that problem to make two same specie live different lifespan ? It modulated the ordering of the bilayer phospholipids' fatty acids to attain a more small-chain polyunsaturation than long-chain (DBI (Double-Bond Index) and PI (Peroxidizability Index)), rendering the membranes - resistant to Complex I ROS attack damage. And thus, nullifying Complex I problem. Complex I, II and III; together produce all the ROS that is needed to shorten the life of an animal.

Flies (Drosophila) produce 300 times mitochondrial H2O2 release in their head and thorax than a 500 year old Oyster clam (Iceland Arctica Islandica) in its gills, mantle and pedal foot.
They also live a measly 60 days. ROS are the major culprit - creating oxidative stress Inside ETC; and Complex dysfunction; energy deficit and thus death. D.Abele and other marine scientists showed that bivalve that live very different lifespans have divergent Complex I activity (the longest bivalve have the lowest Complex I activity), these are animals that have far more in common with us than a fly or worm.

I'm totally with you though, manipulation of the Redox is the last solution. In fact, that clam Oyster that lives 500 years, not only has low peroxidizable lipids, it maintains an unflinching adequate redox for over 150 years (something humans lose as mitochondrial ROS/mutations/deletions/lesions increase with age and compromise the REDOX - creating an oxidized milieu which is ultra-stressful for the mitochondrias (increasing oxidized glutathione GSSG and losing NAD+/NADPH. GSH is lost from the mitochondrias)).

Mitochondrial glutathione, a key survival antioxidant.

Posted by: CANanonymity at May 6th, 2016 1:37 AM

@Morpheus - Yes I too am wondering how Dr O'Connor's team is getting on. I am wondering if expressing the ND4 gene outside the mitochondria is just much easier compared to the other 12 protein coding genes? It has been almost 10 years since Corral-Debrinski's lab demonstrated ND4. I would have though a few of the other genes would have been demonstrated in vitro by now?

On the other hand, the delay in the other 12 could just be due to a lack of decent tools such as cell lines properly null of each of the 12 genes?

Posted by: Jim at May 7th, 2016 2:20 PM

@Reason, do you have any inside information on progress by Dr. O'Connor's lab?

Posted by: Morpheus at May 7th, 2016 7:10 PM

@Morpheus: Yes, but unfortunately I can't say anything on that topic at this time.

Posted by: Reason at May 7th, 2016 7:17 PM

Post a comment; thoughtful, considered opinions are valued. 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.