Velocity of behavioral evolution

post by Dr_Manhattan · 2014-12-19T17:34:36.217Z · score: 3 (4 votes) · LW · GW · Legacy · 10 comments

This suggested a major update on the velocity of behavioral trait evolution.

Basically mice transmitted fear of cherry smell reliably into the very next generation (via epigenetics). 

www.newscientist.com/article/dn24677-fear-of-a-smell-can-be-passed-down-several-generations.html#.VJRgr8ADo

This seems pretty important.

10 comments

Comments sorted by top scores.

comment by Scott Alexander (Yvain) · 2014-12-20T04:41:12.031Z · score: 11 (11 votes) · LW · GW

Smell is interesting because it's way overrepresented genetically. Something like 5% of most animals' genomes are just a whole bunch of olfactory receptor genes, each for a different individual smell. So it should be unusually easy to do epigenetics with it. Just say "Express the gene for cherry smell more" and then the mice have a stronger reaction to it.

This doesn't mean that any more complex behaviors can be inherited epigenetically. In fact, it might be that nothing else is as suitable to epigenetic transmission as olfaction.

comment by HalMorris · 2014-12-21T00:31:58.716Z · score: 1 (1 votes) · LW · GW

So if I understand what you're suggesting, mice might have inherited a self-gene-modification facility that for smells passing certain criteria (highly associated with a threat somehow), can splice into the genome a representation of a receptor for that smell directly engineered from the molecular structure that "is" that smell. By modifying the genome, it seems we must mean modifying the genome in some or all sperm cells in males, and some or all egg cells in females.

Alternatively, mice sperm or egg cells might contain a previously unknown organelle into which the organism somehow routes samples of really bad smells that would play an active role in structuring the olfactory area of the embryo mouse brain predetermining their reaction to that molecule. This might predict a "washing out" effect over some number of generations.

If such a special case of Lamarkian modification is remotely plausible, it seems impossible to generalize to any sorts of trait other than smell perception.

I vaguely remember some other cases where internal stem cell structure other than the DNA played some role given as an example in Miriam Solomon's Social Epistemology as examples of challenges to Darwinism -- but they also seemed to come down to freakish ungeneralizable phenomena.

comment by Lumifer · 2014-12-19T21:46:22.522Z · score: 3 (3 votes) · LW · GW

An interesting paper. If confirmed, has the potential to shake things up.

A fun factoid I learned from it: mouse sperm (sic!) can smell odors (" the odorants ...enter the circulatory stream and activate odorant receptors that are expressed on sperm")

A sad factoid I learned from it: "A recent study used a social defeat procedure in mice and found paternal transmission of depressive-like behavior in subsequently conceived adult offspring"

And a giggly expression: "sexually inexperienced and odor naive ... male mice" :-)

comment by alienist · 2014-12-20T02:22:03.820Z · score: 1 (1 votes) · LW · GW

I believe I remember seeing results along these lines, the thing was these kinds of epigenetic effects would flush themselves out over several generations.

comment by gedymin · 2014-12-20T11:59:20.593Z · score: 1 (1 votes) · LW · GW

I agree. For example, this page says that: "in order to provide a convincing case for epigenetic inheritance, an epigenetic change must be observed in the 4th generation. "

So I wonder why they only tested three generations. Since F1 females are already born with the reproductive cells from which F2 will grow, the organism of a F0 exposes both of these future generations to itself and its environment. That some information exchange takes place there is not that surprising, but the effect may be completely lost in F3 generation.

comment by Lumifer · 2014-12-19T17:52:25.922Z · score: 1 (1 votes) · LW · GW

Would be important if it gets replicated. Anyone has a non-gated link?

comment by gwern · 2014-12-19T20:27:33.507Z · score: 12 (12 votes) · LW · GW

Anyone has a non-gated link?

You can find a link yourself, it's not terribly difficult.

The New Scientist link is dead - the site is spitting out errors for me - so details are a bit hard to find.

But the keywords seem clear; searching 'epigenetics fear mice' in Google Scholar and restricting to 2014 (since this is 'news', after all) shows a number of hits which look likely. The first one is from January and sounds different, but the fifth one is a Nature blurb from December 2014 which is just right; it's paywalled so you can't see the whole thing, but they at least do provide a citation:

Dias, B. G. & Ressler, K. J. "Parental olfactory experience influences behavior and neural structure in subsequent generations". Nature Neurosci. http://dx.doi.org/10.1038/nn.3594 (2013)

which takes us to http://www.nature.com/neuro/journal/v17/n1/abs/nn.3594.html where the abstract sounds correct:

Using olfactory molecular specificity, we examined the inheritance of parental traumatic exposure, a phenomenon that has been frequently observed, but not understood. We subjected F0 mice to odor fear conditioning before conception and found that subsequently conceived F1 and F2 generations had an increased behavioral sensitivity to the F0-conditioned odor, but not to other odors. When an odor (acetophenone) that activates a known odorant receptor (Olfr151) was used to condition F0 mice, the behavioral sensitivity of the F1 and F2 generations to acetophenone was complemented by an enhanced neuroanatomical representation of the Olfr151 pathway. Bisulfite sequencing of sperm DNA from conditioned F0 males and F1 naive offspring revealed CpG hypomethylation in the Olfr151 gene. In addition, in vitro fertilization, F2 inheritance and cross-fostering revealed that these transgenerational effects are inherited via parental gametes. Our findings provide a framework for addressing how environmental information may be inherited transgenerationally at behavioral, neuroanatomical and epigenetic levels.

Except this is from 2013. What's going on? Well, re-googling the New Scientist URL, Google snippets says it was published 'Dec 2, 2013', and several other hits were also published Dec 2013. So it seems OP is posting some old news. No matter, it's clearly the right paper.

Paywalled still, though. (The Readcube thing Nature just announced might work, but screw that!) Going back to Google Scholar: GS usually provides the fulltext if available as a little text link in the upper right corner of each hit, and in this case it was just a '[HTML]'. But I've noticed that sometimes they seem to privilege a paywall, and particularly for 'hot' topics with political implications (such as epigenetics) there will often be jailbroken or preprint PDFs on obscure domains which you need to click on 'All n versions' to get a listing of.

In this case, 'All 16 versions' immediately turns up a bunch of PDFs:

The first PDF link is broken, but the second one works and seems like a clean final version.

Had Google Scholar not found any PDFs, you could have retried in Google Search. They aren't always in sync, and in general search you may find a site which obscures a PDF download but still provides it. (There's a particularly frustrating sleep journal which is coded in awful ASP and provides downloads of all its papers, but in such an obscure Javascript - I think - way that Google Scholar has no idea about it; you can't even get a download link!)

And if that turned up nothing at all, you could then have gone to Libgen/Scihub, where the paper is already available. This PDF is then easy to copy over to PDF.yt, Dropbox, or you can just link Libgen.

Had the GS PDF links and Libgen both failed, you could have taken the dead PDF links and tried them in the Internet Archive to see if it caught a copy (it often has, if the PDF was visible enough for GS to learn it); if GS/Libgen/IA all fail, you could then try a request on Reddit, and if that failed too, you could try the LW paper request thread. (It's rare for all that to fail for any recent research; if it does, the topic is probably so obscure you will need to either start investing money or simply give up as not worth further effort.)

I hope that will help you find future papers. If you'd like more examples of effective searching, I have a long list of LW examples.

comment by Lumifer · 2014-12-19T20:44:03.606Z · score: 3 (3 votes) · LW · GW

Thank you. I found the paywalled abstract at Nature, but then stopped. I don't use Google Scholar often enough, maybe I should...

comment by AABoyles · 2014-12-19T21:12:03.116Z · score: 2 (2 votes) · LW · GW

You definitely should! It's excellent.

comment by ChristianKl · 2014-12-23T09:39:36.038Z · score: 0 (0 votes) · LW · GW

This suggested a major update on the velocity of behavioral trait evolution.

Only once the study got replicated. A single study shouldn't cause a major update.