eFfing Fossil Friday: Frozen Femur

A 45,000 year old human femur from Siberia provides new information about genetic mutation rates and modern human origins. As Quiaomei Fu and colleagues report in this week’s issue of Nature, this seemingly simple leg bone carries so much information, not because of its gross anatomy, but because of the ancient DNA it preserves.

The femur wasn’t discovered by paleontologists, but by an artist/historian looking for fossils around the Irtysh River. The bone came from from a site called Ust’-Ishim, only some 650 km north of the snowy capital where I work in Kazakhstan:


The site in question, Ust’-Ishim is marked by the yellow star. The red and blue sites to the southeast are other Upper Paleolithic sites. Okladnikov (3) and Denisova (4) have also yielded fossils preserving ancient DNA. Modified from Fu et al. figure 1.

The bone was directly radiocarbon dated to around 45,000 years ago. With a fairly precise age of the bone, Fu et al. could estimate the rate at which genetic mutations arise, by counting the number of new mutations in recent humans that aren’t shared by the Ust’-Ishim femur. This led to an estimate of around 0.43×10−9  new mutations per site per year. This is a relatively low rate compared to estimates based on geologically older fossils, but consistent with more recent estimates that directly compare parents and offspring.

The Ust’-Ishim individual had levels of Neandertal ancestry comparable to living Eurasians (~2.3% of the genome), but there is no evidence of any Denisovan ancestry. Because this individual lived closer to the date of modern-Neandertal admixture, the Neandertal segments of its genome are longer than in modern people (recombination over generations breaks these regions apart into shorter segments). Knowing about recombination rates, Fu et al. could infer that admixture between Neandertal and modern human populations occurred between 50-60,000 years ago.

This eFfing Friday fossil provides more tantalizing evidence for DNA-bearing human fossils just across the Kazakhstan border. With Ust’-Ishim to the north, Denisova and Okladnikov caves to the east, and Teshik Tash to the south, my colleagues and I are very keen to find similar sites here on the KZ side.

Reference: Fu et al. 2014. Genome sequence of a 45,000-year-old modern human from Siberia. Nature 514: 445–449. doi:10.1038/nature13810.

Dawn of Paleoepigenomics

It was only a matter of time. In the 1990s scientists started extracting, sequencing and analyzing mitochondrial DNA from Neandertal fossils. In the 2000s they made major advances in obtaining and analyzing ancient nuclear DNA, which is much trickier than mtDNA. In just the past year, paleogeneticists pushed the envelope in sequencing truly ancient DNA, announcing hominin and horse genomes from 400 and 700 thousand years ago, respectively. As I mentioned a few months ago, the burgeoning field of paleogenomics is revealing things about human evolution that could hardly be dreamt of only a few decades ago.

But world of DNA is so much more than just ceaseless sequences of four letters, and the field of ‘epigenetics’ has emerged to investigate the complex way that chemical alterations to DNA structure (not sequence) affect gene expression. Melding epigenetics & paleogenomics, David Gokhmen and colleagues report in Science, “Reconstructing the DNA methylation maps of the Neandertal and the Denisovan.” For a review of what DNA methylation is and does, check out this Scitable overview. In short, DNA methylation is part of the reason why not all of your genes in your genome are expressed at all times throughout your body, even though all of your genes are physically present in all of the cells of your body. Methylation plays an important role in turning genes on or off during development. It’s nuts. Now, the structure of DNA breaks down over time after an animal dies, obscuring original methylation patterns. But the decompositoin process is becoming better understood, including patterns at methylated vs. unmethylated sites. As Gokhmen et al. write, these patterns “may serve as a proxy for the levels of methylation in ancient DNA.”

This brilliant insight allowed Gokhmen and colleagues to identify some 2000 genomic regions in bone cells that differed in methylation between a living human, a Neandertal and a Denisovan (2000 less than 1% of all regions). One such region was the HOXD cluster, which is known to be involved in embryonic limb development. Neandertals and Denisovans were more methylated than humans at the HOXD9 and HOXD10 loci. Whether and how these epigenetic differences might be responsible for anatomical differences between these populations is not at all clear yet. But Neandertals are known to differ from humans in some aspects of arm and leg anatomy – authors point out that Neandertals generally have larger and more robust joints but shorter limbs. They state, “together, these findings suggest that the HOXD cluster might have played a key role in the recent evolution of human limbs.”

Importantly, “Denisovans” are only known from 2 teeth and part of a finger bone, no other limb fossils are known (or at least published) for this ancient population. This leads us to a prediction – if the similarly hypermethylated HOXD sites in Denisova and Neandertals are functionally important, then Denisovan limb fossils, if ever found, should be more like Neandertals than like humans. If this prediction is borne out, this would provide evidence of specifically how HOXD9-10 affect limb development, and how HOXD epigenetic regulation has changed in human evolution. This hypothesis can be tested, but only with the discovery of the right fossils (i.e., genetically attributable to Denisovans). Well, the functional importance of hyper/hypomethylation at these sites could probably also be assessed with transgenic mouse experiments…

There is truly remarkable work being done in paleogenomics – and now paleoepigenomics – which will probably begin to form the basis of some exciting new human evo-devo research.

Is it worth seeking a genetic basis for math genius?

The topic this week in my Human Variation and Race class is intelligence. We’ve read about and discussed what intelligence is, how it is quantified, and the extent to which ‘intelligence,’ however defined, is biologically and/or environmentally determined. Intelligence (test score) has been shown to be heritable, meaning that a proportion of the variation in IQ test scores in a population can be explained by genetic variation. But that is not the same as saying that it is genetically determined. Similarly, complex traits such as intelligence, behaviors, and diseases almost never have a simple genetic basis – a common theme over at the Mermaid’s Tale, one that seems too rarely heeded. So you can imagine my surprise and delight at finding this news piece just published in Nature: “Root of maths genius sought: Entrepreneure’s ‘Project Einstein’ taps 400 top academics for their DNA.” Of course “roots” meant “genes.”

Apparently, bioinformatics entrepreneur and multimillionaire Jon Rothberg has set out to identify the genetic bases of peak mathletics, by analyzing the genomes of hundreds of mathematicians and physicists. Good luck, buddy! My initial reaction was to be appalled that an educated biologist these days could be such a flagrant biological determinist. What’s more, when approached about participating in the study, mathematician Curtis McMullen asked about the ethics of the project and its outcomes: “The uniform answer to my questions was that ‘we are not responsible for how the information is used after the study is completed.'” Ew. The project as briefly described reeked of some eugenics programme.

My prediction is that if this study takes off, Rothberg & buddies will be horribly disappointed. Assuming they are able to identify any genetic variants, these will probably only explain a small amount of variation in “maths genius.” Which itself is problematic, since there is probably not a single manifestation of math genius, and even if there were a single way to be a math genius, there may be several genetic pathways relating to the phenotype (not an uncommon finding of many genome-wide association studies). But hey, it seems to be Rothberg’s own money going into the study, so why not.

But then, if my prediction were to hold, this wouldn’t necessarily be a failure – it would point to an important role of society and learning environment in shaping individuals’ mathematic capability. And then maybe big money could begin to be diverted to more productive programs investigating and improving how people learn, rather than to large scale projects seeking simple answers when there isn’t necessarily any reason to expect them in the first place.

We like turtles (‘s genomes)

June 2013, Volume 45 No 6 pp 579-714

Jonathan the zombie isn’t the only one who likes turtles. These heroes-in-a-half-shell adorn the cover of the current Nature Genetics, as two species of turtle have just joined the Genome Club (Wang et al. 2013; paper’s free!).

This definitely not one of those genome sequencing studies alluded to recently by John Hawks, that’s “too boring for journals.” Wang and colleagues didn’t just sequence the genomes of soft-shell and green sea turtles ‘just cuz.’ Rather, they use these copious data to address several questions, most interesting of which relate to embryonic development.

First, analysis of gene expression during embryonic development supports what the authors refer to as a “nested hourglass model” of development and gene expression. The hourglass shape serves as analogy for variation across related species over developmental time: there is great variation (in both morphology and gene expression) in the earliest stages of development, then species are more similar at a given developmental stage (the “phylotypic period”), and thereafter variation increases again. This phylotypic period (which I don’t believe is unanimously agreed upon) is arguably the most conserved developmental stage in evolution – all vertebrates, for example, simply must pass through this stage to become good vertebrates. Plus, several studies have found that evolutionarily younger genes tend to be expressed before and after this amorphous phylotypic stage, while more ancient genes are expressed during this time. As the authors state

“According to the recently supported developmental hourglass model … the changes underlying major adult morphological evolution occurred primarily in the developmental stages after the period … that serves as the basic vertebrate body plan.”

So the turtle data generally support this model. However they mention a nested hourglass, because they found evidence of an additional bottleneck, a second hourglass, of conserved gene expression when comparing turtles with their close relative the chicken. In other words, “the most conserved developmental stage changes depending on distantly related species are that are being compared.” So since turtles and chickens are more closely related to one another than to many other vertebrates, they might share another conserved developmental stage. Incidentally, both also make for good soup.

Wang and colleagues also looked for genes relating to some of the unique aspects of turtle anatomy, examining what parts of the genome seem to get kicked up after the phylotypic period. It doesn’t take a trained eye to see that these animals are kinda weird in that their bodies are encased in a flagrant shell, with a carapace on top and plastron on the bottom. Now it turns out this carapace is actually formed from what should, in most other vertebrates, become vertebrae and ribs. So by studying the earliest development of these structures, Wang and colleagues could examine the molecular bases of this carapacial deviation.

Fig. 5 from Wang et al., showing Wnt protein expression in turtle embryos. In a), only Wnt5a is expressed in the ‘carapacial ridge’ during its earliest development. Fig c) is a cross-section indicated in b) showing this expression. NT=neural tube, NC=  notochord. The scale bar is 0.5 mm. Tiny!
The authors were able to identify over 200 miRNAs, and implicate the signalling protein Wnt5a, in the development of the “carapacial ridge” (see the arrows in fig. c above), the embryonic precursors to the carapace. Interestingly, Wnt5a is involved in the development of limb buds (e.g, those big purple circles in the red square in a) above). The precise role of Wnt5a and the miRNAs in turtle shell development has yet to be determined, so this study really sets the stage for future investigations.
ResearchBlogging.orgSo there you have it, a pretty cool paper combining genomics with developmental biology, among other things. And so to close, for your bemusement, here’s a video I shot last week at the awesome Kansas City Zoo, of a turtle attempting to make embryos like in the figure above (sorry for the poor quality). Hang in there, little buddy!
They like tuhtles!
Wang Z, Pascual-Anaya J, Zadissa A, Li W, Niimura Y, Huang Z, Li C, White S, Xiong Z, Fang D, Wang B, Ming Y, Chen Y, Zheng Y, Kuraku S, Pignatelli M, Herrero J, Beal K, Nozawa M, Li Q, Wang J, Zhang H, Yu L, Shigenobu S, Wang J, Liu J, Flicek P, Searle S, Wang J, Kuratani S, Yin Y, Aken B, Zhang G, & Irie N (2013). The draft genomes of soft-shell turtle and green sea turtle yield insights into the development and evolution of the turtle-specific body plan. Nature genetics, 45 (6), 701-6 PMID: 23624526