Scientists: Surgeons not out of a Job Anytime Soon---Or Why that Appendicitis Still Hurts.

By Douglas L. Theobald, Assistant Professor of Biochemistry, Brandeis University

“Its major importance would appear to be financial support of the surgical profession.”
—Alfred Sherwood Romer and Thomas S. Parsons, The Vertebrate Body (1986), p. 389.

A recent science news article from the Associated Press reports on a novel hypothesis (Bollinger et al. 2007 JTB in press) concerning the possible function of the human vermiform appendix. Given how much creationists dislike all things vestigial, the Bollinger paper will undoubtedly be paraded on anti-evolution websites with grandiose claims about how “researchers are declaring that they have found a purpose for the human appendix”—and the paper itself unfortunately provides several tasty quote-mines ready to be plucked.

The new hypothesis, to be published in the Journal of Theoretical Biology, was proposed by a group of immunologists and surgeons from Duke University Medical Center. Bollinger et al. suggest that the structure and location of the human appendix has been specifically modified for an adaptive function, namely for housing and preserving beneficial gut bacteria during certain pathogenic infections that otherwise would clear the bowel of all enteric bacteria. Evidently, individuals who had an appendix would have a selective advantage over those individuals who had either no appendix or who had a larger appendix with a larger opening. Presumably after, say, an epidemic of amoebic dysentery, individuals with such an optimally sized appendix would be better able to repopulate the gut with good bacteria and to recover (this is an adaptationist paraphrase; the authors don’t use explicit evolutionary terms). The authors were led to this hypothesis after they noticed that human appendixes (yes, that’s the proper plural for the anatomical structure) contained significant bacterial biofilms, which are notorious for allowing bacteria to withstand all manner of malign assaults, including vigorous mechanical cleaning and chlorine bleaching.

There are two important points to be made about the Bollinger et al. paper, both of which have been completely lost in the wash of media hype. First, this is purely a hypothesis paper, and, aside from cursory physiological plausibility, as yet there is no empirical evidence for the hypothesis whatsoever. A novel hypothesis was proposed; nothing has been scientifically demonstrated. And, in practice, it will be extremely difficult to test this hypothesis. ¹. Second, even if the authors are correct—that this is a bona fide adaptive function for the appendix—-there is absolutely no basis for the claim that the appendix should no longer be considered a vestige. Quite to the contrary, the proposed function offers even more evidence for the vestigiality of the human appendix.

The proposed adaptive function is no silver bullet

First, the good news. Given the facts, the Bollinger hypothesis is plausible. In fact, from my reading of the literature concerning the possible functions of the appendix, their hypothesis, though imperfect, seems by far the most likely for an adaptive hypothesis specifically. All the other currently favored adaptive hypotheses for the function of the appendix are also immunological. But these others are problematic for two big reasons: (1) the evidence for functionality is wholly based on animal models, like rabbits and mice, whose “appendixes” are not homologous with the human appendix, and (2) they propose a function that is evidently also served by the rest of the cecum/colon. In other words, the function proposed is something that the appendix might do, but the function is not specific to the appendix per se. In contrast, this new Bollinger et al. adaptive hypothesis is the best going because, unlike the others, it is based on human physiological data and because it proposes a reasonable function specifically for the appendix as a distinct structure from the rest of the gut.

However, adaptive explanations are not the only possible explanations. Other plausible, non-adaptive explanations for the appendix exist, and there are some significant problems and questions here for any adaptive explanation that purports to explain the continued existence of the human appendix.

For example, without modern medicine, the mortality rate due to appendicitis would be between 3-10%. This is extremely strong selection against having an appendix. If selection is the only thing at play in maintaining the appendix, then, all else equal, the proposed function must have an even greater positive selection pressure to compensate. But it is hard to see how a “protected safe haven” of bacteria could be so beneficial. Presumably, during an infection that clears the bowel of bacteria, some small fraction of bacteria would be left. Is that fraction really so small compared to the amount in the appendix, the interior volume of which has been compared to the size of a matchstick? The proposed function would also be most effective when intestinal illnesses like dysentery or cholera are frequent. Was this the case for Holocene hunter-gatherer societies, which likely wouldn’t have the water-contamination problems characteristic of modern epidemics? Not just humans, but all hominoid apes have a vermiform appendix (that’s gibbons, orangutans, gorillas, chimps, and humans)—and so the proposed function presumably existed during the last 20 million years of ape evolution. Are severe diarrheal illnesses common in apes?

Ultimately, in order to firmly establish empirical evidence for an adaptive function for some structure, evolutionary biologists stipulate that several necessary requirements must be met:

• Phenotypic variability: If there is no variation in a given structure, then selection has nothing to choose from.
• Heritability: There can be strong selection for a trait, but if the trait is not inherited, it won’t increase in frequency in subsequent generations (at least not due to selection).
• Differential fitness (like mortality) based on phenotype: If all the variants of a structure have the same impact on fitness (e.g., none of them help or hinder, relatively, their bearers in reproductive terms), then there is no selection. And finally,
• Non-random change in frequency: It’s possible that the first necessary three requirements are fulfilled, yet the structure still doesn’t evolve in a non-random, selection-driven manner (the classic example is in small populations, where genetic drift dominates and can override selection).

So what is the relevant evidence regarding the appendix?

• How much variation is there in the size of the appendix? Quite a bit. Typically, the human appendix ranges anywhere between 2-7 inches in length, but it is not uncommon to find one under an inch or nearly a foot long. There are many documented, though rare, cases of congenitally absent appendix, where a person is born completely lacking one. So far so good.
• Are variations in the structure of the appendix heritable? Though it is reasonable to think so (typical heritabilities for morphological traits are around 1/2), to my knowledge nobody has ever looked at this specifically.
• Is there differential survival among people who lack an appendix versus those who have one? For example, are people who lack an appendix more likely to die from intestinal illnesses? One way to look at this would be to study whether individuals who’ve had appendectomies recover less often from, say, amoebic dysentery or cholera. Nobody has done this analysis yet, and in fact this would be particularly hard to test. Even if a correlation was found between appendectomy and mortality due to dysentery, this wouldn’t establish that lacking an appendix was the problem. It’s likely that people who are susceptible to appendicitis are generally more susceptible to intestinal problems overall. Ideally we’d want to look at people who had healthy appendixes removed incidentally, which would mostly eliminate possible artifactual correlations between appendicitis and mortality.
• Finally, do we see evolution in appendix size after a selection event? For instance, after an epidemic of cholera, are fewer people in the population born without appendixes, or does the average length/volume of the appendix increase? Nobody knows.

Again, there are other plausible, yet non-adaptive, explanations for the appendix too. The appendix may simply be a detrimental product of historical contingency, similar to other things like crammed wisdom teeth, birth canals narrower than a baby’s head, slipped discs, tail bones, and prostate trouble. Developmental constraints can prevent the complete elimination of a structure—perhaps the mutation that would finally dispose of the appendix is correlated with development of the rest of the gut, and would eliminate the colon too. Perhaps we’re stuck in a local minimum of sorts, unable to get to the global minimum. For example, some researchers have proposed that the appendix would decrease in size, eventually being eliminated, were it not for the fact that small appendixes are more prone to appendicitis (they get blocked and infected easier)—so there is antagonistic negative selection that prevents complete elimination.

All of the above considerations underscore just how difficult it is to establish a true adaptation, how high the evolutionary bar is set for empirical demonstration of function, and how outrageous it is when we see headlines like “Appendix protects good germs” or “Scientist’s [sic] find appendix’s job”.

The appendix can have a function and be vestigial

Bollinger et al. imply in several places (five exactly, including the abstract) that the human appendix should no longer be considered a vestigial structure. This appears to be due to a common misunderstanding of the vestige concept.

A vestige can have important functions—complete non-functionality is not (and never has been) a necessary requirement for a vestigial structure. “Vestige” is a word that has Latin roots and literally means “footprint”, a trace of something. The key, defining element of a vestige is that it has lost an important function. A vestige can only be identified via comparative analysis, as a reduced and rudimentary structure compared to its homologs, one that lacks the complex functions usually found for that structure in other organisms. In discussions with Bill Parker (senior author on the Bollinger et al. JTB paper) and Seth Borenstein (AP reporter), I discussed analogies with bird wings/forelimbs that I think are very instructive.

Everyone agrees that an ostrich’s wing is a vestige—ostriches have wings, which are meant for flight, yet ostriches cannot fly. Nonetheless, ostrich wings may have several important functions: they are used in courtship displays for securing a mate, they shield young birds from the intense desert sun, they enable very quick lateral maneuvers when running at speeds up to 40 mph, and two of their wing digits terminate in fiercely sharp claws that are used in defense. While possibly very important, all of these functions were likely present in the flighted ancestors of the ostrich.

A similar, yet tellingly different, situation is found in the penguin, which has “wings” that have become highly modified for underwater swimming. A penguin’s forelimbs, its flippers, clearly evolved from wings in an ancestral bird, yet a penguin can no longer fly. However, and here is the important difference, the penguin’s flippers have gained an important function that its ancestors did not have—the penguin is an incredibly efficient and skillful underwater swimmer. Penguin flippers are homologous to the wings of birds that can fly, but they are not really vestiges.

Now here are the important lessons:

• Ostrich wings are vestigial because they have lost an important function, because they are organs with a particular structural arrangement that evolved for a different purpose—a purpose they no longer serve.
• Ostrich wings are vestigial even though they may perform important functions.
• Penguin flippers are not considered vestigial, primarily because they have gained an important function that is not found in the ancestral state (and they have been highly modified for that function).

How does this relate to the present Bollinger hypothesis? Analogous to the situation in both ostriches and penguins, the human appendix has lost an important function, namely cellulose digestion (for background and detailed explanation, see “The vestigiality of the human vermiform appendix”). If the Bollinger hypothesis is correct, then the appendix also performs an important immunological function. But is this function “new”, or was it present in the ancestral, cellulose-digesting primate cecum? If the function is new, analogous to the penguin case, then the appendix can’t really be considered a vestige. If the function is old, analogous to the ostrich case, then the appendix is still a vestige. Here is how William Parker puts it:

There is virtually a 100% chance that if we are correct regarding the function of the human appendix, then the cecum of other animals fulfills the same function as the appendix of humans. In fact, much of our data are taken from non-human studies in a variety of species.

It is apparently a case of an organ being good at two functions (bacterial preservation and digestion) evolving to more efficiently carry out only one of those functions (bacterial preservation) while losing the other function (digestion) completely.

In other words, we have here a case that is perfectly analogous to the vestigial ostrich’s wings.

One final note of irony. The authors state that their proposed function is no longer needed in modern society. So, if correct, this would mean that the appendix is doubly vestigial.

• R. Randal Bollinger, Andrew S. Barbas, Errol L. Bush, Shu S. Lin, and William Parker (2007) “Biofilms in the large bowel suggest an apparent function of the human vermiform appendix.” Journal of Theoretical Biology, in press.
• Romer, A. S. and Parsons T. S. (1986) The vertebrate body. Sixth edition. Saunders College Pub.: Philadelphia, PA.
• Also see: The vestigiality of the human vermiform appendix
• Also see: Citing Scadding (1981) and Misunderstanding Vestigiality

Acknowledgments

I extend a warm well of gratitude to the many cantankerous PT regulars who gave much help and input for this post, including, in no particular order, PZ Mayerz, Larry Moran, Pete Dunkelberg, Steve Reuland, and Mark Isaak. I have paraphrased many of their very comments. I also thank Bill Parker for allowing me to quote him, and for amiable discussion.