trp genes has been shown to encode a chan-nel with the precise ion-conduction proper-ties expected of store-operated channels.Indeed, in most (but not all) cases, the chan-nels produced by experimental expression oftrp genes appear to be nonselective cationchannels, whereas the capacitative Ca2&-entry channels are thought to be among themost selective Ca2& channels6.

There are seven close mammalian rela-tives of the fruitfly trp gene, and there is also amuch larger family of more distantly relatedgenes. This larger family can be divided intothree subfamilies, based on similarities in thestructures of the encoded proteins7. The fruitfly TRP protein and its seven closerelatives are ‘short’ TRP channel proteins.Members of a second group are ‘long’ TRPchannel proteins.

The third group comprises ‘osm’ chan-nels, so called because they appear to be reg-ulated by physical stimuli such as osmoticconcentration, heat or mechanical stress.Two members of this family — called ECaCand CaT1 — were first identified as a result ofsearches for channels involved in transport-ing Ca2& across epithelial cells8,9, such asthose that line the gut. Perhaps, from the ini-tial characterization of these channel pro-teins, Yue et al.2 recognized that they mighthave properties similar to those of a class ofstore-operated channels, the Ca2&-release-activated Ca2& (CRAC) channels6, that havebeen studied intensively by electrophysio-logical means. Although their molecularcomponents are not known, these channelsare among the most selective of known Ca2&

channels6, being able to distinguish Ca2&

from Ba2& ions. CRAC channels also exhibitcharacteristic behaviours under a variety ofexperimental conditions. For example, theyare regulated by changes in cytoplasmic Ca2&

levels, and show dramatic changes in sel-ectivity and conductivity in the presence oflow extracellular concentrations of divalentcations.

Yue et al.2 have now expressed DNA cod-ing for the CaT1 protein in a mammalian cellline, and show that several of the propertiesof the encoded proteins closely resemblethose of CRAC channels. Notably, as well asthere being qualitative similarities betweenCRAC channels and CaT1, single CaT1channels conduct monovalent ions in quan-titatively similar ways to CRAC channelswhen there are no divalent cations in thebathing medium.

The authors also show that the expressedCaT1 channels have the signature propertyof capacitative Ca2&-entry channels — theyare activated by depletion of intracellularCa2& stores. Interestingly, the authors seethis only when lower than maximal amountsof CaT1 are expressed. This may reflect a lim-iting quantity of other cellular factors need-ed to regulate the channel. In other words,CaT1 may not be regulated directly by deple-

tion of intracellular Ca2& stores, but ratherindirectly, through such cellular factors.

The weight of evidence supports Yue etal.’s conclusion that CaT1 (or perhaps, insome instances, its close relative ECaC) con-stitutes the ion-conducting pore of CRACchannels. But there are also store-operatedCa2&-conducting channels, known fromelectrophysiological studies, that have prop-erties distinct from those of CRACchannels5. It is possible, and even likely, thatother channel proteins — such as membersof the more studied ‘short’ TRP family10,11 —form part of these channels. In the nearfuture, I anticipate continuing progress inthe search for the complete molecular defini-tion of the capacitative Ca2&-entry channels,as well as a solution to the mystery of howthey are regulated. ■

James W. Putney Jr is in the Laboratory of Signal

Transduction, National Institute of EnvironmentalHealth Sciences, National Institutes of Health,Research Triangle Park, North Carolina 27709,USA.e-mail: putney@niehs.nih.gov1. Putney, J. W. Jr Capacitative Calcium Entry (Landes Biomedical

Publishing, Austin, TX, 1997).

2. Yue, L., Peng, J.-B., Hediger, M. A. & Clapham, D. E. Nature

410, 705–709 (2001).

3. Berridge, M. J. Nature 361, 315–325 (1993).

4. Birnbaumer, L. et al. Proc. Natl Acad. Sci. USA 93, 15195–15202

(1996).

5. Putney, J. W. Jr & McKay, R. R. Bioessays 21, 38–46 (1999).

6. Parekh, A. B. & Penner, R. Physiol. Rev. 77, 901–930 (1997).

7. Harteneck, C., Plant, T. D. & Schultz, G. Trends Neurosci. 23,

159–166 (2000).

8. Peng, J.-B. et al. J. Biol. Chem. 274, 22739–22746 (1999).

9. Hoenderop, J. G. J. et al. J. Biol. Chem. 274, 8375–8378 (1999).

10.Boulay, G. et al. Proc. Natl Acad. Sci. USA 96, 14955–14960

(1999).

11.Freichel, M. et al. Nature Cell Biol. 3, 121–127 (2001).

12.Randriamampita, C. & Tsien, R. Y. Nature 364, 809–814 (1993).

13. Irvine, R. F. FEBS Lett. 263, 5–9 (1990).

14.Berridge, M. J. Biochem. J. 312, 1–11 (1995).

objective — that is, there are different meansto the same end — so tradeoffs between thegases can critically affect the overall cost.

In regard to their concentrations in theatmosphere, the equivalents among green-house gases are relatively straightforward.They all affect radiative forcing: the re-radia-tion of heat by greenhouse-gas moleculesback towards Earth’s surface, usuallyexpressed relative to a preindustrial histori-cal baseline. Radiative forcing is transformedby complex feedbacks — for example, cloudformation — into features such as averagesurface temperature.

In principle, one can calculate the radia-

news and views

NATURE | VOL 410 | 5 APRIL 2001 | www.nature.com 649

In designing policy to check global warm-ing, it is useful to think in terms of anaggregate of various greenhouse gases.

The Kyoto Protocol of 1997, for example,uses this idea in prescribing limits on emis-sions by industrialized countries. The proto-col takes in six gases, aggregated into CO2

equivalents by applying the concept of ‘glob-al warming potential’ (GWP); this is a purelyphysical measure adopted by the Intergov-ernmental Panel on Climate Change1 andinspired by a similar concept used in the context of ozone2.

In their paper on page 675 of this issue3,Manne and Richels bring economics intoconsideration by use of a model that inte-grates physical and economic systems. Theylook at tradeoffs, the principle that a nationcan emit more of one gas if it emits less ofanother, in particular at tradeoffs betweenmethane and nitrous oxide — the two majorgreenhouse gases other than CO2 — andCO2. They show that such tradeoffs meet dif-ferent policy aims, in terms of intended effecton the climate over a certain time, than thosemet by GWPs; that they are highly sensitiveto the details of the specific aims; and thatthey are likely to have to vary considerablyover time.

Reducing greenhouse-gas emissions canonly be done at a cost in goods and servicesforgone. That cost varies across time accord-ing to the gas concerned. There is a continu-um of emission trajectories for the severalgases that will achieve any given climate

Global change

Time, money and tradeoffsDavid F. Bradford

Measurement of the total emissions of greenhouse gases impliesspecifying ‘tradeoffs’ of one against another, as in the Kyoto Protocol.In that process, economics has to be taken into account.

AP

In tackling global warning, the economicperspective matters. But does the United Statesmean business? (See News section.)

© 2001 Macmillan Magazines Ltd

tive forcing of given concentrations of the sixgreenhouse gases at a given point in time.And one can calculate the amount of CO2 inthe atmosphere that, taken alone, wouldresult in the same radiative forcing. Thiswould then be the ‘CO2 equivalent’ concen-tration. But it is not obvious how to defineCO2-equivalent emissions of differentgreenhouse gases: such emissions affect gasconcentrations in the future, as well as today,because each of them has a different lifetimein the atmosphere.

The GWP of a gas is a measure of such anequivalence. The starting point is a baselinetrajectory of radiative forcing over time. If, atsome moment, an extra unit (say a tonne) ofa gas is injected into the atmosphere, the tra-jectory is perturbed from that moment on.One measure of the resulting warming effectfrom that moment to some specified timehorizon is simply the integral of that pertur-bation in radiative forcing. The GWP of a gasis the ratio of that integral to the samemeasure (out to the same horizon) associat-ed with injection of a unit of CO2.

In general, natural scientists have beenattracted to the GWP concept because of itspurely physical quality. Although econo-mists have argued that the tradeoffs cannotbe inferred from physical properties alone,but have an inherent economic and policydimension in terms of targets set4–7, the mes-sage has been slow to be accepted in the scientific community8,9.

Using a computer model that has beenused for various purposes for several years,Manne and Richels3 show both that the eco-nomic perspective matters and that it is possi-ble to implement it. They emphasize thatthere remains much uncertainty about theparameters of their model, and policy objec-tives. The important point about their resultsis less the specific tradeoffs that they derivethan the cost-effectiveness logic they apply.

From a technical point of view, climatechange is simpler than many environmentalchallenges: here we are dealing with a set ofwell-mixed pollutants, each of which has thesame effect. The policy interest in the emis-sions of these gases results from their effectson climate change, but this depends on onething: the trajectory of radiative forcing.That is, we can completely characterize theconsequence of an incremental tonne emit-ted of a greenhouse gas, holding constant allother emission flows through time, by theperturbation in the trajectory of radiativeforcing that it induces. The perturbation thatresults from an incremental tonne varies a lotfrom gas to gas (and with the ambient atmos-pheric conditions, although this effect is gen-erally ignored in these exercises).

To say that emission of an incrementaltonne of one gas has the same implicationsfor policy as emission of x tonnes of anothergas means that we have to assign a value tothe change in radiative forcing at different

times in the future. There is no way to avoidthis step. Using the physical GWP involves animplicit evaluation: a bit of extra radiativeforcing at any time up to the chosen horizonhas the same (negative) value as the same bitat any other time within that span; an extrabit beyond the horizon has zero value. This isclearly wrong.

The grand aim of optimizing policy onclimate change involves placing a value onthe costs or benefits of such change — say of a2 7C rise in temperature. That is a highly con-troversial step. But suppose we had somehowsolved that grand problem. The solutionwould involve a particular path of radiativeforcing. Presumably, if we have really solvedthe grand problem, the emissions of the vari-ous greenhouse gases minimize the cost ofattaining that path. Manne and Richels haverecognized that we do not know the overallcosts and benefits of climate change: insteadof solving the grand problem they havesolved a cost-effectiveness subproblem.

They take as given a desired limit on globalaverage temperature, and use their model tosolve the cost-minimizing way to achieve thespecified target by controlling emissions ofthe various greenhouse gases. (This involvesdetermining a path of radiative forcing, aswell as the cost-minimizing paths of green-house-gas emissions.) Associated with asolution to this problem will be ‘shadowprices’ which answer the question of howmuch the cost of meeting the desired policyobjective would be reduced if we wereallowed to emit one extra tonne of gas g attime t. The ratio of that quantity for gas g tothat quantity for CO2 is the tradeoff we areafter, and which Manne and Richels calculate.

The significance of Manne and Richels’analysis is twofold. First, they have brought apowerful quantitative model to bear on theproblem of aggregating greenhouse gases inthe implementation of climate policy. Sec-ond, the general point that they emphasize isstill far from generally understood. I oftenencounter the view that, arbitrary as theymay be, GWPs constitute a reasonable shot atthe set of tradeoffs that are appropriate forKyoto-style regulation. But this is nonsensewithout a concept of what ‘appropriate’means, which in turn forces us to confrontthe specifics of getting the tradeoffs right. ■

David F. Bradford is in the Woodrow Wilson School,Princeton University, Princeton, New Jersey 08544-1013, USA.e-mail: bradford@princeton.edu1. Schimel, D. et al. in Climate Change 1995: The Science of

Climate Change (eds Houghton, J. T. et al.) (Cambridge Univ.Press, 1996).

2. Lashof, D. A. & Ahuja, D. R. Nature 364, 529–531 (1990).3. Manne, A. S. & Richels, R. G. Nature 410, 675–677 (2001).4. Eckaus, R. S. Energy J. 13, 25–35 (1992).5. Reilly, J. M. & Richards, K. H. Environ. Res. Econ. 3, 41–61

(1993).6. Schmalensee, R. Energy J. 14, 245–255 (1993).7. Wallis, M. K. & Lucas, N. J. D. Int. J. Energy Res. 18, 57–62

(1994).8. Smith, S. J. & Wigley, T. M. L. Clim. Change 44, 445–457 (2000).9. Smith, S. J. & Wigley, T. M. L. Clim. Change 44, 459–469 (2000).

news and views

650 NATURE | VOL 410 | 5 APRIL 2001 | www.nature.com

Daedalus

Warping spaceSome while ago, Daedalus devised amethod of warping space itself. He used abig capacitor, across which an a.c. electricfield maintained an a.c. displacementcurrent. All dielectrics maintain such acurrent, as their electrons are shifted bythe electric field; a magnetic field imposedon the dielectric therefore exerts a forceupon it. Sadly, a resonant system with themagnetic field created by the same coilsthat generate the electric one fails to work (the resonance comes out wrong), so the magnetic field has to be speciallygenerated.

The original idea was to use the deviceas an aero engine, propelling air as thedielectric. But Daedalus soon realized thatthe system has a remarkable property.Suppose the dielectric is a pure vacuum.This sustains a displacement current withthe best of them. Yet the magnetic field,which imposes its force on any current-carrying conductor, is now exerting thatforce on space itself.

Thus Daedalus’s gadget is an ideal wayof studying the ‘flexibility’ of pure space.Cosmology currently holds that the entireUniverse is expanding, like a balloon withmarkings on it being inflated — themarkings in this case being galaxies, whichall seem to be receding from one another.This expansion may be an energetic relic ofthe Big Bang, or it may be maintained bysome unrecognized force. Daedalus nowplans to measure that force.

His scheme is to set up a line of hiselectromagnetic thrusters, all pushing onspace in the same direction, and to shine alaser beam across it. When the thrustersare turned on, the beam (travelling in thespace above them) will be deflected.Interferometric measurements on thebeam will thus calibrate the thrustersagainst the resulting beam deflection.

From the results, Daedalus hopes tomeasure the force sustaining the observedexpansion of the Universe, and to relate itto that exerted by the Big Bang. Of course,space-warp technology has already beendeveloped so enthusiastically by science-fiction writers that Daedalus sees no pointin entering the business himself. Instead, hehopes that the measurements show extremesensitivity in the beam. This would implythat the galaxies are expanding purelyballistically, as suggested by the Big Bangtheory. But a small force, either positive ornegative, may be needed to square anexactly closed Universe with the Big Bangtheory as currently understood. Suchfundamental measurements are alwaysworth making. David Jones

© 2001 Macmillan Magazines Ltd