AST ROPHYSICS

Stars fight back Galaxies contain fewer stars than predicted. The discovery of a massive galactic outflow of molecular gas in a compact galaxy, which forms stars 100 times faster than the Milky Way, may help to explain why. See Letter p.68 P H I L I P F. H O P K I N S

T

he past two decades have seen a revolution in our understanding of the origins and evolution of the cosmos. The standard hypothesis of a Universe domi­ nated by ‘cold dark matter’ has been extraordi­ narily successful in explaining a vast number of astronomical observations. But unsolved puzzles abound when we consider what cosmologists call small scales: namely, single galaxies. According to the simplest cosmologi­ cal models, nearly all the baryonic matter — the ‘normal’ matter in the Universe — should sit at the centres of galaxies in the form of stars. But baryonic matter apparently does no such thing. On page 68 of this issue, Geach et al.1 describe the discovery of a galaxy caught in the act of violently expelling its baryonic mat­ ter, revealing a process that may be crucial to solving this mystery. The problem is an old one. Regardless of the nature of dark matter, many diverse observa­ tions indicate that only a few per cent of the normal matter in the Universe is in the form of stars and gas concentrated inside galaxies. If we start from the simplest assumptions, with

a certain amount of gas in the early Universe, and evolve it forward in time according to the laws of gravity and chemistry, the gas will eventually condense into increasingly dense objects, forming galaxies and, within those galaxies, stars. Inevitably, this leads to the pre­ diction that nearly all baryonic matter should be in stars. However, only part of it is. In galaxies such as the Milky Way, in which the conver­ sion of gas into stars is known to be relatively efficient, only 10–20% of the normal matter that gravity should have dragged in has actu­ ally been retained inside the galaxy (mostly in stars). The rest is thought to be outside the galaxy, in the diffuse intergalactic medium. How might this occur? Theorists have long speculated that feedback mechanisms may be important. Stars are not simply passive spec­ tators. When massive stars form, they soon explode as supernovae, injecting tremendous energy into the surrounding gas. Because this energy is much larger than the gravita­ tional energy holding that gas in place, it is clear that small galaxies should be able to eject most of their normal matter in a galactic-scale outflow after forming just a small number of stars2. However, in more-massive galaxies

Figure 1 | A galactic outflow.  This image of starburst galaxy M82 shows its disk of gas and stars (blue) and a perpendicular outflow of ionized gas (red). 4 4 | NAT U R E | VO L 5 1 6 | 4 D E C E M B E R 2 0 1 4

© 2014 Macmillan Publishers Limited. All rights reserved

(those bigger than our own), this energy source seems to be insufficient for such gas ejection. One idea is that the feedback in these sys­ tems may come from the energy released as matter falls (accretes) into the supermassive black holes — with masses from one million to ten trillion times that of the Sun — that are known to lie at the centres of these galaxies3,4. These accreting black holes shine as quasars or their less luminous analogues, active galactic nuclei (AGNs), and are among the most ener­ getic objects in the Universe. Geach and colleagues’ team has been working to uncover the effects of feedback on galaxies. But it is a challenging task. Although many observations have shown that galactic out­ flows (Fig. 1) are ubiquitous when galaxies are rapidly forming stars5, it has been difficult to measure the mass of these winds, and hence the momentum and energy required to form them. The observations have largely indicated the signatures of galactic outflows in spectro­ scopic features linked to the wind’s absorption of galactic light, because the wind lies between the galaxy and the observers. But turning those results into a mass requires detailed knowledge of where exactly the wind lies, and what frac­ tion of the wind is in the precise atomic- and molecular-gas species and ionization state associated with the spectroscopic features. With their new observations of a galactic outflow, Geach et al. have obtained a much more direct tracer of the mass of an outflow than has been possible so far. By measuring the emission luminosity of carbon monoxide, which is known to be an accurate tracer of the total mass of dense molecular gas under interstellar conditions, the authors realized that most of the material in this outflow was ‘hiding’ in the cold molecular phase, allowing several properties of the outflow to be deter­ mined. Measuring the outflow rate of any gal­ axy, especially one so far away (the light from the galaxy reported here was emitted when the Universe was approximately half its current age) would be a great achievement, but what the authors found was much more remarkable. The galaxy, dubbed SDSS J0905 + 57, drives a wind with an incredible velocity. The speed of the material flying out from the galaxy reaches 2,500 kilometres per second, much greater than the speeds of about 200–500 km s−1 typi­ cally seen around more mundane galaxies. The outflow rate is roughly 100 solar masses per year, the same order of magnitude as the rate at which the galaxy is forming stars (and 100 times higher than the rate of star formation in the Milky Way). If the galaxy can maintain this prodigious outflow, it could empty its res­ ervoir of molecular gas in 10 million years. The wind’s velocity places it among the most extreme known. How could such an outflow be powered? The authors see no AGN in the system, and the energy in supernovae seems insufficient to drive the outflow. But another remarkable

J. GALLAGER (UNIV. WIS.), M. MOUNTAIN (STSCI), P. PUXLEY (NSF)/HUBBLE HERITAGE TEAM STSCI/AURA/ESA/NASA

RESEARCH NEWS & VIEWS

NEWS & VIEWS RESEARCH property of the galaxy may suggest an answer. SDSS J0905 + 57 is extremely compact: half of its huge star-formation rate comes from a region less than 100 parsecs across (the Milky Way is about 20,000 parsecs across). With so much light being emitted from such a small region, the momentum being imparted by the starlight itself on any nearby gas becomes a highly powerful force. The authors show that this ‘radiation pressure’ could indeed accel­ erate the wind to such a high velocity that it would sweep up most of the molecular gas. This would also represent a gentler way of pushing on the gas than through supernovae, which generate strong shocks. Such a mecha­ nism would avoid destroying the molecular bonds in the gas. A similar process leads to the Eddington limit — the maximum luminosity of a star. What the authors see here is a truly ‘Eddington-limited galaxy’. Such a radiation-pressure mechanism represents an exciting and unusual channel for feedback. Although radiation pressure has been proposed as a potential feedback chan­ nel6, it has only just begun to be included in models, and has remained observationally elu­ sive. That may change rapidly now. Of course, uncertainties remain. Perhaps the outflow in SDSS J0905 + 57 is, after all, powered by a supermassive black hole shining as an AGN. AGNs are notoriously fickle. Because black holes are so small, the brightness of AGNs can fluctuate dramatically on short timescales. It is therefore hard to rule out the possibility that an AGN ‘lit up’ the galaxy a few million years ago and launched the observed winds, only to fade away before we could see it. At present, we have Occam’s razor (the principle of simpli­city) as our guide, reminding us that we need not invoke anything other than the observed starlight to explain the winds. How­ ever, future observations by the authors could test this hypothesis definitively by building a statistical sample of galaxies that have similar properties to those of SDSS J0905 + 57. In any case, this object represents a remarkable new class of system, one with violent, massive outflows of dense molecular gas. Geach and colleagues caught a galaxy in the act of ‘fighting back’ against gravity, and forever changing its own future. ■

N EUR O BI OLOGY

A molecular knife to dice depression Chronic stress can cause depression in some individuals, but leaves others untouched. Engagement of a molecular pathway controlling the production of tiny RNA snippets might help to explain the difference. See Article p.51 G E R H A R D S C H R AT T

S

tress and adversity can put the balance of our lives at risk — the death of a loved one, loss of a job or other financial or per­ sonal problems can have detrimental effects on the human psyche. Although some peo­ ple become depressed in such circumstances, others are more resilient and can successfully cope with difficult experiences. Resilience was first documented in the 1970s, but the molec­ ular underpinnings of this phenomenon are still mostly elusive1. In this issue, Dias et al.2 (page 51) report that behavioural resilience in mice is promoted by the β-catenin protein, through control of the enzyme Dicer1. Research over the past decade shows that a complex interplay between genetic and envi­ ronmental factors underlies mood disorders, and points to some key molecular pathways involved in such conditions3. This work has also led to the development of antidepres­ sants, but such drugs are effective in only about

a

b

Depression

β-Catenin

Nucleus

two-thirds of people with depression4. A better understanding of the molecular pathways active in resilient individuals could inform us of alternative treatment strategies5. β-Catenin controls brain function6, and molecules regulating β-catenin activity have been implicated in depression7. Dias and col­ leagues analysed the role of β-catenin in resil­ ience, using a well-established mouse model of depression — chronic social defeat stress (CSDS)8. Briefly, male mice are repeatedly exposed to males from a physically superior strain. The encounter regularly ends with the defeat of the test mice. Although the fights do not usually result in severe physical injury, the repeated defeat leaves psychological scars that ultimately manifest in depressive behaviour. For example, mice that have experienced CSDS will largely avoid social contact, a symptom also observed in people with depression9. The authors focused on a specific brain region, the nucleus accumbens (NAc), which is part of the forebrain. Although the NAc is

Medium spiny neuron

Resilience

Other effectors?

Cytoplasm

Dicer1 β-Catenin

Dicer1

mRNA

Antiresilience proteins

miRNAs

Philip F. Hopkins is at the California Institute of Technology, Pasadena, California 91125, USA. e-mail: [email protected] 1. Geach, J. E. et al. Nature 516, 68–70 (2014). 2. Dekel, A. & Silk, J. Astrophys. J. 303, 39–55 (1986). 3. Croton, J. et al. Mon. Not. R. Astron. Soc. 365, 11–28 (2006). 4. Hopkins, P. F. et al. Astrophys. J. 625, L71–L74 (2005). 5. Heckman, T. M., Armus, L. & Miley, G. K. Astrophys. J. Suppl. 74, 833–868 (1990). 6. Murray, N., Quataert, E. & Thompson, T. A. Astrophys. J. 618, 569–585 (2005).

Figure 1 | A resilience switch.  a, Dias et al.2 report that D2-type medium spiny neurons are not activated in ‘depressed’ mice. As a consequence, β-catenin protein remains in the cytoplasm in these cells, unable to enter the nucleus, and the Dicer1 gene is thus inactive. ‘Anti-resilience’ proteins may therefore be produced from messenger RNA that would otherwise have been inhibited by microRNAs (miRNAs) generated by the Dicer1 protein. b, In resilient mice, β-catenin enters the nucleus of activated neurons, thereby turning on Dicer1 transcription. Elevated levels of Dicer1 protein increase production of miRNAs and possibly other effectors of resilience. This might, in turn, inhibit the production of anti-resilience proteins, because of binding and inhibition of mRNA by miRNAs. 4 D E C E M B E R 2 0 1 4 | VO L 5 1 6 | NAT U R E | 4 5

© 2014 Macmillan Publishers Limited. All rights reserved

Astrophysics: stars fight back.

Astrophysics: stars fight back. - PDF Download Free
944KB Sizes 3 Downloads 5 Views