Quirks of Human Anatomy
by Lewis I. Held, Jr.

purchasing information


101 Unsolved Puzzles in Evo-Devo

Note to users: Evo-devo seeks to decipher how genomic tinkering has resulted in anatomical alterations via developmental reprogramming. These 101 unsolved puzzles suggest the scope of concerns in the evo-devo field, and are intended as a resource for term paper topics in college courses. They can also be used to spark literature exploration or laboratory research at the graduate level or beyond.  More casually, they can be thought of like New York Times crosswords--that is, riddles (at various levels of difficulty) to challenge keen minds and, perhaps, sharpen them even further.

Numbers in brackets denote references, as enumerated in Quirks of Human Anatomy. See also figures and legends. For context and perspective on these topics see Quirks. Definitions of evo-devo terms may be found in Hall and Olson’s book Keywords and Concepts in Evolutionary Developmental Biology (Harvard Univ. Press, 2003). Puzzles have been organized into a dozen categories, within which the order of listing is arbitrary:
I. Origins (10 puzzles).

II. Evolvability (10 puzzles).

III. Genome (10 puzzles).

IV. Symmetry (8 puzzles).

V. Asymmetry (13 puzzles).

VI. Digits (4 puzzles).
VII. Teeth (6 puzzles).

VIII. Limbs (6 puzzles).

IX. Sex (7 puzzles).

X. Reproduction (8 puzzles).

XI. Nervous system (15 puzzles).

XII. Behavior (4 puzzles).
I. Origins:

Puzzle 1: Why (and how) did hair evolve in mammals?

Hair is a ubiquitous feature of mammals [1623,2417]. Its selector gene appears to be Foxn1 [2299], which encodes a transcription factor of the winged-helix family [1327,1860]. We don’t yet know how proto-mammals came to express Foxn1 in the hair follicle (a novel cis-enhancer?) [2299], nor how Foxn1 captured keratin genes as targets therein [1239,2313], nor, finally, how Foxn1 acquired an added role in hair pigmentation [144,2771].

Puzzle 2: Why (and how) did heterodonty evolve in mammals?

How did the different classes of mammalian teeth originate? Reptiles already had all the signals needed for tooth diversification in the right areas of their jaw (BMP in front, FGF in rear, etc.) [476], but their teeth were uniformly peg-shaped [2027]. In the lineage that led to mammals, mutations must have somehow enabled tooth buds to (1) “hear” those pre-existing signals [2488,2780] and (2) respond to them by turning ON cusp-patterning genes [340]. After this co-option by teeth of jaw signals [1695], further shape changes were probably relatively easy [333,1298].

Puzzle 3: Why (and how) did tears evolve as a distress signal?

We do not yet know how our emotions got linked to our lacrimal glands [1476]. The same question can be asked about (1) our chin, which, in some people, quivers before crying, and (2) our throat, which gets a “lump” [2737]. The latter link is also interesting since it usually precludes speaking [612]. Crying resembles laughing insofar as it is partly contagious [612] (cf. [1935]).

Puzzle 4: Why (and how) did the pituitary gland evolve?

Why does our pituitary gland split apart and rejoin after migrating along a peculiar route [1348,2192], and why did it evolve in association with our oral region in the first place [356,1191,1637]? The answer may be traceable to the D-V axis inversion that the chordate body underwent ~500 MYA and the subsequent migration of the chordate mouth down the head to a ventral location [456,1463,1464].

Puzzle 5: Why (and how) did the sex comb evolve in flies?

The “sex comb” (so-named because it resembles a hair comb and is present in only one gender) is a row of bristles that may help males grip females during mating [1874], though it could just be an ornament [2055].

Developmentally, the comb arises as a transverse row (t-row) that rotates ~90° [1138,2567]. Evolutionarily, its bristles became thicker, blunter, darker, and more numerous as it underwent alterations in various lineages [1425]. The taxonomic distribution of sex combs indicates that they originated near the base of the subgenus Sophophora (after its divergence from other subgenera but before the splitting of melanogaster and obscura species groups) [1425] ~62 MYA [103]. Genetically, the sex comb and other dimorphisms depend on a hierarchy of control genes [457,993]—beginning with ones that count the number of X chromosomes (sic, not the X:A ratio!) [727].

The capture of the distalmost t-row by dsx (doublesex) evidently occurred via a new link between dsx and the Hox selector gene Scr (Sex combs reduced) [135,1997], which governs the foreleg-bearing body segment [134]. How the enslaved t-row became a comb is actively being researched [1007,2124,2630].

Puzzle 6: Why (and how) did 5-fold symmetry arise in echinoderms?

Echinoderms somehow evolved radial, 5-fold symmetry as adults [2415], but they retain bilateral symmetry as larvae [2536]. Strangest of all, they metamorphose by an asymmetric process as bizarre as having a wart on your left cheek that gradually expands to become an umbrella [681,1766].

Puzzle 7: What was the antecedent of the neural crest?

Precursors of the neural crest have been found in pre-vertebrate chordates [1286] but not in other phyla that are our nearest relatives (other deuterostomes) [2500]. In one of those pre-vertebrates (the urochordate E. turbinata) certain cells scatter from the neural tube to become pigment cells [1287], and in another (the cephalochordate amphioxus) sheets of cells migrate from the edges of the neural plate toward the midline and form epidermis [122]. Thus, the “migrant worker” strategy predates the origin of the vertebrates, but the itinerants were not nearly as versatile as their vertebrate counterparts.

The pithier question, therefore, is: how did these “proto-crest” cells acquire so many more duties in vertebrates [141,1900]? One idea is that the neural border in basal chordates was primed—or “preadapted” [442]—for further roles [1191,2302] since amphioxus expresses many crest-specific genes there already [951]. Border cells could have been released from the epithelium via subsequent mutations that allowed delamination, migration, and differentiation [1356,1952,2365]. They may have gained access to diverse histological pathways (e.g., cartilage formation) by co-opting existing genetic circuits [1735,2282]. To trace the genealogy of our crest, we will need to learn a lot more about the urochordates [1060,1516] since they turn out to be closer to us phyletically than the cephalochordates [627,874,1192].

Puzzle 8: What was the antecedent of the vertebrate eye?

The earliest known vertebrates had a head with paired eyes [1454], but the “eyes” of their protochordate forebears were situated along the midline [1460]. Urochordates have a single eyespot in the larva [2188,2368], while cephalochordates have four different kinds of light detectors [59,1461,1463,2800]: two with rhabdomeric-type photoreceptors (dorsal ocelli and Joseph cells) and two with ciliary-type photoreceptors (frontal eye and lamellar body). The latter two organs appear be the antecedents of our lateral and pineal eyes respectively, but this conjecture is still quite tentative [1463].

Puzzle 9: How complex was the Urbilaterian eye?

The Urbilaterian was the last common ancestor of all bilaterally symmetric phyla [195,1393]. Was its eye (or eyespot) complex enough to form images? Did it have a lens? Was it a simple eye (as in vertebrates) or a compound eye (as in arthropods) [90]? We don’t know. Moreover, we are equally ignorant about the Urbilaterian had segments [443,731,1226], limbs [1761,2092,2096], genders [737,2623], or other complex traits [119,195,392,604,886,908,2663]. Tracing the clues about these features constitutes one of the most engrossing detective stories in all of science ... or in all of literature for that matter!

Puzzle 10: How did the vertebrate heart evolve?

The evolution of the vertebrate heart is an epic story that we are only now beginning to unravel morphologically [777,2062,2402] and genetically [438,481,582,1933,2280]. There are more twists and turns in the construction of the mammalian heart than in any Shakespearean play [7,308]! How did the assembly process ever get to be so complex?

II. Evolvability:

Puzzle 11: What aspects of genome circuitry dictate a trait’s evolvability?

“Evolvability” denotes the ease (or difficulty) of specific evolutionary changes in anatomy [589,1143,2422,2716]. For example, Darwin was astounded at the plumage varieties attainable by artificial selection using the ordinary rock pigeon [559,560,974,1141]. Dog breeds likewise dazzle us by their wide span of shapes and sizes [560,792,1951]. Evidently, feather patterns and body shapes are easy to alter in these respective species, and body size knew no bounds in either sauropods [2272] or whales [188].

We don’t yet understand enough about genomic circuitry to explain why certain groups of animals resist anatomical changes due to internal constraints [1330,2264], nor to fathom how such constraints have been overcome in other lineages [1689,2095,2262]. All we can say with confidence is that some (intra-module) traits are strongly linked within the genome and others (inter-module) are weakly linked [1393,1400].

Puzzle 12: Why have certain areas of Morphospace remained vacant?

A “Morphospace” is the set of all anatomies that are theoretically attainable by any given species (or larger taxonomic group) given an unlimited amount of time to evolve [959,2600]. One oft-cited example of a Morphspace concerns the mollusc shell [1687]. The spectrum of naturally occurring shell shapes (from conch to nautilus to clam) is produced as a function of only three developmental parameters: (1) the rate of growth of the shell’s mouth, (2) its rate of revolution about a vertical line, and (3) its rate of translation along that line [2132]. These internal variables define the axes of an imaginary cube, wherein most mollusc species can be plotted as single points (x, y, z) [465,2133]. Not all regions of this Morphospace are occupied [959,2600], presumably due to the contingent conditions (i.e., historical constraints) that governed how the various lineages of molluscs happened to evolve [2325,2499,2729].

One trait that seems to have gotten stuck at a single point in its Morphospace is reptile teeth, which remained conical (like those of crocodiles) for more than 100 million years [1357], despite the fact that dental heterogeneity (molars, canines, etc.) would have offered tremendous advantages for chewing different types of food [1587]. Did the right mutations simply never occur [1309,1484], or was it just too hard to concoct the right binding site combinations for the relevant trans-acting transcription factors [1687,2325]? No one knows.

Many other cases of stasis are just as intriguing [959,2824,2861], especially with regard to why certain crannies in “niche space” have stubbornly resisted invasion for eons immemorial [227,482]. Ecologists have toyed with these irritating riddles via game theory [2692], but the ultimate answers are hiding in the genome [793,1303].

Puzzle 13: Why is the number of neck vertebrae so constant in mammals?

Why do mammals typically make seven cervical vertebrae (cf. pterosaurs [426]) despite a wide range of lifestyles [840,2824], whereas birds and reptiles (e.g., sauropods [746]) somehow managed to escape any such restriction [843,1850,1873,2131]? Sloths offer an intriguing exception to the mammalian rule, as described recently by Buchholtz, E.A. and Stepien, C.C. (2009). Anatomical transformation in mammals: developmental origin of aberrant cervical anatomy in tree sloths. Evol. Dev. 11, 69-79.

Puzzle 14: Why have no simians evolved thumbless 5-finger hands?

Our thumb’s uniqueness is attributable to a distinctive “Hox code” [1557]—namely, “d13 ON; d12 OFF”. Consistent with this idea, mouse thumbs can be converted to forefingers by forcing them to express d12 [1406]. Amazingly, a woman with this same thumbless, 5-finger trait in both hands was seen at a birth clinic in Austria in 1957 [1133]. She did not consider her anomaly a disability. On the contrary, she said that it helped her play piano more easily! Her newborn baby looked exactly the same, so the trait is probably genetic. The general term for a transformation of one body part into another is “homeosis” [151].

Among monkeys and apes there is a wide spectrum of hand shapes [2076], but there is one consistent trend: when the lengths of the four fingers change, they tend to do so coordinately—as if they were a coherent module apart from the thumb [2161]. The interdependence of those fingers (and their independence from the thumb) is understandable in terms of the Hox code. That code also explains how evolution could truncate the thumb in a few primates without altering any other digits [2317,2552] (cf. thumb loss in cloven-hoofed artiodactyls [2085]). The only mystery, given (1) the dexterity of the patient mentioned above and (2) the simplicity of the mutation that causes this trait, is why a thumbless, 5-finger hand never evolved in any simian species.

Puzzle 15: Why have there never been any “Batman” mutants?

Bats exhibit extreme limb divergence. Their fingers became as long as their entire body to span their umbrella-like forewing, but their toes stayed small [1869]. Trivial mutations may have been instrumental in this reconfiguration [513]. All it took to enlarge the fingers, apparently, was an excess of BMP2 secretion late in hand development [2328]. Their wing membrane itself may also have arisen just as simply [1372] since interdigital webbing is actually the default condition in tetrapods [2570]. Our own hands would be as webbed as a bat wing (or a duck foot [2591]) were it not for cell death, which carves gaps in the skin between our digits [1970,2924]. One possible factor in the wing-leg disparity of bats has been ruled out: their wings do not get a head start in their growth [202].

If it is really so easy to evolve a bat wing, then why hasn’t a “Batman” mutant arisen by now in our own species? Webbed digits are found in certain syndromes [1711,2512], but body-length fingers have never been documented in the medical literature [1427,1638,1939,2582]. The answer to this mystery may come from further studies of bat wing development at the molecular and morphological levels [428,513,1372,2887].

Puzzle 16: Why have there never been any “centaur” mutants?

One aspect of anatomy that is easy to alter experimentally is the number of limbs. An extra limb can be elicited by merely exposing the flank of a chick embryo to a dose of FGF proteins [489,1758,1921]. Such appendages can be winglike or leglike or a patchwork thereof, but they are never a blend that splits the difference [1248,1922,2560]. What this means for humans is that we couldn’t easily evolve an “arg” (arm-leg intermediate), despite having all the needed ingredients to make such a limb in our genome [473]. Nevertheless, it should have been trivial to evolve extra legs via the capture of an FGF gene by a thoracic transcription factor [1577]. Thus, it seems surprising that no centaurs or other multi-legged vertebrates ever evolved [275,556,2562,2812].

Perhaps such mutants did arise but could not walk due to a lack of needed locomotory circuitry in their brains [692,1954]. Indeed, whenever extra legs are induced, they have a mirror-image polarity [2409,2543,2565], so an extra pair of (backwards) legs would pose a serious “pushmi-pullyu” dilemma for their bearers (a reference to the two-headed beast in Hugh Lofting’s Dr. Dolittle stories).

Puzzle 17: Why have there never been any “Methuselah” mutants?

The fact that we are mortal appears to be a lamentable spandrel [1121,2218,2219,2647,2824], rather than a manifest destiny, as most people assume [672,1722,2732]. Death is not inevitable? Correct! Each of us only dies because natural selection has consistently favored mutations that enhance our ability to reproduce at a young age, even if those same mutations cause us to die after our fertility wanes.

As a result of this Faustian tradeoff (known as antagonistic pleiotropy [1721,2207,2822,2827]), we each carry “time bomb” alleles that will eventually kill us [455]. This mutational constraint might explain why there are no “Methuselah” mutants (life span >200 years) [785,1932,2523]: they would have to suppress too many adverse genetic side-effects all at once to live so long.

Puzzle 18: Why have there never been any “über-furry” mutants?

If hominin hair was suppressed by a single genetic change, then reversion mutations should have occurred by now to create babies with ape-like coverings of hair, and we should know about it. Does the absence of über-furry people refute the Simple-mutation Hypothesis? Not necessarily! A popular conjecture along these lines is Bolk’s Fetalization Theory [956]. It holds that we are like fetal apes in many respects—not just in the sparseness of our hair. (Newborn gorillas have hair on the head, but the rest of the coat does not appear until later [232,598].) If hair pattern is an inextricable part of a larger nexus of changes, then it may be too hard mutationally to reverse it without undoing vital aspects of our development. In other words, atavistic fetuses might indeed arise, but they may die in utero before they make hair.

Proof that a single mutation could have stripped our ancestors of most of their hair has recently come from an unexpected corner of the research world. The genetic basis of hairlessness in the Mexican hairless dog has now been ascertained. The trait is due to a frameshift insertion (a 7-base-pair duplication in exon 1) in the Foxi3 gene [679], which belongs to the same Forkhead box family as the Foxn1 gene that controls hair formation all over the body [1327,2299]. Interestingly, this breed still has plenty of hair on its head (as we do), as well as on its feet and tail. Why this null mutation should leave hair on these sundry (distal?) parts remains to be determined.

Puzzle 19: Why hasn’t evolution made sperm heat-tolerant by now?

As in other mammals, human testes move from their initial location (abdominal cavity) to a cooler, more superficial location (scrotal sac) because sperm cannot become fertile at our core body temperature of 98.6 degrees F [2160,2383,2824]. Why hasn’t any mammal been able to evolve sperm that can function at high temperature [2417,2824]?

Puzzle 20: Why don’t we have the ability to survive without oxygen?

Why must people die when deprived of oxygen? Based on the prevalence of hibernation, estivation, and facultative anaerobiosis among animals [76,1675,2505], the answer is unclear [1171]. Sea turtles, for example, can hold their breath for at least three hours [1600]. Evolution, it would seem, could have given us the means to survive episodes of choking, drowning, or suffocation [1316,2229], but it did not. Why didn’t it? Presumably, the rarity of asphyxiation among primates (by drowning, etc.) reduced the marginal advantage that any salvational mutations might have had to a negligible level.

III. Genome:

Puzzle 21: Why are Hox genes colinear with anatomy?

One unsolved mystery of the Hox clusters is their colinearity [347,416]: the zones of Hox expression along our spine correspond to the order of the genes within each cluster—as if there were a little man (the “homeobox homunculus” [1135]) reclining along each of the respective complexes [318,642]!

Puzzle 22: Is the rate of mutation itself a selectable trait at different loci?

Can the genome target variability to certain compartments to make them mutational hot spots [366,367,1064,1180,1989,2230]? For example, was facial variability selected for in our hominin ancestors as a means of individual identification [11,1141,2903]?

Puzzle 23: How do transcription factors mesh to regulate transcription?

Promoter logic [2850] involves fitting transcription factors together like jigsaw-puzzle pieces into a configuration [373,1326,1729,2672] that stimulates or represses transcription [826,1428,2517]. Evo-devo is just beginning to scratch the surface of how such factors dovetail sterically to implement Boolean logic [240,902,1413,2850] (e.g. [1541]). Solving this riddle should reveal, once and for all, how evolution writes cellular commands in genetic language using protein grammar [572,656,1137,2855].

Puzzle 24: How does our genome activate certain genes at certain times?

How genomes control the timing of gene action is one of the biggest outstanding mysteries in evo-devo.

One interesting example is hair graying, which occurs at different ages in different families. The only clue we have so far about this trait is that its time of onset in mice is dependent on the dosage of the Notch gene [2314]. In flies, pigmentation must be finely tuned temporally because the yellow gene gets turned ON in different body parts in an invariant sequence [1851,2739].

A different example concerns interdigital tissue, which apparently measures how long it is exposed to Shh [645,1094] and responds by emitting a proportional dose of BMP [1045]. Are these cells really measuring time duration directly?

Puzzle 25: Why are are the breast and prostate so prone to cancer?

Given the randomness of how target genes get recruited in general [897,2835], we might expect men and women to differ in the expression of more than just the genes that overtly affect anatomy. Indeed, microarrays have recently revealed thousands of sexual expression differences [713,1251,2186]. Some of the disparities occur in organs that look monomorphic—e.g., kidney [2185], liver [2760] and muscle [2874]. Many of these links are probably adaptively neutral [1369,1741,2206] (cf. fly genes [300,1062,2084]), but this does not mean that they are clinically negligible. On the contrary, some of them probably contribute to a whole spectrum of sex-linked diseases [139,161,339,1000,1854], and this covert network of haphazard connections might also explain why certain quiescent organs are counterintuitively so prone to cancer (e.g., breast [354] and prostate [2745]) [842,2917].

Puzzle 26: Why does acne mainly affect the face?

Acne develops when sebaceous glands get clogged by excess shedding of glandular lining in response to high testosterone levels that arise during puberty [1260]. Acne is thought to be a spandrel of those high hormone levels. Why the face should be more affected than other areas, however, is a mystery [1103]. Perhaps the answer lies in how the genome allocates the intensity of tissue responses via the density of cell receptors.

Puzzle 27: How did the circadian hormone melatonin acquire its targets?

Intriguingly, the hormone melatonin is made from the neurotransmitter serotonin [203], and the two enzymes that are needed for the conversion had actually evolved in the cephalochordates [1396]. Hence, the third (pineal) eye was already preadapted physiologically to becoming a gland. We do not yet know (1) how the release of melatonin came to depend on day-night rhythms [1426,1727] nor (2) how the hormone entrained an apt set of target tissues to respond accordingly [2704]. The answer to the latter question may lie in genomic circuits that control the pattern of receptor expression.

Puzzle 28: Why does anatomy change during domestication?

A stunning instance of genome integration was recently revealed during an experimental domestication of foxes [338,1093]. Namely, a suite of anatomical traits emerged as a result of the artificial selection for docile behavior! The traits included floppy ears, piebald pigmentation, shorter tails, shorter legs, and underbites or overbites [2635]. How many of our own anatomical quirks might likewise have arisen as side effects of selection for adaptive behavioral traits?

Puzzle 29: Why does the iris muscle come from ectoderm vs. mesodern?

Our iris is the only muscle in the body that comes from ectoderm instead of mesoderm [191,584]. How on earth did the ectoderm get the password to unlock the “muscle vault” in the genome?

Puzzle 30: How does the iris manage to regenerate the lens in newts?

The iris, which normally develops independently of the lens, can regenerate a lens in newts if the lens is artificially removed [1125,1273,2639]! This ability must be a spandrel of how eye parts are wired in the genome, but we have no clue about how it evolved [344,1056].

IV. Symmetry:

Puzzle 31: What is the cellular basis for Bateson's Rule?

William Bateson (1861-1926) discovered an odd geometry of abnormally branched legs in animals as different as cockroaches and salamanders [819,1135]. Such legs, he found, always manifest new planes of mirror symmetry that obey what is now called “Bateson’s Rule” [150,304]. Why should limb development be limited to a predictable subset of morphologies whenever disturbances occur?

Puzzle 32: What is the cellular basis for monozygotic twinning?

Based on the numbers of chorions (C) and amnions (A) in monozygotic human twins, 31% of them (2C2A) must split before ~day 4, 65% (1C2A) between ~days 4 and 7, and 3% (1C1A) after ~day 7 [230]. At the ~5-day peak of this distribution the embryo itself has only ~20 cells [2859], so the deaths of just a few cells in the middle of the array could theoretically bisect the cluster. If each half restores its axes (like a half-magnet replacing a missing pole), then it could go on to make a whole baby.

A superficially similar transect scenario has been seen in an unrelated context. When limb buds of tadpoles are infected by trematodes, the parasites carve the leg buds into islands, and each fragment then goes on to make an entire leg, so that the resulting frogs end up with as many as 12 hindlegs [220,1306,2339]! Extra legs can also be induced in healthy frogs by implanting inert beads as partitions [2340]. What could cause a subset of cells to die in a human embryo at 5 days after fertilization?

Puzzle 33: How does twinning occur in armadillos, and how did it evolve?

In the armadillo genus Dasypus [1345] the litters typically consist of identical twins in D. kappleri, identical quadruplets in D. novemcinctus, and identical octuplets in D. hybridus[839]! These powers of two imply that the splitting must involve bifurcations. How does it happen, and why is it so regular?

Why did embryo splitting become the norm in the Dasypus genus and nowhere else in our entire phylum [2510]? The reason, it appears, is a geometric peculiarity of reproductive anatomy. The armadillo uterus cannot hold more than one egg at a time. If the one-egg constraint evolved in the ancestral stock, but ecological conditions later changed so that more pups per litter conferred a survival advantage (cf. [1120]), then embryo splitting might have been the only option available to increase litter size [839]. What were those ecological conditions?

Puzzle 34: What causes the various anatomical quirks of conjoined twins?

One out of every ~100,000 births results in conjoined twins [2440]. Most pairs are remarkably symmetrical. (Asymmetric exceptions [238,2434] include “parasitic twins” [678,1524].) They look as if someone got stuck while trying to enter a mirror, but, to continue the analogy, there is a perplexing spectrum in (1) the angle at which they entered the mirror and (2) the extent to which they traversed it before getting stuck. A similar spectrum of symmetric twinning occurs in other vertebrates as well [1345].

Conjoined twins are monozygotic, so they must start life as a single embryo that splits in two [2440]. It would thus seem reasonable a priori to suppose that the anomaly is due to incomplete splitting [1345,1616]. However, that idea fails to account for (1) why heads can meet at skewed angles, (2) why “Janus” faces can form perpendicular to the body axis, or (3) why certain contact sites predominate [2435]. These trends are more easily explained by secondary fusions after initial splitting and total separation [231].

Among the other quirks of conjoined twins (aside from symmetry) that are in need of mechanistic explanations are (1) the higher frequency of side-to-side fusion than any other angle of contact [2435], (2) the higher frequency of V-V vs. D-D fusions [1339], (3) the fusion of half-faces [1151,1345,2436,2411] in head-to-head conjoined twins [1152,1616], and (4) the diversion of optic nerves (to form inter-twin chiasms) inside the latter fused heads [2688].

Puzzle 35: What is the cellular basis for epithelial fusion?

Bilateral fusions are critical for the assembly of our gross anatomy. They close the neural tube along our dorsal midline [514] and the abdominal wall along our ventral midline [278]. If mistakes occur during these mergers, then the clinical consequences are profound—namely, anencephaly [1433], spina bifida [685,1965], and umbilical herniations [91,1644].

From the standpoint of development, epithelial fusions raise deep mechanistic issues [576]. Topologically, the process is as eerie as clapping your hands together, only to discover that they will not come apart because your palms have vanished, leaving only a seamless sheath of skin around both hands. How two epithelia become one is best understood for the palate [1017,1855], where both cell migration and cell death are known to mediate fusion in response to the growth factor TGFß3 [29].

We are still left to wonder how the “Merge with me?” overtures and the “O.K., let’s do it!” responses evolved in the various organs where fusions occur. Aside from the neural tube, abdominal wall, and palate, other (smaller-scale) examples of bilateral fusion include our heart [2453], sternum [1560], uterus [91], urethra [2332], and much of our face [485].

Puzzle 36: What causes the loss of tissue at symmetric fusion planes?

If conjoined twins arise when identical twin embryos bump into one another, then why can’t collisions happen at any angle and produce fusion geometries that are markedly asymmetric? Collisions may indeed be random at first [1036], but any resulting disparities are evidently eliminated by death of the mismatched tissues or by the stunting of their growth [1616,1771]. The two-headed animals that grab the media spotlight from time to time presumably develop in this way (cf. [348,911,1095]). However, this presumption is only inferential. Expermental evidence is sorely lacking.

Frontal fusions arise via contact of oral membranes [1152,1616]. Amazingly, the fused head can have two faces that point sideways [1151,2436]. Structures often vanish at the fusion plane (small inset) [1616], resulting in cyclopia or fused ears on one side of the head or the other [1345,2411].

Side-to-side fusions typically culminate in a single pair of legs, with loss of tissue at the fusion plane. The right twin often exhibits situs inversus due to trans-twin leakage of signals that stifle the nodal gene [1529,2884]. Walking is difficult due to separate brain control [678,1751]. This Y-shaped anatomy (2 heads, 1 rear) can be artificially induced at high frequency (80%) in frogs by centrifuging fertilized eggs [214].

A comparable phenomenon has been studied in Drosophila, where randomly induced cell death causes the fly’s two forelegs to fuse. The resulting “mermaid” leg is missing varying amounts of tissue at the midline, but it is invariably mirror-symmetric [2069]. Interestingly, this tendency for global (organ) symmetry to emerge from local (cellular) disparities may also explain Bateson’s Rule [1135], which governs the symmetry planes of branched (≈ conjoined) appendages [304]. How do cells iron out those initial disparities?

Puzzle 37: How do our left and right legs grow to the very same length?

Certain aspects of our anatomy are astounding from an engineering standpoint. Our two legs, for example, attain equal lengths despite growing independently for decades [106] even though any slight asymmetry at the outset should be amplified greatly by the end [747,2844]. The precision of symmetric growth is as enigmatic mechanistically as it is elegant morphologically [498,1044,1497], though recent evo-devo findings have begun to demystify how we achieve this feat [537,668,2682].

Puzzle 38: What is the role of FGF in establishing jaw polarity?

The plane of symmetry between our upper and lower jaw appears to have been superimposed on a prior architecture (dorsal/ventral gill arches) [2364], much like a new plane of symmetry was shoehorned into arthropod legs to create the mirror-image (upper/lower) surfaces of insect wings [1448].

If the latter case is any guide, then the new symmetry might depend on a diffusible signal [1137] from the boundary between the two halves [2780] that dictates inverted polarities [2898] (cf. also the boundary between the mes- and met-encephalon [2488]). Indeed, FGF8 appears to serve this very role [761]. It diffuses from the maxillary/mandibular boundary [477,1561], but whether it tells each half of the jaw which way to face is not known.

V. Asymmetry:

Puzzle 39: Why is the aorta on the left in mammals but on the right in birds?

The development of our aorta involves the unilateral destruction of an aortic arch [2876]. Interestingly, we make an aorta from our left 4th arch and erase our right one [2311], while birds do the opposite [1838]. Since mammals and birds both evolved from reptiles, whose arches stay nearly symmetrical, we are left to ponder why the left was commandeered in one clade and the right in another [1488,1838]?

Puzzle 40: Why (and how) did asymmetry evolve in protostomes?

How asymmetry arose in protostomes is unclear [528,1216,2255,2429], nor is it known how often they recruited different signaling pathways for different asymmetric organs [13,1625,2430,2571].

Although nodal and its effector Pitx2 were thought to only operate in deuterostomes, a recent report reveals that these genes are also expressed on opposite sides of oppositely coiling snails! See Grande, C. and Patel, N.H. (2009). Nodal signalling is involved in left-right asymmetry in snails. Nature 457, 1007-1011.

Puzzle 41: What is the adaptive value of heart asymmetry?

The human heart arises via pretzel-like contortions of an initially symmetric tube [7,308]. The evolutionary issue is whether these choreographed deviations from the midline are adaptive (physical constraints of organ packing?) [1976,1980,2678] or merely arbitrary (side-effects of morphogenesis?) [1838].

Puzzle 42: What is the inductive signal that causes fly testes to coil?

Fly testes change from oval to spiral shape (with an invariant chirality) after contact with their respective vasa deferentia [2467,2468], but the inductive signals emitted by the vasa have not yet been identified.

Puzzle 43: Why do both testes of D. melanogaster coil sinistrally?

Fly testes form spirals. At first glance, the spirals look as if they might be symmetric, and in most fly species they are [2467]. However, in D. melanogaster they both coil around their vas deferens like the fingers of a left hand curl around its thumb [1745,2467]. In other words, they are not mirror images! Rather, they’re like a man with two left hands. How is this asymmetry generated? Why does it exist? Another quirk in D. melanogaster is a 360° rotation of the genitals during development, which twists the ejaculatory duct sinistrally around the rectum [919,1216]. Why does this happen?

Puzzle 44: Why do both tusks of 2-tusk narwhals have sinistral grooves?

Male narwhals have a tusk that is up to 2.6 m long (possibly used for jousting [2399]). The tusk is a modified upper left incisor [1876]. It is cone-shaped with helical grooves that always twist sinistrally [1388]. Rarely, the right incisor grows out to form a second tusk, and in every such case its grooves are also sinistral [2146]—a situation as strange as a man with two left hands. How does this bizarre asymmetry develop?

Puzzle 45: What is the cellular basis for snail shell chirality?

The only external asymmetry that has been investigated genetically to any useful extent is helical coiling in snail shells [97,1156,2297]. Dextral shells are the norm and sinistral ones are rare in most gastropod species [976,984], though a few whole species are sinistral [97,2196,2297]. Why is dextral coiling so prevalent [538,850,1305,2684]? Mating between mirror-image snails can occur [102,1210], and chirality turns out to be due to a single gene that acts before fertilization [586,1031,2196,2520,2658]! How does the maternal gene product (a cytoplasmic RNA or protein) bias the spiral cleavage of the zygote [813,1085,1156,2363,2741]?

Although nodal and its effector Pitx2 had been thought to only operate in deuterostomes, a recent report reveals that these genes are also expressed on opposite sides of oppositely coiling snails! See Grande, C. and Patel, N.H. (2009). Nodal signalling is involved in left-right asymmetry in snails. Nature 457, 1007-1011.

Puzzle 46: Do Schwann cells wrap all axons with the same chirality?

Our sweat glands coil dextrally regardless of whether they’re on the left or the right side of our body [510,2024]. A similar question can be asked about Schwann cells: do they wrap axons with a uniform chirality, regardless of which side of the body they’re on?

Puzzle 47: How do single-celled ciliates inherit their chirality?

Single-celled protists called “ciliates” use cilia for propulsion and put more of these “oars” into the water than a Roman trireme. For them, ciliary chirality poses enough puzzles of polarity, asymmetry, and geometry to fill several books [851,1861], and Joseph Frankel has written extensively about their looking-glass world [804-806]. For a review see Frankel, J. (2008). What do genetic mutations tell us about the structural patterning of a complex single-celled organism? Eukaryotic Cell 7, 1617-1639.

Puzzle 48: Which morphogen is wafted to our left side to activate nodal?

Mammalian embryos have a lawn of ~250 cilia whirring away at our midline [1164,2376]. The “Gulf Stream” that they create wafts certain unknown signaling molecules from right to left [1846,2548]. The rising intensity of those signals on the left side eventually turns ON a master gene called nodal [772,2352], which forces tissues on our left side to develop differently from the default condition (nodal = OFF) on our right side [1067,2136]. Nodal exerts its control via subordinate genes termed the “nodal circuit”[1620]. The terminal effector of that circuit is the transcription factor Pitx2, which directly turns downstream genes ON or OFF during the development of the heart [848] and other organs [554,1254,1559,2377] so as to produce the overt asymmetries of our viscera.

The overall causal chain is: (1) the chirality of our amino acids determines (2) the handedness of our proteins, which dictates (3) the clockwise spin of our midline cilia, which forces (4) the leftward movement of the overlying fluid, which activates (5) the transcription of nodal on the left side of our body, which leads to (6) the delegation of Pitx2, which finally controls (7) the asymmetric development of our viscera. This basic mechanism must be ancient since lawns of rotating cilia are also seen in birds, amphibians, and fish. However, it turns out that the chain of command differs among classes because evolution has added symmetry-breaking gadgets to the nodal circuit in series or in parallel since the classes diverged from one another [1959,1976]. The developmental sequence of the contraptions in each class is now being ascertained [1528], after which it should be possible to deduce the evolutionary order in which they arose [1574].

Candidates for the morphogen that is wafted to our left side, which is not yet known [1898,2569], include Sonic hedgehog [1966] and retinoic acid [2754].

Puzzle 49: Does the configuration of our gut really need to be so orderly?

Why did the gut of vertebrates become asymmetric? Part of the answer concerns size [242,955,1238]. The ocean is a fish-eat-fish world, and the bigger you are the more likely you are to survive (all other factors being equal). However, a body with a straight gut cannot grow beyond a certain size because the supply of nutrients (proportional to gut area) cannot keep up with the increasing demand (proportional to body volume) [511,1049]. In order for Urbilaterians (~0.2 mm long) to reach our size (~2 m tall), their gut had to grow much longer than their body [512,2806]. Our intestines are ~10 times the length of our torso. To pack such a “fire hose” into the body cavity requires that it deviate from the midline. Historically, any mutations that let the gut become asymmetric (e.g., those which created a coelom [462,1762]) would have proven profitable to their bearers because they would have permitted a virtually unlimited increase in size.

The deeper mystery is why our gut doesn’t just coil haphazardly to fill the available space. Instead, it undergoes a ritualized sequence of twists and turns [2311] that normally culminates in a clockwise colon (ascending right, crossing, and descending left) [16,582].

The use of nodal to fold the gut neatly as it lengthened evolutionarily may have been a safer alternative than letting it meander randomly, since haphazard looping can tie the gut in knots, leading to lethal sigmoid volvulus [459,1801]. Here we may have a partial answer to the riddle of why our viscera look so orderly despite being so asymmetric.

Puzzle 50: Why do homeotically transformed ribs tend to be asymmetric?

In a man’s skeleton that was donated ca 1882 to Amsterdam’s Vrolik Museum [1938] the most anterior vertebra of every region was shifted in identity by one unit. Similar shifts are seen in Gbx2, Gdf11, or Cdx mutant mice [371,834] or in embryos treated with RA (retinoic acid) [499,1368]. “Frame shifting” of identity relative to merism also occurs in digits [2718]. Asymmetries in rib formation are surprisingly common in human axial homeoses [845]. They might be due to insufficient RA [276].

Puzzle 51: How did the chordate mouth migrate ventrally after inversion?

One abiding mystery of the D-V axis inversion that our chordate ancestors underwent ~500 MYA is how the mouth later migrated down the head to its current ventral location [456,1463,1464].

Interestingly, the protochordate amphioxus is asymmetric as a larva but symmetric as an adult [2250]. As a tadpole, its mouth is on the left, and it has a single row of gill slits on the right [1990]. During metamorphosis these asymmetries disappear [632]. What makes this transformation even more intriguing is its relevance to one of the greatest mysteries in evo-devo—namely, how the body flipped upside-down near the base of the chordate clade [1463]. It is conceivable that the migration of the mouth during amphioxus metamorphosis recapitulates the inversion itself [1464]. The recent sequencing of the amphioxus genome [874,2100] gives us a new playground where we can explore this aspect of our ancient past.

VI. Digits:

Puzzle 52: Why is the number of digits so constant in tetrapods?

Why do tetrapods typically make five fingers [148,844,1848,2362]—sometimes fewer [268,2345,2504,2805] but never more? What is the genetic nature of this constraint?

Puzzle 53: Why is the digit initiation sequence so constant in vertebrates?

Embryologically, the thumb is the last digit to form in all land vertebrates except salamanders, where, strangely, it arises first [825]! How is possible for a pattern that is so strongly conserved in evolution (i.e., the 5-digit tetrapod hand/foot) to manifest opposite sequences of element initiation in different clades?

Puzzle 54: How is our thumb encoded in our genome?

The thumb’s unique identity allows it to develop differently (“dissociate”) from the other digits. Amazingly, the thumb’s code (d13 ON; d12 OFF) dates back ~400 million years to our fish ancestors [581], who had no bona fide hands or feet [1302]. How the code works is unclear [379,841,2825,2904], but we do know that it must act indirectly via genes that affect anatomy directly.

Models of digit patterning are compared in ref. [2547], and models of digit identity are critiqued in refs. [1797,2534]. The leading model assumes that thumbness is specified digitally (no pun intended) by a combinatorial code of ON/OFF states (“d13 ON; d10-d12 OFF” [1406,1557]), but an analog mechanism is also possible [641,1794]—i.e., thumbness could be dictated by a lower anterior dose of total Hox protein. Thumb identity may be the universal default state [439,2216] because removal of Shh converts all digits to thumbs [1557]. However, removing Gli3—an Shh transducer that binds Hox proteins [432]—evokes odd shapes and more digits [2742].

Another debate concerns the role of BMP paralogs. Lowering the dose of several BMP genes fails to show the threshold effects predicted by the leading model [131], but the basic idea is still viable since various other Noggin-sensitive (BMP, etc.) genes could be playing a morphogen role [2534]. Finally, there is the nagging question how Shh is transported and interpreted [644,645].

Puzzle 55: How does each finger and toe acquire a plane of symmetry?

Of all the bones in the human body, our fingers and toes are unique insofar as each of them has its own internal plane of symmetry that is separate from the body midline [1188,2489]. All or our other symmetric bones—e.g., vertebrae—straddle the midline. This feature may be a legacy of the genetic gadgetry that was co-opted to create digits in our amphibian ancestors [2723,2742].

VII. Teeth:

Puzzle 56: How do baby molars become adult premolars?

As children, we have no premolars. Our two juvenile molars are replaced by adult premolars [1866]. How is such a homeotic shift in identity implemented genetically?

Puzzle 57: What is the identity code for our canines and premolars?

Our teeth come from a meristic series of similar tooth buds [1917,2643]. We have about as many types of teeth (four) as vertebrae (five): incisors, canines, premolars, and molars [1300]. The individuation process is similar too [1696]. Opposing gradients (of BMP4 and FGF8) establish zones of expression for homeobox genes (albeit outside the Hox complexes per se [2346]) [1561,1770], and the resulting code dictates different identities [476,2021].

Incisors arise in response to high levels of BMP4 (via Msx1 and Msx2), while molars are elicited by high levels of FGF8 (via Barx1 and Dlx2) [2643]. Consistent with this logic, incisors can be transformed into molars by blocking BMP4, which results in activation of Barx1 [2644]. These data come from mice, but humans develop similarly [1553]. Indeed, a girl with molars replacing incisors was seen at a dental clinic in Minnesota ca1990. Codes for our premolars and canines remain to be determined because the key studies of tooth development were done in mice, which lack these two types of teeth [1346,1695,2488].

Puzzle 58: How are the cusp patterns of our teeth encoded genetically?

How do each of our premolars and molars make a certain number of cusps during development [1299,1346,1770]? How did reptile teeth, which are peg-shaped [2027], evolve the ability to make more than one cusp per tooth in our mammal ancestors [333]? Why did it take them more than 100 million years to discover how to do so [1357]?

Puzzle 59: How is tooth number encoded genetically?

How does each mammal species make a characteristic number of teeth [11,1346,1770]? This puzzle is part of the larger mystery of how the genome manages to enforce a fixed number of elements within meristic patterns in general [151,383,1135,1724,2777]. As for how our single row of teeth evolved from multiple rows (e.g., sharks), recent data implicates the transcription factor Odd-skipped related-2. See Zhang, Z., Lan, Y., Chai, Y., and Jiang, R. (2009). Antagonistic actions of Msx1 and Osr2 pattern mammalian teeth into a single row. Science 323, 1232-1234.

Puzzle 60: Why do certain teeth (e.g. rodent incisors) never stop growing?

One knotty question that evo-devo is starting to address is why certain teeth grow continuously (e.g., rodent incisors) while most do not [1769,2590].

Puzzle 61: What is the genetic basis for jaw patterning?

FGF8 diffuses from the maxillary/mandibular boundary [477,761,1561], but whether it tells each half of the jaw which way to face is not known.

Perpendicular to the FGF8 gradients are BMP4 gradients [2172,2364] that appear to enforce a “Dlx code” for upper vs. lower jaw [636,1292]: the homeobox genes Dlx5 and Dlx6 are ON in the lower jaw but OFF in the upper jaw, presumably because the upper jaw’s BMP4 gradient never attains the threshold needed to turn them ON. If both genes are inactivated, then the lower jaw develops like an upper jaw that is upside-down [189,633]. The inversion confirms the notion of a “zone of polarizing activity” (ZPA [2193]) at the boundary [635].

Given their unique Dlx identities, the upper and lower jaw have been able to acquire various differences in their teeth [2486]. In humans, the disparities include (1) a slight shift (“occlusional offset” [1573,2154]) that allows a tight fit between our upper and lower teeth [50,2768], (2) bigger upper incisors, and (3) extra cusps on our lower molars [1004]. More extreme upper/lower inequalities are seen in a few mammal jaws (e.g., elephants) [2178] and many bird beaks [1864]—including those of the Galapagos finches [8,2770].

A revealing study in this regard was done on dog muzzles by Charles Stockard, who found overbites and underbites in various hybrids [2490]. These mismatches probably stem from disparate downstream genes (linked to the Dlx code by artificial selection [792]) for jaw shape in different breeds [2406,2556]. Indeed, hybridization wreaks havoc on co-adapted gene complexes in general [11], which may be why discrete species (and their quirky isolating mechanisms) evolved on earth in the first place—i.e., to protect the functional integrity of finely tuned genomes [531]. Underbites are also seen in some dolphin species [1669]. A few genes are known to affect upper vs. lower teeth differently within the same shape class [4,2486].

Despite the salience of these upper/lower discrepancies, they are the exception, not the rule in terms of evolutionary trends [635]. The two dentitions have remained similar in most mammal lineages even as dental formulae changed [2027]. Is this coevolution of upper and lower teeth due to some sort of yoke that makes it easier to alter both jaws at once [1400]? Or does natural selection enforce similarity by weeding out any large deviations with no help at all from any covariation constraints? Unfortunately, we don’t yet know how the “jaw module” is wired in the mammalian genome [635], nor what genes therein have been captured by the Dlx5/6 selectors in different lineages over time [1936,2780].

VIII. Limbs:

Puzzle 62: Why can salamanders regenerate their limbs, but we can’t?

Salamanders regenerate their limbs completely after amputation [864]. We obviously do not. How wonderful it would be if we had their talent! It’s not that we can’t regenerate any lost parts [949,2538]. Our liver, for instance, has remarkable powers of regeneration [55,2577]. Why not our arms or legs [847]? After all, amphibians were our ancestors. Have mammals forgotten the recipe [1072]? Has the mechanism atrophied (like our appendix) from disuse?

Some of the factors that may be preventing regeneration in mammals (vs. salamanders) are: (1) an inability of our muscle cells to re-enter S-phase when prodded by the thrombin pathway [285,1910], (2) an inability of our neurons to regrow their axons fast enough to support blastema initiation [434,1449], and (3) a silencing of our Hoxc6 gene at the completion of limb development [284]. The greatest impediment, though, appears to be the tendency of our skin to scar [1829], rather than to heal with a rejuvenated wound epidermis [2492].

Herein lies the greatest medical relevance of the entire evo-devo field [285,2268,2492]. Researchers are actively exploring every conceivable means of coaxing mammalian appendages to regrow [206,2566,2881]. Someday they will crack the cryptic passwords that salamanders use to re-boot their limb-growth software. When that day comes—and it may come sooner rather than later—amputees will be able to discard their prosthetics and grow back their own arms or legs [1187,1829]. Nor is limb regeneration the only riddle that might eventually yield to the withering glare the evo-devo searchlight [291,1185]. Efforts are also underway to decipher how planaria regenerate their central nervous system [404] in the hopes of enabling our spinal cord to do the same [160,421,436,494]. If we could solve the mystery of human CNS regeneration, then paraplegics might someday abandon their wheelchairs and walk again [937,2098].

As farfetched as such cures may have sounded a decade ago, they have already left the world of science fiction and entered the realm of science [1419,2189,2603]. For the next generation of researchers the clinic looms just over the horizon [105,1013,1243,2491,2909], and the first rays of hope are lighting their way [1527]. Our keenest students are already charting their careers to seek those goals.

Puzzle 63: How did evolution reshape our foot for bipedal walking?

Our feet differ from those of other primates insofar as our toes are shorter and our big toe is not opposable [1813,2317,2539], plus our foot is arched [1366]. Hominins may have begun to walk bipedally on branches as orangutans still do [1905,2602], but the selective pressures on our feet certainly intensified once we came down from the trees onto the grasslands [10,2530]. The biomechanical forces that came into play are well understood [54,2075], and the fossil trail of the foot’s history is well documented [1088,2737]. What remains unclear, however, is how the reshaping was implemented developmentally.

Puzzle 64: How did evolution reshape our pelvis for bipedal walking?

One obvious distinction between men and women is the width of our hips [525,2551]. The grace of our bipedal gait (vs. the waddling of chimps) was achieved by reducing hip breadth [1579,1581], but the female pelvis can’t get too narrow because the baby’s head has to traverse its opening [5,1496,2320]. These opposing pressures eventually reached a compromise [2081,2626]: the pelvis is proportionally wider in women [525], but just enough so that the baby’s head can barely squeeze through [10,502,1807]. The genetic basis for this change remains obscure, though, again, it must have been simple since some pelvic dimorphism is already evident in the earliest hominins [2048,2220,2553].

Puzzle 65: How did evolution reshape our spine for bipedal walking?

One dimorphism of humans that is not seen in our ape relatives concerns our vertebrae. Our backbone’s angle relative to Earth’s gravitational field changed radically when our hominin ancestors became bipedal ~7 MYA. In response to the new weight distribution, our spine changed from a bridge-like arch to an S-shaped curve [2456]. Pregnant females suffered an added burden from the weight of the fetus in their womb, and a dimorphism eventually evolved that lets them adjust their center of gravity by lumbar lordosis (bending of the lower back): women have smaller dorsal wedging angles in their L1-L4 vertebrae than do men, and their L3 in particular has a manifestly different geometry (wedge vs. block) [2795].

How was the reshaping achieved genetically? We don’t yet know. Presumably, it entailed a new link between the master gene for femaleness and a target gene(s) for vertebral morphogenesis. Whether the link was forged via a novel enhancer site remains to be determined [2164], as does the role of gonadal hormones as intermediate effectors [686,1737,2160,2342]. Whatever its nature, the link must have been facile since it arose fairly soon after bipedalism made lordosis useful: it is already evident in the spines of Australopithecus africanus [2795].

Puzzle 66: Why are there phyletic differences in the roles of Tbx4 and 5?

In chick embryos, forelimb vs. hindlimb identities can be switched by misexpressing Pitx1 [1567], Tbx4, or Tbx5 [2205,2559]. In mice Pitx1 is also effective [623], but Tbx4 and Tbx5 are not [1113,1765,1844]. This phyletic difference is not yet understood [1208,1728,2503].

Puzzle 67: What is the role of the Hox code in arm vs. leg identity?

Aside from Pitx1, Tbx4, and Tbx5, Hox genes are also involved in assigning identities along the body axis [318,1902,2258], though the exact code is unclear [623]. Many other genes are expressed differently in arm vs. leg [1245,2382], some of which are undoubtedly downstream effectors.

IX. Sex:

Puzzle 68: What is the adaptive significance of female orgasm?

A passionately disputed issue in evo-devo is whether human female orgasm is merely a spandrel (i.e., an incidental side effect) of natural selection for male orgasm. See refs. [972,1209,1318,1609,2734]. (Don’t try debating this with your spouse!)

Puzzle 69: What factors led to the Adam's apple of adult human males?

Evolution lowered the larynx in species like lions [2782] and deer [783,2826] where males roar to impress females and outdo rivals [781] (cf. trumpeter swans [2798,2913]). It is possible that the hominin larynx likewise descended—at least initially—to exaggerate the body size of the suitor, with the ability to speak coming along a lucky side-benefit [779,783,1183,1546,1887,2607]. The human “Adam’s apple” attests to the role that vocalization must have played in courtship [889]. Men’s voices are an octave lower than women’s, which women may have found sexy [889,890]. We don’t yet know how this trait arose in our lineage, but a similar dimorphism has been analyzed in frogs [2608] at both the genetic [400] and hormonal [1647,2609] levels.

Puzzle 70: Why is sex determination so needlessly complex genetically?

The puzzle that dimorphic organs pose is how their genes “decide” to turn ON or OFF [21,65,244,327] so that they make an anatomy appropriate to their gender. Looking at the situation from the gene’s perspective, the deliberation that governs its decision can be phrased as a simple first-person imperative:

IF (Input #1) “I am in organ O” {mediated by a region-specific selector gene(s)}
AND (Input #2) “I am of sex S” {mediated by a gender-specific selector gene(s)},
THEN (Output) “I will turn ON (or OFF).”

An example from Homo sapiens might be a growth gene that behaves as follows so that the pelvis grows bigger in females than in males.

IF (Input #1) “I am in the pelvis”
AND (Input #2) “I am female”,
THEN (Output) “I will turn ON (at higher volume, faster rate, or longer duration).”

Presumably, both inputs (site and sex) involve transcription factors that are jockeying for binding sites in the cis-regulatory region adjacent to each target gene [1365,1518,2444].

The best evidence for a deep conservation of reproductive development all the way back to Urbilateria [1039,2814] comes from a different gene called doublesex (dsx) in flies [121,526], mab3 in nematodes [2880], and Dmrt1 (Doublesex and mab3-related transcription factor 1) in vertebrates [1200,1377]. The gene encodes a transcription factor belonging to the zinc-finger class of DNA-binding proteins [725,2911].

Mutations in doublesex can transform male and female flies into sterile “hermaphrodites,” whose reproductive tract has both types of plumbing side-by-side in jumbled disarray [721]. In normal fly larvae of both genders, the genital disc contains a pair of rudiments [2270]: the male and female genital primordia (MGP and FGP) [722,948]. Wild-type males repress FGP (leaving MGP to make tubes), while wild-type females repress MGP (leaving FGP to make tubes) [733,2269]—a strategy similar to that in humans except that the suppression involves prevention of rudiment growth in the first place rather than destruction of rudiments after growth.

Dmrt1 is expressed differently in males vs. females broadly within the vertebrate subphylum [1378,2899]. Of the 8 putative Dmrt paralogs in humans [1418], only Dmrt1 (on chromosome 9) seems critical [1957], though much remains to be learned about this whole circuit [221,270,2333]. Deletion of Dmrt1 in humans can transform XY individuals into hermaphrodites (or females) [1957]. Given that the dsx gene of insects regulates its target genes via differently spliced mRNA isoforms (dsxM and dsxF) [2240], it is natural to ask whether the same is true for vertebrates [96]. We do not yet know the answer [1200], but human Dmrt genes are subject to sex-specific splicing, and the exon structure of the testis isoform does match dsxM in flies [2899].

Given what we know about master genes, one might expect the most conserved regulator to be at the top of the sex determination hierarchy, but doublesex is near the bottom in both flies and humans. Below are the core components of the fly [457,2064] vs. human [2056,2333] pathways as we currently understand them (Sxl = Sex-lethal; tra = transformer; M and F superscripts = functional male or female splicing isoforms) [1422]. The master gene for maleness (under our XY switch) is Sry (Sex-determining region [on the] Y), and it has been known for some time [2056,2756], while its presumptive counterpart (under the XX switch [221])—R-spo1 (R-spondin1)—was only identified recently [2808]. R-spo1 is a diffusible signal that may use Wnt-pathway transcription factors to control its target genes. Note that Xs and Ys of humans and flies are not homologous.

Flies: XY (default, no Sxl). ... dsxM —> male target genes.
XX —> SxlF —> traF —> dsxF —> female target genes.
Humans: XY —> Sry —> Sox9 —> Dmrt1 —> male target genes.
XX (default, no Sry). ... R-spondin1 —> female target genes.

Here then is the dilemma: over the eons it was the genes at the upper echelons that changed freely [993], while dsx/Dmrt1 endured for >500 MY [448,1411,1753]. For example, usage of Sxl is confined to the drosophilid lineage within dipterans [2338,2622], and usage of Sry is restricted to the therian lineage (placentals and marsupials) within mammals [2291], while Sox9 is more broadly conserved (among tetrapods [2403] and perhaps beyond [618]). Other mediators that the fly cascade has haphazardly enlisted (not shown) include the JAK/STAT [107,1101,2758,2901], Notch [2012], and proneural [466,2860] signaling pathways.

In terms of the analogy of target genes being caught by selector-gene fishermen, this fluidity of upstream factors forces us to envision the fisherman himself as a fish who could be hooked (from above) by an assortment of lines over time, while keeping all of the fish that he’s caught up to now on his own line(s) [2064,2506,2814]. Such “retrograde” retooling of developmental pathways (vs. anterograde additions [2815] or intercalary insertions [878]) is not unique to sex determination [1603,1933,2851,2855] (e.g., upstream triggers for the nodal circuit [108,1528]), but it is more prevalent there than anywhere else [897,993,2238,2632]. Why?

The answer may be that dsx (and Dmrt1?) enforces its diverse dimorphisms by managing multiple target genes in various organ-specific developmental pathways, whereas each of its upstream regulators mainly controls a single subordinate gene [2899]. In other words, the cascade is a simple chain down to dsx, but then it fans out extensively [84,928]. Extricating such a pleiotropic gene from the cobweb of interactions that it has acquired (by random fishing) and replacing it with a substitute can’t be achieved by piecemeal tinkering. In the extent of its entrenchment, dsx resembles the genes of the Hox complex, which have likewise insinuated themselves into the inner workings of a host of embryonic pathways to reconfigure the shapes of serially homologous organs [726] (e.g., the revamping of wing anatomy by Ubx in order to craft a haltere [2474,2763]).

Puzzle 71: What is the genetic basis for our sexual dimorphisms?

Men and women differ in various features. One dimorphism of humans that is not seen in our ape relatives concerns our vertebrae. Women have smaller dorsal wedging angles in their L1-L4 vertebrae than do men, and their L3 in particular has a manifestly different geometry (wedge vs. block) [2795]. How is this dimorphism genetically? We don’t yet know (cf. Puzzle 65).

A more obvious distinction between men and women is the width of our hips [525,2551]. The grace of our bipedal gait (vs. the waddling of chimps) was achieved by reducing hip breadth [1579,1581], but the female pelvis can’t get too narrow because the baby’s head has to traverse its opening [5,1496,2320]. These opposing pressures eventually reached a compromise [2081,2626]: the pelvis is proportionally wider in women [525], but just enough so that the baby’s head can barely squeeze through [10,502,1807]. The genetic basis for this change remains obscure, though it must have been simple since some pelvic dimorphism is already evident in the earliest hominins [2048,2220,2553]. Separate mutations must have targeted fat deposits to female hips [888]—a site unique among primates [1580]—to accentuate the hip-to-waist ratio that men find attractive [2236,2537].

Our most glaring dimorphism, of course, is our adult body size [889,2208]. Males are typically bigger than females in all apes except gibbons [1513,1615], and the same is true for mammals in general [922,1491,2114]. Various social factors may have fostered sexual selection for unequal size in our lineage (e.g., mating systems, suitor rivalry, or division of labor) [1580,2045,2381], but none has yet been tested rigorously [72,219,809]. As for the molecular causes of size disparity [116,537] the chief suspects are growth hormone [1542] and IGF-1 (Insulin-like Growth Factor 1) [2221]. IGF-1 played a major role in the vast range of body size among dog breeds [2533].

Much more is known about the genetics of dimorphisms in flies than in humans [1424,2830] for the simple reason that experimenting with flies is devoid of ethical restrictions. It is likely that at least some of what we learn in that model system will eventually apply to us [537]. The traits that have been analyzed most intensely are abdominal pigmentation [1294,1423,2093,2476,2837], wing spot localization [266,939,2094], wing size inequality [1202], and sex comb formation [1425,2567]. In three of the four cases (wing size is the exception), the dimorphisms have been traced to changes in cis-regulatory enhancers [537].

A major difference between humans and flies is that we have circulating gonadal hormones [221,2160], while they do not [2469] (but cf. [2351]). Hence, it may have been easier for our forebears to make an organ dimorphic by just expressing androgen or estrogen receptors on its cell surfaces [235,400,2342] instead of having to fiddle with individual target genes [21,65,327,2299].

Puzzle 72: How different are the brains of men vs. women?

How different are men and women mentally, and to what extent are the disparities hard-wired genetically [161]? We don’t yet know.

In mice, surprisingly, the brain’s default state appears to be Male, not Female (as has been assumed for the body as a whole [221]) [2442]. In flies, much progress has been made in dissecting courtship behavior [197,2175,2375,2891] and male combat [660], and those findings might have some relevance to us, given that (1) we share dsx/Dmrt as a regulator [2273], and (2) we (like flies) express some direct, cell-autonomous (hormone-independent) effects of our XX vs. XY constitution within our brains [649] (cf. marsupials [2157]), notwithstanding our reliance on gonadal hormones [221].

The virtue of studying fly (vs. human) neurobiology is that some clever experiments can be done. For example, when dimorphic neurons were placed under the control of a light trigger, the researchers could turn the male’s courtship song ON or OFF with the flip of a switch and thus localize the generator circuit within the brain [470,673,2892]!

Puzzle 73: Why are mate-choice fads so ephemeral?

Believe it or not, peahens recently changed their mind for some unknown reason [2804]. They no longer prefer peacocks with fancier fans [2555]! Extravagant tail feathers were just a passing fad? Apparently so! Another example of a faded fad is the tail of the swordtail fish: females of at least one species (Xiphophorus birchmanni) have lost that lovin’ feeling for the male’s glorious sword [2846]. O cruel fate! O fickle genes! Balding men know all-too-well how crestfallen these male fish and fowl must feel. Baldness may have once been as sexy as beards still are. Early female hominins apparently found the bald spot attractive, but at some point their descendants ceased to be turned on by it [1750]. Unfortunately, the rate at which baldness genes are being purged from our species is apparently much lower than the rate at which mutations are changing the criteria for sex appeal in the female brain [965] (cf. [161,2151]).

The nagging questions are: (1) how do female brains get rewired genetically, and (2) why on earth should their He’s-so-hot! arousal circuitry be so dependent on whims?

Puzzle 74: What is the genetic basis for hair patterning?

How did humans lose so much of our fur to become the only naked ape? We don’t yet know the genetic changes [2009], but we do have a few clues based on extra-hair syndromes [111,154,170,1618,1783,2787]. The most dramatic anomaly is Ambras Syndrome [153,2550], where affected individuals—men and women—grow luxuriant hair all over their face, including the forehead and nose [153,765]. Only ~50 cases have been reported since the Middle Ages [153]. The earliest known instance (Petrus Gonsalvus) was documented in a portrait ca 1582 [1524,2134,2787]. A few such people were displayed as curiosities at carnivals in the 1800s [239,1452] under stage names like the Dog-faced Boy and the Lion-faced Man [678].

Ambras Syndrome appears atavistic [239,765,2074], but it can’t be because apes don’t have furry noses [2683]! Other hypertrichosis mutations cause ectopic hair in equally odd places: on the ears [1505,2471], the neck [260], the elbows [171,2058], or on the palms and soles [1262]. The latter two locations also defy an atavistic interpretation because no primate has fur there. Since there is no rhyme or reason to these assorted spots, it’s hard to know what to make of them.

Another important clue has come from an unexpected corner of the research world. The genetic basis of hairlessness in the Mexican hairless dog (revered by the Aztecs for 3700 years) has now been ascertained. The trait is due to a frameshift insertion (a 7-base-pair duplication in exon 1) in the Foxi3 gene [679], which belongs to the same Forkhead box family as the Foxn1 gene that controls hair formation all over the body [1327,2299]. Interestingly, this breed still has plenty of hair on its head (as we do), as well as on its feet and tail. Why this null mutation should leave hair on these sundry (distal?) parts remains to be determined. Several key questions remain about how our genome controls hair distribution:

1. What selector genes (or cis-enhancers?) target hair growth to the sites where it develops (before and after puberty) [2093]? For example, what is the “area code” of the armpit? Flies use a single genetic locus—the Achaete-Scute Complex (AS-C)—to demarcate the territories where bristles can form throughout the body [1137], but no comparable headquarters has yet been found in our genome [1565]. Do we have one?

2. Do genes regulate hair formation (1) positively (with nakedness as a default) via region-specific activation or (2) negatively (with full hair cover as a default) via region-specific suppression? Flies dictate bristle sites positively by means of ~8 AS-C cis-enhancers [1137], which use “OR” logic, so that it was easy for evolution to add or delete them like plug-ins. What about humans?

3. What integration, if any, exists between (1) genes that dictate hair pattern in males and (2) genes that dictate mate choice in the female brain [71,117,1943]? Do they coevolve somehow [46,912] or take steps alternately and independently [1313,1719]?

X. Reproduction:

Puzzle 75: Why (and how) did separate sexes evolve in chordates?

Having testes and ovaries together in one body is a freakish aberration for humans [410,1773,1774], and the same is true for vertebrates in general [378,1788], but it is the norm for a third of all animal species excluding insects [262,585,1280]! Darwin was aware that bisexuality had been documented in some primitive chordates (viz., tunicates [1086,1516] but not amphioxus [1471]), and in Descent of Man he conjectured that our ancestors were hermaphrodites at that stage of our evolution [892]. If our distant ancestors were, at some point, hermaphrodites, then this situation begs the question of why (and how) our more recent ancestors split themselves into two separate bodies—a process at least as profound as any creation myth in any religion [349-352,979]

Puzzle 76: How long ago were our ancestors hermaphroditic?

Were all bilaterians originally hermaphrodites [2623]? Studies of germ cells in various phyla are consistent with the idea but are inconclusive [736,737], and comparative genomics has not reached a point where such a question can be meaningfully addressed [2814]. Thus, there is no definitive answer yet.

Puzzle 77: Why do germ cells migrate long distances to get to the gonad?

Bilaterian gonads typically import their germ cells [923,1393,1453] after the latter have migrated far (using conserved genetic circuitry [2640])—a strange odyssey that may have a simple evo-devo explanation [630,736,737,2623]. Nevertheless, the process seems wasteful.

Puzzle 78: Which is more evolvable: teat number or litter size?

Aristotle observed that teat number rises with litter size among mammal species {Parts of Animals: Book 4: Part 10: pages 688a32ff [137]} (cf. [904,2005,2006,2355]), and Darwin reasoned that our forebears must have had more than one pair of teats [564] (Vol. 1, p. 37). The riddle posed by our last remaining pair, therefore, is: which trait decreased faster in our evolutionary past—teat number [164,1376,2751] or litter size [396,652]?

Here we meet a deeper issue in evo-devo: how fast can any given trait evolve? Evolvability depends upon (1) the intensity of selection pressure acting on the trait and (2) the ease with which the circuits that control that trait can be altered within the genome [2790]. As litter size increases, selection pressure must increase on teat number to allow the added offspring to suckle [155], but what about when litter size decreases [1106,2318]? Are extra teats harmful? If not, why should their number ever decrease?

Puzzle 79: What is the genetic basis for lactation?

Mammary glands originated ~310 MYA in a branch of synapsid reptiles [1913] whose soft-shelled eggs were prone to desiccation [1914]. The initial solution to this dehydration problem, apparently, was for the mother to secrete sweat from apocrine skin glands on her belly during incubation. Later, the oozing fluid was augmented with antibacterial antibiotics [163,927]—setting the stage for hatchlings to partake of it as well. Once neonates came to depend on it, the proto-milk was supplemented with nutrients to form true milk [2707]. This plausible scenario is our best guess for how natural selection fostered the conversion of ventral sweat glands into fully functional mammary glands [2801].

Monotremes offer a likely snapshot of this early stage of mammary evolution [2649]. They lack nipples, and their young lick milk from the fur around the gland [69,293,2802]. Nipples arose later (~150 MYA [2232,2807]) in therian mammals [758]. They apparently evolved from hair follicles [1018,1148,1260,1740].

Because mammary glands and nipples were co-opted from gender-neutral structures (sweat glands and hair follicles respectively), the existence of nipples in males is most parsimoniously explained in terms of their having arisen by a mutation(s) that inserted them into both sexes. They have presumably persisted in male mammals because, although useless, they are relatively harmless [2774].

Are male nipples entirely useless? Is there no species where the father suckles the young? The only putative case is the Malaysian fruit bat Dyacopterus spadiceus [651]. In 1994, 10 mature males were reported to have yielded milk upon palpation, but the amount per male (~0.005 ml) was two orders of magnitude less than that obtained from one lactating female (0.35 ml) [801], so even here it is unlikely that the male nipples are functional [2003].

The worthlessness of extant male nipples begs one last question: were they ever useful? Maybe not. In the echidna (a monotreme) the mammae are as well developed in males as females [2649], but there are no reports of paternal lactation. Sex-limited mutations (“modifiers” [530]) must have arisen in therian clades to yield their anatomical (and functional) dimorphisms. Given how easily a male breast can be hormonally induced to lactate (i.e., gynecomastia [1613]), it is likely that those changes relied on hormonal effectors, but the nature of their control circuitry remains unknown.

Puzzle 80: How do non-primates avoid tubal pregnancies?

Eutherian viviparity evolved from reptilian oviparity by making embryos stick to the uterus before reaching the vagina [1610,2725], but they can also stick prematurely to the oviduct. In humans, such mistakes can lead to “tubal pregnancies” that doom the embryo and threaten the life of the mother [497]. Non-primates stop such embryos from developing [524], but we do not yet know how they manage to do this.

Puzzle 81: Why do kiwis have absurdly large eggs?

The egg of a kiwi (a small flightless bird) comprises 25% of the mother’s body weight [915,2549]—the equivalent of a 120-pound woman giving birth to a 30-pound baby! What possible factors in ratite evolution could have led to such a grotesque disparity? See refs. [342,343,969,2071].

Puzzle 82: Why do male flies have a (useless) muscle that females lack?

No one has any idea why male flies should have an abdominal muscle that females lack. The muscle, which is apparently useless, predates the genus (<62 MYA [103,837,1942]) and may go back even before the dawn of dipterans (<250 MYA [836]). Amazingly, this male-specific muscle can actually be coerced to develop from female cells [1385] since it is induced by a neighboring motor neuron whose gender is the deciding factor!

XI. Nervous system:

Puzzle 83: What is the neural basis for handedness preferences?

About 1 in 10 of us are left-handed [1715,1808,2022]—a frequency that has not changed in more than 10,000 years [522,750]. Curiously, handedness varies independently of visceral asymmetry: only about 10% of people with situs inversus are left-handed [1713]. Hence, these traits must be under separate genetic control [511,512,1630]. Indeed, the circuitry for L/R asymmetry appears to have multiple branchpoints within the genome [208,441,732,1528].

Our annoying inability to use both hands with equal grace can be blamed on our hemispheric lateralization [79,369,872]. Brain lateralities of some kind [79,261,574,872,920] must also dictate left/right behavioral preferences in fish [603,1203,1554,1714], frogs [207,1630,1845], snakes [2142], birds [659,933,1229,2666], rats [1474], humans [112,519,520,750,2879], pre-human hominins [856,914], and non-human primates [361,1068,1117,1517,1803]. Indeed, lateralized brain function has even been found in honeybees [1526]! It remains a mystery [369,790,1945,2001].

Puzzle 84: What is the neural basis for our ability to read?

How our brain processes the squiggles on a page such as this into meaningful concepts remains a riddle [2842]. Written language is only about 5000 years old [1909], so it must have emerged as a side effect of our intellect [2651], rather than being a driving force. Dyslexia, which impairs reading [2842], is a developmental disorder [838] that may stem from the misrouting of axons across the midline of the brain [1080].

Puzzle 85: What is the neural basis for our love for music?

“Musicophilia”, the title of a book by Oliver Sacks [2249], denotes our love for music [1532]. No one knows (1) why we like music so much [1314,2769,2900], nor (2) why it can move us to tears [1271], nor (3) why tastes vary to so greatly from person to person and culture to culture [1320,2037]. Monkeys manifest no preference for music whatsoever [1475,1701], so it is unclear how our affinity for it arose [867,1700,1744]. Could it just be a by-product of how our brains are wired for language [127,211,2037]? Darwin argued precisely the converse—that language actually evolved from music [292,780]!

Puzzle 86: What is the neural basis for autistic savantism?

People with Williams Syndrome have astonishing aptitudes for music [1337], despite being profoundly retarded mentally [1520,2249]! Other autistic savants exhibit equally astounding talents but are likewise deficient in ordinary thinking or social skills [913,2423,2624,2625,2667]. No one knows how savant brains have been rewired to achieve their feats so innately. Nor do we yet understand this see-saw phenomenon where one set of skills apparently must be sacrificed in order for another to flourish.

Puzzle 87: What is the neural basis for claw asymmetry in lobsters? In lobsters, either the left or the right claw can dominate. Which claw becomes the larger crusher (vs. the smaller cutter) is fixed before adulthood based on usage [987,1564]. Both claws become cutters if they are underused, but an unknown neural circuit prevents them both from becoming crushers [985,986]. How does it work?

Puzzle 88: How does our brain recognize faces so instantaneously?

The human face is considerably variable from person to person [1544], but what is more amazing is our knack for detecting such differences with just a glance [305,523,2598]. How do we do it? And why can’t we even begin to describe (in words) how we do it? Moreover, why should everyone look different in the first place? Is it an adaptation for identification or social bonding?

Finally, is facial variability itself a selectable trait [11,1141,2903]? Can the genome actually target variability to certain compartments [367,1180,2230]? These are all outstanding issues in evo-devo [366,1064,1989].

Puzzle 89: Why do Siamese cats have crossed eyes?

Siamese cats, like pure-white albinos in various mammal species [1026,1027,2465] (including humans [67,1507]) have crossed eyes [1323,2306]. This trait is attributable to an excessive crossing of retinal axons at the optic chiasm [1282,2306], and it is ultimately traceable to the enzyme tyrosinase, which is defective in these cats [1912]. Tyrosinase is known to catalyze melanin formation (hence their black tips [1911,2447,2614]), but why should a loss of pigment cause a misrouting of retinal axons [1284,1387,2464]? The etiology of this peculiar pleiotropy is currently under investigation [1022,1495,1833].

This example is emblematic of a whole host of other enigmatic etiologies that are still begging for mechanistic explanations. For adventurous young evo-devotees, such syndromes offer an endless cornucopia of brain teasers as enticing as trying to solve a Sherlock Holmes mystery or a Sunday New York Times crossword puzzle [414,1137].

Puzzle 90: How would a third eye’s axons behave at the optic chiasm?

If our pineal gland were to somehow regain its ability to form an eye, then we could actually reacquire a third eye [694]! That eye should work fine since surgically implanted third eyes can weave their nerves into existing pathways [504,1895]. How the axons of a median eye would act when they reach the chiasm, however, is unclear [752].

Puzzle 91: How are instincts hard-wired in our nervous system?

How did evolution wire men’s brains so that they perceive an hourglass shape (the female form) as alluring [161]. Can natural selection make any shape erotic? A pear? A Volkswagen? If so, how long does such reprogramming require [1811]?

A related question concerns the degree to which fear is innately linked to the perception of certain shapes or movements. For example, primates appear to have a hard-wired fear of snakes, which makes sense given our arboreal origins [1250], but it begs the deeper question of how evolution was able to program our brains thusly (cf. [2274])! What other old circuits still steer our behavior [1412,1555,2091,2254]? Are we just pawns of our genes [1543,1835,2169]?

Puzzle 92: How quickly is evolution able to rewire our brains?

The human female breast is proportionally larger than in any other primate [650], so this trait must have emerged in only the past few million years. How long did it take for evolution to rewire mens’ brains to be aroused by a pair of pendulous globes of fatty tissue with little pink knobs on them?

How much neural rewiring was needed to convert a quadrupedal simian to a bipedal hominin? Maybe not much [385,2730]: (1) a single mutation in an axon-guidance gene can make a mouse hop like a rabbit [1257] (cf. kangaroo rats [1351]), and (2) a human syndrome (due to one allele?) was found in Turkey where affected adults revert to walking on all fours [2564]! Can evolution rewire locomotion faster than arousal cues?

Puzzle 93: Why (and how) did intelligence evolve in hominins?

What drove our split from other ape species and ultimately led to our idiosyncratic intellect? According to the “East Side Story” [224,515,516], our odyssey began with the uplift of Africa’s rift valley, which left an abiding rain shadow over the eastern plateau (our home then) ~8 MYA. As the climate became drier, the forests dispersed into smaller fragments, and intervals between the remnants increased [2849]. Those apes who found themselves in this predicament were impelled to travel farther across open land in pursuit of food.

On such terrain, walking on two legs (or running [267]) is more efficient than ambling about on all fours [1522,2451]. If modern apes are any guide, then we can safely assume that the apes of that period already walked upright for brief errands. Under the newly arid conditions, natural selection would have rewarded any mutations that enhanced this ability to permit foraging on longer treks [50]. As we became more bipedal, the force vectors acting on our skeleton would have shifted [1534], and selection would have favored any structural adjustments that made striding more graceful [10,2286]. Among the adaptive responses that emerged as a result were a shifted foramen magnum [2278], a reshaped pelvis [1579], and an unrotated big toe [1366].

The adoption of bipedalism freed our hands for other uses [938], including (eventually) the grasping of weapons for hunting [224]. Meat added a rich new source of nutrition in the face of diminishing returns from the sparser fruits of the shrinking forests [1522]. The trend toward omnivory is well documented in the fossil skulls of our hominin ancestors [386,2014]: their heavy facial muscles (needed for chewing leaves) began to dwindle after bipedalism arose [11,224,1522]. According to the “Man the Hunter” scenario [1508], it was hunting that set the stage for the evolution of bigger brains [1120,2455]. Meat supplied the metabolic fuel for brain growth and brain maintenance, while cooperative hunting behavior established a selective environment that rewarded even the slightest increments in our ability to plan, socialize, and communicate with others [255].

Nature displays a consistent trend in brain:body ratios. Brain size increases at about the 2/3 power of body size among vertebrates [1295], though there is appreciable scatter, and mammals exhibit an exponent closer to 3/4 [59,1108]. The extent to which a species departs from the regression line of its affiliated group (in our case, primates) is called its “EQ”—the encephalization quotient [612,624,1109].

Hominins began veering away from our chimp cousins (brain volume: ~400 cc) with our greater reliance on bipedalism ~7 MYA [168,169,895,2847], but our EQ remained marginal (Australopithecus spp.: ~5 MYA; ~450 cc) [1383,2610] until ~2 MYA (Homo habilis: ~650 cc) [743,1476] when it began to rise fairly steadily [612,638,1706,2309] up to our current level (H. sapiens: ~0.2 MYA; ~1300 cc) [654,1222,1706,2749,2796]. Overall, our brain tripled in size while the chimp brain stayed the same [385,484,1324,2182].

During the phase of greatest brain expansion (2 MYA to now), our body size remained virtually constant [612]. Evidently, brain:body allometry is just a trend [1341]—not an inviolable constraint [1197,2309,2514]. Otherwise, our EQ could never have increased to the extent that it did [33,259]. The trend might reflect a sort of default state [770] that prevails in the absence of selective forces for (or against) particular brain functions [1442-1445].

When did we cross the Rubicon to a human level of self-awareness [1360,2309]? How quickly did it happen [1534]? Is the concept of a discrete brute/aesthete threshold justifiable [1197,1594]? No one knows [332,1295,2515].

The mere correlation between brain size and cognition does not prove causation [386,1650]. The idea that size alone was the critical factor in pushing us over the sentience threshold could be tested by a simple experiment, but one which is so utterly unethical that it should never be done: to create (by genetic engineering) a 1000-lb. chimp, whose brain should attain 1300 cc (based on extrapolating the primate trend line of brain:body size) [612], and see if the monster is smart as we are [2456].

Would such an über-chimp become the first Pan sapiens? Probably not! Our neocortex is not just an scaled-up version of the primate prototype [2181,2290,2515]: the layout of our cortical areas has been revamped in a mosaic manner [146,2670]. Among other neurological quirks, we have a larger-than-expected prefrontal cortex but a smaller-than-expected visual cortex (based on the primate allometric regression) [770,1198,1534,2181].

Our neocortical distortions presumably occurred by heterogeneous growth in a balkanized patchwork of (1) regional growth-factor morphogens [147,1907,1908,2237], (2) selector genes [73,1633,2121], and (3) transcription factors [236,1393,1533,2509] (cf. the cerebellum [2397]). It probably did not occur in as straightforward a way as Gould imagined [956]—i.e., by merely extending the period of brain growth [654,1717].

Unfortunately, we can’t hope to make much headway on dissecting human brain evolution without first deciphering these regional changes within the brain at the genetic and developmental levels [1908]. To do that we must turn to evo-devo [910,1899].

Alas, having foresworn the best tools of experimental genetics for ethical reasons just stated, evo-devotees have had to resort to approaches that are correspondingly weaker [638]. Not surprisingly, their contributions have been disappointingly meager [1697]. Their chief approach has been: (1) find inherited disorders for the talent of interest (intelligence in general and of language in particular) [723], (2) identify the defective genes responsible [411], (3) study the molecular nature of their mutational lesions [225], and (4) use comparative genomics to assess how they have changed vis-à-vis primates [63,2394,2743]. About half a dozen candidate genes have been analyzed in this way [411,776].

Clearly, we have more questions than answers about the origin of human thought [910,1224,2015]. There is much to learn and much to think about.

Puzzle 94: Why can’t chimps speak?

Unlike human adults, human babies can drink (suckle) and breathe at the same time because their larynx has not yet descended [612,1547]. Their epiglottis is brought into contact with their nasal passage so that air flows directly to the lungs [478] while milk flows around the region of contact [1469]. This proximity restricts their vocal range [2278].

Not surprisingly (given its safety), the larynx of other mammals resembles that of babies more than that of human adults [533,1470,1887]. This trend might explain why chimps can’t speak [779,1546] despite their ability to learn English [2073]. An alternative to this mechanical explanation is a neurological explanation that is based on the evo-devo principle that “Larger becomes more interconnected” [2515,2516]. To wit, as different parts of the brain enlarge, their surplus of extra axons may automatically be displaced to new targets [610,612]. Our neocortical axons would have thus invaded regions of the medulla where they gained direct (conscious) control over motor neurons of the face, jaws, tongue, and larynx [2358-2360,2515] (cf. [1046]). If this latter argument is valid, then chimps would still be unable to speak even if their larynx descended as low as ours [779,1546,2515,2516].

Puzzle 95: How did superhuman vision evolve in sea gypsies?

One possible instance of runaway selection is the superhuman vision seen in tribes of sea gypsies in Southeast Asia [918]. Children of those tribes have been diving for food from the sea floor for thousands of years, and their underwater acuity is twice that of Europeans! This sharper acuity has been traced to an astounding ability on the part of these youths to (1) constrict their pupils and (2) alter the focal lengths of their lenses. However, until adequate controls are done (e.g., testing tribal members raised elsewhere) we can’t rule out a non-evolutionary explanation: it is equally possible that the children’s eyes are changing as they grow in response to the optical demands of their diving habit [2789]. Phenotypic plasticity is common in animal development and may be instrumental in facilitating evolution when ecological pressures persist over many generations [2790].

Puzzle 96: Why is the iris intrinsically photosensitive?

The iris is intrinsically photosensitive—i.e., it constricts on its own [2641]! How it does so is unclear, though we do know that it detects light using some sort of cryptochrome rather than an opsin pigment.

Puzzle 97: How did 11-cis retinal become the chromophore for vision?

Surprisingly, all metazoans use the same sensor in their photoreceptors [539,1478]. This gadget first evolved in prokaryotes [2449]. It relies on (1) the Vitamin-A derivative “retinal” (or a variant [932]) to absorb photons [1019,1723], (2) an “opsin” protein to monitor shape changes in retinal [176,1993,2288], and (3) a signal-relay chain of downstream effectors [787,1466,1971]. The evolutionary riddles posed by this universality are: (1) how long did it take for prokaryotic genomes, by random mutation, to stumble upon an opsin-like protein that could cradle a chromophore and transduce its twitching [621,876,2044,2241]?, (2) why was 11-cis retinal recruited instead of some other photo-active agent [932,1847,2709]?, and (3) how did it get linked to a particular transduction pathway [763,2521]?

XII. Behavior:

Puzzle 98: To what extent are our emotions just arbitrary spandrels?

Why are humans ticklish on our bellies and in our armpits [1097]? Why is laugher so contagious [612,1664.1935], especially among adolescents [419,2894]? To what extent are our emotions in general just accidental side effects of neural wiring that was selected for other sorts of functions [982]?

Puzzle 99: Why are we the only species where individuals kill themselves?

The most enigmatic of all human behaviors from the standpoint of evolution is suicide. It would only make sense if it enhanced the survival chances of one’s offspring or some other close relative [445], but it patently does not in most cases [2646]. Suicide might be a non-adaptive side effect of self-awareness (cf. Hamlet’s soliloquy) [740,1901,2587,2916], which deepens during adolescence [2329]. It has been linked to the same serotonin circuitry as clinical depression [279,1032,1439]. A disorder that seems like slow suicide is anorexia nervosa [1030]. Is anorexia also a dysgenic side-effect of our fragile self-image? What other animal starves itself in the presence of plentiful food?

Puzzle 100: When did dogs start to bark in their evolution from wolves?

Darwin was fascinated with dog barking since it has no obvious precedent among wild wolves: “The habit of barking, however, which is almost universal with domesticated dogs, ... does not characterize a single natural species of the family [562] (p. 27).” No one knows how barking evolved, except to observe that it arose during domestication [1489,1742,1806].

Puzzle 101: What is the adaptive value of yawning?

Yawning appears to be universal among vertebrates [2090]. It is associated with fatigue but also occurs upon waking [74], and it is often accompanied (enigmatically) by stretching. Like laughing, yawning is contagious [2090]—even among ostriches [2281], where preening and dustbathing are also contagious [1972].) What possible value (if any) does yawning offer as a social signal [1935]?

top of page

Lewis I. Held, Jr. is Associate Professor in the Department of Biology at Texas Tech University.

© 2009 Thomas B. Brody, Ph.D.