EFTA00610311.pdf

REPORTS' Fig. 3. Drag coefficient as nos 4. M. D. Powell, P. J. Vickery, 1. A. Reinhold, Nature 422, a function of wind speed. CD 279 (20031. 5. E. D. Fernandez et oL, J. Geophys. Res. 111, C08013 is shown for an observation- 10.1029/2005)(003018 (2006). based resistance coefficient, 6. I. LA:saki. Ginis,I. Hara, J. Armos. SW 61,2334 (2004). r = 0.02 an 54. The red 7. J. A. 1. Bye, A. O. Jenkins, I. Geophys. Res. 111, C03024 open circles are the eval- 10.1029/2005jC003110 (2006). uated CD from the current B. K. Emanuel,?. Armes. Sri. 60, 1420 (2003). 9. M A. lAncheR, W. J. Teague, E. prow, D. W. Wang, and nind Observations, the Geophys. Rn. Lela 32, L11610 10.1029/2005GL023014 solid red line is a fitted (2005). quadratic curve to the CD 10. 0. w. Wang, O. A. Mitchell, W I. Teague, E. faros, estimates, and the red M. S. Hulbert, Some 309, 896 (20051. dashed lines are the 95% IL W. J. Teague, E. prom, 0. W. Wang, O. A. Michell, confidence limits for this I ). Phys. OreoftogL, m press. 12. W. J. league, E. Jarosz, M. R. Carnes, D. A. Mitchell, quadratic curve. The black P. J. H09311, Coot Shelf Res. 26, 2559 (2006). dotted tines represent the 13. J. F. Price, t. 8. Sanford, G. Z. Fcenstall, Phys. window for CD reported in Oreonogr. 24, 233 (1994). (6), whereas the blue dots 14. Matenals and methods are available as supporting represent CD reported in (4). material on Science Online. 15. G. T. Mitchum, W. Sturges,I. Plot Deettooge. 12,1310 Downloaded from www.sciencemag.org on November 2, 2008 (1982). as w a0 40 ea 16. 5. 1. LenD, L Phys. &mow. 24, 2061 (1994). Mind Wiwi (wf0, m C') 17. 5.1. LenD, L Phys. Greasy,. 31, 2749 (2001). 1B. J. M. Mart Phys. Ocean*. 32, 3101 (2002). 19. E. L Andreas, ). Phys. °alma". 34, 1029 (2004). speeds below 30 m are somewhat noisy as a then steadily decreases as the wind speed 20. We thank M. S. Hulbert, A. J. Quaid, and W. A. Goode for mowing support. We also thank the crews of the result of measurement uncertainty and the need continues to rise. Our values for CD are in a research vessels Seward Johnson I and II. This sot was to calculate a velocity derivative, which tends to range of Co values found using meteorological supported by the Office of Naval Research as a part of enhance noise. However, they consistently observations (4) for wind speeds greater than the Naval Research Laboratory's basic research protect show a decreasing trend of Co for wind speeds 32 m 5-I but are higher for lower wind speeds. 'Mope to Shell Energetics and Exchange Dynamics (SEED)' under program element 0601153N, through the greater than 32 m s-I, the lower threshold for a These differences may be attributed to uncertain- Minerals Management Service Environmental Studies category I hurricane on the Saffir-Simpson Scale. ties in the wind measurements and the applica- Program Technology, and by the Minerals Management It is also apparent that the CD values are weakly bility of the simplified ocean dynamics at the Service technology Assessment and Research Program on dependent on the choice of the resistance co- lower wind speeds. Nurticane Ivan. efficient and are larger for increasing values Supporting Online Material of r. The drag coefficient estimates evaluated References and Notes wvrtscierxemag.orgfcgikontentilull/315/5819/1707/DC1 for r 0.1 cm s n are, on average, 20% greater 1. K. Emanuel, Notate 436, 686 (2005). SOM Tea 2. 5. E. Larsen et oL, In Mod Stress Over the Ocean, Fig. S1 than those calculated for r 0.001 cm s I from I. S. F. Hoes,Y. rota, Eds. (Cambridge UMV. Press, References Erg- 3- New Wet 200U, chap. 7. To produce the best representation of CD for 3. M. A. Damian et of, Geophys. Rex fa 31, 118306 1B October 2006; accepted 10 February 2007 each r, a second-order curve (a function of the 10.10291200461019460 (2000. 10.1126Ndence.1136466 wind speed) was fitted by a least-squares technique to all estimated values of CD. The curves are displayed in Figs. 2 and 3. Addition- ally, the 95% confidence limits for the fitted curve are shown in Fig. 3. The pattern of the CRISPR Provides Acquired Resistance relationship between Co and the wind speed is robust, but the curve coefficients are determined Against Viruses in Prokaryotes by the value chosen for r in Eq. 3. However, all Rodolphe Barrangou,' Christophe Fremaux,2 Hake Deveau,3 Melissa Richards,' curves clearly show an initial increase ofthe drag Patrick Boyava1,2 Sylvain Moineau,3 Dennis A. Romero,' Philippe Horvath's' coefficient and monotonic decrease as found by recent studies (3—B) after reaching a maximum Clustered regularly interspaced short palindromic repeats (CRISPR) are a distinctive feature of the value at —32 m s-I. Some of these studies (3, /9) genomes of most Bacteria and Archaea and are thought to be involved in resistance to bacteriophages. imply that the decreasing drag at high winds We found that, after viral challenge, bacteria integrated new spacers derived from phage genomic seems to be related to the spray, foam, and sequences. Removal or addition of particular spacers modified the phage-resistance phenotype of the bubbles from breaking waves that reduce the cell. Thus, CRISPR, together with associated cos genes, provided resistance against phages, and drag and allow the hurricane to slip over the sea. resistance specificity is determined by spacer-phage sequence similarity. With the nearly full water-column ocean cur- rent measurements. the only unknown term left in the simplified equation of motion is the wind acteriophages are arguably the most injection, restricting the incoming DNA, and stress. Thus, the behavior of the drag coefficient (CD) can easily be estimated for a range ofstrong winds. Despite the fact that the drag coefficient is B abundant biological entity on the planet (1) Their ubiquitous distribution and abundance have an important impact on micro- abortive infection systems. These antiviral bar- riers can also be engineered and manipulated to better control phase populations (2. 3). evaluated dittbrently here, estimates of Co bial ecology and the evolution of bacterial Numerous bacteria have been selected by determined "bottom-up" reasonably replicate genomes (2). Consequently, bacteria have devel- humans and used extensively for lemrentation the values determined "top-down" in recent oped a variety of natural defense mechanisms and biotechnology processes. Unfortunately. do- studies (3-7). Results from our research show that target diverse steps of the phage life cycle, mesticated bacteria used in industrial applications that CD peaks at a wind speed near 32 in s-I and notably blocking adsorption, preventing DNA are often susceptible to phage attack including wxfw.sciencemag.org SCIENCE VOL 315 23 MARCH 2007 1709 EFTA00610311 I REPORTS genera and species widely used as dairy cultures Nine phage-resistant mutants were generated in the CRISPR I locus of the various phase- (4). Accordingly, the industry has devised various independently by challenging the WT strain with msistant mutants revealed similarity to sequences strategies to combat phage based on strain di- phage 858, phage 2972. or simultaneously with found within the genomes of the phages used in versity, bacteriophage-insensitive mutants, and both (12), and their CRISPR loci were analyzed. the challenge (Fig. 2 and fig. S2). Interestingly. plagnids bearing phage-resistance mechanisms. Differences were consistently observed at the similarities were observed throughout the phage Streptococcus thermophilus is a low G+C CRISPRI locus, where I to 4 additional spacers gentles. in most finictional modules, both on the Gram-positive bacterium and a key species ex- were inserted next to the 32 spacers present in coding and noneoding strands. No particular se- ploited in the formulation of dairy culture sys- the WT strain (Fig. l). The addition of new quence, gene. or functional gmup seemed to be tems for the production of yogurt and cheese. spacers in response to phage infection seemed to targeted specifically. Them results reveal that, on Comparative genomics analyses of closely be polarized toward one end of the CRISPRI lo- becoming resistant to bacteriophages, the CRISPRI related S. thennopinlus strains have previously cus. This is consistent with previous observations locus was modified by the integration of novel revealed that genetic polymorphism primarily of spacer hypervariability at the leader end of the spacers, apparently derived from phage DNA. occurs at hypervariable loci, such as the cps and CRISPR locus in various strains (9. 13). Se- Surprisingly, we observed that some strains qac operons, as well as two clustered regularly quence analysis of the additional spacers inserted were resistant to both phases, whereas others interspaced short palindromic repeats (CRISPR) loci (5-7). CRISPR loci typically consist of sev- eral noncontiguous direct repeats separated by caa5 cast cash cas7 repeetfspecer region ORF stretches of variable sequences called spacers and Downloaded from www.sciencemag.org on November 2, 2008 are oftentimes adjacent to cuts genes (CRISPR- . ''5 7 e 9 l0 11 12 13 14 16 16 17 It 19 20 21 22 23 24 26 26 27 26 29 30 31 31`r associated). Although the function of CRISPR loci has not been established biologically, in silico analyses of the space's have revealed se- Sensitivity to 0658 Sensitivity to 02972 fie'la tIO te. BP PM HO I .0,10. .0.10, I quence homology with foreign elements, includ- 1 1 I.!• 1 1 1 1 1 1 ing bacteriophage and plasmid sequences (7-9). L 4421 WT Based exclusively on in silico analyses, several .i412. Wrom*SIS2 hypotheses have been put forward proposing roles for CRISPR and am genes, which include LL .41C Wrats.S3 providing immunity against foreign genetic ele- ments via a mechanism based on RNA inter- 0.#141;011 Wroem'S. ference (/0). CL .41•2I Wroem'S5 We analyzed the CRISPR sequences of vari- ous S. thernwpItitus strains, including closely Cc 44I Wron,7246 related industrial strains and phage-resistant var- iants (fig. SI). Differences in the number and Wro2072' type of spacers were observed primarily at the EL WT4.20,2•40 CRISPRI locus. Notably, phage sensitivity ap- 2977411610.611412 peared to be correlated with CRISPRI spacer • I .<5*.Z.54. 1• 21 WT,m6. content. Specifically, spacer content was nearly Wrimwte9,24.513311 identical between parental strains and phase- I fa.ii.;C resistant derivatives, except for additional spacers Fig. 1. Streptococcus thermophilus CRISPR1 locus overview, newly acquired spacers in phage- present in the latter. These findings therefore resistant mutants, and corresponding phage sensitivity. The CRISPR1 locus of DGCC7710 (WI) is at suggest a potential relation between the presence the top. The repeat-spacer region of WT is in the middle: repeats (black diamonds), spacers of additional spacers and the differences ob- (numbered gray boxes), leader (., white box), and terminal repeat (T, black diamond). (Bottom left) served in the phase sensitivity of a given strain. The spacer content on the leader side of the locus in phage-resistant mutants is detailed, with This observation prompted us to investigate the newly acquired spacers (white boxes, 51 to 514). (Bottom right) The sensitivity of each strain to origin and Auction of additional spacers present phages 858 and 2972 is represented as a histogram of the efficiency of plaguing (EOP), which is in phage-resistant mutants. the plaque count ratio of a mutant strain to that of the wild-type. First, we tested the hypothesis that CRISPR loci are altered during the natural generation S9 S11 43 412 $5' 41 44 413 of phage-resistant mutants. A phage-host model system was selected, consisting of a phage- 4,858 sensitive wild-type S. thermophilus strain widely Sti 57 S10 sr, Se used in the dairy industry, DGCC77I0 [wild eitpsW noar packaging morphogenes Mil morphogenesis replication "riser type (WT)j and two distinct but closely related lesis regulation*" ; virulent bacteriophages isolated from industrial 49S S3 S12 SS SP S41 S13 yogurt samples, phase 858 and phase 2972 (//). 02972 'Danisco USA Inc., 3329 Agriculture Drive, Madison, WI 314 $7 410 42' se se 53716, USA. iDanisco France SAS, Bate Postale 10, F-86220 1 kb Dange.Saint.Romain. France. -tepartement de Bi3chimie et Fig. 2. S. thermophilus phase genome maps with the position of sequences similar to the acquired de fAiaolnolegie, Faculte des Sciences et de Genie, Groupe de Recherche en Frolcoie Buccale, lactate de Medecine CRISPR1 spacers of the phage-resistant mutants. Spacers shown above and below the genome maps Dentaire, Felix Reference Center for Bacterial indicate that the spacer matches a sequence on the (+) and on the (—) strand, respectively. An Viruses, Universite Laval, G1K 7Pa Quebec, Canada. asterisk indicates the existence of a SNP between the spacer sequence and that of the phage *To whom c hould be addressed. Email: genome (fig. S1). The genome sequences of phage 2972 (accession number AY699705) and phase 858 are 93% identical. 1710 23 MARCH 2007 VOL 315 SCIENCE www.sciencemag.org EFTA00610312  REPORTS' were resistant only to the phase used in the chal- To determine whether CRISPR spacer con- these observed modifications establish the link lenge (Fig. I). The phage-resistance profile tent defines phase resistance, we altered the between the CRISPR spacer content and phase seemed correlated to the spacer content, such CRISPR I locus by adding and deleting spacers resistance. that strains with spacers showing 100% identity (/2) and tested subsequent strain sensitivity to In the process of generating strain to sequences conserved in both phases were phases. All constructs were generated and inte- WTinsx' sls2ACRISPRI, we created resistant to both phases, such as spacers 53, S6, grated into the S. thermaphilas chromosome with WTosso-sis2::pR, a variant that contains the inte- and S7. In contra* when nucleotide polymor- the system developed by Russell and Klaenhanuner gration vector with a single repeat inserted be- phians were observed between the spacer and the ( /4). We removed the spacers and repeats in the tween the cats genes and the native CRISPR I locus phase sequence [from I to 15 single-nucleotide CRISPRI locus ofstrain WT,/,,014-sis2 and replaced (Fig. 3). Unexpectedly, strain WT,/,1, 3x' sis2::pR polymorphisms (SNPs) over 29 or 30 nucleo- them with a single repeat without any sparer (12). was sensitive to phase 858, although spacers tides], the spacer did not seem to provide re- The resulting strain Art,„„s- I s2ACRISPR l was SI and S2 remained on the chromosome (Fig. 3). sistance, such as spacers SI, S2, S4, S5, and S8 sensitive to phase 858, which indicated that the Similarly, the WT029724S4::pS1S2 construct lost (Fig. I and fig. S2). In addition, when several phase resistance of the original phage-resistant the resistance to phase 2972, although spacer spacers were inserted (S9 to 514), phase re- mutant (WT.:45045152) was probably linked to S4 is present in the chromosome (Fig. 3). These sistance levels were higher. These findings indi- the presence of SI and S2 (Fig. 3). results indicated that spacers alone did not cate that the CRISPR I locus is subject to dynamic Further, to address the critical question of provide resistance, and perhaps, that they have and rapid evolutionary changes driven by phase whether adding spacers provides novel phase to be in a particular genetic context to be exposure. Altogether. these results reveal that resistance, we replaced the CRISPRI locus of effective. Downloaded from www.sciencemag.org on November 2, 2008 CRISPR loci can indeed be altered during the strain WT02972. with a version containing only Although initial work suggested involvement generation of phase-resistant mutants and also spacers SI and S2 (12) and tested whether the in DNA repair (15), the current hypothesis is that establish a link between CRISPR content and phase sensitivity was affected. Remarkably, the eves genes (5, 16) are involved in CRISPR- phase sensitivity. These findings suggest that the resuhing strain WT,/,29724s4::pS1S2 gained re- mediated inununity (JO). Consequently, we in- presence ofa CRISPR spacer identical to a phase sistance to phase 858, which suggested that activated two car genes in strain WT,t45,,,esis2 sequence provides resistance against phases these two spacers have the ability to provide (12): cars (00G3513) and am7. which are equiv- containing this particular sequence. phase resistance de novo (Fig. 3). Altogether, alent to No66571.0O657 and str06601sta0660, respectively (6, 7). The cars inactivation re- sulted in loss of the phase resistance (Fig. 3), COCOS. and perhaps Cas5 acts as a nuclease, because it Cast 0986 caul.. ORF the contains an IINII-type nuclease motif. In con- trast, inactivating clis7 did not alter the resist- rot, ance to phase 858 (Fig. 3). Interestingly, we ant case cas7\ ORF were repeatedly unable to generate CRISPR I phage-resistant mutants from the car? knock- out, perhaps because Cas7 is involved in the 19 CI synthesis and/or insertion of new spacers and toss cast cast cas7 poRi ORF additional repeats. When we tested the sensitivity of the phage- resistant mutants, we found that plaque formation cast cast ORF was dramatically reduced. but that a relatively Iv• small population of bacteriophage retained the ficV ability to infect the mutants. We further analyzed V. pau cat cast 44 997.....: 0 phase variants derived from phage 858 that retained the ability to infect Ari.Nom SI S2. In par- ticular, we investigated the sequence of the ge- 114±. nome region corresponding to additional spacers VI. C4SIC ;646 4D SI and S2 in two virulent phase variants. In both cases, the genome sequence of the phase var- Sensitivity to ease Sensitivity to 02972 iant had mutated, and two distinct SNPs were 19, 114 1,1 ir I ly .0. tO• i i i I I I i t identified in the sequence corresponding to L WLesesINN spacer SI (fig. 53). N. WiessesleteiCRISPR1 Overall, prokaryotes appear to have evolved a nucleic acid-based "immunity" system where- NI. WT •ets2::pR by specificity is dictated by the CRISPR spacer IV. WT,„ 44::pS1S2 content, while the resistance is provided by the V. WTeases122::pcas.5— Cas enzymatic machinery. Additionally, we spec- ulate that some of the cats genes not directly vi. WT •sis2::pcas7— providing resistance are actually involved in the Fig. 3. CRISPR spacer engineering, cos gene inactivation, and corresponding phage sensi- insertion of additional CRISPR spacers and re- tivity. I, mutant WTItts' s 2; II, mutan WT,pess• sis2ACRISPR1 in which CRISPR1 was deleted; peas. as part of an adaptive "immune" response. Ill, mutant WToesan '::pR in which CRISPR1 was displaced and replaced with a unique repeat; Further studies are desired to better characterize IV, WT.2972' 5°::p5152, mutant of strain WT4,25,72' 5° in which CRISPR1 was displaced and re- the mechanism of action and to identify the placed with a version containing 51 and 52; V, favass.sis2::pcos5— with cosy inactivated; VI, specific function of the various my genes. This Wfsesa' sis2::pcos7— with cas7 inactivated. pORI indicates the integrated plasmid (12). The phage nucleic acid-based system contrasts with amino sensitivity of each strain to phages 858 and 2972 is represented at the bottom as a histogram of acid-based counterpart in eukaryotes through the efficiency of plaguing (EOP). which adaptative immunity is not inheritable. www.sciencemag.org SCIENCE VOL 315 23 MARCH 2007 1711 EFTA00610313 I REPORTS The inheritable nature of CRISPR spacers sup- References and Notes 16. 0. H. Haft, J. Selengut, E. F. lIcogodin, IL E. Nelson, ports the use of CRISPR loci as targets for evo- 1. M Brenban, F. Rotates, fiends Microbic[ 13, 278 rad Cookout. Biol. 1, e60 (2005). (2005). 17. P. M A. Greener), A. E. Ilunuhoten, D. van Soolingen, lutionary. typing, and comparative genomic studies 2. S. Clbarn-Chennoull, A. &won, fa.-L Diroann, H. Bdosav, J. O. A. van EmbOen, mat microbe:et 10, 1057 (1993). (9, 17-19). Because this system is reactive to 1. Beetroot 186, 3677 (2004). 18. E. F. mengodm at al, J. &trend. 187, 4935 (2005). the phage environment, it likely plays a sig- 3. ). M. Stoma, T. R. Klaenhammer, Nor. Rev. Mktobiol. 4, 19. R. 1. Deem, E. F. Mongoohn, I. 8. Emerson, K. E. Nelson, nificant role in prokaryotic evolution and ecol- 395 (2006). Beetenot 188, 2360 (2006). 4. H. Brasses, AMA Rev. Mirtabtot. 55, 283 (2001). 20. R. W. Hendrix et of., Proc. Nod. Arad. Sc!. II.S.A 96, ogy and provides a historical perspective of 5. R. Jansen, J. 0.A. van Embrien, W. Gaasva, L. M. Scholl, 2192 (1999). phage exposure, as well as a predictive tool for Mot Whatnot 43, 1565 (2002). 21. J. S. Godde, A. Bickerton, J. MaL Eva. 62, 718 phage sensitivity. The CRISPR-cos system may 6. A. gelatin er a., Not Madinat. 22, 1554 (2004). (2806). 7. A. Solaa, B. (Moguls, A. Sorokin, S. D. Ehrlich, 22. We thank L. Bayer, C. Vos, and A..C. Codti.Mcovoisin accordingly be exploited as a virus defense mech- Mkrobiofogy 151, 2551 (2005). of Caruso) !moaner', as well as J. Laboot6 and animn and also potentially used to reduce the 8. F. J. M. Mona, C. Ellez-VIUasencr, I. Garda-Manfnez, 0. Tremblay of Universne Laval for technical supporL and dissemination of mobile genetic elements and E. Soda, 1. Ala! for. 60, 174 (2005). E. Beth Hansen for diuussion and critical review of the acquisition of undesirable traits such as anti- 9. C. Poured, G. Salvignol, G. Yergnaud, AlIClabiology 151, the manuuript. Also, we thank 1. R. Klaenhammer for 653 (2005). providing the integration system. This work was biotic resistance genes and virulence markers. to. K. S. Makarova, N. V. Cabin, S. A. Shabalina, Y.1. Wolf, supported by holing from Danisco erS. Also, S. AL would From a phage evolution perspective. the inte- E. V. Koomn, &of. Thera 1, 7 (2006). like to acknovAedge support from the Natural Sciences grated phage sequences within CRISPR loci may 11. C. Levesque et at, Appl. fortiori. MktobioL 71, 4057 and EnMneenng Research Council of Canada (NSERC) also provide additional anchor points to facilitate (2005). Cnscovery Program. Sequences sere deposited in recombination during subsequent phage infec- 12. Information on matenals and methods for the generation GenBank, accession numbers Et434458 to 0434500. Downloaded from www.sciencemag.org on November 2, 2008 of phage-resistant mutants, engineering of CRISPR tions, thus increasing the gene pool to which spacers (figs. 54 and 55), and inactivation of cm genes is Supporting Online Material phages have access (20). Because CRISPR loci available on Selena Online. iwm.sciencemag.orgfcgikonteouluW315/5819/1709/DC1 are found in the majority of bacterial genera and 13. R. K Wiest'', P. Redder, R. A Gantt, K. flrigger, Materials and Methods are ubiquitous in Archaea (5, 13,21), their study erchoro 2, 59 (2006). Fags. 51 to 55 14. W. M. Russell. 1. R. Klaenhammer, Ant &Sion aunotoot References and Nam will provide new insights into the relation and 67, 4361 12001). codirected evolution between prokaryotes and 15. K. S. Makarova, L marina N. V. Grishin, I. B. Rogow, 29 November 2006: accepted 16 Fetguany 2007 their predators. E. V. KOOMII, maker ones Rel. 30, 482 (2002). ronzurnerde.rustoo Arubidoµsis putative GPCR protein (GCRI ) has A G Protein-Coupled Receptor Is a been characterized in plants (17-20), and no ligand has been defined for any plant GPCR. Plasma Membrane Receptor for the To identify previously unrecognized GPCR proteins in Arubidopsir, we started by searching Plant Hormone Abscisic Acid the Arabidopuis genome and found a gene (GCR2, GenBank accession code AtIg52920) encoding a putative GPCR. Transmembrane Xigang Liu,1*3 Yanling Yue,' Bin Li,3 Yanli Hie/ Wei Litz Wei-Hua Wu,3 Ligeng Ma" structure prediction suggests that GCR2 is a membrane protein with seven transmembrane The plant hormone abscisic add (ABA) regulates many physiological and developmental processes helices (fig. SI, A and B). The subsequent cel- in plants. The mechanism of ABA perception at the cell surface is not understood. Here, we lular localization analysis confimied its plasma report that a G protein—coupled receptor genetically and physically interacts with the G protein membrane localization in the transgenic plant a subunit GPA1 to mediate all known ABA responses in Arobidopsis. Overexpressing this receptor root (fig. SIC). GCR2—yellow fluorescent pro- results in an ABA-hypersensitive phenotype. This receptor binds ABA with high affinity at tein (YFP) is detected in the membrane fraction physiological concentration with expected kinetics and stereospecifidty. The binding of ABA to the isolated from the GCR2-YFP transgenic plant. receptor leads to the dissociation of the receptor-GPA1 complex in yeast. Our results demonstrate Similar to GCRI (19), GCR2 is mostly asso- that this G protein—coupled receptor is a plasma membrane ABA receptor. ciated with the membrane fraction (fig. SID). Furthermore. even after washing with detergent bscisic acid (ABA) is an important In contrast, several earlier experiments had sug- or a higher pH buffer, GCR2 is retained with the A hormone that mediates many aspects of plant growth and development, particu- larly in response to the environmental stresses gested that extracellular perception is critical for ABA to achieve its functions (7-9). Thus, other ABA receptors, especially plasma membrane- membrane fraction, suggesting that GCR2 is an integral membrane protein (fig. SID). One feature of the GPCR is its ability to (1-3). Several components involved in the ABA localized receptors. may be the major players for interact with G protein to form a complex. To signaling pathway have been identified (4). Two perceiving extracellular ABA and mediating the confirm the physical interaction between GCR2 recent reports have shown that the nuclear RNA classic ABA signaling responses. and Ger. we used four different approaches to binding protein flowering time control protein Ligand-mediated signaling throughG protein- detect their interaction. We first used surface (FCA) (5) and the chloroplast protein Mg coupled receptors (GPCR5) is a conserved plasmon resonance spectroscopy to investigate chelatase II subunit (6) are ABA receptors (6). mechanism for the extracellular signal percep- the interaction between GCR2 and GPAI. For tion at the plasma membrane in entrap/one this purpose. we expressed and purified recom- 'National Institute of Biological Sciences, 7 Science Park Road, organisms (10). The GPCR-mediated signaling binant GCR2 and GPAI proteins in bacteria 2.tiongguancun Life Science Park, Beijing 102206, China. pathway plays a central role in vital processes (fig. S2). This in vitro assay clearly indicated /Laboratory of Molecular and Cellular Biology, Hebei Normal such as vision, taste, and olfaction in animals that GPAI is capable of binding to GCR2, where- University, Shipathuang, Hebei 050016, China. 'State Key (11). IlOwever. the higher plant Arubidopsir as no binding activity was detected between Laboratory of Plant Physiology and Biochemistry, College of Biological Sciences. China Agricultural University, Beijing thulium has only one canonical Ga (GPAI) GPAI and bovine serum albumin (BSA) (fig. 100094, China. subunit, one G0 subunit, and two Gy subunits S3, A and B). The dissociation binding con- 'To whom coat nOence should be addressed. Email: (12-16). The significance of these subunits in stant (Kd) for GCR2 and GPA I is 2.1 x 10-9 M plant systems is poorly understood; only one (fig. 53C). 1712 23 MARCH 2007 VOL 315 SCIENCE www.sciencemag.org EFTA00610314