Impaired adult neurogenesis leads to depression – is it realistic?

depressionAbout a year ago we published a paper linking adult neurogenesis to depression. A causal sort of ‘linking’, right? I mean, we found that, when adult neurogenesis was eliminated, mice had elevated glucocorticoids in response to stress and showed depressive-like behaviours1. So doesn’t this mean that impaired adult neurogenesis could lead to depression in humans, in the real world?

Well, it could…and we did end our paper with the following:

Because the production of new granule neurons is itself strongly regulated by stress and glucocorticoids, this system forms a loop through which stress, by inhibiting adult neurogenesis, could lead to enhanced responsiveness to future stress. This type of programming could be adaptive, predisposing animals to behave in ways best suited to the severity of their particular environments. However, maladaptive progression of such a feed-forward loop could potentially lead to increased stress responsiveness and depressive behaviours that persist even in the absence of stressful events.

We had to end it somehow – I was just happy that after 3 years of work we were DONE2! But our final speculation makes it clear that, while this chapter may be done, the story is not. And this fact was rightly pointed out in a recent commentary by Lucassen et al. in Molecular Psychiatry3, where they continue the discussion and bring up some good points. Here is a loose elaboration on some of the outstanding issues they bring up.

Is a feed-forward cascade plausible? In other words, is it possible that stress reduces neurogenesis, which leads to a hyperactive HPA response, which further reduces neurogenesis, thereby additionally increasing the stress response etc etc, eventually damaging the brain and leading to depression? As Lucassen et al point out, stress typically reduces neurogenesis by only ~30% which is much smaller than the 100% reduction seen in our transgenic mice. Could a 30% reduction be enough to initiate this vicious cycle? What if it was chronic? People have looked in the past and not observed HPA alterations after smaller (but also equally large) reductions in neurogenesis, so the answer might appear to be negative. But there are many important differences between studies, including when stress hormones were measured (e.g. baseline or after stress) as well as factors such as life history, genetic makeup, and stressor controllability, as is suggested in the commentary. Of course, in reality, chronic stress has multiple effects throughout the hippocampus (and brain) and so the development of depression is certainly due to additive effects (this is both their sentiment and mine). And so perhaps in the real world a more modest reduction in neurogenesis does have the potential to tip the scales towards depression, if it is coupled with other (which?) pathologies.

How could so few adult-born neurons regulate the HPA axis? Depending on my mood, my thoughts on this question fluctuate quite a bit. Half of the time I think “This is ridiculous” and the other half of the time I look at the evidence and think “Heck yeah. Maybe4.” (see our recent review for more detailed arguments on the heck yeah side of things). In our study we reduced neurogenesis for up to 12 weeks, preventing about 50,000 cells from being added or 10% of the total population. More important than sheer numbers, however, is the ever-increasing evidence that adult-born neurons are different from mature neurons – more plastic, more excitable, uniquely neuromodulated. Some of the most intriguing evidence comes from one of the authors themselves who has shown that even 4 month old neurons (but potentially older?) undergo extensive structural modification following learning whereas perinatal-born cells do not. We have estimated that ~40% of the total granule cell population is added in adulthood in the rat and so my point is that in the real world we have to consider that there are probably cumulative effects of adult neurogenesis over years, and even decades in humans. And so the population of adult-born cells, and its functionality, may not be so small in the end.

Do new neurons sense detect stress through glucocorticoids? One common assumption is that glucocorticoid receptors are necessary for inhibition of the HPA axis…but must that always be the case? By 1-2 weeks of age the majority of adult-born neurons do express MRs and GRs and therefore certainly could directly detect glucocorticoid levels and initiate a shutting down of the HPA axis. But perhaps new neurons process stressful information that is relayed through glutamatergic pathways. In this scenario new neurons might not be required for sensing glucocorticoids and initiating negative feedback. Instead, after being stressed, they could be required for biasing the animal away from the negative experience (akin to perceiving a change in context, similar Opendak & Gould‘s proposal for reconciling stress and memory hypotheses of new neuron function) which might reduce CNS drive on the HPA axis. This opens the door to the possibility that the HPA and behavioural roles of adult neurogenesis are somewhat dissociable, in which case a role for new neurons in the development of depression is not (only) through the feedforward cascade hypothesis but through direct effects on behaviour. This might also help explain the inconsistent links between stress hormones, hippocampal volume, and depression that plague the stress-depression literature.

So, we showed that adult-born neurons are required (in mice) for normal stress responses and emotional behaviour. Our final speculation was, well, just that – a leap towards the next big question. Maybe realistic, maybe not. Maybe a component of the real picture. In any case, the route has been mapped.


Lucassen PJ, Fitzsimons CP, Korosi A, Joels M, Belzung C, & Abrous DN (2012). Stressing new neurons into depression? Molecular psychiatry PMID: 22547116

Snyder JS, Soumier A, Brewer M, Pickel J, & Cameron HA (2011). Adult hippocampal neurogenesis buffers stress responses and depressive behaviour. Nature, 476 (7361), 458-61 PMID: 21814201


1You thought this footnote was going to be some comment on human depression vs. depressive-like behaviour in animal models but instead I just wanted to mention that, now that I’ve moved back to Canada, I feel compelled to use ‘behaviour’ instead of ‘behavior’ and it feels silly because the same people read this stuff no matter which country I’m in when I write it.

2Is there an emoticon for dusting the dirt off your hands, blowing smoke away from the tip of a revolver, or enjoying a refreshing beverage after chopping a bunch of wood or giving birth? Insert it here.

3One of the authors having actually commented on this blog (!), and who is 1st author on what looks to be a very interesting paper that is related to this whole discussion.

4I may be a pessimist. Realist? A guy with a healthy amount of skepticism?

5The photo? It’s a depression.

4 thoughts on “Impaired adult neurogenesis leads to depression – is it realistic?

  1. Hi Jason,
    Again, another great post! I wish I would have the time to read your blog more often. Regarding the very interesting questions your paper generated and our comments on them (we were not the only ones, of course), I have to say that as “One of the authors having actually commented on this blog (!), and who is 1st author on what looks to be a very interesting paper that is related to this whole discussion” (he, he) I though I may give my comments on the discussion you started in your blog (very interesting one, indeed!).

    First, your paper is an excellent one, so I am just going to focus on the last sentences of your discussion, where you propose this feed-forward loop between stress and the inhibition of adult hippocampal neurogenesis (AHN). Secondly, for the record, this is my own personal opinion on the topic and it does not imply to any extent that it is shared by any other author of the paper(s) I am going to mention (or is it quote, in a blog too?).

    Coming to the discussion points you highlighted, I fell the “How could so few adult-born neurons regulate the HPA axis? “ question is the most trivial (or obvious, I should say) one, because it is a point that could be made to every single paper that tries to deal with adult neurogenesis and its functional implications. Whatever thing these “newborn cells” are meant to do (if any, because that is still questionable, of course) they do it in their “fewness”. Perhaps they are “nodes” within hippocampal circuits? It is one of the theories but, who knows?
    So I will better focus on the other two points, as follows:
    Do new neurons sense/detect stress through glucocorticoids?
    As you discussed, although we still can not rule out indirect effects of the glucocorticoids on AHN (possibly trough glutamatergic pathways) or even less, AHN-independent effects on behavior, we have recently published a paper ( where we provide reasonably strong evidence that newborn “neurons”, or at least a fair subset of them, (actually neuroblasts) do sense GC levels directly through the glucocorticoid receptor (GR). Further, when we removed the GR from them, the cells displayed a very interesting repertoire of morphological and physiological changes, suggesting that the GR is a key player in their (correct?) functional and morphological integration. Lastly, these changes were paralleled by impaired contextual freezing during fear conditioning, suggesting again a link to behavior (see previous discussion point).
    So I would say, yes, they do sense HPA activity (directly) and this activity influences AHN, consistent with your observations (and those of others of course, whose work I won’t intend to review here but it could be a nice topic for a future post of yours?).

    So finally, we come to the last (and most interesting, I find) discussion point: what is that HPA activity does to these cells? Or in your words:
    Is a feed-forward cascade plausible?
    This is to me the most difficult to digest of your conclusions, not because I don’t like (I love it, actually) but because, as we discuss in our Commentary in Mol Psych (see Lucassen et al in your post), it seems largely unsubstantiated by the previous literature.
    What do I actually mean by that? Well, the general concept in the stress field seems to be, as far as I understand it, that the effects of stress on AHN even when strong and chronic, may be largely reversible (see as a simple example Suppressed proliferation and apoptotic changes in the rat dentate gyrus after acute and chronic stress are reversible. Heine VM, Maslam S, Zareno J, Joëls M, Lucassen PJ. Eur J Neurosci. 2004 Jan;19(1):131-44.).
    So the question seems to be how to incorporate your feed-forward loop concept into these previous observations. It is a challenging questions and I would be very interesting in hearing your thoughts about it.
    From my side, one possibility is that although proliferation recovers, in each one of these cycles of inhibition of AHN induced by high glucocorticoids, the cells develop some sort of “cellular adaptive memory” that helps them to develop a “ type of programming (that) could be adaptive, predisposing animals to behave in ways best suited to the severity of their particular environments” as you propose in your article.
    Does that sound plausible enough to you? Interested? I will be presenting some of our new data that provide some preliminary experimental support for these ideas in the SFN nanosymposium 720 “Postnatal Neurogenesis: Regulation and Temporal/Spatial Patterns” @ room 273 on October 17 Wed from 8 – 11:30 a.m., so you and everyone else interested in the discussion is welcome to be there.
    Take care,

    1. Carlos – thanks once again for your comments! For the record, I’m definitely not attached to the feedforward loop idea. I think you make a good point which is that we probably need to think about the functional significance of neurogenesis not in the simple terms of whether there are fewer or more new neurons. The transient nature of the stress-induced suppression of neurogenesis might suggest that a feedforward cascade is unlikely, but the truth is that we don’t know much about the circuit functions of new neurons (or even how hippocampal neurons more generally regulate stress responses). I like the idea you mention that maybe stressful events alter new neuron function (without necessarily altering the actual number of new neurons present). Maybe this is a more realistic mechanism by which stress could contribute to depression through neurogenesis – e.g. stress creates a dysfunctional population of plastic cells that then hijack the hippocampus and alter behaviour.

Leave a Reply