Evolution Icon Evolution
Intelligent Design Icon Intelligent Design

Revealing Darrel Falk’s Overstatements about Limb Bones in Fish Fins

Casey Luskin
Image: A zebrafish, by yourgenome, via Flickr.

In his critical review of Return of the God Hypothesis for BioLogos, to which Brian Miller and I have been responding (herehere, and here), biologist Darrel Falk argued that Stephen Meyer “does not fully appreciate the power of gene duplication and mutation in generating new proteins and changing the way that gene regulatory networks function.” We saw in a previous post that Meyer discussed these topics extensively in both his latest book and his previous book, Darwin’s Doubt. He showed how they do not explain the origin of animal body plans. Falk never addresses Meyer’s specific criticisms of both processes as mechanisms of evolutionary innovation. But he does argue that one “mystery” that seems to be solvable is how fins evolved into limbs. Here’s what Falk writes:

One of the mysteries that, according to Meyer, “Neo-Darwinism fails to explain” is the evolutionary transition from the fins of fish to the limbs of land animals (p. 303). This, and other challenges like it, is simply no longer the mystery he thinks it is. In fact, Gerd Müller, that first speaker at the 2016 [Royal Society] meeting, wrote:

“When natural selection affects such kinds of systems, the resulting phenotype variation does not need to be gradual and continuous. In fact, simulations of the dynamical behaviours of gene regulatory networks in evolution demonstrate that bistable changes are more likely to occur than gradual transitions.”

A Tractable Problem?

Falk goes on to say that he’s hopeful that the evolution of fins into limbs is a tractable problem because it has an “active research program” and “the experiments are still going full bore.” We’ll get to that shortly, but first let’s see the full context of Meyer’s original comment about Gerd Müller and the origin of fins and limbs in Return of the God Hypothesis:

What exactly does neo-Darwinism fail to explain? A phenotype refers to the visible form of an animal’s or plant’s anatomy. Müller was therefore saying that standard neo-Darwinian theory has failed to explain the origin of the new and complex anatomical features and structures that have arisen throughout the history of life. That would include novel animal architectures such as the arthropod, chordate, and molluscan body plans; new anatomical structures such as wings, limbs, eyes, nervous systems, and brains; and new specialized organs such as the vertebrate liver, digestive system, and kidneys. In short, neo-Darwinism fails to explain the origin of the most important defining features of living organisms, indeed, the very features that evolutionary theory has, since Darwin, claimed to explain. 

(Return of the God Hypothesis, pp. 195-196)

Not a Misrepresentation

Is this a misrepresentation of Müller’s position? Not at all. Müller has been quite clear that modern neo-Darwinian evolutionary biology fails to address the origin of novelty, as he co-wrote with Stuart Newman:

[C]oncern with the “gene” has overwhelmed all other aspects, and evolutionary biology today has become almost synonymous with evolutionary genetics. 

These developments have edged the field farther and farther away from the second initial theme: the origin of organismal form and structure. The question of why and how certain forms appear in organismal evolution addresses not what is being maintained (and quantitatively varied) but rather what is being generated in a qualitative sense. This causal question concerning the specific generative mechanisms that underlie the origin and innovation of phenotypic characters is probably best embodied in the term origination, which will be used in this sense throughout this volume. That this causal question has largely disappeared from evolutionary biology is partly hidden by the semantics of modern genetics, which purports to provide answers to the question of causation, but these answers turn out to be largely restricted to the proximate causes of local form generation in individual development. The molecular mechanisms that bring about biological form in modern-day embryos, however, should not be confused with the causes that led to the appearance of these forms in the first place. 

(Gerd B. Müller and Stuart A. Newman, “Origination of Organismal Form: The Forgotten Cause in Evolutionary Theory,” Origination of Organismal Form (MIT Press, 2003), p. 3)

They go on to list various “open questions concerning morphological evolution,” including what he calls the “Novelty” question: “How are new elements introduced into existing body plans?” They even say that “Open questions” include “Do new structural elements arise from mutations?” He further states that neo-Darwinism has “no theory of the generative”:

Although this theory can account for the phenomena it concentrates on, namely, variation of traits in populations, it leaves aside a number of other aspects of evolution, such as the roles of developmental plasticity and epigenesis or of nonstandard mechanisms such as assimilation. Most important, it completely avoids the origination of phenotypic traits and of organismal form. In other words, neo-Darwinism has no theory of the generative. As a consequence, current evolutionary theory can predict what will be maintained, but not what will appear. [Emphasis Added.]

(Müller and Newman, 2003, p. 7)

Thus, it seems entirely appropriate that Meyer cited Müller to argue that “standard neo-Darwinian theory has failed to explain the origin of the new and complex anatomical features and structures that have arisen throughout the history of life.”

Citing a Post-Return of the God Hypothesis Paper

As noted, Falk strikes a hopeful tune that evolutionary biology will be able to explain the origin of novel features in the history of life, focusing on the purported evolution of fins into limbs. After presenting the quote from Müller above which says “phenotype variation does not need to be gradual and continuous” and “dynamical behaviours of gene regulatory networks in evolution demonstrate that bistable changes are more likely to occur than gradual transitions,” Falk writes:

This is wonderfully illustrated by a paper that came out two months ago in the journal, Cell. The authors showed how two mutations changed the bones in the fin of a zebrafish into tiny bones which are likely the equivalent of the radius and ulna, two of the main bones in the limb of a land animal. With just those two mutations, not only were the bones produced, but they became attached to muscles — the beginning of functionality. Furthermore, their formation was influenced by a latent pattern of gene expression already present in fish. We know that this pattern is likely the same one used in the development of limbs in mice. 

The paper that Falk refers to is titled “Latent developmental potential to form limb-like skeletal structures in zebrafish,” and it was published this past February in Cell — long after Return of the God Hypothesis was finalized and sent to the publishers. It doesn’t seem very fair of Falk to accuse Meyer of failing to address a paper that was published after his book was written. Even so, a closer look shows that the paper does not show what Falk claims it shows. 

Lack of Homologous Bones in Fins and Limbs

The investigators in the study induced mutations in the developmental genes vav2 and waslb which are involved in regulating appendages in vertebrates — fins in fish and limbs in tetrapods. The investigators claimed that these mutations generated “new bones” in the fin, which “integrate into musculature, form joints, and articulate with neighboring elements.” Yes, they did produce a couple new bones in the fin of a zebrafish, but the authors did not say that they are homologous to bones in tetrapod limbs.

Before we appreciate the modest results of this study, let’s understand the bone structure of fish fins and tetrapod limbs. Below (left) is a diagram of a fin of a living teleost, or a ray-finned fish, compared with a tetrapod limb (right), with bones labeled, modified from Figure 1 of the paper:

Modified from Figure 1 of “Latent developmental potential to form limb-like skeletal structures in zebrafish,” Cell, 184: 1-13, by M. Brent Hawkins, Katrin Henke, Matthew P. Harris (February 18, 2021), with permission from Elsevier. Figures not necessarily to scale.

In the teleost (ray-finned) fish fin at left, there are three main types of bones: proximal radials (green), distal radials (red), and fin rays (grey). On the right we see a tetrapod limb (human). Moving from bottom to top, it shows the typical vertebrate limb pattern: one bone (humerus, orange) followed by two bones (radius and ulna, brown), followed by multiple small wrist bones (carpals, blue), followed by elongated finger or toe bones, known as phalanges (purple). Yano and Tamura (2013) explain how difficult it is to identify corresponding homologous bones between modern fish fins and tetrapod limbs:

In fact, it can be difficult to find corresponding elements between the fin skeleton and limb skeleton in extant vertebrates. Whereas the limb skeleton is composed of endochondral bones (endoskeleton), the fin skeleton consists almost entirely of fin rays (exoskeleton), with a poor underpinning endoskeleton (Tamura et al. 2008). Because of the vast morphological differences between fins and limbs, some regard them not as homologous organs in the classical (morphological) sense but as organs with ‘deep homology’, which means they arose by the modification of pre-established genetic regulatory circuits (Shubin et al. 2009).

We’ll investigate the meaning of “deep homology” and Yano and Tamura (2013)’s citation of Shubin et al. (2009) in just a moment. For now, suffice to say that although limbs and fins are generally considered homologous structures, it’s difficult to identify specific bones within the two structures that are actually homologous.

Mutating a Fin to Create…What?

For the paper in Cell, the researchers mutated developmental genes involved in fin growth and, in certain cases, were able to produce two new bones. But did they produce anything that looks like, functions like, or even resembles a vertebrate limb? Not at all. The mashup below shows the best of the results that were obtained by the study:

Reprinted Portions of Figures 2, 6, and 7 of “Latent developmental potential to form limb-like skeletal structures in zebrafish,” Cell, 184: 1-13, by M. Brent Hawkins, Katrin Henke, Matthew P. Harris (February 18, 2021), with permission from Elsevier.
The far left (what I’m calling Figure 7A) shows a normal fin of a ray-finned fish. Figure 7B shows a stylized drawing highlighting two new bones (colored orange) which the paper claims were produced by their study. The photos at right (Figures 2C, 6B, and 6F) shows various fish fins that are claimed to exhibit the bones in Figure 7B. As can be seen, they have not produced a tetrapod limb or anything like it. At best what they’ve done is add two bones to a fin’s limb that have no clear relationship or functionality, if any, to actual bones in a tetrapod limb. Figures not all at same scale.

Not What the Paper Says

Falk claims that those two new bones are “likely the equivalent of the radius and ulna” from a tetrapod limb. But that is not what the paper says. Nowhere does the Cell paper claim that these two bones produced by mutating fish developmental genes are the equivalent of the radius and ulna in tetrapod limbs. Rather, the authors acknowledge that these newly constructed “bones” lack clear homologues real tetrapod limbs:

Due to the morphological disparity and differential retention of ancestral components of the appendage skeleton in teleosts in contrast to tetrapods (Figure 1A), we cannot assign a homology relationship between intermediate radials and any specific element of the limb. Rather, we propose a ‘‘deep homology’’ (Shubin et al., 2009) between middle elements of these appendages, whereby Hox11 genes establish an intermediate regionality in which long bones can develop. This highlights the flexibility of developmental programs to traverse morphological boundaries and the potential to blur the distinction between them. [Emphasis added.]

Falk further claims that these bones are “attached to muscles” and represent “the beginning of functionality.” Since muscles tend to grow around bones, the fact that the bones are attached to muscles isn’t strange. But “the beginning of functionality”? What kind of function are we talking about? Have any tests been done to see if fish with these mutations and extra bones have any new kind of functionality? No, they haven’t been done. These mutants were killed in the lab and studied. There is no evidence that there’s any new functionality added by these bones, much less any functionality that is like that of a tetrapod limb. In other words, there’s no evidence that these mutations would yield any kind of a functional advantage that could provide a benefit towards survival and reproduction, were they to occur in the wild. 

These bones represent the interesting results of experiments in mutating developmental genes in fish, but it’s not clear what relevance there is to explaining how vertebrate limbs — i.e., vertebrate limbs we actually observe in living organisms today — evolved. Falk accuses Meyer of understating progress in this field, but in reality, Falk over-extrapolates the import of these experiments. 

Deep Homology or Deep Trouble for Neo-Darwinism?

As noted above, the authors of the Cell paper (as well as Yano and Tamura, 2013) cite Shubin et al. (2009) to claim that the common developmental pathways that produce these mutant bones in fish fins and normal bones in tetrapod limbs show “deep homology.” But Shubin et al. reveal that the very concept of “deep homology” was born out of data that was unexpected under a neo-Darwinian evolutionary paradigm: 

One of the most important, and entirely unanticipated, insights of the past 15 years was the recognition of an ancient similarity of patterning mechanisms in diverse organisms, often among structures not thought to be homologous on morphological or phylogenetic grounds. In 1997, prompted by the remarkable extent of similarities in genetic regulation between organs as different as fly wings and tetrapod limbs, we suggested the term ‘deep homology’ to describe the sharing of the genetic regulatory apparatus that is used to build morphologically and phylogenetically disparate animal features.

In other words, there are some structures — they give examples of bird wings vs. fly wings, insect legs vs. vertebrate legs, or insect eyes vs. jellyfish eyes vs. vertebrate eyes — that have no obvious homology but use similar genes for their construction. Neo-Darwinism did not predict this data, which is why Shubin et al. called this discovery “entirely unanticipated.” From a design standpoint, we might have expected this data under the idea of common design, where there is re-usage of common genetic programs in widely diverse organisms. Could the same thing be going on here, where similar genetic programs were intelligently designed in fish and tetrapods to control fin or limb growth, even though there is not necessarily an evolutionary link between these structures? 

“A Historical Continuity”?

Shubin et al. (2009) say that deep homology “implies a historical continuity, but in this case the continuity may not be so evident in particular morphologies; it lies in the complex regulatory circuitry inherited from a common ancestor.” In other words, similar genes are being used to control and regulate the development of widely different structures — structures which otherwise have no obvious evolutionary relationship. They call these “cryptic classical homologies, when morphological data alone are inadequate to make the case for homology” — again, meaning that similar genes are being used to construct structures that are morphologically distinct. As they put it, “morphologically disparate organs whose formation (and evolution) depends on homologous genetic regulatory circuits.”

This is the sort of “deep homology” that the authors of the Cell paper are talking about here. They aren’t saying that they’ve constructed bones found in vertebrate limbs. They’re saying that they’ve mutated genes to induce the creation of something that shares “deep homology” with unspecified bones in tetrapod limbs. But as the originators of the term “deep homology” point out, saying two structures have “deep homology” could be like saying they’re as different as a fly’s leg and a whale’s fin, but are nonetheless regulated by similar underlying genetic pathways. Citing “deep homology” does not mean you’ve actually accounted for the origin of a structure. It means that we’ve found that the same genes control the origin of very different structures — a reusage of genetic components “entirely unanticipated” under neo-Darwinism but unsurprising under ideas like common design.

To appreciate how unexpected this kind of data is from an evolutionary perspective, consider what Shubin et al. (2009) say about the usage of similar genetic programs to control the development of eyes:

Are all eyes homologous in the classical sense of being diverse forms of the same structure? Have different eyes instead evolved entirely independently and just happened to use similar genetic components? Or have different eyes evolved in parallel as elaborations of structures and genetic regulatory mechanisms present in common ancestors? Weighing up these possibilities requires a deeper examination of eye development and a broader survey of taxa. It turns out that far more striking developmental similarities exist between diverse taxa than would ever have been expected.

This data thus shows “far more striking developmental similarities exist between diverse taxa than would ever have been expected” under their evolutionary view. Again, from a design standpoint, “deep homology” means common basic developmental programs being reused in different organisms, despite widespread morphological diversity. 

“Latent” Abilities or Front-Loaded Intelligent Design?

Though Falk overstates the paper’s findings, it offers intriguing comments about the implications for evolution. Recall that the title of the paper is “Latent developmental potential to form limb-like skeletal structures in zebrafish.” The paper repeatedly uses the word “latent” in describing these genetic programs (all emphases added):

  • “revealing in fins the presence of latent developmental programs once thought unique to limbs”
  • “Teleost fishes have latent limb-like Hox programs that can elaborate the endoskeleton”
  • “Our findings reveal a latent, limb-like pattern ability in fins that is activated by simple genetic perturbation.”
  • “Our results demonstrate that zebrafish retain the capacity to form structures with an intermediate regional address specified by hox11 paralogs, similar to the limb zeugopod. This latent capacity is activated by simple genetic changes, revealing a novel appendage-patterning pathway that may have played unrecognized roles in the fin-to-limb transition and skeletal development at large.”
  • “endogenous limb-like developmental programs latent in fishes”
  • “Our results reveal latent or emergent properties of development within vertebrate appendages to form elaborate, articulated skeletal structures.”

For the record, the word “latent” is defined as follows: 

Latent is an adjective that you use to describe something that is capable of becoming active or at hand, though it is not currently so.

I’m highly skeptical that this paper shows fish have the “latent” ability to produce something like a tetrapod limb. But let’s say they’re right, and they do. This would raise the question, Why did fish have the latent capability to produce limbs? Under Darwinism, evolution has no foresight and is blind to future needs. But this sort of pre-adaptation, as it is sometimes called, fits with intriguing models of “front-loaded” design where organisms appear pre-adapted or front-loaded to produce certain features, even though they are not yet needed for an organism to survive and reproduce. This study certainly doesn’t show that fish are pre-adapted to produce limbs. However, the interpretation of the results as showing “latent” genetic capabilities has teleological overtones that are compatible with intelligent design.