Friday, July 11, 2014

The anti-Lamarckian gut reaction: keep the bar high--but your minds open

Biologists tend to ridicule Jean-Baptiste Lamarck for his version of evolution, expressed in his 1809 Philosophie Zoologique.  Laughing at him is one of our favorite sports.  Of course most of those who do the laughing never bothered to do any actual reading of Lamarck's famous book, but who's gonna sweat the details?

So what was it that he said, and why was it so risible?

Larmarckian Inheritance
Jean Baptiste de Lamarck
About 50 years before Darwin's famous theory was published, Lamarck explained the diversity of complex organisms by what we term the inheritance of acquired characteristics.  In his theory, traits that were used by an organism were transmitted to its offspring, and traits that were not used were not transmitted.  Very gradually, organisms would develop, refine, and elaborate useful traits.  Lamarck may have been wrong but not entirely so, given the data and attitudes of the time, and he was seeking a material explanation for biological complexity and its origins.  Good overviews can be found in the prefatory material to the 1984 English translation of Philosophie Zoologique published by the University of Chicago, and by SJ Gould's 2002 The Structure of Evolutionary Theory.  Other authors seem typically more derogatory and pejorative and less clearly acknowledging what Lamarck actually said.

The point is not to defend Lamarck, but to see why his views evoke such gut negative reactions.  The modern purely materialist theory of evolution (largely of Anglo-American origin) stresses randomness of inherited change due to external forces of mutation, and the brutal screening by natural selection to favor those genotypes that confer advantageous traits on their bearers.  Self-satisfied in what really is somewhat our own arrogant dogma (see our series on the mythology of natural selection, which begins on Monday of next week), we brook not even the slightest breach in our own dogma.  Still, in the way he tried to explain things, Lamarck does seem to have been mistaken, and his ideas (which were in the air at the time) in a way set the table for a better kind of material explanation for evolution and the traits of organisms, due to Darwin, Wallace, and a few others who ventured correct partial statements of things.

Lamarck's idea was that what organisms do during their lives, in response to the challenges of their environment, is somehow materially transmitted to their offspring.  This is an inner rather than outer source for the variation: systematic habit-induced physical change rather than externally screened random variation.  It didn't help Lamarck that the Soviet era genetics tried to appropriate Lamarckian rather than Darwinian evolution for various ideological and sociopolitical reasons, nor that some of Lamarck's contemporary rivals denigrated him, nor that Darwin found basic holes in the idea (despite holding a very similar theory of inheritance). 

Our point here is that there are some very good reasons based on 19th and 20th century biology, to hold the view that Lamarck's mechanisms don't hold much water, and this is why when one argues for anything that might seem even the slightest bit Lamarckian, the ridicule begins and the burden of proof is raised to a much higher level than when biologists venture their routine Just-So stories about how the elephant got his baggy skin.

What we do know
One thing we know very well is that gene usage does involve experience-based feedback onto cells' genomes. Cells use a subset of their genomes, and which subset is dictated by their context--by aspects of the cell's local environment.  Indeed, cells are loaded with environment-sensor (like receptor molecules) that monitor the world outside the cell and adjust gene expression accordingly.  This is clear, experience-based modification of the genome.  Of course, the modification is epigenetic:  it is not a change of the code book (the DNA sequence) itself and that is the key.  It is a change in the DNA molecule that affects how it's used, but not the sequence-based code.  To oversimplify, it is the binding of a molecule to a specific sequence-based bit of DNA near a gene that affects whether that gene is used by the cell at that time. When circumstances change, the molecule may leave the DNA, changing whether the gene is used or not.  Because the molecule binds to specific DNA sequence, epigenetic change does involve DNA sequence, but doesn't change it.

Second, in many if not most multicellular organisms the experiences of life affect its body tissues.  If a vertebrate does hard work, its muscles and joints may gain improved strength--something all of us know very well.    But if cell structures are produced by genetically coded molecules, the code itself isn't changed by the experience.  More importantly, an entirely separate line of cells, the germ line (sperm, eggs, pollen, etc.) is separated from the rest of the body's cells early in development.  Pumping iron may pump up your pecs, but it doesn't alter the relevant genes in your germ cells.  Even if it would be a good thing, the basic idea for over a century, supported by lots of evidence, has been that there is no way for a (say) muscle-specific genetic change to be built into the genome of a muscle cell, much less into a germ cell.

Instead, if pumping iron is good for your reproductive success, then those who by good mutational luck carry muscle-related genetic variants will reproduce better, passing the screen of natural selection and proliferating their good genes into the future.

Arnold Schwarzenegger in Pumping Iron; Wikimedia
The problem with a Lamarckian genetic alternative would be that pumping iron selectively mutates the muscle-related gene and when it turns out to be useful then engineers that same mutation in his germ cells. There is no known mechanism by which a specific mutation in a specific useful gene can be engineered into a germ cell, just because it happened to be useful in a muscle cell.  If it exists it will have transformative effect on biology, and that is why the standard of proof is so high, and skepticism so great against anything that seems like such a claim.

In fact, there are many examples in the literature of trans-generational transmission of epigenetic states. Each has been blasted and for essentially this reason:  biologists are generally not willing to open a breach in their Darwinian selectionist firewall.  Since anything truly Lamarckian could threaten the most solid bits of biological-evolutionary theory, it is entirely appropriate that a high bar and burden of proof be maintained.  But if it's true, it really is no threat to the state of the world; it's just knowledge of a new mechanism for at least short-term adaptability of complex species. In fact, it was anticipated way back into the late 19th century.

Current contentious examples
We write this because a report has just appeared in Science that claims something that seems like Lamarckian inheritance.  In this example, a starvation induced in laboratory mice when they were pregnant led to an epigenetic change in various genes that, the authors say, led to risk of diabetes, and that risk also was characteristic of the next generation, the grandchildren, even though their mothers were not subject to starvation. As described in a commentary in 10 July Nature, the experiment is in effect a test of the aftermath of a winter of starvation in Denmark in 1944-45.  Children conceived during this time were born underweight, and experienced health problems which their children, too, seemed to go on to experience.  

This report was just the most recent of several related to claimed epigenetic transmission of chronic disease-related traits that have appeared.  An even eerier recent report claimed that male mice exposed to fear-stimulus in the presence of a specific odor, were conditioned to hyper-express an odor-receptor gene that detected that order, but then this same gene was marked for over-expression in their offspring, and their grand-offspring, even though there was no further fear-conditioning.  (We blogged about this here.)  Gene expression induced by experience in the nose and then the same gene primed in sperm cells to be expressed in the males' offspring--and maintained in the sperm-line for a third, grandchild, generations?  Are you kidding?  

And there are other reports of similar multigenerational epigenetic transmission, some of them in controlled experimental settings like these examples.

The key reason for the strong skepticism at reports like these is that a specific gene in a germ cell line, a cell not directly affected by the environmental factor, is modified by that experience and the modification is then transmitted.  This might not seem like a problem, except that, genome-wide, the epigenetic state is generally highly programed for embryonic development, with sperm and egg genomes are subjected to heavy genome-wide epigenetic reprogramming before conception.

The idea that an experience-based epigenetic responsive state can get into the germ line and be transmitted for several generations is a threat to a Darwinian dogmatist--emotionally, it's like trying to get a biblical literalist to accept that Genesis might be at least a bit metaphoric.

Science is about learning new things
But we shouldn’t just defend dogma by being dogmatic!  There really shouldn't be any problem at all with this kind of multi-generational transmission--if it's true.  If it is, then we just have learned of a new mechanism of adaptability by organisms by which they change their biological state to reflect their circumstances (like shivering when it's cold, or an adrenalin rush when frightened).  Somehow, the body would know which gene was modified epigenetically by experience, and finds that same spot in the genome of a sperm or egg cell and makes the same modification.  That this could be transmitted to future generations could be a fine adaptive mechanism because circumstances might not change, and organisms would be epigenetically prepared from birth to meet them.  If true or shown somehow to be general, we'll all say that this is a marvelous aspect of evolutionary adaptation, and how could we have missed it!  

One can conjure up various ways this might happen….except for the minor detail that we don’t actually know of any such mechanism!  That doesn’t mean it can’t exist but it’s proper that one must find it before the results will be accepted.  Still, the more results of this sort that are reported, the harder should be the search for something that a century of work suggested didn’t exist.  Can all these reports be wrong?  If not…..what can make them true?  If so, are they wrong for some murky methodological artifact?  At some point, sneering should stop and hard work to find the mechanism should start.

We've tried to outline what the controversy is and why it is reasonable to be very skeptical of these reports, that seem so 'Lamarckian' even if we take into account that he was writing, or guessing, based on the state of knowledge 200 years ago.  If indefinite proliferation of an epigenetic change is ever proven, then it will show that this sort of inheritance, even if not affecting DNA sequence directly, is a part of evolution. It won't be exactly Lamarckian, but it won't be exactly Darwinian either.  It will be a remarkable revelation whose discovery we'll celebrate.

But if it's warranted that we give these reports a very hard look, that should not require that we cling to a cartoonesque oversimplification of the current dogma, that everything about everything is the result of Darwinian style competitive natural selection.  And with this comment, we'll return to writing our series about selection and the origin and genetic basis of complex traits.


Holly Dunsworth said...

Added to my intro bioanth course reading list. Covers Lamarck & epigenetics just enough and so well. Thank you!

Ken Weiss said...

A couple of tweeters have mentioned August Weismann who provided evidence for the separation of the germ line from the 'soma' (body cells). We jus didn't mention him by name, but that was basically what the problem was that we discussed here. Weismann famously cut mouse tails off for several generations leading to no inborn shortening.

But if evolution of complex traits is slow, as basically everyone accepts to be the general 'rule', then Weismann's experiment would be totally unconvincing in itself (indeed, even if Lamarck were correct, not all traits need follow his path and indeed, the mice did not lose tail-ends by their own habits.

Nonetheless, with some important modifications from what he did a century and more ago, the germline is still largely isolated from the effects on the other kinds of cells in the organism. Unless or until some clever work shows that we've been wrong all this time in some important ways. As we said, it seems unlikely at present, but if true, well, then we'll know more about biology than we do now.

Ken Weiss said...

I want to add another afterthought. Single-celled organisms don't have separated germ and somatic cell lineages. Some organisms reproduce in non-'standard' ways: slime mold exists mainly as separated cells, but they aggregate under some conditions to form a partly differentiated temporary organism that has some 'reproductive' cells that are shed. Sponges and some other species have cells that can differentiate into germ cells, rather than clearly separated cell types.

Indeed, advanced plants like the trees in your backyard, have stems that grow but differentiate to form germ cells (in their flowers), again a different kind of separation of cell type.

What is consistent, even in single-celled organisms, is that we don't know how experience can lead to adaptively-specific DNA sequence mutations that would then be inherited.

So, if the living world is a mix of 'strategies', the basic idea Lamarck advocated which in his stage in the age of science was plausible in principle, we still know of no general way that acquired characteristics based on habits can be introduced into genomes. Thus, even with various specialized relationships between body form, mutations, response to environments and so on, 'real' Lamarckian inheritance has not been demonstrated. Not yet, at least, if it ever will be.

Sto-ology said...

I write about this in my book. There are several weak theories in evolutionary biology that have been lingering way too long - Hamiliton's Kin Selection was recently dumped by EO Wilson and its weak cousin - Dawkin's selfish gene was undone before it even hit as a meme (his only contribution - aka trend or social virus) by Lynn Margulis endosymbiosis theory 30 years pervious, which he recently finally acknowledged - right before she passed. Dawkins is a zoo-ologist and spews out such negative nonsense - it is amazing that the media still gives him a voice. His followers are mostly angry atheists. Other weak theories in evolution are A few by guru Stephen J. Gould - such as Punctuated Equilibrium - does a good job at explaining the problem of the fits and starts in the evolution of species but a weak explanation of why. How mutations occur over time is a total mystery and there are varying schools of thought. Studies have shown phylogenetic mutations occurring anywhere from 3-50,000 years. There are issues I have with Darwin's 'solid' theory - such as the core randomness of all mutations. The word random means ignorance to me. I am of the Einstein school - God does not play dice with the universe even though most current paradigms support that randomness is the norm - at least for now.

Peace and Properity

Sto-ology said...

Nice balanced article. At this point - with over 10 years of Epigene studies all over the world - it seems pretty obvious to me - an amateur scientist - that the methylation of epigenes is the first stage of eventual genetic mutation of the main DNA code. In this case it's all about the environment and repeating either good or unhealthy behaviors patterns in response to it. It's not always about procreation. I propose that the need to breed and survive are absolutes found in every form of life - happens regardless of what creatures do or do not do. Like a constant. Let's face it - other than the strength of the phylogenic tree based on old digs and solid DNA bio-markers - biological evolutionary theory is up for grabs based on all my readings from different experts who totally disagree. I mentioned EO Wilson and there is this brilliant Japanese guy Dr. Nei with his mutation driven evolution concept. It's like the the chicken or the egg conundrum - then there is the new school of RNA folks saying it came before DNA. On and on it goes and I love it - student for life!

Ken Weiss said...

I don't know your identity so may or may not know your book. We deal with some of these issues in our Mermaid's Tale book. I think that the basic power of the Darwin/Wallace realization (and Wallace himself backed away in various ways) is its simplicity.

The kinds of things you raise in terms of various dogmas, including hyper-Darwinism or the claims of alternatives are a problem because they are often claimed AS alternatives. Instead, we think as we'll try to outline next week, that life including the evolution of complex structures is nuanced.

The debate about mutations is another example. A cosmic ray's DNA hit cannot be predicted in any way that is useful to understanding life and evolution even if, in a perfectly Newtonian world everything is predictable in principle. In that case, evolution and selection and so on are just working out of what was built in at the Big Bang.

But for other purposes, including sampling and study design, randomness is more meaningful. Not all DNA sites are equally mutable but relative to function it makes general sense (even if it turns out that there are some exceptions) to refer to it as random.

In that sense, as we'll discuss next week, there is surely a chance (random) aspect to survival and reproduction, and genetic drift surely is part of the story (who gets hit by lightning will usually have nothing to do with his/her genotype), and the degree to which 'selection' is deterministic vs affected by chance is important, and still controversial. Some biologists are arch-determinists and seemingly won't allow anything to be due to drift. But not every nucleotide in DNA can have evolved by deterministic selection, and if fitness is very weak relative to drift (almost an untestable proposition, even if it's important to the issue), then we really ought to stay away from dogmatism about evolution.....

There are deeper issues there, in my opinion, which have to do with the probabilistic nature (or not) of genomic causation and hence of prediction, especially with respect to feedback-dependent things like behavior.

Sto-ology said...

Sounds cool - I will be checking in! I guess my other comments on Epigenes were G-filed?

Sto-ology said...

Sounds cool - I will be checking in. I guess my other comments on epigenes were g-filed?

Ken Weiss said...

Many people want to be the next Darwin or the genius to overthrow him. They're largely trying too hard. But there are many unclear aspects of evolution. Nei (whom I've known for 40+ years) makes important points. I can't say the same for Wilson (in this context), but the stress he and Nowak have on cooperation (also argued by another Wilson, David Sloane) is important--something we stressed even in the title of our book, but for different reasons. Common ancestry was as fundamental insight as any other, but the origin of complex traits,that probably involves various kinds of selection, even if not just as Darwin had it, seems to be involved.

The idea of epigenetics leading to genetic evolution was, in pre-DNA sequence terms, extensively argued. Baldwin made the point in regard to behavior, and later Waddington made it more formal in his idea of 'assimilation'. But it is not yet clear how often these things are important or whether slower, subtler, less clearly understood factors are at work.

Sto-ology said...

I have heard of Waddington - but not Baldwin - thanks! It seems you have been at this for a long time. I really respect your flexibility - in terms of being open to different ideas. What do you think about EO Wilsons breaking away from the dominant but very weak Hamilton's hypothesis of Inclusive Fitness - it seems 50+ years of proofs of altruism outside of not only direct kin but actual species has relegated his limited mathematical models to the dust bin of history. Unfortunately his concepts are well entrenched because of annoying amateur scientists like Richard Dawkings spewing his selfish gene concept built upon the Hamilton foundation.

Ken Weiss said...

The best population ecologists who have actually tried to confirm Hamilton's hypothesis (often wrongly referred to as a 'law') is that it only seems to apply under restrictive circumstances and is not, generally, a 'law' of how Nature works. It is in fact just a statement of fact: if you help your kid, you can give up no more than one more of your own kids for his/her gain of at least two.

To generalize and explain the grossly obvious existence of more altruism than can be related to Hamilton's idea that you'll only help your kin in this restrictive way, terms like reciprocal altruism have been introduced. In a hyper-deterministic, ideologically hyper-Darwinian worldview, you only help your neighbor because (and to the extent that you expect) s/he will later help you. Since such judgments can be fallible, we now have a 'theory' that can be applied to whatever one wants, and hence it departs from anything that should be called science.

EO Wilson did brilliantly wonderful work on ants. He's been a terrific ecologist. He bought into hyper-Darwinian theory with Sociobiology, and should have left off his last chapter (on humans) or his book on humans. Now he's following a new wave, or probably more a fad, about cooperation. We've subtitled our Mermaid's Tale book with the word 'cooperation', whose usage we try to explain there. Wilson and Nowak are generally talking about social cooperation, and I think that has come under attack (e.g., by Nei) even in regard to bees, if I recall correctly. Cooperation is a fact of social life and needs to be studied properly. I can't really judge Wilson/Nowak in any detail since I'm not familiar with it. But Wilson was, as I said above, a brilliant ant ecologist.

I agree with your comments, and I think most serious biologists have long ago recognized the excess simplification (to be kindly) in Dawkins.