Student Report: Neurogenesis, where have you been?

Quick! Change 'peaks' to 'piques' before any of the grammar/spelling cranks see it!

Sounds like an interesting area to explore. But being a pedant sometimes (and a one time editor) peaks should be piques.

Damn! Too late.

Merriam-Webster's on line dictionary includes this definition of the word peak:
a : the highest level or greatest degree b : a high point in a course of development especially as represented on a graph.
Maybe Bright lights was just trying to evoke a visual metaphor of the peaks and valleys on a graph representing the waxing and waning of the reader's potential interest in this subject. No?

Jeez, one might think someone is trying to *provoke* him into actually looking up the meaning of words in a dictionary. Given that the advice of past comenters regarding the use of a spell checker unfortunately wouldn't have helped him in this case.

Main Entry: 2pique
Function: transitive verb
Inflected Form(s): piqued; piqu·ing
Etymology: French piquer, literally, to prick -- more at PIKE
1 : to arouse anger or resentment in : IRRITATE
2 a : to excite or arouse especially by a provocation, challenge, or rebuff b : PRIDE
synonym see PROVOKE

New cell growth isn't everything--your synaptic connections count as well. Even "old" neurons can be dynamic cells, forming or trimming synapses. Remember, too, that a single neuron isn't hooked up to just one other neuron, as is often depicted in textbook illustrations, but, especially in the brain, can form a very complicated network with many other neurons.

But, indeed, forming new neurons is pretty fascinating!

These student essays keep reminding me how little I know about neurology.

"Quick! Change 'peaks' to 'piques' before any of the grammar/spelling cranks see it!"

He's referring to the mountain evidence for this phenomenon....

Personally, I'm surprised that this was ever a question. There's a huge amount of data suggesting that humans don't 'remember' things in the sense of stored data (writing, computers, videotape, whatever) but that instead we simply leave ourselves a few malleable clues and reconstruct memories as we go. We don't actually remember anything; we just think we do.

"Personally, I'm surprised that this was ever a question. There's a huge amount of data suggesting that humans don't 'remember' things in the sense of stored data (writing, computers, videotape, whatever) but that instead we simply leave ourselves a few malleable clues and reconstruct memories as we go. We don't actually remember anything; we just think we do."
I think that is a nice description to a point, but of course stuff is stored to a degree (just search for a memory and you will see). We can also test by stimulating parts of the brain and sometimes getting a memory reported by the subject. I think we probably start off with close to enough neurons by and large for our purposes. Much of the modification I would have thought was in the dendrites which make the connections. Think of your computer, it comes with a hard drive with a limited capacity - but you can put a shitload on there and change it time and time again. But the vagueness described by Nathaniel is also very real especially as you get older. Memories don't come as quickly as before, and you have to feed in quite a few cues before you finally make a link. I would think that in the hardware this would correspond to stimulating more and more neurons till you get the jump.

Ah, neurogenesis, that old point of contention between the likes of Elizabeth Gould and Pasko Rakic (whose name is way too cool).

It's interesting that the olfactory system should be involved... I wonder, however, to what degree neurogenesis in this case has to do with new smells.

Basically, you have a whole bunch of odorant receptor cells in the olfactory epithelium (OE) in your nose. Signals from the OE get organized according to receptor type in the olfactory bulb (OB), which forwards the signals to the olfactory cortex (OC) for processing.

You get a lot of turnover of receptor cells in the OE, and if I recall correctly the populations of various types of receptors can change depending on what kinds of smells you're generally exposed to (ie, wine experts are conditioned so their OEs produce more wine-sensitive receptors). But as far as "new smells" go, I'd think that should be the business of the OC, not the OB.

So I wonder what neurogenesis is doing in the OB... Maybe it's just a health and maintenance issue? The OE and OB make up one of the most direct routes from the outside world to the brain short of a hole in the back of your head.

As a purely speculative matter, one wonders whether there is a link between certain human intellectual faculties (any more specificity than that risks being too speculative) and relatively recent mutations (4-5MY) in primate (especially human) olfactory receptors.

One would think that, ceteris paribus, the more acute olfactory sense enjoyed by our remote ancestors would at least not be disadvantageous, so the frequency of the blunted sense alleles should not have increased to such a pervasive extent unless all else were not, in fact, equal. The hypothetical suggestion is that some other advantage must have been conferred by those mutations, one that more than offsets any potential problems associated with a weaker nose.

This isn't to say the advantage must have been intellectual in nature. For all I know the weaker sense is a side effect of a flatter face and we earned it through sexual selection.

I cannot imagine what form a test of any such hypothesis might take, but maybe that's why I'm not a biologist.

I'm reminded of Douglas Hofstadter here. While much more a philosopher than a neurobiologist, he does have an interesting idea in regards to memory/thinking/consciousness. His argument is that consciousness is more about patterns of activity in the brain, and, as such, can supervene on any adequate hardware. Thus, a thought or memory need not be tied to any single neuron, or even group of neurons, so long as the activation pattern can be maintained. This is basically an analogy to computer systems where identical operations may be performed at different physical locations in the hardware, but produce the same result in the software (there isn't necessarily a single place in a computer's processor/memory relation that is the "addition system"). Actually, the whole system is fairly easy to interpret through a modern computing analogy, so if you know anything about how software supervenes on computer hardware, you have a good idea of where this is going.

Now, to bring it back to a level of empirical data, there were two psychological experiments conducted at roughly the same time that brought back seemingly contradictory results (as a psychology student in my undergrad career, I remember the controversy and the experiments more than who actually performed them, sorry). The name of the experimenters escapes me at the moment (I'll look them up for you if you want, it's in one of a few notebooks scattered out my desk):

One experiment involved training rats to run a maze and then removing portions of the rat's brain. The experimenter would then allow the rat to run the maze again, take out another chunk, so on and so on. The end result was that while removing chunks of the brain would impair functioning of the rat (and, inevitably, was lethal), it did not significantly reduce the rat's apparent memory of the maze. This would seem to argue that memory is distributed across the brain.

The second experiment was one of our favorite electrode stimulus to the brain tests. Basically, an electrode was implanted in the brain, that region got a little zap, and subjects reported their experiences. A lot of them reported the vivid experience of memories. Move to a different location with the electrode, zap, and a different memory or experience. This seems to argue for a localization of memory.

So, here we have a great controversy: two experiments with seemingly contradictory results in the localization/distribution of memory. The resolution? Well, perhaps memory has important relay stations. Basically, stimulate one area and it causes a cascade of activation that "recalls" a memory. But, these localizations may have backups, and so be distributed throughout the brain so that localized damage is not lethal to whole groups of memories. This is quite good from an evolutionary safe-guard view. Or, maybe it's like a telephone call. It is possible to trace the exact route of every call you make through all the switching stations, but two calls placed to and from the same phones may take very different routes depending on the overall activity on the telephone network (a call will be re-routed to another switching station).

Now, I'm just a humble grad student doing evolutionary psychology of religion (PZ, please send us some students with hard science training! The postmodernists are deeply entrenched here!), so this is really my two cents. I bow out to the neurobiology experts, and it's entirely possible that these studies have been refuted by now and neurobiology has moved a long way away from these kinds of ideas. I just feel wary around articles that try to explicitly localize events like this.

I think that is a nice description to a point, but of course stuff is stored to a degree (just search for a memory and you will see).

But probably in a highly compressible format. I can't remember the reference, but I believe neuroscientists have found that our map of the visual field is very sparsely populated. Apparently our perception results in a few highly derived "symbols" that provide detail together with similarly stored symbols when unpacked.

AFAIK another interesting and perhaps related phenomena recently reported is that memory fades every time it is recalled, so we "restore" it every time, with unavoidable change from the intermediate processing.

Memories don't come as quickly as before, and you have to feed in quite a few cues before you finally make a link.

The details depend on how memory is stored and recalled.

Shortterm memory is almost certainly in the form of synapse changes, but AFAIU they see protein changes associated memory which could concern longterm memory. In any case, recall should happen through the use of earlier synaptic network (as Nathaniel describes), and that will change over time. (New synapses formed if often used pathway, old ones removed if not often used, et cetera.)

So it could be problems for old memories, both to be found by associative processes and to be recalled in a firm format.

Ragoth, your statement that the fact that the portions of the brain removed did not impact that particular task does not mean the task is not localized. It is possible that the regions of the brain removed were not involved in the task at hand, or perhaps even involved in memory at all. We do know that there are specific areas of the brain involved in specific aspects of memory in that sort of task. So there is no reason to assume that the results are contradictory at all.

The problem is that the brain does not work at all like a computer system, so any analogy between the two can only be at the most superficial level. Computers are based of general-purpose hardware that can run any program properly designed for that hardware. The brain is not like that, it is a collection of small, extremely specialized pieces of hardware that can only do a very limited set of tasks and cannot change all that much. It can rewire itself to tweak some things, but unless there is catastrophic damage to organism these structures do this specific task consistently throughout an organism's adult lifespan and consistently between adult members of the same species. This sort of general-purpose processing that changes based on the software just does not happen in the brain.

For instance there are particular places in the rat brain at least that form spatial maps of the environment and respond in a specific manner when the rat is in a particular place in a particular room. There are specific structures in the visual system that deal with, for instance, movement of objects horizontally across the visual field, another with objects moving vertically, another with objects moving towards or away from you, and yet a set of others that deal with relative motion of objects compared to the environment in each of those directions. There are specific regions of the brain that deal with extracting specific sorts of location cues from sound, and these regions have a consistent and very specific organization based on sound frequency (and in many organisms the location of the sound as well). These structures do the same thing in all mammals tested and there is an equivalent structure in birds.

But probably in a highly compressible format. I can't remember the reference, but I believe neuroscientists have found that our map of the visual field is very sparsely populated. Apparently our perception results in a few highly derived "symbols" that provide detail together with similarly stored symbols when unpacked.

Basically what the retina does is to extract the borders from the image, places where contrast changes. It is largely insensitive to regions of solid color or brightness. These borders are then further broken down into lines, these lines into edges and corner, and so on. In this manner more complicated shapes are built up as you move along the pathway. For instance research in monkeys have found neurons that respond best to very abstract versions of a fact, two dots for the eyes, a line for the mouth, and a circle for the head. If any of these components are missing or misplaced the neuron does not respond anywhere near as much. What is more, movement and position are analyzed in a completely different pathway traveling along the opposite side of the brain. They don't come back together until the end. How this very diffuse and seemingly highly specialized pieces of information about the visual system come back together to form our unified perception is far from clear.

What is known is that you can have specific defects in very specific tasks due to brain damage. For instance a stroke in a particular region can make you can lose the ability to identify specific faces while stile being able to identify faces in general. When someone with this problem looks at a picture of himself or herself the person can easily identify it as a face, but will vehemently deny that is is his or her own face.

In this manner more complicated shapes are built up as you move along the pathway.

There is that too, with feed forward regulation and all that exciting stuff.

But I was thinking of some more recent specific results, where they IIRC could directly discern our attention (more or less your unified perception, perhaps) on a few visual objects (symbols).

Children say the darnedest things: "...neurogenesis suggests that we can create new neurons while learning new material or having new experiences throughout life (throughout life meaning, past the age of 60)." I suppose it's difficult for a late teen to think that those humans older than 60 are actually living beings, not zombies. Sorry, my cheeks ache from laughing so hard.

Thanks for the information including the link to BioEd. Nice site! I interpreted your "if any of this peaks your interest" to be the colloquialism (peaks your interest/perks your interest - Tow-MAY-toe/Tow-MAH-toe). Certainly not "pique" as that would be out of context with the overall tone of your post.

with no new cell growth after a specific age

Did you mean something else there? We get new cells right up until we die.

I've seen plenty of news stories in recent years that indicates we even get new brain cells (though in decreasing number) throughout our life. (We also lose cells as we age of course, but that doesn't mean that we don't get any new ones.)

"These new neurons apparently are only observed in the olfactory bulb and hippocampus of the brain, which makes sense since you are constantly creating new memories and experiencing new smells."

Prof. PZ, I feel this logic holds true for all the sensory systems. So why do you think that we don't see neurogenesis, for example in the occipital lobe, though we are bombarded with new (in)sight all the time. May be the olfactory neurons are directly exposed to environment and thus have higher wear-tear that they need neurogenesis? Just a wild guess, no scientific basis here.

Hello Animesh, those are the two neurons where adult neurogenesis have been shown to happen. Now, those new neurons can migrate to other areas-including the occipital lobe.

What I find most fascinating about neurogenesis is how our experience/ environment can influence its rate. Oversimplifying perhaps, but it seems so far that physical exercise can enhance the rate, stress reduces it. And mental exercise contributes to neuron migration and survival.

Here's an interesting tidbit--it turns out that hippocampal neurons contain a higher amount of zinc, copper, and iron than is considered "normal" for a cell. Since the hippocampus is involved in memory formation, it's been proposed within the bioinorganic community that we need "zinc to think" (or "copper to think", or "iron to think", depending on your favorite metal). Why this should be so isn't 100% understood. It's known, for example, that an abnormally-high amount of free zinc is stored within the pre-synaptic vesicles, and is released and taken up in the post-synaptic neuron in response to an action potential. However, it's also known that this free zinc isn't really "needed" for cell growth, since if you remove and culture hippocampal neurons, this excess free zinc is lost, even if you grow the cells in zinc-supplemented media. Furthermore, it's thought that this free zinc is important in the biochemistry of Alzheimer's disease, since the amyloid beta plaques are known to be rich in zinc and since it's known that zinc can induce changes in amyloid beta's tertiary structure. So if this zinc is so harmful, why the hell is is there in the first place?

TheBlackCat,

Thanks for the corrections (and yes, I'm honest here. I don't want to be stuck in a mire of untruth any longer than I have to be). My only defense of the rat experiment would be the random areas that were selected in each rat. As such, there was no "program" followed with each rat, and one would have hoped that at least one of these lesions would have contained the memory of the maze, or at least part of the maze, before the rat died.

Actually, I'm in sort of a tough place on this topic and experiencing a lot of cognitive dissonance. Or, I guess you could say that I wear two hats: on the one hand, I have read a lot of texts on the evolution of the mind/brain and know that the brain is modular and not a general information processor, as the Standard Social Sciences Model takes it to be, but in fact has quite specific modules that process very specific forms of data. I also did a lot of work with perception in psychology and I know the various areas of perception and processing in the brain. When I'm working specifically with psychology or evolutionary sciences, this is framework with which I approach the brain, and thus avoid any messy entanglements with the much more speculative philosophy of mind.

However, I'm also a proponent of the possibility of Strong AI, and in this case I don't think that there is anything inherent about the structure of the brain that is absolutely necessary for consciousness. The modular regions with their particular distribution, material, location, and connections are probably not strictly necessary for the existence of consciousness. While some analog may be necessary, I don't think that you have to have organic neurons arranged identically to the structure of the human brain (and I know this is an over-simplification) to produce consciousness. It may be the case that (at least for right now) the brain as it is, is necessary for consciousness. In the future, however, we may see that consciousness can supervene on other mechanical systems. Or, of course, this might not bear out and all us Strong AI types will be sadly disappointed, but forced to concede to the facts of the matter.

After all, if human consciousness arises out of the interactions of neurons (and unless you believe in a soul [which I don't], it really can't be any other way), then why not out of other mechanical or partially mechanical systems? As a gross over-simplification, the neuron's function is to either transmit or inhibit a signal it receives: either it fires, or it does not, and there is a stimulus threshold which determines this activity. Again, a gross over-simplification, but electronic logic-gates provide much the same function. This holds water only so far as whether or not a functional reduction of consciousness is possible. It is also a question of whether or not, in principle, the mechanisms that give rise to consciousness can be reduced to the level of physics, whether or not this actually helps the analysis (which, it really, really doesn't. An explanation of any macro phenomenon at the level of particle physics is close enough to meaningless).

So, our brains: very modular with a great deal of specificity and localization, which may in fact be a key to the evolution of consciousness. Future brains: maybe not?

As I said, I did philosophy of mind and evolutionary psych as an undergrad and am working on a psychological (realist and naturalist) critique of religion in grad school. Neurobiology, specifically, is not my strongest point. However, it is an area I am very interested in, and want to try to work in to my reading with the rest of my grad work. If you have any texts that could recommend, I would greatly appreciate the input.

Thanks.

Hey Alvaro, thanks for the insight. Would like to read the research article if you can share via mail (since most of them are locked behind payment gateways). I hope it tries to answers, why those two regions?

Animesh Sharma is absolutely correct is his (her?) guess. The "olfactory dogma" holds that constant generation of olfactory sensory neurons (OSNs) happens in order to replenish wear and tear. OSNs are the only part of CNS that's directly exposed to the air. The turnover rate is actually pretty fast: the lifetime of a mature OSN is a couple of months.

The really cool thing here, of course, is that all OSNs expressing a given receptor project to the same (pair of) glomeruli in the olfactory bulb. And the axons of new ones find their way to the very same glomerulus.

But all of this is hardly news. The comment that exposure to specific odorants increases the prevalance of corresponding receptors, on the other hand, is news (to me). People have been waving hands, but technical challenges would be very considerable (no need to go into detail here). I haven't been following the field very closely, but I think that if there was a paper showing that, it would have made a big spalsh.

If you have any texts that could recommend, I would greatly appreciate the input.

Principles of Neural Science, Kandel and Schwartz. Ask anyone in neuroscience about "Kandel" or "Kandel and Schwartz" and they can point you in the right direction.

If you are from psychology your probably don't know differential equations. If you do, the next book is great. If you don't, you won't have a clue:

Theoretical Neuroscience, Dayan and Abbott. Once again, "Dayan and Abbot" is well-known and anyone dealing with mathematical aspects of neuroscience will recognize it by that name. If you understand the equations this one will make it clear you just how completely different neurons are from logic gates, and why you often need a powerful computer running for hours just to use a highly simplified computerized version of a single neuron. If you can't understand the math then Kandel and Schwartz will at least give you conceptual understanding but not the mathematical one needed to really understand the differences.

Hello Animesh, I don't have your email.

But let me
1) clarify that no scientist I know of claims that neurogenesis only happens in those areas. as I understand it (I am not a neuroscientist myself), today it only has been shown in those areas. who knows what will be discovered next year? (we can guess)
2) great write up at SfN, including references
http://www.sfn.org/index.cfm?pagename=brainBriefings_adult_neurogenesis

TheBlackCat,

Thanks for the texts. I do know differential equations (I did math and evolutionary bio for electives outside of psych and philosophy & religion; but unfortunately could never work in a class on neuroscience specifically), so I'll check out the Dayan and Abbott along with the Kandel and Schwartz. Again, I really appreciate the corrections and the help. I'm really interested in improving my knowledge in this area. Thanks!

Thanks AO for the information.
Alvaro thanks for that link, I was looking for those references. I am inclined to understand "why it happens there" part of it.

There are some hotly-contested mutterings running around in neuropsychology circles that the brain is really more of a holographic computer than it is a finite state machine, with several loci for different "images" or functions the brain performs. This view seems supported by the fact that in older persons with degenerative brain disorders, memories don't seem to fade in random order, nor do they appear to be there and suddenly *not* there. Rather, memories seem to fade in an order consistent with those less-ingrained fading before more-ingrained ones, and that the memories become harder to recall but don't actually up and leave.

It's an interesting take on the idea - I wish I had some resources for you to look, but Google should be a wonderful friend to you on this one.

I think that is a nice description to a point, but of course stuff is stored to a degree (just search for a memory and you will see).

But probably in a highly compressible format. I can't remember the reference, but I believe neuroscientists have found that our map of the visual field is very sparsely populated. Apparently our perception results in a few highly derived "symbols" that provide detail together with similarly stored symbols when unpacked.

AFAIK another interesting and perhaps related phenomena recently reported is that memory fades every time it is recalled, so we "restore" it every time, with unavoidable change from the intermediate processing.

Memories don't come as quickly as before, and you have to feed in quite a few cues before you finally make a link.

The details depend on how memory is stored and recalled.

Shortterm memory is almost certainly in the form of synapse changes, but AFAIU they see protein changes associated memory which could concern longterm memory. In any case, recall should happen through the use of earlier synaptic network (as Nathaniel describes), and that will change over time. (New synapses formed if often used pathway, old ones removed if not often used, et cetera.)

So it could be problems for old memories, both to be found by associative processes and to be recalled in a firm format.

In this manner more complicated shapes are built up as you move along the pathway.

There is that too, with feed forward regulation and all that exciting stuff.

But I was thinking of some more recent specific results, where they IIRC could directly discern our attention (more or less your unified perception, perhaps) on a few visual objects (symbols).