<<

Dispatch R605 the visual representation can be processing regions are necessary imagery on conscious perception. Curr. Biol. 18, 982–986. described as truly pictorial. Moreover, for imagery. 8. Inouye, T. (1909). Die Sehstorungen bei only the first few visual regions (V1, V2 The new study by Pearson et al. [7] Schussverletzungen der kortikalen and V3) are orientation selective, with provides a much needed contribution Sehsphare nach Beobachtungen an Verwundeten der letzten japanischen Kriege sub-regions that process specific to the category of disruptive (Leipzig: W. Engelmann). stimulus orientations [11,12].Itis evidence. They showed that imagery 9. Sereno, M.I., Dale, A.M., Reppas, J.B., Kwong, K.K., Belliveau, J.W., Brady, T.J., therefore not surprising that binocular disrupted binocular rivalry in an Rosen, B.R., and Tootell, R.B. (1995). Borders rivalry studies that have used orientation-specific manner, of multiple visual areas in humans revealed by orientation grating stimuli have suggesting orientation processing in functional magnetic resonance imaging. Science 268, 889–893. reported modulation of activity in these early visual regions was necessary 10. Slotnick, S.D., and Yantis, S. (2003). Efficient orientation-selective regions [13,14]. for both perception and imagery. It acquisition of human retinotopic maps. Hum. Brain Mapp. 18, 22–29. It follows that the binocular rivalry is notable that even if this evidence 11. Tootell, R.B.H., Hadjikhani, N.K., Vanduffel, W., stimuli used in the new study by was correlational, it would still be Liu, A.K., Mendola, J.D., Sereno, M.I., and Pearson et al. [7] also evoked activity in impenetrable to a mental simulation Dale, A.M. (1998). Functional analysis of primary visual cortex (V1) in humans. Proc. Natl. these regions. When considered in argument (as it would be nonsensical Acad. Sci. USA 95, 811–817. conjunction with the observed to propose participants had any 12. Vanduffel, W., Tootell, R.B.H., Schoups, A.A., and Orban, G.A. (2002). The organization of orientation specific effects of imagery, knowledge of how rivalrous perceptual orientation selectivity throughout macaque the evidence suggests the interactions stimuli and imagined stimuli might visual cortex. Cereb. Cortex 12, 647–662. between imagery and perceptual rivalry interact). Considering these factors, 13. Polonsky, A., Blake, R., Braun, J., and Heeger, D.J. (2000). Neural activity in human occurred within the earliest orientation this evidence can be considered the primary visual cortex correlates with perception selective visual regions. most compelling to date that imagery during binocular rivalry. Nat. Neurosci. 3, 1153–1159. Evidence bearing on the depictive can be pictorial. 14. Tong, F., and Engel, S.A. (2001). Interocular nature of imagery can be broadly rivalry revealed in the human cortical blind-spot classified as either correlational or References representation. Nature 411, 195–199. 1. Felleman, D.J., and Van Essen, D.C. (1991). 15. Kosslyn, S.M., Ball, T.M., and Reiser, B.J. disruptive. Correlational evidence, Distributed hierarchical processing in the (1978). Visual images preserve metric spatial which is abundant, refers to imagery primate cerebral cortex. Cereb. Cortex 1, 1–47. information: evidence from studies of image effects that mirror perception and 2. Slotnick, S.D. (2004). Visual memory and visual scanning. J. Exp. Psychol. Hum. Percept. perception recruit common neural substrates. Perform. 4, 47–60. includes behavioral findings — such Behav. Cogn. Neurosci. Rev. 3, 207–221. 16. Pylyshyn, Z.W. (2002). Mental imagery: in as shifting attention between objects 3. Kosslyn, S.M., Ganis, G., and Thompson, W.L. search of a theory. Behav. Brain Sci. 25, (2001). Neural foundations of imagery. Nat. Rev. 157–182. in an imagined visual scene takes Neurosci. 2, 635–642. 17. Craver-Lemley, C., and Reeves, A. (1992). How progressively longer as distance 4. Kosslyn, S.M., Alpert, N.M., Thompson, W.L., visual imagery interferes with vision. Psychol. increases, as if the image is being Maljkovic, V., Weise, S.B., Chabris, C.F., Rev. 99, 633–649. Hamilton, S.E., Rauch, S.L., and Buonanno, F.S. 18. Farah, M.J., Soso, M.J., and Dasheiff, R.M. scanned [15] — and neural findings (1993). Visual mental imagery activates (1992). Visual angle of the mind’s eye before and (mentioned previously). Such evidence topographically organized visual cortex: PET after unilateral occipital lobectomy. J. Exp. Psychol. Hum. Percept. Perform. 18, 241–246. could be disregarded by symbolic investigations. J. Cogn. Neurosci. 5, 263–287. 5. Klein, I., Dubois, J., Mangin, J.F., Kherif, F., 19. Kosslyn, S.M., Pascual-Leone, A., Felician, O., imagery theorists as epiphenomenal, Flandin, G., Poline, J.B., Denis, M., Camposano, S., Keenan, J.P., Thompson, W.L., however, because it is possible that it Kosslyn, S.M., and Le Bihan, D. (2004). Ganis, G., Sukel, K.E., and Alpert, N.M. (1999). Retinotopic organization of visual mental The role of area 17 in visual imagery: reflects mental operations that have images as revealed by functional magnetic convergent evidence from PET and rTMS. nothing to do with imagery — for resonance imaging. Cogn. Brain Res. 22, Science 284, 167–170. example, a mental simulation based on 26–31. 6. Slotnick, S.D., Thompson, W.L., and knowledge that symbolic imagery Kosslyn, S.M. (2005). Visual mental imagery Psychology Department, Boston College, might follow known physical laws, such induces retinotopically organized activation of Chestnut Hill, Massachusetts 02467, USA. early visual areas. Cereb. Cortex 15, E-mail: [email protected] as it takes longer to scan greater 1570–1583. distances [16]. 7. Pearson, J., Clifford, C.W., and Tong, F. Disruptive evidence refers to (2008). The functional impact of mental DOI: 10.1016/j.cub.2008.06.002 interference of a given mental process either by another process that shares the same neural substrates or by direct disruption of the underlying neural substrates. Such evidence supporting Stabilization: Formins pictorial imagery is relatively sparse. Behavioral work has shown imagined Assert Their Independence vertical lines can impair performance on a perceptual line discrimination task [17], a patient’s imagined (and Mammalian Diaphanous-related (mDia) formins are well known for their perceived) visual field was restricted nucleation and filament elongation activities. They have since emerged as following partial removal of occipital microtubule-binding , and a recent study shows that mDia2 stabilizes cortex [18], and temporary cortical independently of its actin nucleation activity. deactivation of occipital cortex impaired performance on a task Aaron D. DeWard concert to facilitate essential involving imagery of oriented lines [19]. and Arthur S. Alberts changes in morphology [1,2]. This type of evidence is particularly One mechanism governing actin convincing because it shows visual Side by side, microtubules and remodeling includes Rho GTP-binding perceptual processing or visual filamentous actin (F-actin) work in proteins signaling through mDia formin Current Biology Vol 18 No 14 R606

A Mammalian Diaphanous-related formin Domain Structure

GBD GTPase-binding domain DAD tethers DID to FH2

GTP DID Diaphanous (Dia) inhibitory domain autoinhibition Rho DD Dimerization domain

GBD DID DD CC FH1 FH2 CC Coiled-coil

Rho GTPase binding Proline-rich region; Dimerization DAD FH1 Formin homology – 1 domain sterically interferes with binds -actin actin nucleation filament DAD binding to DID and numerous elongation and FH2 Formin homology – 2 domain SH3-domain microtubule stabilization containing proteins DAD Dia autoregulatory domain

B Actin nucleation

FH1 FH1

FH2 Formin-mediated actin FH2 FH2 FH2 monomer addition at barbed end

FH1 FH1 (–) (+) Actin monomer

Microtubule stabilization

C Microtubule stabilization via binding to polymerized MTs? D Microtubule stabilization via tip complex proteins?

FH1 FH2 FH2 FH1 ? EB1?

Dimer versus Monomer

FH1 FH2 FH2 FH1 FH1 FH2 FH1 FH2 (–) FH2

FH2 (+) (–) (+)

FH1 FH2 FH2 FH1 Bundling? Buckling? Tubulin dimer addition APC?

Current Biology

Figure 1. Formins mediate actin and microtubule dynamics. (A) Schematic representation of mDia formin domains. (B) mDia formins nucleate and processively elongate F-actin by promoting the addition of actin monomers to the barbed (+) end. (C) A potential mode of microtubule stabilization by mDia2 binding to the sides of microtubules to mediate stabilization. (D) A mode of microtubule stabilization by mDia2 associating with microtubule-tip-binding proteins APC and EB1 in addition to protecting microtubules from disassembly. effectors [3]. Formins contribute to mDia formins normally exist in an they can stabilize microtubules. Like numerous cellular functions that autoinhibited inactive state, mediated the actin cytoskeleton, microtubules depend on actin dynamics, including by the interaction between their form an array of dynamic filaments , filopodia formation, amino-terminal Dia-inhibitory (DID) and throughout the cell [10]. Previous -mediated carboxy-terminal Dia-autoregulatory studies have shown that microtubule gene expression, vesicle trafficking, (DAD) domains [7,8]. ‘Activation’ of stabilization downstream of Rho and [3]. mDia is induced upon interaction with GTPase signaling is essential for cell mDia formins are a highly conserved GTP-bound Rho family small polarization during migration [11]. family of proteins that nucleate and , which disrupts DID–DAD The formins mDia1 and mDia2 were processively elongate linear actin binding (reviewed in [6]). This later revealed as RhoA effectors that filaments through a canonical formin control mechanism allows for mediate this process [9]. Expression homology 2 (FH2) domain (Figure 1A,B) precise spatial and temporal of activated versions of mDia (lacking [4,5]. The FH2 domain potently regulation of formin-mediated all or part of key autoregulatory assembles new actin filaments by cytoskeletal remodeling. domains) or disruption of the mDia binding to the barbed (plus) end of Several years ago, Gregg autoregulatory interaction was growing filaments to promote actin Gundersen’s group [9] discovered sufficient to induce stable monomer addition; not surprisingly, a separate function of formins not microtubules. While it was clear that this activity is tightly regulated [6]. directly related to actin dynamics: the Rho–mDia pathway played a role Dispatch R607 in microtubule stabilization, the examine microtubule dynamics, The studies by Bartolini et al. mechanism of stabilization was Bartolini et al. [14] observed that suggest that the biochemistry of the still unknown. purified mDia2 modestly mDia2–microtubule association will Subsequent work revealed that mDia slowed microtubule assembly (tubulin likely be complex but may also be binds to two microtubule-tip (or plus dimer addition) at both the plus (+) and markedly different from the molecular end) interacting proteins, EB1 and the minus (2) ends. However, mDia2 mechanisms mediating actin APC, which function downstream of dramatically reduced microtubule nucleation and processivity. In the Rho–mDia signaling to induce stable depolymerization at both the plus and prevailing model for mDia-mediated microtubules. This set of proteins may minus ends, suggesting a cap-like actin nucleation, the FH2 domain form a complex at microtubule tips to activity as a potential mechanism functions as a tethered dimer [6].As carry out their activity [12]. As stable for formin-mediated microtubule mutations abrogating mDia nucleation microtubules contribute to cell stabilization (Figure 1C). Consistent activity are still capable of stabilizing polarization during migration, it was with the cell-based experiments, microtubules, Bartolini et al. [14] tested not surprising then that fibroblast in vitro microtubule stabilization by whether FH2 dimerization was required migration was diminished in cells mDia2 appeared to be independent for microtubule stabilization. A single expressing a mutant version of EB1 of its actin assembly activity, since the mutation was generated in the that was defective in microtubule nucleation-deficient versions of mDia2 truncated mDia2 FH1–FH2 fragment stabilization, suggesting a requirement still reduced microtubule disassembly. (guided by the lasso-post structure for the microtubule end-binding protein. These experiments also demonstrate of the formin Bni1p [6]) to abolish The mechanism of microtubule that the FH2 domain of mDia2 is dimerization. Expression of this mutant stabilization downstream of Rho sufficient for microtubule stabilization was fully capable of inducing stable signaling gained another layer of in vitro. microtubules, indicating that FH2 complexity when mDia1 was shown Bartolini et al. [14] then performed dimerization is not necessary for to regulate glycogen synthase kinase-3 in vitro assays using cold and microtubule stabilization. However, b (GSK3b) through a novel class of dilution-induced microtubule this does not rule out the possibility protein kinase Cs (PKCs) [13]. depolymerization in the presence of that mDia2 dimers can still stabilize Specifically, the effects of RNA recombinant mDia2. In support of their and perhaps bundle or crosslink interference (RNAi) or expression of previous data, mDia2 was shown to microtubules; both possibilities are dominant-negative mutants revealed bind directly to microtubules to protect illustrated in Figure 1C. that GSK3b is regulated downstream against microtubule depolymerization. An interesting feature of the of mDia1 and PKC, but upstream of Microtubule binding was mapped to stable microtubules induced by EB1, to induce stable microtubules. the FH2 domain and to an additional nucleation-deficient mDia2 was that How APC fits into this more complex region on the carboxy-terminal end they frequently appeared ‘knotted’. pathway is not clear, but it is interesting of mDia2, suggesting that there are These stabilized microtubules were to note that APC is a substrate for at least two microtubule-binding sites often located in a perinuclear region, GSK3b and that phosphorylation of in mDia2. without radiating to the cell periphery. APC decreases its ability to interact It is unclear if FH2 domain-mediated A more severe knotted phenotype with microtubules in vitro. microtubule stabilization is distinct was associated with the mDia2 A new study has now established from other potential capping mutant most deficient in its ability to a direct, yet distinct, connection mechanisms mediated by microtubule nucleate actin. Perhaps this reflects between the actin and microtubule tip complexes bearing EB1 or APC [1]. the ability of a formin FH2 dimer to : Bartolini et al. [14] Previous studies have shown that bundle microtubules? Overall, these report that the FH2 domain of mDia2 truncated, interfering EB1 protein experiments suggest that while actin binds directly to microtubules and can effectively block mDia-triggered nucleation activity is separate from that the actin nucleation activity of microtubule stabilization in cells, microtubule stabilization, both mDia2 is not required for its suggesting an important role for EB1 activities likely contribute to the normal microtubule stabilization activity. in this process [12]. To eliminate the distribution of stable microtubules. In a key experiment, two different possibility that the nucleation-deficient Further, the results leave open the mutations were made in the FH2 variants of mDia2 may have possibility that a formin could elongate domain of mDia2 that abolish its ability secondarily lost their ability to bind F-actin while bound to a microtubule. to nucleate actin filaments. The to EB1 and APC, Bartolini et al. [14] The main caveat for interpretation mutations were made in a truncated performed GST pull-down assays to of these data is that a truncated version activated version of mDia2 composed confirm the direct interaction between of mDia2 lacking key autoregulatory of the FH1–FH2 domains previously mDia2 and EB1 or APC [12]. In both domains was used to induce shown to stabilize microtubules when cases, the association with the microtubule stabilization. It is likely that expressed in cells [9]. Importantly, formin remained intact. So while the additional cellular factors are involved both actin nucleation-deficient previous in vitro data suggest that in formin-mediated microtubule mutants of mDia2 were still able to mDia2 independently affects stabilization that are unable to interact induce microtubule stabilization. microtubule dynamics by interfering with this truncated version of mDia2. The effect of mDia2 on microtubule with disassembly, these data suggest Bartolini and colleagues have assembly and disassembly rates could that an mDia2–APC/EB1 pathway provided novel insights into the cellular reveal a mechanism by which formins remains a viable mechanism for function of formins, which lead to stabilize microtubules. Using indirect microtubule stabilization several important questions. First, microscopic imaging approaches to (Figure 1D). does direct mDia binding to Current Biology Vol 18 No 14 R608 microtubules cooperate with an the APC tumor suppressor gene drive 11. Cook, T.A., Nagasaki, T., and Gundersen, G.G. (1998). Rho guanosine triphosphatase mediates indirect mechanism mediated by an the progression to malignant colon the selective stabilization of microtubules mDia–APC/EB1 association? cancer [19]. It is also interesting to induced by lysophosphatidic acid. J. Cell Biol. Alternatively, cells may utilize a direct consider that a potent microtubule 141, 175–185. 12. Wen, Y., Eng, C.H., Schmoranzer, J., Cabrera- or an indirect microtubule stabilization stabilizing agent – taxol – is commonly Poch, N., Morris, E.J., Chen, M., Wallar, B.J., mechanism depending on the specific used in the clinic to treat cancer. While Alberts, A.S., and Gundersen, G.G. (2004). EB1 and APC bind to mDia to stabilize microtubules context. For example, one mechanism the mechanism of microtubule downstream of Rho and promote cell migration. may be invoked in mitotic cells for stabilization is different between Nat. Cell Biol. 6, 820–830. microtubule search-and-capture, mDia and taxol, the potential exists 13. Eng, C.H., Huckaba, T.M., and Gundersen, G.G. (2006). The formin mDia regulates GSK3beta whereas migrating cells may employ for mDia-mediated microtubule through novel PKCs to promote microtubule a different mechanism to establish cell stabilization to be a promising stabilization but not MTOC reorientation in migrating fibroblasts. Mol. Biol. Cell 17, polarity. In either case, it will be therapeutic target. Insight into the 5004–5016. interesting to reveal how cells might mechanism of Rho–mDia microtubule 14. Bartolini, F., Moseley, J.B., Schmoranzer, J., employ the two mechanisms in stabilization will certainly be an Cassimeris, L., Goode, B.L., and Gundersen, G.G. (2008). The formin mDia2 different scenarios. important focus of research in the future. stabilizes microtubules independently of its actin Another important question will be to nucleation activity. J. Cell Biol. 181, 523–536. 15. Ishizaki, T., Morishima, Y., Okamoto, M., identify whether microtubule binding is References Furuyashiki, T., Kato, T., and Narumiya, S. unique for mDia2? The FH2 domain 1. Basu, R., and Chang, F. (2007). Shaping (2001). Coordination of microtubules and the primary amino acid sequences of the actin cytoskeleton using microtubule tips. actin cytoskeleton by the Rho effector mDia1. Curr. Opin. Cell Biol. 19, 88–94. Nat. Cell Biol. 3, 8–14. mDia1 and mDia2 are highly conserved 2. Chhabra, E.S., and Higgs, H.N. (2007). The 16. Harris, E.S., Rouiller, I., Hanein, D., and [5]. Both mDia family members many faces of actin: matching assembly Higgs, H.N. (2006). Mechanistic differences in factors with cellular structures. Nat. Cell Biol. actin bundling activity of two mammalian nucleate and processively elongate 9, 1110–1121. formins, FRL1 and mDia2. J. Biol. Chem. 281, actin with similar efficiency and both 3. Wallar, B.J., and Alberts, A.S. (2003). The 14383–14392. can stabilize microtubules [9,15]. formins: active scaffolds that remodel the 17. Eisenmann, K.M., Harris, E.S., Kitchen, S.M., cytoskeleton. Trends Cell Biol. 13, 435–446. Holman, H.A., Higgs, H.N., and Alberts, A.S. Despite these similarities, mDia2 has 4. Kovar, D.R. (2006). Molecular details of (2007). Dia-interacting protein modulates demonstrated subtle biochemical formin-mediated actin assembly. Curr. Opin. formin-mediated actin assembly at the cell Cell Biol. 18, 11–17. cortex. Curr. Biol. 17, 579–591. differences from its relative. 5. Higgs, H.N. (2005). Formin proteins: 18. Peng, J., Kitchen, S.M., West, R.A., Sigler, R., Previous work has shown that mDia2 a domain-based approach. Trends Biochem. Eisenmann, K.M., and Alberts, A.S. (2007). bundles F-actin whereas mDia1 does Sci. 30, 342–353. Myeloproliferative defects following targeting 6. Goode, B.L., and Eck, M.J. (2007). Mechanism of the Drf1 gene encoding the mammalian not [16]. Also, mDia2 activity alone and function of formins in control of actin diaphanous related formin mDia1. Cancer Res. is inhibited by the shared mDia1/ assembly. Annu. Rev. Biochem. 76, 593–627. 67, 7565–7571. 7. Li, F., and Higgs, H.N. (2005). Dissecting 19. Nathke, I. (2006). Cytoskeleton out of the mDia2-binding partner Dia-interacting requirements for auto-inhibition of actin cupboard: colon cancer and cytoskeletal protein (DIP) [17]. This is surprising nucleation by the formin, mDia1. J. Biol. changes induced by loss of APC. Nat. Rev. Chem. 280, 6986–6992. Cancer 6, 967–974. given that DIP binds to the FH2 domain 8. Alberts, A.S. (2001). Identification of a of both proteins with equal avidity. The carboxyl-terminal diaphanous-related formin structural/biochemical explanation for homology protein autoregulatory domain. Laboratory of Cell Structure and Signal J. Biol. Chem. 276, 2824–2830. Integration, Van Andel Research Institute, these differences remains unsolved. 9. Palazzo, A., Cook, T.A., Alberts, A.S., and 333 Bostwick Ave., N.E., Grand Rapids, Clearly, more mDia1/mDia2 Gundersen, G. (2001). mDia mediates Rho-regulated formation and orientation of Michigan 49503, USA. comparison–contrast experiments stable microtubules. Nat. Cell. Biol. 3, 723–729. E-mail: [email protected] addressing these topics are needed, 10. Bartolini, F., and Gundersen, G.G. (2006). as well as experiments using more Generation of noncentrosomal microtubule arrays. J. Cell Sci. 119, 4155–4163. DOI: 10.1016/j.cub.2008.06.001 divergent formin family members. The observations by Bartolini et al. [14] raise the possibility that formins not only act as actin assembly factors but also as microtubule-binding Social Learning: Nectar Robbing proteins to cross-link the two cytoskeletal components. The mDia Spreads Socially in Bumble Bees formins are now poised to play a central role in uniting the mechanisms controlling both actin and microtubule Social transmission of learned behaviour is well documented in vertebrates dynamics. An additional candidate for but much less so among invertebrates. New research shows that nectar this role is the microtubule-binding robbing can spread socially among bumble bees, even in the absence of protein APC. In a previous collaborative nectar-robbing models. effort between the Gundersen and Goode labs [18], APC was shown to David F. Sherry corolla tubes and floral spurs to obtain affect F-actin dynamics through direct nectar. Because of their relatively binding and bundling of F-actin. Bumble bees are robbers. Along large size, bumble bees, along with The impact of these studies may also with acting as legitimate carpenter bees and flower-piercing provide insight on the contribution of pollinators — collecting nectar birds, are the major nectar-robbers that microtubule stabilization and actin and pollen and transporting pollen insect-pollinated plants contend with dynamics in diseases such as cancer. between flowers — bumble bees also [1]. Bumble bees are also sophisticated mDia1 appears to harbor tumor circumvent floral structures designed learners, capable of learning novel suppressor activity [18] while defects in to ensure pollination by biting into flower-handling techniques [2],