Editors' Choice

Science  26 Apr 2019:
Vol. 364, Issue 6438, pp. 348
  1. Neuroscience

    Why some are more anxious than others

    1. Peter Stern

    Anxiety is a common, but highly variable, response to stress.


    Like many other human behavioral traits, the level of anxiety each person perceives varies. Little is known about how individual differences in brain chemistry and brain circuit activity correlate with variations in anxiety. Berry et al. measured self-reported anxiety in healthy adults and investigated its relationship with brain dopamine function and functional connectivity within brain circuits implicated in anxiety regulation. Individual differences in anxiety were associated with variation in dopamine release in the amygdala and rostral anterior cingulate cortex of the brain. There were also hints of a relationship between dopamine release and functional connectivity between these two areas. These findings provide insights into the relationship between brain activity and neurochemical signaling underlying individual differences in affective function.

    J. Neurosci. 39, 2735 (2019).

  2. Psychology

    High replicability in personality psychology

    1. Tage S. Rai

    According to the Big Five theory of personality, personality traits can be organized into five primary dimensions, including extraversion, agreeableness, conscientiousness, neuroticism, and openness to experience. These dimensions are associated with myriad life outcomes, such as job satisfaction or health. Soto conducted a replication study of 78 previously published personality–life outcome findings after high-profile failures by others to replicate studies in other areas of psychology. Of personality–life outcome effects, 87% replicated successfully with effect sizes that were 77% as large as those in the original studies. Replicability was predicted by features of the original studies and the replication studies. These results suggest that findings in personality psychology are more robust than in other subareas of psychology.

    Psych Sci. 10.1177/0956797619831612 (2019).

  3. Chemical Bonding

    Measuring anion-π interactions

    1. Jake Yeston

    A benzene ring hosts a cloud of electrons spread around its circumference. However, if fluorines replace the hydrogens on the periphery, the ring face counterintuitively attracts negative anions. This anion-π interaction contributes to the structures of noncovalent assemblies but has been challenging to quantify precisely. Anstöter et al. measured photoelectron spectra of gas-phase iodide-hexafluorobenzene complexes, prepared photolytically in molecular beams. By supplementing their measurements with theoretical calculations, they extracted a bond strength of 51 kilojoules per mole, attributed predominantly to correlation energy.

    J. Am. Chem. Soc. 141, 6132 (2019).

  4. Heart Disease

    Targeting cardiomyocyte cell death

    1. Beverly A. Purnell

    Narrowing of blood vessels, arrythmias, and congenital heart defects are all signs of heart disease. Because of the high prevalence of these conditions, developing treatments to minimize death of heart muscle cells (cardiomyocytes) is needed. To understand cardiovascular disease, Fiedler et al. used human stem cell–derived cardiomyocytes and mouse models. They noticed an increase in a protein kinase called MAP4K4 in defective heart tissue. Lack of oxygen induces MAP4K4 to signal cell death. Cardiomyocyte viability and function can be rescued by blocking MAP4K4 with the small-molecule inhibitor DMX-5804, and in mice, infarct size is reduced within an hour of MAP4K4 inhibition.

    Cell Stem Cell 24, 579 (2019).

  5. Cancer

    Redefining a high-risk leukemia

    1. Paula A. Kiberstis

    Acute erythroid leukemia (AEL) is an aggressive, poorly understood disease that has been difficult to distinguish from other blood cancers that are associated with a better prognosis. To determine if genomic information could improve diagnostic accuracy, Iacobucci et al. comprehensively analyzed 159 patients with AEL. Their study not only identified mutations and mutational patterns that were specific to AEL but also revealed that 45% of the patient samples had alterations in cell signaling pathways that could be therapeutically targeted. Of note was the discovery of activating mutations in NTRK1, a kinase for which there is a clinically approved pharmacological inhibitor. Treatment of a mouse model of AEL with this drug inhibited growth of leukemic cells.

    Nat. Genet. 51, 694 (2019).

  6. Cellular Microbiology

    TB provides HIV's little helpers

    1. Stella M. Hurtley

    Macrophage tunneling nanotubes, visualized by scanning electron microscopy


    It is well known that tuberculosis (TB) can exacerbate HIV pathology, and exactly how it does that involves a variety of factors. Souriant et al. studied the interactions between the two pathogens in cell culture. They found that macrophages induced by TB infection were particularly susceptible to HIV infection in vitro. One class of TB-induced macrophages called M(IL-10) cells were more prevalent during coinfection with HIV [or simian immunodeficiency virus (SIV) in a nonhuman primate model]. It appears that the TB infection created an environment in which macrophages produced so-called tunneling nanotubes (TNTs). These cellular protrusions connect cells, allowing cell-to-cell transfer of viruses. The TNTs promoted viral spread without directly affecting viral entry, viral activation, or viral restriction factors. Thus, TB appears to promote a microenvironment within patient tissues that facilitates HIV production.

    Cell Rep. 26, 3586 (2019).

  7. Machine Learning

    Synthetic data, privacy, and the law

    1. Brad Wible

    Machine learning can synthesize “almost-but-not-quite replica data” based on real data, facilitating research and data sharing while protecting privacy of the real data, but inconsistent data protection laws can stymie use of this approach. Removal of key information from data can enhance privacy, but this limits data utility and fuels an arms race between deidentification and reidentification. Instead, a generative adversarial network can synthesize data that mimic a protected dataset for analytical purposes but are less likely to reveal any actual private information. Bellovin et al. recommend amendments to privacy statutes that are often too absolute and fail to recognize the protections and analytical potential of this approach.

    Stanf. Technol. Law Rev. 22, 1 (2019).