Editors' Choice

Science  04 Apr 2003:
Vol. 300, Issue 5616, pp. 19

    Crystals via Liquids

    Crystallization is used widely to isolate and purify molecules, but the early stages of nucleation and growth are not completely understood. Molecular dynamics simulations have suggested that a liquid-like state precedes crystal formation. A liquid phase has been observed to appear during crystallization of macromolecules, but evidence that small molecules behave similarly has been limited. Bonnett et al. observed liquid phase separation during crystallization of a methoxyacrylate from a water-methanol mixture as it cooled. Small drops of solute-rich liquid formed, and then crystals began to grow at the expense of the drops. This behavior is due not to kinetic effects but to a feature of the phase diagram that allows liquid-liquid phase separation before the onset of crystallization.—JFU

    Chem. Commun.2003, 698 (2003).


    Snow Job

    The largest source of year-to-year variability of winter weather in the extratropical Northern Hemisphere is an oscillating pattern of atmospheric pressure that has poles near Iceland and in the subtropical North Atlantic, commonly called the Arctic Oscillation/North Atlantic Oscillation (AO/NAO). The AO/NAO directs the tracks of storms crossing the Atlantic, affecting temperature and rainfall in North America, Europe, and northern Asia. It has been thought that the pattern of interannual variability of the AO/NAO is random and a consequence of the internal dynamics of the climate system. However, it also has been noted that, over the past 30 years, the AO/NAO has exhibited a trend; this has led to the proposal that it is modulated by soil moisture, secular changes of atmospheric greenhouse gas concentrations, sea-surface temperatures, the stratosphere, or sea-ice cover. Saito and Cohen add another ingredient to this olio by showing that continental-scale snow cover in North America and Eurasia varies with the atmosphere, leading the atmosphere by several months and influencing the AO/NAO on annual to decadal time scales.—HJS

    Geophys. Res. Lett.30, 1302 (2003).


    Quick Decisions on Catalysts

    The development of new polymerization catalysts, especially for specific copolymerizations of two distinct monomers, can be an expensive trial-and-error process. Boussie et al. use microscale, high-throughput primary screening methods to identify 10 candidate catalysts from about 400 trials. Analysis of these screens was carried out by rapid gel permeation chromatography and Fourier-transform infrared spectroscopy methods. From these leads, a hafnium(IV) complex was found to polymerize 1-octene to high conversions. Secondary screens of a 96-member focus library identified a high-temperature ethylene-1-octene copolymerization catalyst.—PDS

    J. Am. Chem. Soc. 10.1021/ja020868k (2003).


    Mapping Virus History

    Southeast Asia is attracting attention as a source of potentially dangerous emerging infections, not least influenza, but also dengue, Nipah, and possibly the causative agent of severe acute respiratory syndrome (SARS). The most important cause of epidemic encephalitis worldwide is the mosquito-borne Japanese encephalitis virus (JEV), a flavivirus allied to West Nile virus. Solomon et al. have probed the geographical distribution of JEV strains across Asia since the first reported outbreak in Japan in the 1870s. All five genotypes identified occur simultaneously in Indonesia and Malaysia, whereas only evolutionarily more recent strains have spread out of this region. The Japanese epidemic was devastating and affected people of all ages, as flaviviruses do when they reach an immunologically naïve population. By contrast, in Malaysia, it is rarely epidemic, suggesting a long-standing association with humans here and pointing to the Indonesia-Malaysia region as the nidus of JEV.—CA

    J. Virol.77, 3091 (2003).


    A Star is Born

    The first billion years of the universe [redshift (z) greater than 6] was a time of great darkness and commotion. After inflation, there was mostly hydrogen and helium gas expanding into darkness. Residual ionization and primordial density fluctuations from the Big Bang led to clumping of this molecular gas and dark matter. The first sources of light, probably stars or quasi-stellar objects (QSOs), began to form around z ∼ 30, and most of the hydrogen was reionized by z ∼ 7. Observers are beginning to discover some of the earliest QSOs (z ∼ 6), and these ancient objects provide the most luminous markers for checking models of the formation of the early universe.

    Using the Hubble Space Telescope, Freudling et al. have obtained near-infrared spectra of three QSOs (5.7 < z < 6.3). The Fe/Mg abundance ratios are similar to or greater than that of the Sun, suggesting that massive galaxies (as opposed to galaxy mergers) with solar or supersolar Fe/Mg appeared within the first billion years. If the QSOs were triggered by supernovae, then it would have taken 0.5 to 0.8 billion years from the formation of the supernovae progenitors for the gas to become enriched enough to supply solar or supersolar Fe/Mg. Formation of the progenitors at z ∼ 30 to 10 would be consistent with recent results from the Wilkinson Microwave Anisotropy Probe, indicating that star formation began about 0.2 billion years after the Big Bang.—LR

    Astrophys. J.587 Rapid Release 21 March 2003 (astro-ph/0303424).



    Rather than trying to sample all of chemical space in searching for drug leads, designers have deconstructed the problem and compiled combinatorial libraries of fragments. This approach facilitates synthesis, especially when the target can be used as a template to favor some fragments over others. Braisted et al. have selected for weakly binding fragments by capturing them via disulfide tethering to the target. In one such instance, joining a fragment to a previously identified lead improved the inhibitory potency of an interleukin-2 antagonist by 50-fold. Erlanson et al. have extended this approach by screening for two weakly binding fragments that together conform to the linear active sites of caspases, the enzymes of apoptosis.—GJC

    J. Am. Chem. Soc. 10.1021/ja034247i (2003); NatureBiotechnol.21, 308 (2003).


    The Genetics of Overeating

    While it is generally accepted that genetics plays a role in obesity, the culprit genes have been difficult to identify because this disorder typically arises from the combined effects of multiple genes and environmental factors. The MC4R gene, which encodes the melanocortin 4 receptor, is one of the few genes causally linked to rare, monogenic forms of obesity. New work reinforces the idea that MC4R mutations are strong contributors to the development of morbid obesity induced by hyperphagia (overeating) and suggests that these mutations may be more common than previously thought.

    In a study of 500 patients with severe childhood obesity, Farooqi et al. found that nearly 6% had mutations in MC4R and that the severity of their clinical phenotype correlated with impairment of receptor function. Independently, Branson et al. found that sequence changes in MC4R were especially prevalent in obese subjects who engaged in binge eating, although the functional effects of the sequence changes were not tested. Recent mouse data from Weide et al. support the hypothesis that hyperphagia, rather than reduced metabolism, is the primary disturbance leading to early-onset obesity in the MC4R-deficient state.—PAK

    N. Engl. J. Med.348, 1085; 1096 (2003); Physiol. Genomics13, 47 (2003).


    A Copper Economy

    Copper is an essential element, but free copper is toxic, so bacteria have developed regulatory systems, such as the pco system in Escherichia coli and the homologous cop system in Pseudomonas syringae. Four proteins, Pco/CopA, B, C, and D, are expressed under the control of a copper-inducible promoter. Recently, Peariso et al. used x-ray absorption spectroscopy (XAS) to show that the small periplasmic protein PcoC binds both Cu(I) and Cu(II), using different sets of ligands.

    Arnesano et al. have determined the solution structure of CopC bound to Cu(I), and show that the binding sites for Cu(I) and Cu(II) in CopC are 30 angstroms apart. Cu(II) is bound by two histidines, an aspartic acid, and a glutamic acid, whereas Cu(I) is bound by a histidine and three methionines. A shift in redox state causes the copper ion to migrate between the sites.

    How does CopC contribute to copper homeostasis? CopA is a periplasmic copper oxidase, and CopB and CopD are outer- and inner-membrane proteins, respectively. CopA and CopB contain methionine-rich repeats that, on the basis of the CopC structure, are likely to bind Cu(I). CopC may act as a molecular switch that facilitates either Cu(II) import to the cytoplasm via CopD, or Cu(I) export via CopB. Conversely, CopA may oxidize Cu(I) bound to CopC, or Cu(I) delivered from CopC might catalyze the oxidase activity of CopA.—VV

    J. Am. Chem. Soc.125, 342 (2003); Proc. Natl. Acad. Sci. U.S.A.100, 3814 (2003).