Institute of Organic Chemistry and Biochemistry, Academy of Sciences of the Czech Republic

A chemist realizes that popularity is no measure of strength.

Urea, the water-soluble organic compound found in mammalian urine, has been known for its ability to denature — or unfold — proteins for more than 100 years. To this day, it is among the most widely used protein denaturants. So one could be forgiven for taking it for granted that we know in gory detail what happens when we pour urea into a protein solution. But, alas, nailing down the individual molecular interactions between urea and the chemical groups at a protein's surface is exceedingly difficult.

Experiments and simulations suggest that urea interacts primarily with amide groups in the protein backbone, but every such group in a given protein has its own local environment, leading to fuzzy signals in spectroscopic studies. Paul Cremer's group at Texas A&M University came up with a good means by which to address the problem. They employed a popular protein proxy, poly(N-isopropylacrylamide), in which all of the amide groups are chemically equivalent (L. B. Sagle et al. J. Am. Chem. Soc. 131, 9304–9310; 2009). Using infrared spectroscopy combined with measurements of hydrophobic collapse, they showed that urea interacts only weakly with this polymer.

Essentially, Cremer and colleagues' measurements suggest that one needs buckets of urea to see any effect. This is exactly the same situation as that observed for proteins, in which high concentrations of urea are necessary for denaturation. Thus one of the most common denaturants is actually a shockingly weak one. In fact, the strength of its interactions with the protein is little greater than those of harmless water molecules.

In the end, the key to the denaturating mechanism may be the fact that urea is a larger molecule than water — which has subtle entropic consequences — rather than that the two have different hydrogen-binding abilities.

Posted by Brendan Maher at 07:10 PM | Permalink | Comments (0) | TrackBacks (0)

University of Missouri

An ecologist marvels at animals that learn to eavesdrop.

All through college I resisted getting glasses, but I finally succumbed for my first field trip to Peru; I was determined to see everything. Upon arrival, however, I realized that good vision was scarcely enough. One morning, I walked through the forest with an ornithologist, the late Ted Parker, famed for having learned the songs of thousands of birds. Although we saw few of the singers, he knew the source of each fluted phrase, monotonous chant and raucous outburst.

Attending to the signals of other species — visual, auditory and so on — is useful not only to ecologists, but also to the predators that eavesdrop on their prey. In some cases, such behaviour is hard-wired; however, this seems unlikely for species that exploit a wide range of prey.

Recent research has revealed a more flexible strategy. Martinus Huigens of Wageningen University in the Netherlands and his colleagues studied a tiny wasp that parasitizes butterfly eggs (M. E. Huigens et al. Proc. Natl Acad. Sci. USA 106, 820–825; 2009). The wasp, Trichogramma evanescens, learns, after a single experience, to exploit the hosts' chemical-communication system to find and hitchhike on a mated female, disembarking when the butterfly lays her eggs.

Prior research had revealed only one other such case, in the bat Trachops cirrhosus, which learns the calls of poisonous and edible frogs (R. A. Page and M. J. Ryan Curr. Biol. 16, 1201–1205; 2006).

These examples suggest that learning which communicative signals to follow may be a common feature of the evolutionary race between predator and prey. It is doubtful that any bat or wasp can retain as many signals as a legendary ornithologist, but it seems that the drive to learn them has a long history.

Posted by Brendan Maher at 12:34 PM | Permalink | Comments (0) | TrackBacks (0)

GFZ German Research Centre for Geosciences, Potsdam, Germany

A geochemist learns that mountain building does not accelerate rock weathering.

Mountain building has been deemed essential for stabilizing Earth's climate over the scale of millions of years. As tectonic forces push mountains into the sky, they provide fresh rock surface that is degraded by the physical effects of rain and temperature change, and by chemical weathering as carbon dioxide is dissolved in rain to form carbonic acid. Atmospheric CO2 is thus consumed to convert rock into soil, which happens fastest where erosion rates are highest, exposing new rock to be weathered. Rivers then transport this carbon to the oceans where it is disposed of as carbonate sediment.

But many Earth scientists have questioned this story. Wouldn't periods of high tectonic activity, such as the rise of the Himalayas, provide enough rock to consume so much CO2 that the planet would turn into an ice house?

Yet the consequences of mountain building are perhaps less profound than expected. The amount of weathering over time can be accurately measured on hill slopes using new geochemical methods that combine solute loss from soils with radioactive isotopes formed by cosmic rays to determine how long it takes for rocks to break down into soil. Two recent papers modelled the implications of this approach numerically. Surprisingly, the prediction is that weathering decreases rather than increases at erosion rates typical of high, active mountains (K. L. Ferrier and J. W. Kirchner Earth Planet. Sci. Lett. 272, 591–599; 2008; E. J. Gabet and S. M. Mudd Geology 37, 151–154; 2009). So hill-slope weathering in the Himalayas might do no more to withdraw CO2 than any actively eroding, mid-altitude mountain range found worldwide.

Perhaps geochemists have been looking in the wrong place. Does the CO2-consuming mineral decomposition thought to occur on high slopes actually happen on the floodplains below large, active mountains? We might need to take a closer look at these areas before we really understand the geological carbon cycle.

Posted by Brendan Maher at 12:36 PM | Permalink | Comments (0) | TrackBacks (0)

Howard Hughes Medical Institute, Boston University, Massachusetts

A bioengineer gets schooled by Escherichia coli.

The ability to learn from situations and to predict certain outcomes sets us apart from many living things. It prevents many of us from chasing balls into busy streets or placing bottles of ethanol near Bunsen burners. Still, it didn't stop thousands of US researchers submitting applications for the National Institutes of Health's Challenge Grants — funded by President Barack Obama's economic stimulus package — despite an expected success rate little better than one or two per cent.

To enjoy the benefits of learning and predictive behaviour, we usually assume that you need a nervous system or at least a neuron. So it was surprising to read that Saeed Tavazoie at Princeton University, New Jersey, and his colleagues have demonstrated that bacteria can learn and exhibit anticipatory behaviour (I. Tagkopoulos et al. Science 320, 1313–1317; 2008). They show computationally and experimentally that Escherichia coli can learn temporal correlations between environmental stimuli — for example, that an increase in temperature is followed by a decrease in oxygen levels — allowing the bacteria to predict and prepare for future environmental changes.

The researchers show that this associative learning is accomplished by rewiring of biochemical networks. Strikingly, they also show that, like many of us, E. coli quickly 'unlearn' (in fewer than 100 generations) what they had learned in a new situation.

Now we know that bacteria can be taught such tricks, it will be interesting to see if we can use novel combinations of environmental stimuli to train microbes to efficiently convert biomass into energy sources, such as hydrogen or butanol. By providing E. coli with such an educational stimulus package, we may be able to boost the global economy.

Posted by Brendan Maher at 07:10 PM | Permalink | Comments (0) | TrackBacks (0)

Medical Research Council, Cambridge, UK

A biologist looks at the effect of a dynamic nuclear environment on gene expression.

In many organisms, including animals, genes are arranged linearly on chromosomes. But this linear order is largely meaningless during transcription, when RNA is made from DNA. Instead, a very different three-dimensional arrangement of genomic regions emerges in which structural flexibility and ability to reorganize become crucial to gene expression. Some regions loop out dynamically, moving far from their neighbours. Genes can participate in 'transcription hotspots' in close association with genes from other chromosomes. But the question remains as to what leads this dance. Is the chromosomal reorganization a cause or a consequence of transcription?

Using Hox clusters — groups of genes important in development — Wendy Bickmore of the Medical Research Council in Cambridge and her colleagues start to answer this question. Hoxb and Hoxd have very different environments in terms of their location on the chromosome and expression of their neighboring genes. The authors found that during tissue differentiation, Hox genes loop out and undergo active transcription. This reorganization then spreads from the Hox locus into adjacent genomic regions, but does not necessarily affect transcription of neighbouring genes (C. Morey Genome Res. doi:10.1101/gr.089045.108; 2009).

The authors propose that on activation, structural changes alter the constraints on genes' expression, allowing them to loop out and explore a much larger transcriptional environment within the nucleus. The team concludes that positioning outside of a chromosomal region is important for, but not a driver of, transcriptional activation.

These findings have broad implications: first, dynamic reorganization of chromosome territories is necessary but not sufficient for activation. Second, this reorganization is associated with modification to DNA's structural packaging, which can permanently alter a cell's nature.

Posted by Brendan Maher at 07:26 PM | Permalink | Comments (0) | TrackBacks (0)