Introns

Illusions as neuro-signsNicholas Wade

Orientation is a dimension muchstudied in visual neuroscience. Manysingle cells in the visual cortexrespond to lines or edges in specificorientations, and in monkey cortexthese cells are specific for lineswithin ±20° of the preferredorientation.

One class of illusion, involvingperception of orientation, can directus towards interpretations of neuralfunction (in this series, referred to asneuro-signs). The horizontal parallellines in the image below do not

appear so — they look to bealternately inclined clockwise andanticlockwise. If the horizontal linesintersected the curved onesorthogonally, they would appearparallel. Thus, the illusion occurswhen horizontal lines intersect near-horizontals. This indicates that theperception of orientation is basednot only on the orientationselectivity of a particular cell, butalso on interactions with cells withsimilar orientation selectivities.

The illusion could point to one ofseveral neuro-signs: the range ofactive orientation-selective cellsmight be biased by the surroundingcontours; the receptive field (theregion of the retina that can changethe activity of a group of nerve cellswhen light falls on it) properties ofparticular cells might be modified bytheir immediate context; or cellswith similar receptive fieldorientations could inhibit oneanother.

In addition to the simultaneousbiasing of perceived orientation,there are tilt after-effects:observation of inclined lines forsome seconds results insubsequently viewed vertical linesappearing tilted in the oppositedirection. There are two distortionsin the circular image above — tiltand displacement. The ‘Zöllnerillusion’ is the apparent tilt of thefine vertical lines because of theirintersection with the thickerinclined lines. Indeed, the verticalline segments do not even seem to

be aligned, an effect that is calledthe ‘Poggendorff illusion’.

Both illusions could be due tothe same process: the interactionbetween orientation-selective cellsresults in acute angles appearinglarger than they are. The shortvertical line segments then lookinclined anticlockwise, which inturn makes them appear misaligned.

Usually, orientation illusions arebased on interactions betweencontours of similar orientation. Thecoloured image below is composedof only vertical and horizontal lines,although the verticals do not appearso. Some squares of the

checkerboard pattern are darkerthan others, and they form ahorizontal zig-zag pattern. Thisresults in the lighter squaressurrounded by three darker ones

Magazine R509

looking smaller than those with twoor no neighbouring dark squares,which in turn leads to the splayedappearance of the vertical lines.

Typically, orientation selectivityis deduced from studies of monkeycortical cells and the properties ofillusions from human psychophysics,then correlations between the twoare sought. Such simplemanifestations of neuro-signs are,however, only correlations, and thelink between correlation andcausation is notoriously tenuous,particularly when crossing species.Increasingly, studies are measuringboth perception and neurophysiologyin the same animals so that this linkcan be strengthened. The neuro-signs will direct us more clearly whenneural interactions betweenorientation-selective cells to complexpatterns containing manyorientations are better understood.

Address: Department of Psychology,University of Dundee, Dundee DD1 4HN, UK.

Gazetteer

The Sanger Centre

What is it famous for? Being one ofthe main players in the race tosequence the human genome. TheSanger Centre has generatedone-third of all the sequenceproduced to date in the HumanGenome Project, and is the biggestsingle provider of public domainsequence information in the world —with the emphasis on the public.

Where is it? In the back of beyond, onthe Wellcome Trust Genome Campusat Hinxton in rural Cambridgeshire,UK (see http://www.sanger.ac.uk). Thecentre was established jointly by the

Wellcome Trust and the UK MedicalResearch Council (MRC) in 1992. Itis, naturally, named after the doubleNobel laureate Fred Sanger, whodeveloped a method for reading DNAsequences essentially the same as thatused today.

How did it come to be? A happycoincidence of expertise from thenematode genome sequencingproject (pioneered by John Sulston atthe MRC Laboratory of MolecularBiology in Cambridge and BobWaterston at Washington Universityin St Louis, USA) and an injection offunds from the UK’s Wellcome Trust(see Curr Biol 1998, 8:R330).

What is it like? It’s a hi-tech ruralidyll. Surrounded by 55 acres ofparkland, the state-of-the artbuilding features a staircaserepresenting a DNA spiral and anelectronic display of the number ofbases sequenced, which ticks over atdizzying speed. At the heart of theoperation is an enormous roomcontaining row upon row ofautomated sequencers, which feeddata to banks of computers.

Is there any danger of feeling isolated?The 350 Sanger Centre staff sharethe Genome Campus with 150members of the EuropeanBioinformatics Institute and theMRC Human Genome MappingProject Resource Centre. And —perhaps in a bid to become the ColdSpring Harbor of Europe — HinxtonHall, the eighteenth century countryhouse at the centre of the originalestate, has been turned into aconference centre. (In place ofbeaches and fresh lobster, it offers alake and a good pub in the nearbyvillage.) There are also plans toincrease the critical mass of biologistsby creating a ‘biopark’ on the site.

Is it financially stable? In1995, the Wellcome Trustmade a commitment to fundthe sequencing of one-sixthof the human genome (at

least 500 million bases) at theSanger Centre, to the tune of £95million. This May, the Trustannounced an additional £110million over the next seven years, toallow the Centre to sequenceone-third of the genome by 2005.

What’s the hurry? The major playersin the Human Genome Project (aloose conglomerate of like-mindedgenome research organisations thatincludes the US National Institutesof Health and the MRC) have deepconcerns about the wholesalepatenting of sequences, which manybiotech companies argue they mustdo to protect their investment insequencing. The Sanger Centre’sphilosophy of open access tosequence information has broughtits director, John Sulston, intoconflict with the commercial world,most notably recently with CraigVenter’s new company, formed togenerate a partial sequence of theentire human genome by 2001 (seeCurr Biol 1998, 8:R440).

How does the sequence information getpublished? New sequences from theCentre are directed to the nucleotidesequence database of the EuropeanBioinformatics Institute daily at2.00am local time. The EBI web site(http://www.ebi.ac.uk) sees roughly100,000 hits a day.

What will the Centre do once the humansequence is completed? Carry onsequencing. Ongoing projectsinclude Caenorhabditis elegans,Neisseria meningitidis, Plasmodiumfalciparum, bits of Drosophilamelanogaster … and on ad infinitum.

R510 Current Biology, Vol 8 No 15

This is the third in a short series ofarticles by Nicholas Wade, describingsome common visual illusions.