3
Dispatches Symmetry Breaking: Scaffold Plays Matchmaker for Polarity Signaling Proteins Many cell types can spontaneously polarize even in the absence of specific positional cues. In budding yeast, this symmetry-breaking polarization depends on a scaffold protein called Bem1p. A recent study defines Bem1p’s molecular function during symmetry breaking. Benjamin D. Atkins 1 , Satoshi Yoshida 1 , and David Pellman 1,2 Cell polarization and asymmetric cell division are central mechanisms to regulate the developmental fate of dividing cells. Cell polarization can be guided by internal or external spatial cues, such as internal landmark proteins or nearby cells. Polarization can also occur randomly in the absence of such cues, by a spontaneous ‘symmetry breaking’ mechanism. Cell polarization via symmetry breaking is an example of pattern formation by amplification of stochastic fluctuations, an idea first proposed more than 50 years ago by Alan Turing. Budding yeast has been one of the leading model systems for the study of spontaneous polarization because most, if not all, of the polarity proteins have been identified and because gene replacement and GFP-tagging allow the visualization of the dynamics of endogenously expressed proteins. Despite intensive study and significant progress, key aspects of the underlying molecular mechanisms remain poorly understood. A recent Current Biology article by Lew and colleagues [1] has unraveled one of these mysteries: the molecular function of a scaffold protein called Bem1p. As in many other systems, cell polarity in budding yeast is regulated by the Rho GTPase Cdc42p, a so-called ‘master regulator’ of cell polarity [2]. Newly born G1 cells are round and have an unpolarized actin cytoskeleton. To form the daughter cell, or bud, the actin cytoskeleton and secretory machinery must be polarized, enabling the directed transport of vesicles, proteins, and RNAs to the emerging bud. The key step in polarizing the cell is to cluster and activate Cdc42p. In haploid cells, Cdc42p polarization is biased to occur adjacent to the previous site of cytokinesis by cortical landmark proteins that are interpreted by a GTPase module containing the Ras-related GTPase Rsr1p (Figure 1, top; for review, see [3]). Symmetry breaking enables cell polarization in the absence of cortical landmarks or Rsr1p: active Cdc42p spontaneously clusters at a single but randomly located site (Figure 1, middle). Cdc42p symmetry breaking is independent of polymerized actin or microtubules but critically requires the scaffold protein Bem1p as well as Cdc42p’s ability to hydrolyze GTP [4]. Subsequent actin polymerization and vesicle trafficking reinforces the asymmetric clustering of Cdc42p [5–7]. Without cortical landmarks, BEM1 becomes essential for viability (Figure 1, bottom). The specific role of Bem1p in symmetry breaking was not known. However, it was appealing to think that Bem1p might contribute to positive feedback amplification of Cdc42p signaling because a wealth of theoretical and experimental data suggests that such amplification is central to spontaneous Cdc42p polarization [8,9]. Addressing the function of the scaffold Bem1p during symmetry breaking has been difficult because Bem1p has multiple protein–protein interaction domains and a plethora of interacting partners. By analogy to other scaffolds, Bem1p could function by increasing the local concentration of proteins necessary for symmetry breaking, or by orienting proteins properly to facilitate catalysis, or by allosterically regulating the activity of a binding partner, as was recently shown for the mating pheromone MAP kinase scaffold Ste5p [10]. Kozubowski et al. [1] have now found that the only essential function of Bem1p during symmetry breaking is to bring the Cdc42-activating Previous division site Cdc42p cluster Actin cytoskeleton Cue Scaffold + + + Landmark proteins Current Biology Figure 1. Cell polarity establishment during yeast budding. (Top) In normal wild-type cells, cortical cues dictate the site where Cdc42p is clustered. (Middle) In the absence of cortical cues, Cdc42p spontaneously breaks symmetry and polar- izes at a random site in a Bem1p-dependent manner. (Bottom) In the absence of both cortical cues and the scaffold protein Bem1p, Cdc42p fails to cluster and cells fail to break symmetry. Current Biology Vol 18 No 24 R1130

Symmetry Breaking: Scaffold Plays Matchmaker for Polarity Signaling Proteins

Embed Size (px)

Citation preview

Current Biology Vol 18 No 24R1130

Dispatches

Symmetry Breaking: Scaffold Plays Matchmaker forPolarity Signaling Proteins

Many cell types can spontaneously polarize even in the absence of specificpositional cues. In budding yeast, this symmetry-breaking polarizationdepends on a scaffold protein called Bem1p. A recent study defines Bem1p’smolecular function during symmetry breaking.

Benjamin D. Atkins1,Satoshi Yoshida1,and David Pellman1,2

Cell polarization and asymmetric celldivision are central mechanisms toregulate the developmental fate ofdividing cells. Cell polarization can beguided by internal or external spatialcues, such as internal landmarkproteins or nearby cells. Polarizationcan also occur randomly in the absenceof such cues, by a spontaneous‘symmetry breaking’ mechanism. Cellpolarization via symmetry breakingis an example of pattern formation byamplification of stochastic fluctuations,an idea first proposed more than50 years ago by Alan Turing. Buddingyeast has been one of the leadingmodel systems for the study ofspontaneous polarization becausemost, if not all, of the polarity proteinshave been identified and becausegene replacement and GFP-taggingallow the visualization of the dynamicsof endogenously expressed proteins.Despite intensive study andsignificant progress, key aspects of theunderlying molecular mechanismsremain poorly understood. A recentCurrent Biology article by Lew andcolleagues [1] has unraveled one ofthese mysteries: the molecular functionof a scaffold protein called Bem1p.

As in many other systems, cellpolarity in budding yeast is regulatedby the Rho GTPase Cdc42p, aso-called ‘master regulator’ of cellpolarity [2]. Newly born G1 cells areround and have an unpolarized actincytoskeleton. To form the daughtercell, or bud, the actin cytoskeletonand secretory machinery must bepolarized, enabling the directedtransport of vesicles, proteins, andRNAs to the emerging bud. The keystep in polarizing the cell is to clusterand activate Cdc42p. In haploid cells,

Cdc42p polarization is biased to occuradjacent to the previous site ofcytokinesis by cortical landmarkproteins that are interpreted by aGTPase module containing theRas-related GTPase Rsr1p (Figure 1,top; for review, see [3]).

Symmetry breaking enables cellpolarization in the absence of corticallandmarks or Rsr1p: active Cdc42pspontaneously clusters at a singlebut randomly located site (Figure 1,middle). Cdc42p symmetry breaking isindependent of polymerized actin ormicrotubules but critically requires thescaffold protein Bem1p as well asCdc42p’s ability to hydrolyze GTP [4].Subsequent actin polymerization andvesicle trafficking reinforces theasymmetric clustering of Cdc42p [5–7].Without cortical landmarks, BEM1becomes essential for viability(Figure 1, bottom). The specific role of

Bem1p in symmetry breaking was notknown. However, it was appealingto think that Bem1p might contributeto positive feedback amplification ofCdc42p signaling because a wealthof theoretical and experimental datasuggests that such amplification iscentral to spontaneous Cdc42ppolarization [8,9].

Addressing the function of thescaffold Bem1p during symmetrybreaking has been difficult becauseBem1p has multiple protein–proteininteraction domains and a plethora ofinteracting partners. By analogy toother scaffolds, Bem1p could functionby increasing the local concentrationof proteins necessary for symmetrybreaking, or by orienting proteinsproperly to facilitate catalysis, or byallosterically regulating the activity ofa binding partner, as was recentlyshown for the mating pheromoneMAP kinase scaffold Ste5p [10].Kozubowski et al. [1] have nowfound that the only essential functionof Bem1p during symmetry breakingis to bring the Cdc42-activating

Previous division site Cdc42p cluster Actin cytoskeleton

Cue Scaffold

+ +

– –

+

Landmark proteins

Current Biology

Figure 1. Cell polarity establishment during yeast budding.

(Top) In normal wild-type cells, cortical cues dictate the site where Cdc42p is clustered.(Middle) In the absence of cortical cues, Cdc42p spontaneously breaks symmetry and polar-izes at a random site in a Bem1p-dependent manner. (Bottom) In the absence of both corticalcues and the scaffold protein Bem1p, Cdc42p fails to cluster and cells fail to break symmetry.

DispatchR1131

guanine nucleotide exchange factor(GEF) Cdc24p and a Cdc42 effector(p21-activated kinase, PAK) into closeproximity.

Bem1p contains two amino-terminalSrc homology 3 (SH3) domains anda carboxy-terminal Phox and Bem1(PB1) domain (Figure 2). The secondSH3 domain and the PB1 domain areknown to be necessary for symmetrybreaking [4]. Bem1p’s PB1 domaininteracts with Cdc24p [11], whereasthe second SH3 domain interacts withmultiple proteins, including effectorPAKs and other polarity regulators.Kozubowski et al. [1] whittled downthis list through a clever series ofgenetic experiments, obtainingresults that strongly implicated PAKsas the relevant targets of Bem1p’ssecond SH3 domain in symmetrybreaking.

To test whether these interactions(between the second SH3 domain andPAK, and between the PB1 domainand the Cdc42p GEF) were sufficientto explain the role of these domains insymmetry breaking, the authors tookwhat could be described as a syntheticbiology approach, using modularfusion proteins (alternatively, Ludditesmight call it a molecular biologyapproach). Directly fusing Bem1p toa PAK (Cla4p) rendered the secondSH3 domain of Bem1p dispensablefor symmetry breaking, whereas fusionof Bem1p to Cdc24p rendered thePB1 domain dispensable. Importantly,mutant versions of Bem1p that cannotbind either to Cdc24p or to Cla4pwere unable to break symmetry evenwhen both mutants were expressedin combination, suggesting thata single Bem1p molecule must beable to bind both proteins. These datastrongly suggest that Bem1p’s role insymmetry breaking is to build a ternaryGEF–Bem1p–PAK complex.

The coup de grace was to makea shotgun wedding between Cdc24pand Cla4p; Kozubowski et al. [1]directly fused these proteins andfound that this fusion protein was ableto completely bypass the requirementfor Bem1p in symmetry breaking.Thus, Bem1p’s essential function asa scaffold during symmetry breakingis to locally concentrate Cdc42p’sGEF and a PAK in close proximity.The authors integrate these findingswith prior work to suggest a newmodel for the molecular eventsunderlying spontaneous polarizationof Cdc42p. They suggest that the

GEF–Bem1p–PAK complex, via theCdc42/Rac-interactive binding (CRIB)domain of PAK, binds GTP–Cdc42pgenerated at low levels at randomlocations. This binding activates PAK,which in turn might activate the GEF.Local GEF recruitment would increaseGTP-loading of nearby Cdc42pmolecules, providing a mechanism forpositive feedback (Figure 2). Oneattractive feature of this model is thatit explains why nucleotide cyclingplays an essential part of signalamplification and why a ‘GTP-locked’Cdc42p mutant (which cannot interactwith the GEF) fails to break symmetry[4]. An unknown factor is the extentto which this mechanism increasesthe amount of GTP–Cdc42p in thecell, since measuring endogenousCdc42p activity in budding yeast hasproven to be a technical challenge.

There are several aspects of theseresults that either are surprising orimpact on controversies in theliterature. First, in contrast to the datapresented here, previous studies hadsuggested that PAKs were notessential for polarization [12,13].Kozubowski et al. [1] argue thatpreviously studied conditional PAKalleles did not completely abolishPAK activity. In support of thisinterpretation, they find that severalexisting conditional alleles of Cla4pretain sufficient activity at therestrictive condition for spontaneousCdc42p polarization. Second, althoughit is known that Cla4p phosphorylatesCdc24p [14,15], it has beencontroversial whether PAKphosphorylation positively ornegatively regulates GEF activity. Theresults of Kozubowski et al. [1] seemmost neatly to fit a model where PAKphosphorylation activates Cdc24p;and it is intriguing that mathematicalmodeling predicts that the presenceof a GEF-activating GTPase effectorcan dramatically increase the efficiencyof GTPase nucleotide cycling [16].However, whether Cdc24p is therelevant PAK target for symmetrybreaking has not been demonstrated;biochemical studies of Cdc24pactivity are an important missingpiece of this puzzle. Thus, at this point,the competing model that PAKphosphorylation dissociates Cdc24pfrom the complex, thereby terminatingpolarized growth [15], cannot beexcluded.

Is this symmetry-breaking pathwayunique to yeast, or could it represent

a conserved polarity-generatingmodule? Obvious Bem1p homologuesare not recognizable in highereukaroytes. However, in highereukaryotes several Cdc42 GEFs candirectly interact with PAKs via theirSH3 domains. Interestingly, theauthors show that an artificialGEF–SH3 fusion with a similararchitecture to one of thesemammalian GEFs is able to promotesymmetry breaking in the absence ofBem1p in yeast. This furtherunderscores the importance of theGEF–PAK interactions duringsymmetry breaking and raises thepossibility that the core mechanisminvolved in yeast symmetry breakingcould be conserved.

If this were the case, whyhave yeast evolved an extracomponent — the Bem1p scaffoldprotein — to bring the GEF and PAKtogether? One explanation could bethe need to utilize Cdc42p in multipledifferent complexes at differentstages of the cell cycle: Bem1p couldfavor the interaction between Cdc42pand PAK during cell polarization overother effectors involved in otherprocesses. Another possibility is thatBem1p enables additional layers ofregulation of the complex. As hubsof signaling networks, scaffolds areideal targets for regulatory input; forinstance, cyclin–CDK complexesphosphorylate the mating MAP kinasescaffold Ste5p to restrict pheromonesensitivity to G1 phase [17]. Sincecell polarity is tightly coordinated withthe cell cycle in budding yeast, it ispossible that the GEF–Bem1p–PAKcomplex is undesirable in stages ofthe cell cycle where the actin

Bem1 SH3-1 CI PX PB1

Cdc24p(GEF)

Phospholipids

?

SH3-2

PAK

Cdc42p

Potential signal amplification loop

Current Biology

Figure 2. Potential role of the scaffold Bem1pin Cdc42p signal amplification during cellpolarization.

See text for details. SH3, Src homology 3domain; CI, Cdc42-interacting domain; PX,Phox homology domain; PB1, Phox and Bem1domain; GEF, guanine-nucleotide exchangefactor; PAK, p21-activated kinase.

Current Biology Vol 18 No 24R1132

cytoskeleton is depolarized, suchas during mitosis [15].

Overall, how similar is symmetrybreaking in different organisms?Positive feedback seems to beimportant in all cases, and smallGTPases as well as lipids areimportant regulators [18,19]. GEF–PAKcomplexes are observed in manyorganisms and, at least in somecases, these complexes have roles inpolarity regulation [20]. Although thedegree to which these complexesserve conserved functions remains tobe determined, Kozubowski et al. [1]have significantly advanced ourunderstanding of the molecularmechanism of spontaneouspolarization in one of the mostwell-studied model systems. Thisshould amplify positive feedbackamong scientists, providing newavenues for experimentalists to testtheoretical models.

References1. Kozubowski, L., Saito, K., Johnson, J.M.,

Howell, A.S., Zyla, T.R., and Lew, D.J. (2008).Symmetry breaking polarization driven bya Cdc42p GEF-PAK complex. Curr. Biol. 18,1719–1726.

2. Etienne-Manneville, S., and Hall, A. (2002). RhoGTPases in cell biology. Nature 420, 629–635.

3. Park, H.O., and Bi, E. (2007). Central roles ofsmall GTPases in the development of cellpolarity in yeast and beyond. Microbiol. Mol.Biol. Rev. 71, 48–96.

1

1

1

1

1

Convergent Evolutioa Family from the D

New molecular evidence shows that Hawathe similar looking Australasian honeyeatcase of convergent evolution. These nowrepresenting the only complete extinctiontimes.

Irby J. Lovette

Convergent evolution — theindependent origin of similarcharacteristics in separate evolutionarylineages — provides biologists withsome of the most beguiling illustrationsof adaptation [1,2], but convergencecan also bedevil systematists when itobscures the true relationshipsamong similar-looking organisms [3].A subtle — and far more challengingto recognize — form of convergenceinvolves traits that have arisen

is‘HCcemUpoiba

4. Irazoqui, J.E., Gladfelter, A.S., and Lew, D.J.(2003). Scaffold-mediated symmetry breakingby Cdc42p. Nat. Cell Biol. 5, 1062–1070.

5. Irazoqui, J.E., Howell, A.S., Theesfeld, C.L., andLew, D.J. (2005). Opposing roles for actin inCdc42p polarization. Mol. Biol. Cell 16,1296–1304.

6. Wedlich-Soldner, R., Wai, S.C., Schmidt, T.,and Li, R. (2004). Robust cell polarity isa dynamic state established by couplingtransport and GTPase signaling. J. Cell Biol.166, 889–900.

7. Marco, E., Wedlich-Soldner, R., Li, R.,Altschuler, S.J., and Wu, L.F. (2007).Endocytosis optimizes the dynamic localizationof membrane proteins that regulate corticalpolarity. Cell 129, 411–422.

8. Altschuler, S.J., Angenent, S.B., Wang, Y., andWu, L.F. (2008). On the spontaneousemergence of cell polarity. Nature 454,886–889.

9. Goryachev, A.B., and Pokhilko, A.V. (2008).Dynamics of Cdc42 network embodiesa Turing-type mechanism of yeast cell polarity.FEBS Lett. 582, 1437–1443.

0. Bhattacharyya, R.P., Remenyi, A., Good, M.C.,Bashor, C.J., Falick, A.M., and Lim, W.A. (2006).The Ste5 scaffold allosterically modulatessignaling output of the yeast mating pathway.Science 311, 822–826.

1. Ito, T., Matsui, Y., Ago, T., Ota, K., andSumimoto, H. (2001). Novel modular domainPB1 recognizes PC motif to mediate functionalprotein-protein interactions. EMBO J. 20,3938–3946.

2. Cvrckova, F., De Virgilio, C., Manser, E.,Pringle, J.R., and Nasmyth, K. (1995). Ste20-likeprotein kinases are required for normallocalization of cell growth and for cytokinesis inbudding yeast. Genes Dev. 9, 1817–1830.

3. Weiss, E.L., Bishop, A.C., Shokat, K.M., andDrubin, D.G. (2000). Chemical genetic analysisof the budding-yeast p21-activated kinaseCla4p. Nat. Cell Biol. 2, 677–685.

4. Bose, I., Irazoqui, J.E., Moskow, J.J.,Bardes, E.S., Zyla, T.R., and Lew, D.J. (2001).Assembly of scaffold-mediated complexes

n: Raisingead

iian honeyeaters did not evolve fromers, but instead represent a strikingextinct birds form their own family,of an entire avian family in modern

ndependently in closely relatedpecies, a process often termedparallel evolution’. A study of extinctawaiian songbirds in this issue ofurrent Biology [4] provides anotherompelling example of parallelvolution involving a host oforphological and behavioral traits.ntil now those birds’ actualhylogenetic affinities had beenbscured and led to incorrect

nferences about when and how landirds colonized the Hawaiianrchipelago.

containing Cdc42p, the exchange factorCdc24p, and the effector Cla4p required forcell cycle-regulated phosphorylation ofCdc24p. J. Biol. Chem. 276, 7176–7186.

15. Gulli, M.P., Jaquenoud, M., Shimada, Y.,Niederhauser, G., Wiget, P., and Peter, M.(2000). Phosphorylation of the Cdc42 exchangefactor Cdc24 by the PAK-like kinase Cla4 mayregulate polarized growth in yeast. Mol. Cell 6,1155–1167.

16. Goryachev, A.B., and Pokhilko, A.V. (2006).Computational model explains high activity andrapid cycling of Rho GTPases within proteincomplexes. PLoS Comput. Biol. 2, e172.

17. Strickfaden, S.C., Winters, M.J., Ben-Ari, G.,Lamson, R.E., Tyers, M., and Pryciak, P.M.(2007). A mechanism for cell-cycle regulation ofMAP kinase signaling in a yeast differentiationpathway. Cell 128, 519–531.

18. Fivaz, M., Bandara, S., Inoue, T., and Meyer, T.(2008). Robust neuronal symmetry breaking byRas-triggered local positive feedback. Curr.Biol. 18, 44–50.

19. Sasaki, A.T., Janetopoulos, C., Lee, S.,Charest, P.G., Takeda, K., Sundheimer, L.W.,Meili, R., Devreotes, P.N., and Firtel, R.A.(2007). G protein-independent Ras/PI3K/F-actin circuit regulates basic cell motility.J. Cell Biol. 178, 185–191.

20. Li, Z., Hannigan, M., Mo, Z., Liu, B., Lu, W.,Wu, Y., Smrcka, A.V., Wu, G., Li, L., Liu, M.,et al. (2003). Directional sensing requires Gbeta gamma-mediated PAK1 and PIXalpha-dependent activation of Cdc42.Cell 114, 215–227.

1Department of Pediatric Oncology,Dana-Farber Cancer Institute, HarvardMedical School, Boston, MA 02115, USA.2Howard Hughes Medical Institute,44 Binney St, Boston, MA 02115, USA.E-mail: [email protected]

DOI: 10.1016/j.cub.2008.11.005

Enigmatic Hawaiian NectivoresThe Hawaiian islands comprise theworld’s most remote archipelago, andprovide an interesting ‘naturallaboratory’ for evolution. They areknown to have been naturally colonizedby songbirds only six or seven times[5,6]. One such colonization event bya finch-like ancestor initiated theadaptive radiation of the iconic groupof native Hawaiian songbirds, thehighly diverse Hawaiianhoneycreepers. A differentcolonization led to the more modestradiation of the now extinct Hawaiianhoneyeaters, which included fourspecies in the genus Moho (Figure 1)and one species of Chaetoptila, plusone or two additional speciesrecovered only from fossil deposits [5].

These Hawaiian honeyeaters werewell known to native Hawaiians — whocalled them ‘O’os — and to the earlybiologists who worked in thearchipelago [7]. ‘O’o species were