In This Issue

Decoding patterns of gene expression in cancer

Heat map representation based on surprisal analysis of the mRNAs and miRNAs in the disease signature for lung cancer.

To identify the most aggressive cancers for targeted treatment, gene expression microarrays represent an invaluable tool. However, analysis of expression data tends to emphasize low-abundance transcripts that undergo large fold changes over highly expressed genes with relatively smaller fold changes. To identify cancer-specific signatures while taking into account both abundance and fold change, Sohila Zadran et al. (pp. 19160–19165) adapted a thermodynamic approach called surprisal analysis to analyze mRNA and microRNA (miRNA) expression data from ovarian, prostate, breast, and lung cancers. The authors identified distinct sets of mRNAs and miRNAs whose relative tissue-specific expression levels remained consistent across cancer patients and healthy individuals, defining a balance state in which each transcript could be assigned a thermodynamic weight based upon its abundance. Surprisal analysis led to the identification of RNAs that deviate from a balance state, resulting in expression patterns that distinguish between cancer and noncancer patients. To validate their findings, the authors experimentally silenced the single gene determined to have the greatest thermodynamic influence for each type of cancer in a corresponding cancer cell line, and found that in each case, cell proliferation was reduced. According to the authors, the pattern of deviations from the balance state generates a cancer-specific disease signature that may be used to identify targets for therapy. — C.B.

Carbon sequestration technology may be linked to Texas earthquakes

Injecting liquefied CO₂ into deep rock formations provides an attractive method for carbon sequestration, yet the method increases fluid pressure within rock layers. To determine whether these pressure changes could cause earthquakes, Wei Gan and Cliff Frohlich (pp. 18786–18791) examined the active Cogdell oil field in western Texas, which experienced 18 earthquakes of magnitude 3.0 or greater between 2006 and 2011. Many of these earthquakes occurred within 2 km of wells used to inject supercritical CO₂ into petroleum-bearing formations. Injections of CO₂, employed to enhance oil recovery, began in 2004 and far outpaced the amount of liquid pumped out, building up pressure within the deep limestone. The increased pressure may have reduced friction on preexisting faults, allowing them to slip more easily. A similar rash of earthquakes struck the Cogdell field between 1975 and 1982, when engineers injected water into deep wells, to enhance oil production. Injection of supercritical CO₂ does not always cause earthquakes, according to the authors, and detailed modeling is needed to understand why earthquakes occur in some situations but not others. — P.G.

Specialized neurons help primates quickly detect snakes

Humans and other primates can rapidly detect snakes, even within a crowded visual scene. Quan Van Le et al. (pp. 19000–19005) probed this ability and found a group of neurons in the primate brain that responds selectively and rapidly to images of snakes. The authors used microelectrodes to measure the response to images of snakes and other natural stimuli from 91 visually responsive neurons in the medial and dorsolateral pulvinar regions of the brain of two macaques. The pulvinar subdivisions chosen by the authors are primate-specific and appear to be involved in visual attention and fast processing of threatening images. While showing the monkeys images of snakes, monkey faces, monkey hands, and simple geometrical shapes, the authors measured the ratio of neurons that fired for each stimulus and the response magnitude and firing latencies of the neurons. The authors report that snakes prompted responses by the largest proportion of neurons (41%) and elicited the strongest, fastest responses. The findings provide a neurobiological explanation for why primates have heightened visual sensitivity to snakes and help researchers better understand the evolutionary pressures that have shaped our primate lineage, according to the authors. — B.A.

Pancreatic islet transplantation without immunosuppression

Schematic view of the chamber system for pancreatic islets.

Transplantation of pancreatic islets—clusters of cells that produce insulin—is a promising approach for the treatment of type-1 diabetes, but the approach is limited by the lack of donor organs, the need for immunosuppressive medications to prevent the immune system from rejecting the transplanted cells, and the progressive loss of islet function over time due to inflammation and low oxygen levels at the transplantation site. To overcome these challenges, Barbara Ludwig et al. (pp. 19054–19058) developed an implantable, semipermeable chamber system for islets that shields the cells from the host’s immune system but allows diffusion of glucose and insulin and delivery of oxygen to the transplant. The authors implanted the device in the abdomen of a 63-year-old male patient with type-1 diabetes who was completely insulin deficient, and assessed the transplant’s function and stability for 10 months. The patient displayed modestly elevated levels of C-peptide, a by-product of insulin production, and the levels of insulin and C-peptide rapidly increased following glucose injection, demonstrating that the islets were functional. Furthermore, the recipient displayed no signs of graft rejection or immune sensitization, even without the use of immunosuppressive agents, the authors report. The findings suggest that the chamber system may prove useful for future cell-based therapies, according to the authors. — N.Z.

A genetic switch to combat drug-resistant tuberculosis bacteria

More than a million people die every year from infections with the bacterium Mycobacterium tuberculosis (Mtb), largely due to the difficulty of targeting Mtb cells that lie dormant in the body during chronic tuberculosis infections. Jee-Hyun Kim et al. (pp. 19095–19100) engineered a dual-control (DUC) genetic switch capable of silencing Mtb gene activity and identifying proteins required for the pathogen’s growth and persistence. Upon exposure to a single inducer—either anhydrotetracycline or doxycycline—the DUC switch simultaneously represses transcription of a specific Mtb gene and degrades the encoded protein. Finding that the switch resists phenotypic reversion and quickly depletes proteins in replicating and nonreplicating Mtb, the authors used it to target an Mtb protein known as nicotinamide adenine dinucleotide synthetase (NadE). Depletion of NadE rapidly killed replicating and dormant Mtb cells in infected mice and in vitro, the authors report. The findings demonstrate that the DUC switch can be used to probe the essentiality of Mtb proteins under various conditions, and unveil the NadE protein as a drug target that may help shorten the duration of current TB drugs. According to the authors, the DUC switch could help speed the development of treatments for active and dormant Mtb infections. — A.G.

Sequence of biomarker changes in early onset Alzheimer’s disease

Medial and lateral left cortical gray surface showing differences between carriers and noncarriers in Pittsburgh Compound (A), FDG (B), and cortical thickness (C) at 15, 10, 5, and 0 years before predicted symptom onset. Regions with significant increases are shown in red/yellow and decreases in blue/cyan.

In families with early onset autosomal dominant Alzheimer’s disease (ADAD), a form of the dementia tied to mutations in a handful of genes, the age of disease onset is thought to be conserved among family members and, thus, predictable. Tammie Benzinger et al. (pp. E4502–E4509) previously reported the timing and order of changes in established disease biomarkers in a cohort of individuals from families with ADAD mutations recruited for a study known as the Dominantly Inherited Alzheimer Network initiative. The authors extended their study to a larger cohort, including 137 mutation carriers and 92 people without the mutations, examining changes in the biomarkers across the entire brain, in particular the timing and location of amyloid accumulation (imaged using a dye called the Pittsburgh compound), glucose metabolism (imaged using FDG-PET), and tissue atrophy (imaged using MRI). The authors report that most brain regions of mutation carriers first showed a rise in amyloid deposition, then a drop in glucose metabolism, and, finally, atrophy across several cortical regions, with regional variations in the time course of the changes. But the authors also observed exceptions to this order of events in some brain regions, such as the hippocampus, where amyloid accumulation was followed by loss of tissue volume, which preceded a decrease in glucose metabolism. The authors suggest that clinically useful information might be gleaned from local changes in amyloid deposition. According to the authors, the findings underscore the importance of combining multiple biomarkers while monitoring clinical responses to treatments. — P.N.