Is Aging Controlled from the Brain? NPY and ALK5

For more than a decade, Claudia Cavadas of the Center for Neuroscience in Coimbra, Portugal has been on the trail of a signal molecule that comes from a region of the brain associated with timing.  It’s a small protein called Neuropeptide Y, and Cavadas has recently collected the body of evidence that it is a central determinant of aging in the brain and throughout the body.  Whether or not NPY proves to be the Philosopher’s Stone, I think she’s on the right track to be investigating neuroendocrine origins of aging.

The present column also contains a (belated) update on ALK5 and TGFβ.

Since the genetic science of aging began to take off in the 1990s, the biggest surprise has been the extent to which aging is centrally orchestrated. ”Regulated” is the accepted word, but I don’t hesitate to say “programmed”.  For those of us interested in intervening to slow or reverse the process, the burning question is: how is the program implemented?  If the process is centrally orchestrated, where is the orchestra’s conductor?  

We have the same genes when we are old as when we are young, but different genes are turned on and off in different stages of life.  How genes are turned on and off is the science of epigenetics, and we are just beginning to untangle the set of chemical add-ons that bind to the chromosome or to the histones, protein spindles around which the chromosome is spooled.  Chemical signals that turn whole suites of genes on and off are called transcription factors, and these can be big molecules or small, proteins or RNAs, very specific and targeted to a single gene, or aimed more generally at large swaths of the chromosome.

At the least, transcription factors are able to regulate the body’s rate of aging.  But I see evidence that the chemical signals have even more power—the chemical signals tell the body how old it is.  Change the signals, and the body can change its age.

When we think this way, the questions “what are these signals?” and “where do they come from?” become exciting and highly charged.  The most intuitive and logical place to look for a source of age signals is in the brain.  Here are four reasons to look to the brain’s hormone center, the neuroendocrine region, as a source of aging signals.

  • We know that aging is highly plastic and adaptive depending on behavior and environment.  To sense many factors and make a decision about aging, it seems that the combined forces of the nerves and endocrine signal transducers in the brain are best equipped for the job.
  • The hypothalamus is a neuroendocrine region of the brain already known to house the seat of the 24-hour clock that synchronizes our circadian rhythms, the suprachiasmatic nucleus.  
  • The hypothalamus is also responsible for flooding the body with at least one of the transcription factors that increase destructive inflammation late in life (NFkB).
  • In a worm experiment fifteen years ago, genes for aging were modified in different systems of the worm, and the modifications were effective in changing the worm life span only when the genes were modified in the nervous system (not in skin or muscle or digestive system).

This mode of thinking suggests that Cavadas has been digging at the taproot of aging, and that the signal she has identified may be of central importance.  In a recent paper, Cavadas lays out the case for a central role in Neuropeptide Y in dictating the age of the body.  

Accumulating evidence suggests that neuropeptide Y (NPY) has a role in aging and lifespan determination. In this review, we critically discuss age-related changes in NPY levels in the brain, together with recent findings concerning the contribution of NPY to, and impact on, six hallmarks of aging, specifically:

  • loss of proteostasis
  • stem cell exhaustion
  • altered intercellular communication
  • deregulated nutrient sensing
  • cellular senescence, and
  • mitochondrial dysfunction

NPY is a small protein, with 36 amino acids, found in the nervous systems of all higher animals.  It is one of the most abundant neuroendocrine proteins, but since these chemicals are very powerful signal molecules, the body’s total inventory is measured in μg (millionths of a gram).  NPY has various roles related to appetite, anxiety, memory, and circadian rhythm.  Levels of NPY decline with age, especially in certain regions of the brain.  Caloric Restriction also elevates levels of NPY.  High blood sugar levels tend to inhibit NPY.  And there is a bit of evidence that NPY is necessary for CR to extend life span:  Mice in which the NPY gene has been knocked out don’t respond to CR. NPY is also associated with cancer suppression in mice.

Autophagy is the process by which cells break down and recycles damaged molecules.  As we age, autophagy slows down and damaged molecules accumulate:  misfolded proteins, cross-linked sugars, lipofuscin and amyloids.  A central function of NPY is in promoting autophagy, especially in the brain.  

The NPY gene can be read in a way that transcribes only the second half, and the result is a protein that binds to mitochondria and improves energy efficiency, reduces ROS damage.  Some studies suggest that NPY can extend the useful life of stem cells, and delay cellular senescence.  Links to anti-inflammatory chemistry are less well established.


Why NPY probably is not the Philosopher’s Stone

All this suggests a role for NPY in aging, and the possibility that increasing NPY might increase life span.  But it is not easy to increase the body’s transcription of a target protein.  (It is easier to block a protein that is bad than to promote one that is good.)  And my guess is that we will find signals further upstream from NPY that are even more effective points of intervention.  The guess is based on the fact that NPY is a neurotransmitter, an “end-use” molecule.  I suspect that the upstream source of aging will be found in transcription factors, the molecules that bind to DNA and determine which genes are expressed.


Other Signal Molecules from the Brain

TGFβ might be a good candidat for the first brain signal to become a target for anti-aging therapy. TGFβ is a not a transcription factor but a cytokine, a signal protein that affects the energy metabolism and, in particular, inflammatory response.  It comes not from the hypothalamus, but from the hippocampus, another part of the brain, an inch or two underneath.  TGFβ works against us, i.e., we produce more and more of it as we age, and it rallies the inflammatory legions that promote arterial diseases and cancer.  We would want to block its action, and that might not be too difficult.

One of the targets, receptors into which the TGFβ molecule plugs to do its work, is called ALK5.  Jamming ALK5 has been promoted by the Conboy lab at Berkeley and others as a strategy to test further.  ALK5 inhibitor can be injected deep into the body cavity, where it has already been shown to promote new growth in both muscles and nerves in mice.  This function is apparently related to its role as antagonist to TGFβ.  Remarkably, stem cells retain their ability to regenerate new tissue well into old age, but they receive signals telling them to stand down.  Simply changing the signaling environment can make an old stem cell act young.  This has been a major theme of the Conboys’ work.

What good is TGFβ?  The story is complicated.  The molecule can apparently be pro-inflammatory or anti-inflammatory, also pro-cancer or anti-cancer.  This relates to the GDF11 controversy, too.  GDF11 is in the TGFβ family.  To me, the Conboys are a trusted source, and they have systematically built a case that too much TGFβ in later life is a big factor leading to more inflammation and less stem cell activity.

Their ALK5 inhibitor has only been tested for short-term benefits.  The next step is to do life span studies in mice with ALK5 inhibitor.


24 thoughts on “Is Aging Controlled from the Brain? NPY and ALK5

  1. I am happy to say that Josh is working with the Major Mouse Testing project to put Alk-5 to the test in robust lifespan studies to prove this. We are hoping to launch our fundraiser in March this year and believe this and our Senolytic lifespan work will be of great benefit to the field.

  2. I do not believe that there is a central aging coordinator other than the blood. The brain, in particular the hypothalamus’s superchiasmatic nucleus. In experiments this organ responds to telomere attrition signalling to release less GnRH which sets the clocks of subservient tissues ahead. Supplementation with GnRH cause increased lifespan, about 18% like every other intervention targeting a specific ‘clock’ (like mitochondrial dysfunction, DNA lesion accumulation and of course the epigenetic patterns laid down by these clocks and their effectors, anyway, that’s the way I see it. Finally I agree with Josh, that aging does not occur (or he agrees with me) at the cellular level but at both the cellular and organismic levels which interact extensively. Yes, the central nervous system is part of aging (there were interesting experiments where people were exposed to the environments they encountered during their teen years, and that made them more similar to when they were young in terms of performance), but there are older and more fundamental clocks that the superchiasmatic nucleus is informed by.

  3. “there were interesting experiments where people were exposed to the environments they encountered during their teen years, and that made them more similar to when they were young in terms of performance”

    Do you have the link?

    • I’ve read and quoted this first paper in my own, it’s a fantastic study, but the ‘take-home’, to me at least, is that while the hypothalamus does to some extent control aging it’s only one of the ‘aging clocks’, so that while supplying exogenous GnRH does extend lifespan, that extension is only a fractional extension. That inflammation is a major ’cause’ of aging and the diseases of aging seems indisputable to me, of course the ‘test’ would be to eliminate inflammation and see.
      I think the real challenge is to determine to what extend aging is reversible, rather than the extent to which we can stave it off.

  4. Great article as always, however I have to say I am less than impressed with these findings. 18% life extension in mouse is palliative at best. That is probably less than normal variation between and inside mice breeds. This whole IGF1, Nkb, etc pathway that many interventions target is probably just something that makes ageing more miserable but does nothing to the root cause.
    I believe we are still just scratching the surface with these studies. There are rodents that have 500-1000% bigger lifespan than mice., I would expect to find something close to the root cause of aging if we had a transgenic mouse with at least 100% extension in lifespan.

    • What I mean to say is that aging research in general is not going the best possible way. Maybe its lack of funding I don’t know. However the way things are going reminds me to common errors that most people make while debugging software bugs. They usually do heuristic search (try different random things and sometimes they succeed) and they do depth first search (lock on some unusual symptoms even if it has nothing to do with the bug). They rarely do systematic breadth first search, which usually mechanistically leads to solution in case of mechanistic systems like computers.
      What I would do here is that I would take two rodents, mouse and rat or mouse and naked mole rat. Then I would start editing genomic regions cutting from rat genome and paste into mouse genome one by one. If possible in an incremental manner.
      I would probably do it with large chunks of DNA that are orthologous.
      Most probably synteny blocks. There are about the same number of synteny blocks as genes 10-20 thousand. The are usually 300k-1Mbp long. I believe CRISPR would allow us for such editing.
      So with 10-20 thousand animals I would build up a rainbow spanning from mouse to rat. If we are lucky we could see jumps in lifespan (altered genetic program) when specific synteny blocks are added. Then we could narrow down on those genetic regions that show a jump in lifespan when inserted.

      • Problem with that approach would be the same as if you randomly cut a program into chunks of code and tried to run those chucks on their own or envelope them in say in a C+ class, – they would contain incomplete statements that wouldn’t do anything or might do something bad. So you’d have to know where some code ends and others begin. And then you have the problem of one piece of code calling on another – say a function call – and that other piece wouldn’t be there (maybe located on a different chromosome). So I’m afraid that approach would be relying on luck more than the current (unfortunately directed by the wrong model of aging, hence useless) approach.

        • Well, program code is easy to mess up because it has very few entry points (couple of tens at most, but usually 1) and runs sequentially or on a few threads as opposed to DNA, which has many entry points ( at least as many as genes, and possibly even more) and runs on countless threads (as many replisome a nucleus might have).
          But my proposal is to cut between synteny blocks. My understanding is that synteny blocks have remained similar between the two species during evolution – even if they got inverted or translocated to another chromosome – because their edges are selected for cut and paste.
          I believe the cuts could be guided in the least conserved regions (white space) further minimizing loss of function mutations.
          Of course there is a chance for lethal mutations but it would be upper bounder by the lethality odds of KO mice because we would be replacing orthologous regions and not lose function.
          See for example this paper:

          Only 162 synteny blocks between mouse and rat spanning 96% of the mouse genome. Also similar number of breakpoint regions. So its about 300 strains. Virtually nothing compared to the 10 000`s or so KO mice strains. And if we get too many lethal mutations we can go for the 7200 micro rearrangements in a further refinement.

    • Fun fact: The Caduceus of Hermes (the two snakes on a staff with wings) was mistaken for the staff of Asclepius (staff with a snake), who was the god of healing

  5. It seems to be complicated. On one hand, we have hormones that clearly do things that we want to stop.

    On the other, they may have helpful aspects too.

    Another interesting possibility is that apparently there is research that suggests that Senescent cells may help in wound healing.

    Anyways, here is an article on the Atlantic covering this. Josh – let me know what you think. I still believe that getting rid of all senescent cells is the way to go, but that we will have to find a way to activate faster regeneration somehow.

    • Thats a bit misleading. the original article is about SASP, that is the terminal phase of cellular aging and or damage.
      I think its not yet proven how this contributes to organismal aging.
      Clearance of SASP cells (senolytics) has some anti aging effects.
      However aging still occurs despite clearance of SASP cells. At the end of the day chromatin in our cells keeps aging regardless whether terminal SASP phase is entered or not.
      I believe SASP is something that makes ageing more miserable, but targeting the IGF1, MTOR, NKb pathway in different places does not change the fundamental pace the organism and its cells are aging.
      Moreover there are hypotheses that SASP cells are actually supposed to be cleared by the immune system, so their accumulation is not the cause but effect of aging (immunosenescence).

      I think chromatin aging is the key. DNA methylation, histone modifications seems to be the most correlated with chronological and biological aging. Also it can be a logical explanation for the huge variance we see in lifespans among mammal species.
      However thats just the nth hypothesis. We need proof. We need to reproduce the rate change of aging and the reduction of the problem.
      I propose we build a rat from a mouse step by step replacing the mouse genome with rat genome. A rat lives two times longer than a mouse. During this incremental genome replacment we should be able to see some jump in lifespan at some points and then we will get the problem reproduced and the root cause(s) as well.

      • I agree that DNA methylation patterns and other epigenetic signals are the primary mode of aging. Whether we can intervene at that level remains to be seen. In the meantime, it’s now clear that a good deal of the damage is done through signaling from senescent cells, so that just by eliminating senescent cells we can get a rejuvenating boost that’s probably as powerful as any intervention we have to date. Note that this can’t yet be done in “natural” mice or humans, that so far it requires preparing the animal before birth in such a way that the senescent cells have a built-in suicide mode.

        • The more important paper in the same issue of Nature talks about life-extension and shortening and produces the extremely important result that “” that the tested lifespan-altering manipulations change the probability of every possible cause of death in concert and to exactly the same extent.” from the Soustrup article. The Baker II article showed the same sort of lifespan alterations seen in nearly every successful intervention some slight lengthening of life span, some delay of the diseases of aging. To me the Soustrup article explains this phenomena – at least when my model of aging is used, it generates exactly the same scale-able curves as Soustrup discovered. It’s really exciting to me.

          • Can you please help me with a link to the Soustrup paper I couldnt find it in neither Nature issues with the Baker papers. Thansk a lot.

          • The temporal scaling of Caenorhabditis elegans
            Nicholas Stroustrup , Winston E. Anthony1, Zachary M. Nash , Vivek Gowda , Adam Gomez , Isaac F. López-Moyado1 Javier Apfeld1†§ & Walter Fontana
            4 f e b r ua ry 2 0 1 6 | VO L 5 3 0 | N AT URE | 1 0 3 -107

  6. Interessting…. We can already start today

    Caloric restriction or intermittent fasting to promote mitophagy and HIT exercise to increase mitochondrial biogenesis.

  7. Hi Josh,

    The more I read about the hypothalamus-pituitary-adrenal axis, the more it seems that aging is heavily influenced (if not outright controlled) by the communication between the HPA axis and the gonads. It is much more evident in females with their strict 28-day cycles and programmed menopause than in males, I believe. There was a lot of great research done on this back in the 50s-70s, both in USSR (Dilman, Frolkis) and the West (Everitt). I wish more of such research was done today using the new genetic and molecular tools that scientists in the 70s could not even dream of.

    Btw, there’s a great 1977 book by Everitt called Hypothalamus, Pituitary and Aging which is a collection of papers by the leading scientists of the time on the topic. A scan of it is available here:

    Best regards,

Leave a Reply

Your email address will not be published. Required fields are marked *