The Enigma of Lamarckism

Lamarckism (or Lamarckian inheritance) is the idea that an organism can pass on characteristics that it acquired during its lifetime to its offspring (also known as heritability of acquired characteristics or soft inheritance).

– Wikipedia

Many of us have probably been taught that Lamarkian inheritance is anathema. Heresy. But why would that be the case? Is it for theoretical reasons or simply because of a lack of empirical evidence?

Let’s transition to a hypothetical RNA World.

Must it not have been the case that in the RNA world inheritance was strictly Lamarckian? How could it have been otherwise?

So at some point there must have been a transition from Lamarckian inheritance to Darwinian inheritance. Why would that be an advantage and why would vestiges of Lamarkian inheritance not still be evident?

What do readers think?

74 thoughts on “The Enigma of Lamarckism

  1. Must it not have been the case that in the RNA world inheritance was strictly Lamarckian? How could it have been otherwise?

    Why? Why not?

    If you have a case to make, why not stick your neck out and make it?

  2. Must it not have been the case that in the RNA world inheritance was strictly Lamarckian?

    You say that as if it is obvious. But it is far from obvious to me. Can you at least provide some reasoning to support the claim?

  3. Maybe off thread but recently on other blogs the idea occurred to me that the traits domesticated animals uniformly acquire might be a obvious case of Lamarckism.
    The evolutionists scrambled to explain away domesticated creatures traits that could not be from selection on mutations/random or not.
    they tried adrenelinish ideas but even that would be Lamarckism.
    Creatures picking up new traits and giving them to offspring might of been staring at us all along.

  4. The RNA world refers to the self-replicating ribonucleic acid (RNA) molecules thought to have been precursors to all current life on Earth.

    Wikipedia

    If such a molecule acquired, during it’s lifetime, some characteristic that it passed on to it’s offspring, that is, by definition, Lamarckian inheritance.

    What is the alternative, in an RNA world?

  5. Firstly, I’d say that Darwin didn’t propose a mechanism of either inheritance or of variance generation, and Lamarck only proposed the latter.

    The main difference between the two is that Lamarck thought that life circumstances of the parents would tend to produce children with heritable characteristics that better suited them to those life circumstances, whereas Darwin (though he considered Lamarck’s idea seriously) concluded that it wasn’t necessary – that even if variation was produced independently of the current environment, you’d still (through “natural” selection) end up with more of those variants that happened to be better suited than those that happened not to be.

    Which is true.

    So Lamarckian inheritance is not “anathema” – it’s just not part of Darwin’s theory. That doesn’t mean it doesn’t happen, although until relatively recently it wasn’t obvious why or how it might, except in the cultural sense of inheritance.

    We now know of more vectors of inheritance than genetics and culture, including epigenetics, and possibly even mechanisms that jiggle about with DNA sequences in response to environmental stimuli.

    Not all epigenetic effects are beneficial though – how does it make an individual fitter to have epigenetic markers that predispose her to depression, say, inherited from a mother who became depressed?

    However, it’s possible that some real Lamarckian inheritance happens – that things that happen to parents are passed on genetically (or culturally!) to the offspring and better fit them to deal with those things happening to them as well.

    Which is one of the ideas behind the “Third Way” – essentially, the argument is that Darwinian selection can happen at the level of the population as well as at the level of the individual – populations of organisms whose heritable makeup is such that it includes mechanisms that help it to evolve adaptively will be more likely to persist (not go extinct) just as individuals whose heritable makeup help the individual survive and breed will be more likely to leave descendents with that makeup.

    So I wouldn’t say that Lamarckianism is “anathema” to Darwinian evolution – Darwinian evolution can actually potentially account for why Lamarckian mechanisms, if they do exist, could have come about, by extending the theory of natural selection to population level. The same principle would predict that mutation rates would tend to optimise – populations whose genomes reproduce too faithfully, or not faithfully enough will tend to less robust to environmental change than populations whose genomes reproduce with optimal fidelity.

  6. Mung: If such a molecule acquired, during it’s lifetime, some characteristic that it passed on to it’s offspring, that is, by definition, Lamarckian inheritance.

    Not really. That would just be a germline mutation that occurred during the lifetime of the parent. It would only be “Lamarckian” if that mutation was one that would benefit the offspring when faced with the hazard that produced the mutation in the parent.

    What is the alternative, in an RNA world?

    I don’t know, but it isn’t Lamarckian anyway.

  7. Elizabeth: Firstly, I’d say that Darwin didn’t propose a mechanism of either inheritance or of variance generation, and Lamarck only proposed the latter.

    I don’t think it’s true that Lamarck only proposed the latter. He understood that natural selection would weed out some types and not others. What he didn’t get was that that was sufficient. His proposal was more complicated as well as incorrect.

  8. I accept what you say, walto, but what I’d meant was that neither actually proposed a mechanism of inheritance that I am aware of – inheritance of traits was observed, but I don’t think either knew of Mendel, and even Mendel didn’t have a mechanism – he just got a good handle one way in which the mechanism must behave.

  9. Yeah, I think that’s right. And Lamarck had no concept of random variation as a a possible motor–in conjunction with what he DID notice, survival of the fittest.

  10. Mung: If such a molecule acquired, during it’s lifetime, some characteristic that it passed on to it’s offspring, that is, by definition, Lamarckian inheritance.

    That seems to be talking about a mutation. So no, that isn’t Lamarckian.

    Several decades ago, some folk had the idea that planaria stored memories in RNA. It turned out to be false. But if something like that had been true, then that might be basis for Lamarckianism.

    My understanding of Lamarck’s ideas is that some charactistics acquired by a learning process could be passed on.

    The Baldwin Effect seems to be the closest that we come to that.

  11. From the OP:

    Lamarckism (or Lamarckian inheritance) is the idea that an organism can pass on characteristics that it acquired during its lifetime to its offspring (also known as heritability of acquired characteristics or soft inheritance).

    – Wikipedia
    …..
    Let’s transition to a hypothetical RNA World.

    RNA is not an organism:

    In biology, an organism is any contiguous living system, such as an animal, plant or bacterium.

    —Wikipedia

  12. This OP is so confused it is difficult to pick a place to begin.

    All germline mutations or insertions or changes are “acquired,” regardless of the complexity of the organism. Doesn’t matter if it’s a virus or a human. it just so happens that viruses are all germline.

    What Lamark envisioned is that things that happened to an organism at the phenotype level could be passed down to the germline.

    In simple terms, that would violate Crick’s dictum that somatic changes cannot be translated backwards into relevant changes in DNA. This doesn’t mean environmental challenges can’t result in mutations; it just means that “learning” can’t be passed on.

    Epigenetic changes are sometimes passed on for a few generations, but they do not change the genome sequence.

  13. Germline introduces a distinction between germ cells and somatic cells that did not exist in the RNA world.

    And yes, calling the sort of inheritance that did take place Lamarckian is anachronistic, but it clearly states the issue, which is that changes to the “organism” during it’s lifetime were passed on during replication because that’s what replication meant in that context. A 3D shape and a copy of that 3D shape.

    The point is that there must have been a transition from one style of “inheritance/copying” to a completely different style of inheritance, the sort of heredity that we associate with Darwinian evolution.

    But why? What would the advantage have been?

  14. Mung: Germline introduces a distinction between germ cells and somatic cells that did not exist in the RNA world.

    And doesn’t exist in the unicellular world even today. Until meiosis comes along, it’s all “somatic” in effect – which means that at least some mutations will occur during the life time of the “parent” cell, and some will be introduced as it divides. That doesn’t make the former “lamarkian” and the latter not.

    Lamarckian inheritance only really makes sense at the level of the phenotype – when some hazard or opportunity that confronts the phenotype leads to offspring with genetic sequences that will help that offspring, and its descendents, deal better with similar hazard or opportunity.

    So I don’t think it’s especially relevant to RNA world, although it might be. An example might be if an RNA cell, or some protocell with a rudimentary polymer genome lived in an environment in a particular monomer was abundant, and so cells with polymer sequences that included that monomer would tend to reproduce more successfully than cells that didn’t. Jack Szostak has suggested this as a possible way in which cells-with-genomes might have got going.

  15. Lamarckian evolution is about the environment poking useful changes into the genome. It is conceptually opposed to “random” change being selected in a population.

    Is this what Mung has in mind? A process of genetic change that isn’t random change?

  16. Mung,

    The point is that there must have been a transition from one style of “inheritance/copying” to a completely different style of inheritance, the sort of heredity that we associate with Darwinian evolution.

    The transition you’re speaking of is illusory, an artifact of the way you’re looking at things.

    Here’s a better way of looking at it that demonstrates the continuity.

    Inheritance was (neo)Darwinian both in the RNA World and afterwards. That is, mutations occurred and selection worked to weed out deleterious mutations and retain beneficial ones.

    The real transition was not from one mode of inheritance to another. Instead, the change was from the RNA World, in which the phenotype was not distinct from the genotype (since only one molecule was involved) to a post-RNA World in which phenotype became distinct.

  17. petrushka: Is this what Mung has in mind? A process of genetic change that isn’t random change?

    No. I don’t care how the molecules changed, only that they did and that the change to the molecule was inherited.

    I’m trying to focus on the transition that must have taken place. I do see it a change from one mode to another.

  18. Mung: The point is that there must have been a transition from one style of “inheritance/copying” to a completely different style of inheritance, the sort of heredity that we associate with Darwinian evolution.

    The same style of copying is present in both the RNA world and modern biology – template-dependent polymerization of nucleic acids. Evolution by natural selection is a consequence of any such system (new copies can have new mutations, mutants are better, worse, or just as good at making copies as mama; those that are worse make fewer copies in the next generation, those that are better, more) – so “Darwinian” evolution was present in the RNA world too. Whether on top of that there were some forms of Lamarckian evolution is certainly possible, although its hard to imagine anything beyond the trivial (damage to an RNA sequence could affect ribozyme activity and in principle result in new copies having a mutation).

  19. Elizabeth,

    Lizzie, What is your definition of fitter?

    Isn’t that which survives and reproduces fitter, in the neo-darwinian model?

    We have been telling you all along, this is the entire problem with your concept of survival of the fittest. You all just seem to try to talk around this problem.

  20. keiths: The real transition was not from one mode of inheritance to another. Instead, the change was from the RNA World, in which the phenotype was not distinct from the genotype (since only one molecule was involved) to a post-RNA World in which phenotype became distinct.

    Let me try by way of an analogy. How does 3D printing work?

    Do you put the object you want a copy of in some sort of container and the 3D printer constructs a copy by reading off each atom of the original and reproducing it, eventually producing an object of the same shape and composition? I think not.

    In the RNA world all you have are these three dimensional objects. So how were copies of them constructed? You need a 3D copier, not a 3D printer.

    If not, why not?

  21. phoodoo:
    Elizabeth,

    Lizzie, What is your definition of fitter?

    The standard one: better able to produce viable offspring in the current environment.

    Isn’t that which survives and reproduces fitter, in the neo-darwinian model?

    Yes. It’s the entire principle – the better a variant reproduces the more copies of that variant there will be in the next generation.

    We have been telling you all along, this is the entire problem with your concept of survival of the fittest.You all just seem to try to talk around this problem.

    Sorry, what problem are you referring to?

  22. Mung: Let me try by way of an analogy. How does 3D printing work?

    Do you put the object you want a copy of in some sort of container and the 3D printer constructs a copy by reading off each atom of the original and reproducing it, eventually producing an object of the same shape and composition? I think not.

    In the RNA world all you have are these three dimensional objects. So how were copies of them constructed? You need a 3D copier, not a 3D printer.

    If not, why not?

    Well, we don’t have a good model yet. But Szostak’s proposal is that membrane vesicles containing self-replicating polymers evolved in tandem. Vesicle reproduction isn’t difficult, and it can be done without the contents spilling. So that leaves the self-replicating polymer part.

    A polymer with a double strand (like RNA or DNA) can split, and then the two halves attract monomers to complete them. So you then end up with two double strands, with the same sequence. That’s the principle. Under some conditions, osmosis will tend to expand the vesicle, until it splits – two vesicles each containing copies of the polymer.

    I’m paraphrasing of course, but that’s the working basis of the Szostak model, as I understand it.

    There’s more information here. I don’t see that it’s especially Lamarckian, but it could be, given that at that degree of simplicity, the genotype IS the phenotype, essentially. Perhaps that was your point?

  23. Mung:

    In the RNA world all you have are these three dimensional objects. So how were copies of them constructed? You need a 3D copier, not a 3D printer.

    RNA folds, Mung. If you duplicate the sequence, the 3D structure comes for free.

  24. Spiegelman introduced RNA from a simple bacteriophage Qβ (Qβ) into a solution which contained Qβ’s RNA replication enzyme, some free nucleotides, and some salts. In this environment, the RNA started to replicate. After a while, Spiegelman took some RNA and moved it to another tube with fresh solution. This process was repeated.
    Shorter RNA chains were able to replicate faster, so the RNA became shorter and shorter as selection favored speed. After 74 generations, the original strand with 4,500 nucleotide bases ended up as a dwarf genome with only 218 bases. Such a short RNA had been able to replicate very quickly in these unnatural circumstances.
    In 1997, Eigen and Oehlenschlager showed that the Spiegelman monster eventually becomes even shorter, containing only 48 or 54 nucleotides, which are simply the binding sites for the reproducing enzyme RNA replicase.

    https://en.wikipedia.org/wiki/Spiegelman%27s_Monster

    RNA replication mediated by a single enzyme. No 3D printer required.

    Before the creationists jump on this, it is not a likely candidate for early life.

    But it does demonstrate that RNA can self-replicate and evolve.

    Any more questions?

  25. keiths: RNA folds, Mung. If you duplicate the sequence, the 3D structure comes for free.

    Yes, keiths, I know that RNA folds.

    How does it get copied when it’s folded? That’s the question. What sequence are you talking about? It’s a 3D molecule.

  26. Mung: How does it get copied when it’s folded? That’s the question. What sequence are you talking about? It’s a 3D molecule.

    https://profiles.nlm.nih.gov/ps/retrieve/Narrative/PX/p-nid/197

    In 1961, Spiegelman became intrigued by the “strange biological situation” of a recently discovered bacterial virus (phage) called MS2. This phage had no DNA–its genetic material was RNA. How, then, did it complete its life cycle in a cell dominated by DNA? Spiegelman reasoned that the RNA strand injected into its host (in this case, E. coli) must serve directly as its own “translator” for directing its replication.

  27. Mung: How does it get copied when it’s folded? That’s the question. What sequence are you talking about? It’s a 3D molecule.

    It’s a folded 2D sequence.

    Have a look at those animations on the Szostak website. The one at the bottom.

  28. petrushaka, I don’t see how this helps you:

    A few years later, however, Spiegelman’s teams demonstrated that the RNA polymerase coded for by a second RNA phage–Q-beta–was specific to Q-beta RNA and would replicate no other.

    How is this a “self-replicating” molecule?

  29. Elizabeth, from your link:

    Scientists hypothesize that a ribozyme that was capable of making copies of other RNAs, called a replicase, evolved very early in life’s history.

    You have to get from something that is copying itself, to something that is a copy maker.

    When RNA folds on itself it doesn’t make a 2D shape.

  30. Mung: You have to get from something that is copying itself, to something that is a copy maker.

    Hmm. Thing that copies itself to thing that makes copies. I can’t imagine any plausible pathway there. I mean, where’s the copying mechanism to come from for a copy maker if you only have a thing that copies itself?

  31. Mung:

    What sequence are you talking about? It’s a 3D molecule.

    The nucleotide sequence, Mung. RNA molecules consist of sequences of nucleotides just as protein molecules consist of sequences of amino acids.

  32. Mung:
    Elizabeth, from your link:

    You have to get from something that is copying itself, to something that is a copy maker.

    When RNA folds on itself it doesn’t make a 2D shape.

    Not sure of your point. If you have a “1D” strand – a polymer roughly straightened out, it can fold to form a 3D object, and will do if it was a double strand that has split so has a lot of chemically active sites along its length.

    Like you can make a 3D tangle out of 1D wool.

  33. It’s sobering to view the lack of understanding of elementary molecular biology displayed here and at other sites by sectarian critics of evolution.

    Too little education is a foolish thing.

  34. Mung,

    Do you see your mistake?

    Just as protein transcription doesn’t involve the 3D printing of folded proteins, RNA replication doesn’t involve the 3D copying of folded RNA molecules.

    The organism/replicator produces the correct linear sequence of amino acids in the case of proteins and of nucleotides in the case of RNA. Physics takes care of the folding.

  35. No, keiths, I do not see my mistake. But at least now you seem to be getting a grasp on the issue.

    You understand, I take it, that a 3D printer works by translating a one-dimensional sequence of symbols that represents the desired shape that should be constructed. Any debate there? In the modern cell something analogous happens.

    In the RNA world where did the 3D printers come from how did the 3D RNA molecule being copied become a one-dimensional sequence of symbols representing the desired 3D shape?

    RNA, as you know, folds on itself. That’s how it get’s it ability to catalyze a reaction. How do you propose to convert this 3D structure into a linear sequence of symbols, and after you’ve done that, where is your 3D printer?

    Perhaps you could tell us what’s missing from the RNA World site. How do they get to their magical 3D printer?

  36. keiths: The nucleotide sequence, Mung. RNA molecules consist of sequences of nucleotides just as protein molecules consist of sequences of amino acids.

    Sequence is a mathematical term. You’re confusing the map with the territory.

    https://en.wikipedia.org/wiki/Sequence

    But this mistake you make is not uncommon, take the genetic code, for example.

    If nucleotide sequences are real, and amino acid sequences are real, why isn’t the genetic code a “real code”?

  37. david: The same style of copying is present in both the RNA world and modern biology – template-dependent polymerization of nucleic acids.

    Only by fiat.

    In modern biology we’re getting a reasonable grasp on how this is happening. In the RNA world we don’t have a clue.

    The typical story goes something like this:

    There was a self-replicating molecule.
    Replication wasn’t perfect.
    Therefore there was Darwinian Selection.
    And that explains everything.
    Therefore Life

    Was that self-replicating molecule RNA?

    Let’s ignore, for now the hurdles involved in forming RNA nucleotides in the first place.

    Let’s ignore, for now, the hurdles involved in forming an RNA polymer.

    Let’s say that you actually have an RNA polymer. What is it’s shape? Is it so short that it can’t fold? How then can it possibly “self-replicate”?

    And even if it does “self-replicate,” then what?

    How do we get from that mechanism of inheritance to what we see today, which is something utterly other. And why?

  38. Mung,

    1. You don’t need a “3D printer” for proteins and RNA molecules that spontaneously fold.

    2. “Sequence” is both a mathematical and a biological term. LMGTFY

    3. Your knowledge of molecular biology is abysmal.

  39. Mung: Let’s say that you actually have an RNA polymer. What is it’s shape? Is it so short that it can’t fold? How then can it possibly “self-replicate”?

    And even if it does “self-replicate,” then what?

    How do we get from that mechanism of inheritance to what we see today, which is something utterly other. And why?

    The RNA molecule doesn’t have to “fold” to self-replicate.

    Do watch the Szostak animation – it addresses precisely this question, but, in short, the proposal is:

    A single RNA strand can form a double strand without an enzyme in an environment rich in its constituent nucleotides, because each nucleotide on the strand will tend to bind to its matching pair, forming a double strand, where the second strand has the same sequence information as the first (only with the pairs reversed).

    If the temperature rises far enough, the double strand will split into two single strands.

    Although both strands will have the same sequence information (one will “code for” the other), they won’t have the same electrochemical properties.

    One may, in fact, have properties that cause it to fold, while the other does not.

    If the other does not, it will, when the temperature cools again, bind with a new set of nucleotides to form a double strand again, a copy of the original double strand.

    Rinse and repeat (for instance, in a convection current).

    Now, if the sequence is such that the folded version of the single strand forms an enzyme, and the enzyme improves the chances of the other strand forming a double strand again (and one can envisage circumstances in which this might be the case, for instance, if the folded molecule helped its ex-partner to find New Love i.e. form a double strand again), then that sequence would be more often replicated than sequences that didn’t do this.

    And so you would have the adaptive evolution of an enzyme that helped RNA to replicate itself, including the replication of the RNA sequence that codes for the enzyme.

    I’m not saying it’s right – and nor are Szostak et al – but it’s perfectly in accord with the properties of the molecules in question.

  40. Mung: Sequence is a mathematical term. You’re confusing the map with the territory.

    In this case, no, I don’t think so. A “sequence” of nucleic acid is a long molecule in which the bases occur in a “sequence”, just as you could have a “sequence” of beads on a string. That sequence could be written down and the written version would be the “map” of the “territory” that is the string of beads.

    And if, instead of a string of beads, you had a string of, say, magnets of different strengths and orientations, then the string would tend to “fold” to form a specific 3D object.

    Mung: If nucleotide sequences are real, and amino acid sequences are real, why isn’t the genetic code a “real code”?

    And you could, depending on how you wanted to define “code”, then say that the sequence of magnets-on-a-string “coded for” the 3D object that they fold into when you let go the ends. A different sequence will give you a different object.

  41. Here’s the story in the words of the Szostak lab site:

    PREBIOTIC RNA REPLICATION

    Even in the absence of enzymatic catalysts, single-stranded RNAs may have been able to copy strands of RNA through template-directed polymerization. This process is shown in the animation on the left, and is based on experiments performed in Jack Szostak’s Lab (MGH/Harvard) using chemically activated nucleotides.

    This process of non-enzymatic replication, however, is likely to have been slow and error-prone. Eventually, this mechanism of RNA replication is likely to have been replaced by a more reliable catalyst, such as a ribozyme. Scientists hypothesize that a ribozyme that was capable of making copies of other RNAs, called a replicase, evolved very early in life’s history.

    The animation on the lower left shows a theoretical replicase copying a template strand of RNA. While the structure of the replicase shown in the animation is based on an existing ribozyme that is capable of carrying out the basic steps of a replication reaction, a true replicase that is capable of copying an RNA copy of itself has not yet been isolated in a laboratory. Recently, however, the Hollinger group (MRC, UK) discovered an ice-water stabilized RNA polymerase ribozyme that is capable of copying strands of RNA that were over 200 basepairs long – longer than the ribozyme itself – suggesting that a self-replicating RNA is indeed possible.

    Under the proper temperature and salt conditions, double-stranded RNA can undergo strand separation. Since the two strands are complements of each other (and not exact duplicates), only one of the two strands will be able to refold into an active replicase. The other strand can act as a template for further rounds of replication to create more replicases.

    I do suggest you check it out, Mung, not to be convinced, but at least to see what the argument is.

    The animations are very clear.

Leave a Reply