R

E

Jacques Mallet – Tyrosine hydroxylase

the environment with gene expression. Finally, the possibility to transfer genes directly to the CNS by means of adenoviralvectors has opened new possibilities in the treatment of neurologicaldisorders. Selectedreferences 1 Le Douarin, N. (1980) Nature 286, 663-669 2 Black, LB. and Patterson, P.H. (1980) Cum Top. Dev. BioL

15,

27-40 3 Thoenen, H. (1974) LifeSci. 14, 223-235 4 Black, I. et aL (1987) Science 236, 1263-1268 5 Lamouroux, A. et al. (1982) Proc. Natl Acad. Sci. USA79, 3881-3885 6 Grima, B. et al. (1985) Proc.NatZAcad.Sci.USA82, 617–621 7 Horellou, P. et aL (1986) FEBSLett.205, &10 8 Campbell, D., Hardie, D.G. and Vulliet, R.R. (1986) J. BioL Chem.261, 10489-10492 9 Haycock, J.W. (1990) }. BioL Chem.265, 11682-11691 10 Le Boudell?s, B. et al. (1991)1. BioL Chem.266, 17124-17130 11 Sutherland, C. et al. (1993) Eur.J. Biochem.217, 715-722 Acknowledgements 12 Zigmond, R.E., Schwarzschild, M.A. and Rittenhouse, A.R. (1989) Annu.Rev. Neurosci.12, 415-461 Theauthorthanks 13 Grima, B. et aJ. (1987) Nature 326, 707–711 hiscolleaguesin the 14 Le Bourdell&, B. et aL (1988) J. Neurochem. 50, 988-991 laboratoryfor 15 Nagatsu, T. (1991) Neurosci.Res. 12, 315-345 16 Dumas, S. et aL (1991) ]. Chem.Neuroanat.5, 11–18 helpfil discussions 17 Coker, G.T., III, et aZ.(1990) Mol.Brain Res. 8, 93-98 and critical reading 18 Ichinose, H. et aZ.(1994) J. Neural Transm. 8, 149-158 of the manuscript. 19 Horellou, P. et al. (1988) J. Neurochem. 51, 652-655 The work reported 20 Ginns, E.I. et al. (1988) J. Biol. Chem.263, 7406-7410 21 Kobayashi, K. et aZ.(1988) FEBSLett.238, 431–434 here was funded by 22 Haavik,J. et aL (1991) Eur.J. BioL Chem. 199, 371–378 grantsfrom the 23 Haavik, J. et aL (1992) Eur.J. BioL Chem. 210, 23-31 CNRS,INSERM, 24 Nasrin, S. et al. (1994) J. BioL Chem. 116, 393-398 25 Dumas, S. et al. 1. Neurochem. (in press) AFM,IRME, 26 Knappskog,M. et al. (1995) Hum. Mol. Genet. 4, 1209-1212 International 27 Ichikawa, S., Sasaokat, T. and Nagatsu, T. (1990) Biochem. Biophys. Res. Commun. 176, 161@1616 HumanFrontier 28 Birman, S. et al. (1994) J. Biol. Chem.269, 26559-26567 ScienceProgram 29 Schussler,N. et aZ.(1995) J. Neurosci.Res. 42, 846-854 and Rh6ne-Poulenc 30 Lani?ce, P. et aL J. Neurochem. (in press) Rorer. 31 Faucon Bignet, N. et al. (1989) Neurosci.Lett.5, 189-194

32 Richard, F. et al. (1988) J. Neurosci.Res.20, 32-37 33 Faucon Biguet, N. et aL (1986) EMBOJ. 5, 287-291 34 Faucon Bignet, N., Vyas, S. and Mallet,J. (1991) J. PhysioL 85, 105-109 35 Vyas, S., Faucon Biguet, N. and Mallet, J. (1990) EMBOJ. 9, 3707-3712 36 Icard-Liepkalns,C. et al. (1992) J. Neccrosci.Res. 32, 290-298 37 Miner, L.L. et aL (1992) J. Neurosci.Res. 33, 10-18 38 Gizang-Ginsberg,E. and Ziff,E.B. (1990) GenesDev. 4,477-491 39 Yamamori, T. et al. (1989) Science246, 1412-1416 40 Misawa,H., Ishii, K. and Deguchi, T. (1992) J. BioL Chem.267, 20392-20399 41 Kengaku,M. (1993) Mol.Brain Res. 18, 71-76 42 Bejanin, S. et aL (1994) J BioL Chem.269, 21944-21947 43 Cervini, R. et al. (1995) [. Biol. Chem.270, 24654-24657 44 Alfonso, A. et aL (1994) J. Mol. Biol. 241, 627-630 45 Erickson, J.D. et aL (1994) /. BioL Chem.269, 21929-21932 46 Craddock, N. and McGuffin, P. (1993) Ann. Med. 25, 317-322 47 Gottesman, L.L. and Shields,J. (1982) Schizophrenia: The Epigenetic Puzzle, CambridgeUniversityPress 48 Schildkraut,J.J. (1965) Am. J. Psychiatry 122, 509-522 49 Davidson, M. et al. (1987) Am.J. Psychiatry144, 934-938 50 Leboyer, M. et al. (1990) Larrcet335, 1219 51 Korner,J., Fritze,J. and Propping, P. (1990) PsychiatryRes. 32, 275-280 52 Gill, M. et aZ.(1991) J. PsychiatryRes. 25, 179-184 53 Meloni, R. et aL (1995) Lancet 345, 932 54 Byerly, W. et al. (1992) Hum. Hered. 2, 259–263 55 Lim, L. et al. (1993) Am.J. Med. Genet. 48, 223–228 56 Meloni, R. et aZ.(1995) C.R.Acad. Sci. 111318,803-809 57 Bjorklund, A. and Stenevi, U. (1979) Brain Res. 177, 555-560 58 Lindvall, O. et al. (1990) Science 247, 574-577 59 Horellou, P. et al. (1989) Proc.NatlAcad.Sci.USA86, 7233-7237 60 Horellou, P. et aZ.(1990) J. Neurosci.2, 116119 61 Horellou, P. et aL (1990) Neuron 5, 393-402 62 Uchida, K. et al. (1990) Neurosci.Lett. 109, 282-286 63 Fischer, L.J. et al. (1991) Neuron6, 371-380 64 Jiao, S., Gurevich,V. and Wolff,J.A. (1993) Nature362,450-453 65 Le Gal La Salle, G. et al. (1993) Science 259, 988-990 66 Ridoux, V. et al. (1994) Brain Res. 648, 171-175 67 Finiels, F. et al. (1995) NeuroReport7, 373-378 68 Sabat6, O. et al. (1995) Nat. Genet. 9, 256-260 69 Horellou, P. et al. (1994) NeuroRepoti6, 49-53

Interactions between self-motion and depth perception in the processing of optic flow V.Cornilleau-PWs and C.C.A.M.Gielen Movingand actingin a 3D environmentrequiresthe perceptionof its 3D structure.Vision is knownto playa crucialrole in the controlof self-motion,particularlythroughthe changesin the retinal image subsequentto movements of the observer. Reciprocally,signalsrelated to self-motioncan also influenceour visualperceptionof 3D space.These interactionsbetween 3D visualperceptionand self-motion,asdemonstratedbehaviorally,are nowbetter understood thanksto the developmentof computationalmodelsfor processingmoving images.They also V. Comilleau-Pt2+s bear a particular interest in the context of the recent intensiveexplorationof the inferior is at the Laboratoire parietal lobe (IPL) by neurophysiologists. The IPL is now firmly establishedas one site of de Physiologicde la Perceptionet de interactionbetween3D visualperceptionandmotor control.The parallelbetweenbehaviourand [’Action,CNRSneurophysiology leadsto a setof crucial,yet unanswered, questions. Coll?gede France, Paris,France, and Trends Neurosci. (1996) 19, 196-202 C.C.A.M.Gielen is IBSONSET the basis for modern studiesof visual objects. Generally,these two functional roles (proprioat the Laboratoryof perception by describing the visual input as an ceptive and exteroceptive)have been approachedsepMedicalPhysicsand Biophysics, optic flow, rather than a succession of static imagesl. arately in the past. However,several studies, most of Universi~of When an animal moves,optic flowcarriesinformation which are quite recent, have explored the complexity Nijmegen,The related to the motion of the animal, as well as to the and neural substratesof the interactions between selfNetherlands. 3D layout of the environment and the movement of motion and 3D shape.

G

196

TINS Vol. 19, No. 5, 1996

Copyright @ 1996, Elsevier Saence Ltd. All rights reserved. 0166-

2236/96/$15.00

PII: S0166-2236(96)1

OO25-4

V. Cornilleau-P6r&

REVIEW

and C. Gielen – Processing of optic flow

The 3D structureof the environment and the movement of the observerinteractin the productionof optic flow. Fora translation of the 3D imageby the observer, the velocityof the retinalimagevarieswith the distance of the object points (Fig. 1). This variation is termed motion parallax.In computervision, recoveringdepth information from monocular motion parallax has proved to be a difficult task23, whereas it is readily achieved by human observers in many situations4,5. Motion parallaxis thereforea depth cue for the human visual system and, in this sense, presents similarities with binoculardisparity.It also presentsdifferentpropertiesthat areworthmentioning.First,it providesdepth information only up to a certain scale factor (a distant object that is moving quickly can yield the same optic flowas a nearbyobject that is moving slowly).Second, rotationsof an object aroundan axisthat passesthrough the eye displaces the retinal image globally, without providing any depth information (similarly, eye rotations provideno depth information). Finally, when the relative movement between the observerand the Fiq. 1. The optic-flow field produced by a lateral (frontoparallel) translation. The imaqe visualscene is not restrictedto such a rotation, the 3D ve;ocityvorieswith the distanceof objectpoints, decreasingas this distanceincreases. structure of 3D objects can, in principle, be extracted from optic flowz. seems to be the drivingvariable for a range of motor responsesin humansandanimals15-17, althoughits entire Influenceof 3D structureon self-motion independencefrom depth cuesor other visualvariables Studies of both the perception and the control of remains to be demonstrated. self-motion reflect the theoretical interplay between The relationship between optic flow and motor depth and motion. responsesis not alwaysas simple, however,and other The perceptionofselj%otion studies have shown that motor responsesdepend on Self-motion perceptiondepends not only on the variablesthat can alsospecifydepthperceptually.A first motion of the observerwithin a visual scene, but also example is given by posture studies: it is well known on the 3D structure of this scene. This is shown in that an observerstanding in front of a frontoparallel experiments on vection (the sensation of self-motion stationaryplane developsa spontaneousposturalsway. caused by visual motion) and heading (the direction Asthe distance, D, betweenthe observerand the scene of self-motion). It can be illusory, as experiencedin a increases, the amplitude of this sway increaseslsly. stationarytrain, when the train alongsidestartsmoving Becausethe optical expansiondueto head movements away.Vection dependson variousparameterssuch as decreasesas D increases, this was initially interpreted the velocity and areaof visualstimulation,and alsothe as a threshold effect; for largedistances,largerpostural perceiveddepth within the visual scene. For example, sways are required to provide the minimum visual if severalpatches within the visual field have different motion necessaryto detect a change in the position of movements, the patch which is apparently the most the observer(relativeto his environment), and drivea distant is perceived as stationary in space and deter- postural readjustment. Recently, Schoner2”proposed a more elaborated model, taking into account the mines vection6-8. In ‘heading’experimentsthe movement of a subject dynamics of the postural-control system, and the in a visualenvironmentis simulatedon a displayscreen. couplingbetweenvisualmotion and head movements. The subject has to report the direction of self-motion Asa consequence of this coupling,vision inducespos(or ‘heading’)within this environment.Highaccuracies tural responses with amplitudes that decrease as D (about one degree),are usuallyachievedin such a task, increases.The model then predictsthat the stabilityof but again, the simulated depth map is an important postural responses also decreases, in the sense of a factorg-lz.As predictedon theoretical grounds13,the largervariabilityin its phasedelayrelativeto the visual resultsof headingexperimentsindicatethat variationsin stimulation. This prediction was tested by exposing depthfavour‘heading’accuracy,whichis minimalwhen observersto planar surfacesoscillating along a sagittal the simulatedenvironment is a frontoparallelplane. axis, which showedthat the temporal locking of posThe controlof selfmotion tural swayand visual expansion effectively improves The studiesby Lee” andco-workerssuggestthat depth as D decreaseszl.Hence,it is not only the visualexpanis not always a primary variable for the guidance of sion that drives the postural response, but also the self-motion. During a movement of approach with a temporalcouplingbetween self-motion and visualsigconstant velocity, the retinal image expands in all nals. That absolute distances can also be perceived directions. The inverse of this expansion rate is called from this coupling is suggestedby perceptual studies the tau-margin, and is equal to the time-to-contact (seebelow),althoughthe demonstrationthat has been (the delaybeforecollisionwiththe approachingobject). performedis for lateral, rather than antero–posterior, Leeproposedthat the visualsystemextractsthis variable movements in humans. from the optic flow, without inferring the distance of A second example concerns convergenceeye movethe approaching object in order to drive rapid motor ments. Asa targetmovesalong a sagittalaxis, the accuresponses (for example, avoidance of an impending racy of convergence eye movements is highly imobject or landingmanoeuvres).Indeed,the tau-margin provedby the presence of a visual scene surrounding 7ZVSV01. 19, No. 5,1996

197

REVIEW

V. Cornilleau-P6r6s and C. Gielen - Processing of optic flow

A

motion parallaxprovidesonly information about relative depth, and has to be combined with other sensory signalsto yieldabsolutedepth perception.Self-motion, in this sense, abolishes the ambiguity surrounding motion parallax,as proprioceptivesignalsrelativeto the motion velocitycan be usedto scalethe motion parallax. Thus, animals such as the locust or the gerbil produce spontaneous head movements to judge absolute distance#,z6. Similarly, humans are ableto reportabsolute distancefrom active head movements2728(seealsoJ.N. Park, PhD thesis, p. 69, University of Kansas, 1964). However,when a target of limited size is presented B alone rather than in a structuredenvironment, errors ”: : “.:~. . . . s. .. . : . ..... .. . .*.’ $. . . . ..’.”.’f in the estimation of its distance can be very large.The .“ .,. .., :* ., ...... .... .. . .. .-. .... >.:..$.......:., ”.....:: . .. - . .. . .:. ., head movements are then accompanied by apparent . . . ,. . .. “ . .“ .. . . .. ,“. ,* :. t .~,. .;.:...: , ..... ........., .. ..,:”: 3D movements of the object29. 1. .. . :.:. ..: ..,. . . . . , .. ; ., =:. :: . . Using random patterns of dots displayed on an ..:.*. : .’.*.. . . . .. . .:: ::..:. :-. ”W. W ... . .. ..:, .. . : . . ..:. .. . . .. oscilloscope screen, Rogers and Graham4 compared ::;