Suzan Mazur: Is Darwinism still valid? An interview with scientist Stuart Newman
"I don't want to be a Steve Gould," New York Medical College cell biologist Stuart Newman told me recently when I visited him at his lab in the village of Valhalla, a short train ride north along the Hudson River from Manhattan. Newman, an elegant, engaging and somewhat enigmatic man actually went to the same New York City high school as Gould but says he doesn't like being compared. While he considers himself a public intellectual, he also enjoys having a private life and getting lost in art. But some of those precious moments gazing at the composition of a Rubens may be up for grabs because Stuart Newman's now got a seductive theory about the origin of form of all 35 or so animal phyla--"it happened abruptly" not gradually, roughly 600 million years ago via a"pattern language"--which serves as the centerpiece of the"Extended Synthesis." That's the reformulation of the Modern Synthesis or neo-Darwinian theory of evolution kicked off this summer at Konrad Lorenz Institute in Altenberg, Austria by 16 scientists I dubbed"the Altenberg 16."
The impetus for the Extended Synthesis, a graft onto, or a major departure from, the Modern Synthesis (depending on who is describing it), was the overwhelming data generated in recent years that just didn't fit the old formula. Phenomena like self-organization, epigenetics and plasticity intruded in ways that were complementary to, and sometimes contradictory to, natural selection. Then there was niche construction to consider--where organisms invent their habitats (burrows, bird nests, bee hives, etc.) rather than being selected by their fitness to pre-existing ones. And also punctuated evolution, abrupt transitions in the fossil record, and the even more puzzling episodes of stasis.
While much of the evolutionary biology community resists the notion of an evolutionary framework that begins to consider the role of determinants beyond the gene, as the Extended Synthesis does, the momentum of the new synthesis is undeniable (see Google for"the Altenberg 16"). And there are other scientists and philosophers of science--avowed non-creationists--who say the Extended Synthesis does not go far enough in relegating natural selection to a reduced role.
Another emerging view is that like the Darwinian model, which has a historical window of 150 years, the Extended Synthesis leaves out too much of world civilization's influence prior to the Victorian age--the evolutionary language continually being uncovered by archaeologists. But the most contentious issue may be that the discourse about evolution is now in the public domain because of the Internet and that evolutionary science no longer belongs exclusively to the scientific establishment.
Stuart Newman has fostered bonds between science and the people throughout his scientific life and has attempted to make his research findings publicly available. He's championed causes such as nuclear disarmament and awareness of the dangers of human genetic manipulation and bioengineering, addressed audiences about the perils of the commercialization and industrialization of organisms--which he fears will ultimately include humans--and has testified before the Senate.
Newsweek first profiled Stuart Newman's work on the embryo for an international audience a quarter century ago in a cover story called"How Life Begins". I've been writing about his theory of form for several months, but his ideas have been discussed as the way forward over the past few years in books and articles in the philosophy of biology, e.g., J.S. Robert,"Embryology, epigenesis, and evolution: taking development seriously" (Cambridge University Press, 2004) and R.G.B. Reid,"Biological emergences: evolution by natural experiment" (MIT Press, 2007). Innovative younger scientists like Alex Badyaev, Armin Moczek and Isaac Salazar-Ciudad have incorporated Newman's thinking in their own work.
Stuart Newman received his A.B. from Columbia University and his Ph.D. in chemical physics from the University of Chicago. He is co-author with Gabor Forgacs of Biological Physics of the Developing Embryo (Cambridge University Press), and with Gerd Müller co-edited Origination of Organismal Form: Beyond the Gene in Development and Evolutionary Biology (MIT Press).
My interview with Stuart Newman follows.
You and your colleagues initiated a reformulation of the Modern Synthesis or neo-Darwinian theory of evolution at Altenberg's Konrad Lorenz Institute this summer calling it the"Extended Synthesis". The Modern Synthesis has lacked a theory for form and your hypothesis about origin of form now serves as the centerpiece of the Extended Synthesis.
In an April interview Ramray Bhat told me you were trying to"tease out" a pattern language--"dynamical patterning modules," or DPMs--you say was in play 600 million years ago exploring body plans of all of the 35 or so animal phyla and that is still in play, though to a limited extent. Without disclosing too many lab secrets, can you say how you plan to knock out the genes that stabilize development to reveal these DPMs?
Knocking out genes is one experimental strategy we use on cells we have isolated from embryos. We're particularly interested in how the limb develops, because the appendages of vertebrate organisms represent innovations of form that emerge after the origin of the body plans, about 400 million years ago, using what we believe to be the same DPM principles.
If you remove cells from the chicken embryo limb and put them into a dish, they will self-organize to form nodules and stripes of cartilage. And cartilage nodules and stripes are what form the vertebrate limb in the embryo itself. In other words, the limb cells in the dish self-organize into the type of patterns but not the exact patterns you see in a limb.
The cells communicate with each other in the dish--they have freer play there because they're not confined by the thin layer of skin that surrounds them in the embryo. They're not restricted to the geometry of the embryo. The spatial scale and the time course of the patterns they form are very similar to what happens in the limb.
We introduce morphogens--diffusible signal molecules--and adhesive proteins (both molecular components of our DPMs) or we knock these molecules down. We do this by introducing drugs or nucleic acids that prevent the proteins from being made and then look at the changes in the patterns. You can probe the dynamics of pattern formation in this way. And you can then go back and investigate similar problems with similar reagents in the embryo itself.
We're getting a very good idea of how this pattern forms in the limb by going back and forth from the embryo to the tissue culture dish as well as by using mathematical models that simulate these dynamical processes.
You've been able to knock out all of the genes?
We've been working for almost 30 years on the limb. It's not possible to knock out all the genes because the cells would not survive this. We selectively go after molecules we believe to be the key factors that were present when limbs first emerged in vertebrate organisms.
We know that no new genes arose in conjunction with the vertebrate limb. Pre-existing genes were used in a new context. The limb bud is an outpocketing of the body wall. It came into existence as a result of slight changes in the surface of the embryo. This provided (by the action of some of the DPMs) a new"morphogenetic field," the limb bud, which served as the medium for a whole new set of self-organizing dynamics.
The time between establishing the limb bud and establishing the limb pattern was probably not very great. The second followed from the first fairly readily because of the self-organization that was mobilized when the new morphogenetic field appeared.
Continuous gradual changes in growth and shape can happen by the standard gradualist neo-Darwinian modes. But once the tissue reaches a critical size and shape, the genes that were already there start to interact with each other and the local physical environment in new ways because they're in a new context.
This happened in the limb bud 375 million or 400 million years ago and it happened in the animal body 200 million years before that. Something changed in the environment of the single-celled ancestors of modern animals or perhaps a minor genetic change occurred in several different cell populations. The result was clusters of cells where before you had only single cells. And we know that the genes that cause cells to cluster were present way before any cell clusters existed. [emphasis added]
In a recent Nature magazine story, University of Uppsala palaeobiologist Graham Budd was quoted saying he didn't think there was any evidence for plasticity and spontaneous emergence of body plans 600 million years ago as you and your co-author Ramray Bhat propose in your Physical Biology paper. Would you sort of encapsulate your findings to address Graham Budd's comment?
In modern-day organisms there is significant plasticity. You don't have to go back 600 million years to see that. If you take a plant and put it in different soil or a different environment, it can look entirely different.
So we don't need fossil evidence.
The fossil evidence tells us there are different kinds of organisms. The fossil evidence doesn't tell us how to get from one organism to another.
The person who made this criticism doesn't appreciate the plasticity and nonlinear nature of embryonic mechanisms. He is making a claim that the plasticity and sensitivity to the physical environment of developmental mechanisms that we see in modern organisms didn't exist back then [600 million years ago] when genetic programs were much less consolidated, which to me does not make sense at all.
Paleoanthropologist Mary Leakey once lamented to me at Olduvai Gorge about gaps in the fossil record. Paleontologists Niles Eldredge and Steve Gould in their punctuated equilibrium concept talk about gaps of millions of years in the fossil record, arguing that evolutionary jumps happened. The Extended Synthesis says those jumps are sometimes hundreds of millions of years and you've coined the term punctuated evolution to cover this. Extinction accounts for part of these gaps. What else does?
The dynamics of tissues, cells and the molecules they produce are capable of making forms that are very different from one another from the very same set of ingredients. So the question of gaps in the fossil record is not simply a matter of time. It's a matter of an explanatory model that recognizes that a morphological phenotype is not a direct read-out of a genotype.
You see a form and maybe a half million or a million years later you see an extremely different form. And you ask the question--how are these forms related to each other? They seem to be part of the same overall group. Maybe the same phylum or the same order or the same genus, but they look very different from each another.
And if you are a neo-Darwinian, you search for all of the gradations between those two very different structures, since Darwinism is a gradualist model. But what we're saying and what we can show using many examples from modern organisms and many computer simulations of developmental processes, is that you don't need to have continuous gradations between very different forms.
Slight differences in rates or slight differences in interaction parameters between molecules and cells can change an organism that is a continuous unsegmented form to an organism that's segmented. Or take you from a form which is a solid mass of cells to a form that's a hollow group of cells. Or from a form that has one layer to a form that has two or three layers. You don't need all the intermediates because very slight differences and the non-linear interactions among cells and molecules will cause this to happen.
You also say jumps can happen on a smaller scale as in the formation of human vertebrae. Humans belong to one of the 35 or so animal phyla--the chordates, characterized by a backbone - that arose more than 500 million years ago. Would you describe how the backbone forms?
The segmented backbone of a vertebrate organism forms by utilizing a capacity of cells to act like clocks. Cells have molecules in them that get made and get broken down. For some molecules this happens with a fixed period, providing the cell with one or more internal clocks. If a group of cells are near each other, some of their clocks tend to synchronize because the cells interact with each other. And if you have synchrony, then all of a sudden all of the cells march around the clock in the same phase. The synchronization of cellular clocks is one of our DPMs.
If something happens at 9 o'clock, for example, and the cells become more adhesive to each other and less adhesive to their neighbors, then a group of cells will pinch off and become a separate block of tissue. This happens around the clock again and again as the embryo grows.
There's a clock and there's a gradient. A gradient is a source of a morphogen, a diffusible product of the cells. Morphogen gradients constitute another of our DPMs. The interaction between the clock and the gradient will generate segments.
How many segments you get before the whole thing stops depends on at what point the clock fades out, or at what point the gradient becomes too weak, or at what point growth stops. (The increase in tissue mass by addition of cells is another DPM.)
Thus there are several DPMs that contribute to how many segments you get. There are snakes that have more than 300 vertebrae and humans generally have 33 (the number of segments, or"somites" that initially form is somewhat larger). Snake and human segmentation occur by the same mechanism, the difference is how long it proceeds.
Genes don't tell the molecules to oscillate--genes specify the molecules the cells can produce and the molecules interact with each other and with the cells themselves. The non-linear actions and reactions among the molecules and cells, their production and their breakdown, cause them to oscillate with time. The genes are necessary components of these networks of interactions, but to say that it's the genes that make it all run is incorrect. You also can't have it happen without atoms but to say atoms"program" development is an absurdity--what philosophers call a" category mistake." What generate forms are the networks of molecular-cellular-physical interactions that occur within tissues, which we have decomposed into DPMs.
I think your view is that the mechanisms of evolution have themselves evolved and you therefore reject Darwinian uniformitarianism. What are these other mechanisms you refer to?
Darwin took his cue from some of his contemporaries, geologists named Hutton and Lyell. They were his teachers and inspiration. They said that if you look at rocks and very complicated formations and ask how they occur, rather than occurring all at once and just being put there by some creative force in a single swoop, they were built up over time gradually by processes that we see occurring to this very day. Chemical crystallization processes, pressure and other physical processes mediating the movement of rocks against each other, the crushing of rocks, etc. They said geological formations are generated in a uniform fashion.
In other word, the forces are no different now than they were back then, but the forms, being outcomes of the action of the forces over very long times, are quite different. It all happened fairly gradually and uniformly and by the same mechanisms.
One of the insights for which Darwin is most celebrated is the idea that a modern organism that's very complicated didn't have to start complicated, it could have started simply. Moreover, the small scale evolutionary"forces" we see acting today--small gene-associated variations and selection among those variations--can build up incredibly complex things if you just give them enough time.
That represented progress in its own time--to look at very complex things and try to understand them on the basis of the simple effects that we know and understand. But Darwin wasn't entirely correct about this because organisms we encounter in today's world use very intricate, highly integrated genetic mechanisms to produce their forms, the so-called genetic programs for development. Such programs could not have existed when the simplest multicellular organisms first arose. But because the transition between single-cell ancestors and highly diverse multicellular forms happened too abruptly to have been accomplished by familiar incremental mechanisms, uniformitarianism does not seem to be a satisfactory framework for the earlier phase of animal evolution.
Genes work by specifying proteins that affect cell behavior and the activity of other genes. Very complex forms are built up by what looks like very complex programs. Part of these developmental programs is the mobilization of the processes of the physics of materials. The genes mobilize and restrict these physical forces to operate in specific, focused ways, and this is what constitutes developmental programs.
Developmental programs arose over time.
But did they arise gradually?
Most likely they did not. Individual cells are very susceptible to the physical environment and when they are aggregated, as in ancient cell clusters, the physical environment changes, with new forces coming into play. Synchronization, adhesion, chemical gradients and biochemical feedback lead them to organize into variable but characteristic patterns, just like the cells in our dishes.
The theory that we've come with from all this is that in the earlier stages of the development of multicellular organisms, such as the animals, physical forces were more prominent as causal mechanisms than they are in modern organisms. You can actually predict the kinds of physical forces clusters of cells susceptible to and calculate that those physical forces are sufficient to cause some of the clusters to be hollow, multilayered and segmented. Some of them will be elongated and some will have appendages.
It has always been a collaboration between cells (including their genes) and physics. Today the purely physical aspect is less prominent. But back then you could start the whole process by employing genes that existed. . .
You're talking about when?
About 600 million years ago. You could start the whole process by using genes that had evolved for individual cell function, not for the purpose of making complex three dimensional forms.
We don't know exactly what all of these molecules do in individual cells. But it's not controversial that these molecules, referred to as the"developmental genetic toolkit"--existed before there were programs of development.
What made these genes do new things? The context and scale of the multicellular state, allowed the genes that had evolved for other functions to play new"approximate" functions--that is, not yet programmed functions--in these clusters of cells. Doing so they generated a wide range of forms that natural selection could then pick and choose among depending on how suitable these forms were for given ecological niches.
Once the form had appeared?
In this picture you can see there's no real gradualism. Things get formed because of the interaction of gene products and cells and physical mechanisms. You get a whole variety of different forms. Some of the forms survive and some of them don't. But there are no intermediates between one form and another form. It didn't happen gradually. It happened abruptly.And a common ancestor?
Two ancient organismal forms that were hollow, for example, did not need to have to have a common ancestor that was hollow. They could have developed their interior body cavities completely independently by the physical consequences of localization of similar (or even different) proteins products on the surfaces of cells.
Consider segmented organisms. Worms are segmented. Our backbones are segmented. Insects are segmented. Some scientists at one point wanted to say that if you have segmentation, all of the segmented forms had to have a common ancestor. But, in fact, it's generally acknowledged now that animal segmentation probably arose independently in several different lineages, maybe half a dozen times. If that's the case, then you can't trace segmentation as a morphological motif back to a common ancestor.
But what threw people for a loop is that organisms in lineages that didn't seem to have a common ancestor nonetheless use evolutionarily very related ("homologous") genes to generate those functionally similar ("analogous") motifs. Flies, for example, have legs, and so do humans. Paleontology tells us that the worm-like common ancestor of the fly and the human did not have legs. So arthropod and tetrapod legs are analogous, rather than homologous, structures.
But if you look at the genes used to make a fly's leg and those used to make a human leg, many of the same genes are used--which is a big puzzle.
The genes are homologous, but the structures are not. They're analogous. I call this the"molecular homology-analogy paradox." The only way to understand that is to think that the very same genes have been mobilized in completely different lineages to make structures similar to each other even though those structures can't be traced back to a common ancestor.
The resolution of this apparent paradox can be found in the concept of DPMs (dynamical patterning modules), in which there are molecules that are predisposed to mobilizing certain processes of the physical world. When this mobilization occurs in a tissue mass structures get made. For example, if a certain kind of protein on the surface of a cell - a" cadherin"--tends to get sticky under certain microenvironmental conditions, and when this happens cadherin-bearing cells will clump together.
This is the basis of the most fundamental DPM, cell adhesion, which is the sine-qua-non of multicellularity. The single-celled ancestors of the modern animals are known to have had cadherins on their surfaces, even though those cells didn't use the cadherins to stick to each other.
Because the function of a cadherin as a sticky protein is environmentally dependent, you can imagine that somewhere in Australia and somewhere in North America the environment might have changed and these single-celled organisms turned into clumps rather than single cells. But this obviously does not imply that clumpiness was a character that appeared in a common ancestor from which all multicellular forms descended. Rather, all cells that had this protein on their surface had the tendency, upon slight changes in the environment, to mobilize the physical force of adhesion. Without having a common clumpy ancestor, you got individual clumped forms at very distant places. And so on with other things like the genes whose products are involved in forming appendages--the limb of a human and the limb of a fly.
There's a propensity of some of those gene products to cause a sheet of cells they are associated with to bend and protrude in certain ways. So when the context changes in a way that favors that, you'll get an appendage forming and you'll have two different evolutionary lineages--insects and vertebrates--that have appendages and use the same genes for appendages, but had a common ancestor that didn't have appendages.
How much will the scientific establishment now have to reorient in light of the momentum of the Extended Synthesis? Will Darwin go the way of Freud? And will the Extended Synthesis require an extensive rewriting of textbooks?
I believe that the field eventually will have to reorient. I don't by any means think the science that's been done under the Darwinian paradigm will disappear or will be seen to be entirely invalid. But the Darwinian mechanism that's used to explain all evolutionary change will be relegated, I believe, to being just one of the several mechanisms - maybe not even the most important when it comes to understanding macroevolution, the evolution of major transitions in body type.