Nelson: Developmental Systems Drift

Hi Paul, just a reminder that I don’t need–or want–to see a “monster slide deck” from some lecture. I’m inviting you to discuss DSD, especially since you named a prominent and accomplished scientist while making claims about evidence in the area. If you really don’t have anything other than a “monster slide deck,” then there isn’t going to be a discussion. Which is fine, but IMO you should be forthcoming about your level of engagement with the topic.

1 Like

Steve,

I am hoping Discovery can give me a URL today for the “monster deck” pdf, because it’s relevant as context for any discussion of DSD. It is unreasonable to keep you waiting, however, so I’ll just jump in and talk about my discussions with Joanna Masel in April 2019, and about DSD generally. When I have the URL, I’ll post it here.

Background: Change Tan at U Missouri and I presented a poster at the Evolution Evolving meeting (Churchill College), on “The Engima of Taxonomically Restricted Essential Genes of Unknown Function.” This was our abstract:

The advent of automated DNA sequencing in the mid-1990s enabled the development shortly thereafter of transposon mutagenesis screens of entire genomes, to determine what genes (and their protein or functional RNA products) were essential for cell viability under laboratory conditions. Two decades later, these experiments, as well as systematic or targeted gene deletion experiments, have consistently returned an unexpected result: many essential genes are taxonomically restricted in their distribution, not universally, or even widely, shared. Moreover, when first annotated, these genes are most often classified as “unknown function.” Evident already in the initial transposon mutagenesis screens of Mycoplasma genitalium (Hutchison et al. 1999), where approximately one third of the essential genes were listed as “unknown function,” the same “unique, essential, and unknown function” (UEU) signal has been found in all three domains of life. We have begun a project to collect and characterize these UEU sequences, to assess their impact on our understanding of theories of cell function and evolution.

You know how poster sessions work. I’m standing there with my glass of wine in hand, and Masel comes up to chat. She’s an absolute gem: whip-smart, funny, and a leading theoretician on cellular processes which may produce orphans. I’ve read everything she has published on that topic, and one of her ideas, related to stop codon read-through, is potentially revolutionary. She didn’t want to talk about any of that, however.

The property of “essential,” she said, shouldn’t be in your comparative analysis. Dump it. On a long enough time scale, nothing in biology is truly essential. Evolution churns. Developmental pathways in animals, for instance, change all the time. She then mentioned her Arizona colleague Alexander Badyaev, who she noted collects examples of developmental systems drift [DSD], which are ubiquitous. Badyaev is also a theoretician:

OK, so I’m old (really). I remember when DSD was coined as a term, by True and Haag in 2001. The pattern itself, however – conserved (homologous) phenotype, divergent (non-homologous) developmental pathway – was well known to E.B. Wilson and other embryologists in 1895 (107-8; emphasis added):

It is a familiar fact that parts which agree closely in the adult, and are undoubtedly homologous, often differ widely in larval or embryonic origin either in mode of formation or in position, or in both. Innumerable cases will suggest themselves to any embryologist of hollow organs that arise either by invagination or delamination; of paired organs that arise from either single or paired foundations, and vice versa. No one is disposed to question the homology of the spinal cord of a teleost and that of a shark on the ground that one arises as a solid cord, the other as an infolded tube.

Well, I would be disposed to question their homology: those divergent pathways may be evidence of phylogenetic discontinuity. Why is that not the case for Wilson? You know the answer. Which brings us to the core of the issue.

Why would Wilson say, ”No one is disposed to question?” Common descent, of course. DSD emerges as a phenomenon, or, really, a name for a mystery (see below), against the received backdrop of common ancestry: sharks and teleosts stem from the same unknown vertebrate ancestor. That’s a given.

When one lives, however, in two worlds at the same time – in my case, evolutionary theory as world A, crazy ID theorist with YEC upbringing as world B – one can see how assumptions such as common ancestry play a regulative role in theorizing, a role which may be invisible to those living entirely in world A, whose reasoning rests on those assumptions. The same is the case, incidentally, in molecular phylogenetics, with concepts such as long-branch attraction. Another story for another time.

Let’s not be naïve. DSD is a name for a mystery. You mentioned nematodes. If all one knew about nematode development was the canonical cell lineage of C. elegans, would anyone believe me if I said that one of its phyletic kin, the freshwater nematode Tobrilus diversipapillatus, built itself completely differently?

When you mentioned nematodes higher up in this thread, I had to smile, because ever since I fell in love with C. elegans, many years ago, I’ve watched to see how other worms do things differently. DSD everywhere.

Why a name for a mystery? No experimental data showing the successful (meaning viable and stably heritable) perturbation of early development via undirected mutation. If you have some, I’d like to see it (I’m deadly serious: I need to struggle every day against my own confirmation bias). Before he died (October 2010), Van Valen and I talked about this a lot. Leigh knew the evo-devo literature better than anyone, and I would ask him if he’d run across any examples of the heritable modification of cleavage patterns, via mutation, in any metazoan group. He would shrug – the only example he knew was change in the direction of coiling (levo-dextro) in some gastropods. Certainly nothing as dramatic as what we see in C. elegans and its freshwater cousin Tobrillus.

When you’re crazy, like me, you notice how the normal people think. You see what they take as a given. The common ancestry of the Metazoa is a given. Against that backdrop, DSD “explains” why homologous adult anatomies in animals can arise by a wild array of non-homologous developmental pathways.

But of course DSD doesn’t explain anything at all. Mutate nematodes (or flies, or frogs, or mice) and watch what happens. DSD is a name for a mystery.

I apologize again for stringing you along for several days, while I waited for my precious little URL, which I still don’t have (and all the bad about that is on me, not the Discovery staff in Seattle).

E.B. Wilson. 1895. The embryological criterion of homology. Biological Lectures Marine Biological Laboratory. 1894 (Sixth Lecture):101-164.

REQUEST TO MODERATOR: please change the name of this thread to something relevant to its actual content.

Here’s the link to the body plan evolution slide deck:

This is material presented over the past few years at the Discovery Institute natural sciences summer seminar.

I have split out a new topic, with a hopefully more appropriate title (@pnelson, let me know if it should change) and since this looks like a good scholarly exchange I am “promoting” it to the Conversations category, which means @moderators will expect professional conversation and substantive posts. Posts may have some delay as they require moderator approval (as long as everybody “plays nice” it should be quick, but rewrites of trivial or unprofessional content may be requested).

2 Likes

I think it’s very important to note that it’s only in very recent years that we’ve had the technology to start experimentally manipulating gene regulatory networks, so it’s not surprising that experimental evidence of the evolvability of early developmental systems is lacking, and it’s far too early to conclude that these systems aren’t labile.
Common descent is certainly sufficiently well-supported to be included in hypothesis generation on this subject.

2 Likes

Hi Paul, thanks for an engaging response! So, I looked at the “monster slide deck” just now, waded through several slides of uninteresting quotes from people, then arrived at a slide about “Three Pillars of Neo-Darwinian Theory,” in which the third pillar is abiogenesis. That’s inexcusable and you should disclaim that kind of falsehood. I do not expect or want a response to that, but I will say that I will not read your deck any further. It’s not worthy of scholarly consideration.

But fortunately, we have some things that we might be able to discuss below, and even if that “monster slide deck” was any good, it would never substitute for discussion or for citation of credible and peer-reviewed scholarship.

Now. My first reaction to your response is somewhat negative, and it’s: your response is mostly about you. You cite your abstract, you talk about the meeting you went to, you muse about yourself and your commitments. That’s not an accusation, just an observation that explains why there’s relatively little (but not nothing) for us to continue to discuss. I note that in this and other contributions to PS conversation, you like to talk about what others said or wrote, sometimes in previous centuries/millenia, sometimes in uninteresting and irrelevant transcripts of irrelevant conversations. I think this might mean that you love forensic analysis of conversations and the evolution of ideas, and that’s all good, but it can look like a red herring chowder sometimes. FWIW.

But I want to emphasize something positive now, and not just because I am such a nice guy. :smiling_imp: DSD is very, very interesting because it is right in the center of a topic that I think both of us really care about, which is design. Because of that and our mutual fondness for interesting ideas (shared by lots of others reading this), I think it would be great fun to share a beverage with no time pressure and to explore DSD, evolutionary innovation, “essentiality” of genes, and what all of this has to do with design, which is a lot. I’m serious when I say this would be a really good time, and I would say that about exactly one person who works at/for the DI.

You wrote this about your conversation with Prof Masel:

She’s right. And when we have that beverage and convo, we can explore one implication of this, which I allude to above: in many and perhaps most cases, the genes or pathways or signals that underlie a developmental or physiological process are themselves unimportant. Why should neurogenesis be about Notch? Why should Ras be a gateway to MAPK signaling and why should that be so influential in cell division? In some cases, to be sure, we’ll discover a finely-tuned enzyme that controls a biochemical reaction of bedrock metabolic importance. But Notch? Hox proteins? Any transcription factor for that matter? This is where I think the conversation about DSD should go, because the answers are unavoidably about design. About outcomes. Don’t you want to talk about that? I know you do. But instead, we go here:

You are right that common ancestry is a “received backdrop” but wrong to label it as an “assumption.” Common ancestry is settled science, and its alternative is magical insertion of physical entities into history. To ask about design is to enter serious adult conversation; to paint common ancestry as an assumption or as a weakness in thought is to leave it. We’d barely have time to order our beers if that’s what we talked about.

Over that beer, I would beg you to stop talking like this. DSD is a high-level, long-term, dare I say "macro"evolutionary process and concept. The experimental data that would support it would not look like this, and to ask for that kind of data is, well, unprofessional. I say that out of respect for you, because I am suggesting that you should know better.

Here is what DSD implies: that phenotypes, including body plans and large-scale characteristics, can be held somewhat constant while the genetic underpinnings vary. If you agree with me on this, then you should be able to immediately see what kind of experimental evidence we would want, and what kind we could not possibly seek or expect. The literature I have already cited, which you haven’t engaged yet, is that kind of evidence. More notably, myriad aspects of evo-devo give us the tools to picture how DSD can happen.

I am not saying that DSD is not immensely interesting and perhaps even some kind of “mystery” if by that you mean “something we still don’t understand.” (All of the biologists I know, BTW, use that term much more sparingly and circumspectly than you seem to.) I am saying most importantly that it is badly wrong to suggest that because DSD has not been recapitulated in a lab that this makes it “mysterious” and it is even worse to suggest that DSD is outside of any explanatory reach of evolutionary theory. Neither of those proposals is remotely credible.

So, in summary:

  1. DSD is very interesting, and despite our disagreement about some fundamental questions about reality, we could (and should) have an enjoyable and profitable discussion of what it means for design in the biosphere and even in the universe.
  2. Referring to abiogenesis as a “pillar” of “Neo-Darwinian Theory” is the kind of apologetic BS that is grounds for disqualification from scholarly discourse. Please work on that as soon as you can.
  3. Expecting lab experiments that recapitulate macroevolutionary phenomena is almost as irresponsible as the previous point. It’s not worth discussing, and if it persists as an emphasis, it is disqualifying.
  4. Design is cool and DSD belongs in any discussion of the concept. So call me maybe.
7 Likes

Something vaguely similar to DSD has happened on the morphological level. The original gnathostome mandible was a single endochondral bone, the mandibular. In living species, this remains only in persistently cartilaginous form in sharks. A variety of dermal bones, likely originating as modified scales, evolved as extra jaw supports, surrounding the mandibular: dentary, splenial, angular, etc. In some modern species the mandibular remains in the middle of the jaw. In others only one end remains, the articular. And in mammals, the articular is entirely separate from the jaw and has become part of the middle ear. And so have most of the other dermal bones, leaving only the dentary in the jaw.

Now, I ask you: is the lower jaw of a shark homologous to the lower jaw of a mammal? They have none of the same parts, but there is a continuous series of ancestors and descendants, all with mandibles, leading from the one-element mandibular bone jaw, as in the shark, to the one-element dentary bone jaw, as in mammals. Sounds like a very similar situation to DSD.

1 Like

I agree it’s similar, in that the outcome is apparently constrained but the process to get there is not (or is convoluted). But I think DSD is perhaps more specifically about developmental systems–signaling, morphogenesis, etc–and not about evolutionary trajectory.

1 Like

Yes, that’s why it’s similar, not identical. And of course DSD is about evolutionary trajectory, except that it’s a Red Queen’s trajectory, in which all the running keeps you in the same place. (Or it would be if there weren’t already a different evolutionary phenomenon with that name.)

2 Likes

I guessed that you wouldn’t like that slide, and you didn’t. But I left it in the deck anyway, because it’s true. Abiogenesis via natural (undirected, physical) causes undergirds neo-Darwinian theory, because central premises of the latter – e.g., its understanding of molecular similarities – require specific models for the formation, from abiotic sources, of universally-distributed entities like the ribosome:

Most sequence or three-dimensional similarity, statistically above the trivial level of coincidence, surely reflects homology in its true sense of “having a common evolutionary origin.” The only way it would not would be for us to start believing, against the odds, in either many recurrent independent origins of similar genes and proteins or frequent freak conditions for selection to drive two relatively dissimilar (nonhomologous) sequences to the same composition. (Gabriel Dover 1987, 515)

Here (below) is a slide from Jonathan Eisen’s intro bio course at UC-Davis, which he just taught. The lecture in question concerns whether life falls into a nested single tree, or multiple trees (e.g, Common Descent, presented as Neo-Darwinian Pillar #1 in my slide deck):

In other words, theories about the Tree (or Trees) of Life, and theories about the origin(s) of life, are as intimately connected as any two questions can be.

I’ll need to reflect on the rest of your thoughtful comments, i.e., please give me some time to think about them before responding. I appreciate your feedback!

P.S. About the slide deck. Please take note that the original audience for the material (Discovery Institute natural science summer seminar) is interdisciplinary (biology students, physics students, geology, psychology, you name it). The sort of straight-into-the-tall-grass molecular detail you want would lose me most of the audience.

G. Dover. 1987. Nonhomologous views of a terminology muddle. Cell 51:515.

No it doesn’t. The genealogy of all known extant life having passed through the bottleneck of a single common ancestor says nothing about what the first life that came into existence must have been like, or how it subsequently evolved between those first stages, and into the entity we can infer to represent the last universal common ancestor.

5 Likes

No, they don’t. All that’s required is that a ribosome be present in the LUCA, which as you should know is quite far from the postulated origin of life. Now, ribosomes do present some evidence reaching before the LUCA, i.e. suggestions about the RNA world hypothesis. But even that isn’t back to the origin of life. This doesn’t even require a single origin of life, only a single origin with surviving descendants.

3 Likes

It’s not true. As a rejector of Neo-Darwinan Theory, you don’t get to decide what it comprises.

Also not true. Universally distributed entities are tracked to LUCA, but no further. No such specific models are required for evolution of extant life from LUCA. Nor is there any requirement for ribosomes to have developed from abiotic sources, as they could have developed from previous biotic sources.

3 Likes

That’s false, and it is not worthy of further consideration. As long as you present “Neo-Darwinian Theory” in that way, you will be engaging in misinformation.

1 Like

I don’t understand the argument in that slide. There are in fact a number of ways to investigate the period before LUCA. There can be ancient paralogs preceding LUCA. There can be clues, as I’ve mentioned, about preceding situations, as in the ribosome suggesting an RNA world. Nor do I see how the existence of LUCA allows for multiple origins or how inferring it could obscure the existence of real multiple origins.

Nor does it seem to connect the origin of life to phylogenetic inference. The simplest explanation for my confusion is that you’re fooling yourself. But can you elucidate?

1 Like

That nicely illustrates why your claim is wrong. You are ignoring the context of your false slide, which is “New-Darwinian Theory.”

These are examples of scholarly malpractice and they hinder discussion of actual science. You must know that the only link between abiogenesis and evolutionary biology is what many call “methodological naturalism” but what is better understood as an assumption that the biosphere can be explained. That’s the only link, and I was generous to assume that you meant “evolutionary biology” when in fact you made a gross error in referring to “Neo-Darwinian Theory.”

I think maybe I’m not being clear enough here: I have a professional obligation to oppose falsehood and misrepresentation when it comes to matters of scientific import. And I have a strong personal interest in questions of design. My professional obligation makes it a big challenge to justify continued discussion with someone willing to misrepresent basic scientific thought. I was serious when I said that slide could disqualify you from scholarly discourse.

I will have nothing more to say about that.

1 Like

Hi @pnelson, maybe you should make explicit, clear, exacting connections between the mechanisms by which LUCA arose and Developmental Systems Drift. I am having a hard time understanding how any of the issues (known and still unresolved) about abiogenesis tell us anything about DSD and the problems the latter may pose for the evolution of multicellular animals.

Please, no vague generalizations. That would just be rehashing decades of pointless posturing. Rather, I think it would be helpful to have some explanation of exactly how, for example, the properties of the first peptidyl transferase factor into mechanisms and difficulties of DSD?

Again, I beg - specifics, nuts and bolts …

4 Likes

That image refutes your claim that universally-distributed entities such as ribosomes had to develop from abiotic sources, by highlighting the difference and distance between FUCA and LUCA.

2 Likes

Hi Paul,

Trying to get be helpful here. It seems like the last sentence of the slide is overstated. The LUCA does put a constraint on the intermediate region between the FUCA and LUCA insofar as only one tree survives into the blue region. However, the LUCA is as compatible with a single divine miracle FUCA as it is with competing FUCAs of abiotic origin that engaged in a ferocious competition that allowed only one, the LUCA, to emerge.

In other words, the theory of evolution is not the same thing as an OOL hypothesis, and the TOE is not inextricably linked to any particular OOL hypothesis.

To the extent that the slide implies that evolution must include an abiotic FUCA, then, it would not be correct. Perhaps you did not mean to imply that, in which case I would suggest you edit the slide to clarify your intent.

Best,
Chris

4 Likes