Forecasts turn tide on silt

When Hurricane Irene hammered the eastern United States in August 2011, floods sent a glut of silt into New York City’s drinking-water system. The turbid waters rushed more than 100 kilometres through an aqueduct from the Catskill Mountains to the Kensico Reservoir, the last stop before the slurry would have reached millions of taps. For more than eight months afterwards, the city was forced to use an environmentally contentious chemical to rid the water of silt.

With the frequency and intensity of such siltation events on the rise, New York City is about to embark on a pioneering upgrade to its water system. The focus will not be on new dams or silt traps. Instead, starting this November, New York’s reservoirs will be managed by souped-up software that automatically incorporates short-term weather forecasts and seasonal climate forecasts — helping water managers to deal with floods and droughts.

Hydrologists far beyond the Big Apple will be watching closely. New York’s programme depends on a streamflow-forecasting system developed by the US National Weather Service, which aims to implement the system nationwide in the coming years. “This project opens the door to a more quantitative use of seasonal climate forecasts, which will help people make better decisions,” says Andrew Wood, a hydrologist at the National Center for Atmospheric Research in Boulder, Colorado. Australia is now experimenting with a similar forecasting system, and the European Commission Joint Research Centre has also rolled out a flood-prediction system.

But New York will be one of the first cities to connect the forecasts to a water-management system. Its immediate goal is to manage storm runoff to meet water-quality standards without adding costly new infrastructure to the system, a network of 19 reservoirs and 3 lakes that collectively hold more than 2 trillion litres of water. Under normal circumstances, managers handle siltation by holding water in reservoirs and letting silt settle out. But several times since 2005, the city has had to let silty water flow all the way to Kensico, where it was then treated with aluminium sulphate — a chemical that causes silt particles to coagulate and sink. State and federal regulators have raised concerns about build-up of aluminium sulphate in reservoir sediments, and its potential effects on fish and aquatic organisms.

The new system is designed to help the city to cope with major siltation episodes, which are expected to increase as the climate warms. With better warning of impending storms, water managers can drain certain reservoirs ahead of time and alleviate the potential for siltation. The software also takes seasonal forecasts into account, allowing managers to work out various conservation strategies if, for example, drier weather is expected.

“For years, we’ve done this by ourselves, just trying to balance all of this in our heads,” says Jim Porter, who heads water operations for the city’s department of environmental protection. “Hopefully, we can predict out a little further into the future.”

The potential savings are enormous. Coping with the silt problem by building a new intake system at one reservoir or increasing the size of a second reservoir would cost between US$200 million and $500 million. A new filtration plant could run to more than $10 billion. By contrast, the city’s analysis suggests that an integrated reservoir-management system can tackle the problem for roughly $8 million.

But to make it happen, the city first needs better streamflow forecasts. Although the National Weather Service makes streamflow predictions, it has until now done so mainly by comparing current conditions — precipitation, soil moisture, snowpack and streamflow — with historical averages, and then extrapolating the results. This approach assumes that streamflows will evolve as they have in the past under similar circumstances, but does not look ahead to future conditions.

The incoming system — years in the making — combines short-term and seasonal precipitation forecasts, and adds those predictions into the streamflow forecasts. To validate the system, the National Weather Service checked its predictions against historical data.

New York is paying the Weather Service about $1 million to accelerate the process so that the system will be available for use this year. “We now have something that is ready for prime time,” says John Schaake, a hydrologist and independent consultant in Baltimore, Maryland, who helped to develop the streamflow-forecast system.

The system will be available at 5 of the 12 regional US river-forecast centres, although it is unclear when it will become standard nationally. There are budget constraints, and each centre will have to customize the system. “There’s a lot of interest, but the question is how you institutionalize that,” says Kevin Werner, a hydrologist at the Colorado Basin River Forecast Center in Salt Lake City, Utah.

And it is also unclear whether others will follow New York’s lead and hitch the forecasts to a reservoir-management system. Demonstrating that the forecasts improve water management should help to ease doubts, says Daniel Sheer, president of HydroLogics Incorporated in Columbia, Maryland, which is providing New York with the reservoir-management software. “There will be much broader interest in the forecasts if we can show that they work.”

Metabolism: Sweet enticements to move

The formation of new blood vessels from pre-existing ones is a carefully orchestrated dance. A study reveals that the metabolism of sugar by glycolysis contributes to its regulation.

The breakdown of fuel by metabolism is the engine that sustains all cellular activities. But can metabolism also steer and control cellular processes? Writing in Cell, De Bock et al.¹ suggest that the answer is yes, at least in the context of glucose metabolism and angiogenesis — the formation of new blood vessels.

Glycolysis is the cellular process by which glucose is converted into pyruvate. A cell then makes a choice: it can convert pyruvate to lactate, which exits the cell, for a net yield of 2 ATP molecules (the currency of cellular energy transfer) or, in the presence of oxygen, the pyruvate can enter cellular organelles called mitochondria and become fully oxidized, producing a net yield of more than 30 ATP molecules. One would not expect any oxygenated cell to opt out of this mitochondrial bonanza, but some do, in a phenomenon first noted² in cancer cells by Otto Warburg in 1956. Cancer cells probably make this choice because intermediate molecules formed during glycolysis support the synthesis of macromolecules needed for cellular replication³. But do any non-cancerous or even quiescent cells also display the Warburg effect? Endothelial cells, which line blood vessels throughout the body and mediate angiogenesis, do³, but until now little was known about how metabolism affects their function.

De Bock et al. began their study by confirming a previous report⁴ that endothelial cells are highly glycolytic but perform little pyruvate oxidation. The authors then asked the interesting question: could modulation of glycolytic activity have an effect on angiogenesis? To assess this, they altered the amount of phosphofructokinase 2 (PFK2) in endothelial cells. PFK2 is a glycolysis-regulating enzyme that was discovered only in the 1980s, long after all key enzymes of the glycolytic pathway were thought to be known⁵. The related enzyme PFK1, identified decades before PFK2, catalyses the crucial committing step of glycolysis: the conversion of fructose-6-phosphate to fructose-1,6-bisphosphate. PFK2, by contrast, converts fructose-6-phosphate to fructose-2,6-bisphosphate, which is a potent allosteric activator of PFK1 (ref. 6). Activation of PFK2 thus drastically accelerates glycolytic flux through PFK1.

De Bock et al. show that reducing PFK2 levels in endothelial cells not only lowers glycolytic flux, as expected, but also impairs angiogenesis, by reducing the ability of the cells to form tip cells, migrate and form blood-vessel 'sprouts'. Conversely, and importantly, increasing PFK2 levels has the opposite effect: angiogenesis is increased. The authors also show that PFK2 lies downstream of VEGF and Notch, two proteins that are dominant determinants of endothelial-cell characteristics during angiogenesis.

How does PFK2 achieve these effects? Perhaps most interestingly, the authors demonstrate that PFK2 localizes to structures at the margins of endothelial cells called lamellipodia and filopodia. These cellular projections, which contain meshes and filaments of the protein actin, mediate endothelial-cell movement and sprout formation during angiogenesis (Fig. 1). PFK2 activity at this site probably coincides with the cellular position of large complexes of glycolytic enzymes, known as metabolons, which facilitate the channelling of metabolic products from one enzyme to the next⁷. Thus, it seems that PFK2 alters angiogenic capacity by altering glycolytic flux at the site of primary cell motion.

The study is important for several reasons. The findings imply that glucose metabolism can 'steer' the angiogenic process, in addition to simply being its 'engine'. This unveils glucose metabolism as a potential target for pro-angiogenic therapies (such as in patients with inadequate blood supply to the heart or limbs) or anti-angiogenic therapies (for example, to tackle tumours). Metabolic enzymes make good drug targets, so this is an exciting possibility. The study also provides an additional explanation for why endothelial cells perform glycolysis rather than oxidative breakdown of glucose: rapid local generation of ATP can occur in glycolytic metabolons located in the lamellipodia and filopodia, which are too small to accommodate mitochondria and are often found at angiogenic fronts where oxygen is scant.

Like all seminal work, this study generates several questions. Does modulation of glycolytic flux in ways other than through PFK2 also affect angiogenic sprouting? Could non-enzymatic properties of PFK2 contribute to the observed phenomena? Such behaviour has been seen for pyruvate kinase, another key enzyme in glycolysis that was recently found⁸ to be present in the cell nucleus and associated with transcription factors that drive gene expression. Does PFK2 modulate the activities of Rac, Akt and eNOS — key enzymes that regulate endothelial-cell motility — and, if so, how? How do Notch and VEGF signal to PFK2? Does glycolysis regulate migration of other cell types, such as smooth-muscle cells or macrophages, or even cancer cells? And is the pro-angiogenic activity of PFK2 altered when glucose homeostasis is perturbed, such as in diabetes?

These questions aside, De Bock and colleagues' study deepens our understanding of why some cells choose to forego the lucrative use of mitochondria to break down their glucose, even when, as is the case for endothelial cells, the cells are not highly replicative. The authors' findings also introduce a new concept in endothelial biology: that metabolic decisions can regulate the endothelial phenotype, as well as vice versa. It turns out that, much like children, endothelial cells that gorge on sugar become hyperactive.

Onset of deglacial warming in West Antarctica driven by local orbital forcing

The cause of warming in the Southern Hemisphere during the most recent deglaciation remains a matter of debate^1, 2. Hypotheses for a Northern Hemisphere trigger, through oceanic redistributions of heat, are based in part on the abrupt onset of warming seen in East Antarctic ice cores and dated to 18,000 years ago, which is several thousand years after high-latitude Northern Hemisphere summer insolation intensity began increasing from its minimum, approximately 24,000 years ago^3, 4. An alternative explanation is that local solar insolation changes cause the Southern Hemisphere to warm independently^2, 5. Here we present results from a new, annually resolved ice-core record from West Antarctica that reconciles these two views. The records show that 18,000 years ago snow accumulation in West Antarctica began increasing, coincident with increasing carbon dioxide concentrations, warming in East Antarctica and cooling in the Northern Hemisphere⁶ associated with an abrupt decrease in Atlantic meridional overturning circulation⁷. However, significant warming in West Antarctica began at least 2,000 years earlier. Circum-Antarctic sea-ice decline, driven by increasing local insolation, is the likely cause of this warming. The marine-influenced West Antarctic records suggest a more active role for the Southern Ocean in the onset of deglaciation than is inferred from ice cores in the East Antarctic interior, which are largely isolated from sea-ice changes.

Exceptional records of Southern Hemisphere climate change come from Antarctic ice cores^2, 6, 7. Most of these records are from high-altitude sites on the East Antarctic plateau. Questions about the reliability of the two previous deep West Antarctic ice-core records result in those records often being excluded from reconstructions of Antarctic climate^4, 8. Because the climate of West Antarctica is distinct from that of interior East Antarctica, the exclusion of West Antarctic records may result in an incomplete picture of past Antarctic and Southern Ocean climate change. Interior West Antarctica is lower in elevation and more subject to the influence of marine air masses than interior East Antarctica, which is surrounded by a steep topographic slope^9, 10. Marine-influenced locations are important because they more directly reflect atmospheric conditions resulting from changes in ocean circulation and sea ice. However, ice-core records from coastal sites are often difficult to interpret because of complicated ice-flow and elevation histories. The West Antarctic Ice Sheet (WAIS) Divide ice core (WDC), in central West Antarctica, is unique in coming from a location that has experienced minimal elevation change¹¹, is strongly influenced by marine conditions⁹ and has a relatively high snow-accumulation rate, making it possible to obtain an accurately dated record with high temporal resolution.

Drilling of WDC was completed in December 2011 to a depth of 3,405 m. Drilling was halted ~50 m above the bedrock to avoid contaminating the basal water system. WDC is situated 24 km west of the Ross–Amundsen ice-flow divide and 160 km east of the Byrd ice-core site (Supplementary Fig. 1). The elevation is 1,766 m; the present-day snow accumulation rate is 22 cm yr⁻¹ (ice equivalent) and the average temperature is approximately −30 °C. The age of the oldest recovered ice is ~68 kyr. The WDC06A-7 timescale is based on the identification of annual layers to 29.6 kyr ago using primarily electrical measurements (Methods). To validate WDC06A-7, we compare times of abrupt changes in atmospheric methane concentration (Supplementary Information) with the Greenland Ice Core Chronology 2005¹² (GICC05). We also compare the methane variations in WDC with abrupt changes in a speleothem δ¹⁸O record from Hulu Cave, China. The difference in age between the ice and gas at a given depth is calculated using a steady-state firn-densification model and is always less than 500 yr. The age differences between WDC06A-7 and GICC05 and between WDC06A-7 and the Hulu Cave timescale are much less than the independent timescale uncertainties (Supplementary Fig. 6).

We interpret δ¹⁸O of ice (Methods) as annual-mean surface air temperature, as supported by independent estimates of temperature from borehole thermometry¹³. WDC has many similarities with other records (Fig. 1) and resolves Antarctic Isotope Maximum (AIM) events clearly. The late Holocene WDC record shows cooling, suggesting that the increase in δ¹⁸O at Byrd over the past few thousand years resulted from ice advection and thinning¹¹. The abrupt increase in δ¹⁸O ~22 kyr ago at Siple Dome is not observed at WDC. The AIM1 peak and the subsequent Antarctic Cold Reversal (ACR; 14.5–12.9 kyr ago) are more pronounced in WDC than at Byrd and Siple Dome, possibly owing to discontinuous sampling of the Byrd core and thinning of Siple Dome......................................

...............

Cross-neutralization of four paramyxoviruses by a human monoclonal antibody

Broadly neutralizing antibodies reactive against most and even all variants of the same viral species have been described for influenza and HIV-1 (ref. 1). However, whether a neutralizing antibody could have the breadth of range to target different viral species was unknown. Human respiratory syncytial virus (HRSV) and human metapneumovirus (HMPV) are common pathogens that cause severe disease in premature newborns, hospitalized children^2, 3 and immune-compromised patients^2, 4, 5, and play a role in asthma exacerbations⁶. Although antisera generated against either HRSV or HMPV are not cross-neutralizing⁷, we speculated that, because of the repeated exposure to these viruses, cross-neutralizing antibodies may be selected in some individuals. Here we describe a human monoclonal antibody (MPE8) that potently cross-neutralizes HRSV and HMPV as well as two animal paramyxoviruses: bovine RSV (BRSV) and pneumonia virus of mice (PVM). In its germline configuration, MPE8 is HRSV-specific and its breadth is achieved by somatic mutations in the light chain variable region. MPE8 did not result in the selection of viral escape mutants that evaded antibody targeting and showed potent prophylactic efficacy in animal models of HRSV and HMPV infection, as well as prophylactic and therapeutic efficacy in the more relevant model of lethal PVM infection. The core epitope of MPE8 was mapped on two highly conserved anti-parallel β-strands on the pre-fusion viral F protein, which are rearranged in the post-fusion F protein conformation. Twenty-six out of the thirty HRSV-specific neutralizing antibodies isolated were also found to be specific for the pre-fusion F protein. Taken together, these results indicate that MPE8 might be used for the prophylaxis and therapy of severe HRSV and HMPV infections and identify the pre-fusion F protein as a candidate HRSV vaccine.

Pathogen blocks host death receptor signalling by arginine GlcNAcylation of death domains

The tumour necrosis factor (TNF) family is crucial for immune homeostasis, cell death and inflammation. These cytokines are recognized by members of the TNF receptor (TNFR) family of death receptors, including TNFR1 and TNFR2, and FAS and TNF-related apoptosis-inducing ligand (TRAIL) receptors¹. Death receptor signalling requires death-domain-mediated homotypic/heterotypic interactions between the receptor and its downstream adaptors, including TNFR1-associated death domain protein (TRADD) and FAS-associated death domain protein (FADD)². Here we discover that death domains in several proteins, including TRADD, FADD, RIPK1 and TNFR1, were directly inactivated by NleB, an enteropathogenic Escherichia coli (EPEC) type III secretion system effector known to inhibit host nuclear factor-κB (NF-κB) signalling^3, 4. NleB contained an unprecedented N-acetylglucosamine (GlcNAc) transferase activity that specifically modified a conserved arginine in these death domains (Arg 235 in the TRADD death domain). NleB GlcNAcylation (the addition of GlcNAc onto a protein side chain) of death domains blocked homotypic/heterotypic death domain interactions and assembly of the oligomeric TNFR1 complex, thereby disrupting TNF signalling in EPEC-infected cells, including NF-κB signalling, apoptosis and necroptosis. Type-III-delivered NleB also blocked FAS ligand and TRAIL-induced cell death by preventing formation of a FADD-mediated death-inducing signalling complex (DISC). The arginine GlcNAc transferase activity of NleB was required for bacterial colonization in the mouse model of EPEC infection. The mechanism of action of NleB represents a new model by which bacteria counteract host defences, and also a previously unappreciated post-translational modification.

Bacterial colonization factors control specificity and stability of the gut microbiota

Mammals harbour a complex gut microbiome, comprising bacteria that confer immunological, metabolic and neurological benefits¹. Despite advances in sequence-based microbial profiling and myriad studies defining microbiome composition during health and disease, little is known about the molecular processes used by symbiotic bacteria to stably colonize the gastrointestinal tract. We sought to define how mammals assemble and maintain the Bacteroides, one of the most numerically prominent genera of the human microbiome. Here we find that, whereas the gut normally contains hundreds of bacterial species^2, 3, germ-free mice mono-associated with a single Bacteroidesspecies are resistant to colonization by the same, but not different, species. To identify bacterial mechanisms for species-specific saturable colonization, we devised an in vivo genetic screen and discovered a unique class of polysaccharide utilization loci that is conserved among intestinalBacteroides. We named this genetic locus the commensal colonization factors (ccf). Deletion of theccf genes in the model symbiont, Bacteroides fragilis, results in colonization defects in mice and reduced horizontal transmission. The ccf genes of B. fragilis are upregulated during gut colonization, preferentially at the colonic surface. When we visualize microbial biogeography within the colon, B. fragilis penetrates the colonic mucus and resides deep within crypt channels, whereasccf mutants are defective in crypt association. Notably, the CCF system is required for B. fragiliscolonization following microbiome disruption with Citrobacter rodentium infection or antibiotic treatment, suggesting that the niche within colonic crypts represents a reservoir for bacteria to maintain long-term colonization. These findings reveal that intestinal Bacteroides have evolved species-specific physical interactions with the host that mediate stable and resilient gut colonization, and the CCF system represents a novel molecular mechanism for symbiosis.

Gene regulation: Long RNAs wire up cancer growth

The discovery of long non-coding RNAs that control the liaisons between a transcription factor with a key role in prostate cancer and its target genes sheds light on how RNAs dictate information flow in the cell nucleus.

The human genome generates more than 10,000 long non-coding RNA (lncRNA) molecules, yet the functions of only several dozen of these transcripts are known. In a study published on Nature's website today, Yang et al.¹ describe two lncRNAs that bind to, and govern the function of, the androgen receptor — a transcription factor that activates the expression of thousands of genes in response to the hormone androgen. The authors find that inhibition of these two lncRNAs can block the growth of prostate-cancer cells that are resistant to hormone therapy owing to a mutation in the androgen receptor.

Androgen signalling has been recognized as the principle driver of prostate-cancer growth since the 1940s, when castration was first shown to slow the progression of prostate cancer². Consequently, androgen-deprivation therapy, which can involve castration, has been the mainstay of treatment for advanced prostate cancer and achieves excellent response rates. However, in most men the cancer becomes resistant within two years of starting this therapy. It is now clear that such resistance arises by various mechanisms that reconstitute androgen-receptor signalling².

A large proportion of lncRNAs associate with regulators of chromatin (DNA–protein complexes)³, and several are known to 'tag' specific chromosomal regions with particular chromatin marks that modulate the availability of the associated genes for expression⁴. Several prostate-cancer-specific lncRNAs have been identified, and some are associated with distinct subtypes of the disease⁵. In 2012, the US Food and Drug Administration approved the use of the lncRNA PCA3 for the detection of prostate cancer⁶. But, despite the discovery of multiple cancer-related lncRNAs, functions for most of these remain unknown.

Yang et al. report that PRNCR1 and PCGEM1, two lncRNAs seen at high levels in many aggressive prostate cancers, enhance androgen-receptor-associated transcriptional programs to promote this cancer's growth. The authors' ChIRP analysis (a recently developed method of RNA localization⁷) reveals that these lncRNAs localize to distal androgen-response elements on chromatin, co-localizing with the androgen receptor (Fig. 1). In a surprisingly intricate series of events, PRNCR1 interacts with the acetylated carboxy terminus of the androgen receptor and recruits the enzyme DOT1L to methylate the amino terminus of this receptor; this step is necessary for the subsequent association of the androgen receptor with PCGEM1. PCGEM1 then interacts with the protein Pygo2, which can bind to H3K4me3 — a methyl mark on the chromatin-associated histone H3 protein that is prevalent at gene-promoter 

Yang and colleagues further show that PCGEM1 and Pygo2 facilitate 'looping' between enhancer and promoter sequences — a recently recognized role for lncRNAs⁸ — and thus activation of androgen-receptor target genes. Moreover, even a truncated androgen receptor that is always active regardless of the presence of androgen depends on these lncRNAs. And lncRNA inhibition blocks the growth of both androgen-dependent and androgen-independent cancer cells. If these results can be replicated in subsequent studies, they offer a notable advance in the fight against advanced prostate cancer, by highlighting lncRNAs as a class of drug targets.

In addition to their relevance to disease, the current results illuminate several fundamental molecular mechanisms. PRNCR1 and PCGEM1 underscore a new role of RNA interaction with sequence-specific DNA-binding proteins — modification of transcription-factor activity. The liaisons between lncRNAs and transcription factors can program step-wise chemical modifications on transcription factors, gating the successive flow of information from enhancer engagement to target-gene activation.

The insights that these findings provide into how lncRNAs can mediate enhancer–promoter looping are also intriguing. The RNA-mediated recruitment of a protein with intrinsic avidity for a promoter-associated histone mark to distantly located enhancer elements could stabilize DNA looping and promote communication over three-dimensional space. This would mean that, rather than being simple scaffolds, lncRNAs are more akin to a complex computer circuit board, linking together various disparate molecular components and dictating the logical operation of the system.

Yang and co-workers' study also raises a plethora of questions. For example, the researchers report a series of apparently specific RNA–protein interactions, the molecular details of which should be investigated. It is not clear whether all of the components involved in these interactions have been identified, and whether similar mechanisms operate for other steroid-hormone receptors or transcription factors.

Moreover, how PRNCR1 and PCGEM1 levels are controlled remains unclear. The authors' data suggest that defined ranges of PRNCR1 and PCGEM1 levels are crucial for proper androgen signalling, but mechanisms by which lncRNA biogenesis and increases in the levels of these RNAs occur in prostate cancer are not known. With the recent recognition that many lncRNAs are dysregulated in human cancers⁹, understanding how these transcripts are generated and regulated is likely to take a central place in this field of research. The intricacies of the RNA circuits motivate the development of precise RNA-engineering technologies to reprogram nuclear signalling and for cancer diagnostic and prevention purposes, rather than simply modulating lncRNA levels.