Autism: Face-Processing Clues to Inheritance

  • Published on

  • View

  • Download


  • Drosophila melanogaster caused by motion.Cell Tiss. Res. 270, 2535.

    18. Strausfeld, N.J. (1984). Functionalneuroanatomy of the blowflys visual system.In Photoreception and Vision in Invertebrates,M.A. Ali, ed. (New York: Plenum Press),pp. 483522.

    19. Strausfeld, N.J., Douglass, J., Campbell, H.,and Higgins, C. (2006). Parallel processing inthe optic lobes of flies and the occurrenceof motion computing circuits. In InvertebrateVision, E. Warrant and D.-E. Nilsson, eds.(Cambridge, UK: Cambridge University Press),pp. 349398.

    Autism: Face-Processing Cluesto Inheritance

    When reading faces, autistic individuals gain considerably less informationfrom the eyes andmore from themouth. A new study reports that some parentsof autistic children use strikingly similar strategies, providing a crucial clue asto what might be inherited in autism.

    Elizabeth Pellicano

    Autism is a highly heritable conditionand the search for the genes forautism has thus become a vigorousarea of inquiry. Yet the culprit geneshave so far proved elusive, partlybecause of the disorders highlycomplex make-up: it isa developmental conditionencompassing profound difficulties insocial reciprocity and language,accompanied by restricted andunusual interests and activities. Thereis enormous variability in the degreeand form of these symptoms acrossindividuals. Furthermore, although twinand family studies demonstrate thatgenetic factors play a key role in theaetiology of the condition, the patternsof inheritance are far from simple.Rather than there being a single autismgene, autism is influenced by multiplegenes, with multiple genegene andgeneenvironment interactions [1,2]. Itis therefore not surprising that progresshas been slow.

    The research endeavour is furthercomplicated by the difficulty ofmapping the pathway from genes tobehaviour. The same genotype cangive rise to different behaviouralphenotypes, and the same phenotypecan arise from a range of genotypes [3].

    index genetic liability to the disorder inotherwise apparently unaffectedindividuals, and therefore showtremendous promise in the search forsusceptibility genes for a wide range ofcomplex psychiatric conditions [3,4].

    An endophenotypic approach isparticularly suited to the study ofautism. Since the first pioneering twinstudy [5], it has been well-establishedthat a significant portion of relativesshow features that are qualitativelysimilar to the core features of autism,but milder in form. Components of thisbroad autism phenotype include analoof or socially-detached disposition,a rigid personality style, difficultiesinitiating and maintaining friendships,and limited communicative use oflanguage [6,7]. Different candidategenes might be responsible for distinctcomponents of the broad autismphenotype [7,8], with some parentsand siblings carrying a subset of thegenes for autism. The discovery ofendophenotypes, which are held to bemore directly tied to relevant geneaction than are behavioural outcomes,should edge us closer to isolatingcandidate genes that confersusceptibility to autism.

    A paper by Adolphs et al. [9],published recently in Current Biology,points strongly towards distinct

    Current Biology Vol 18 No 17R748Behavioural phenotypes are thereforenot an ideal basis for identifying geneticmechanisms. Attempts to overcomethese difficulties have turned towardsidentifying traits which areunobservable to the untrained eye butwhich provide a direct connection inthe pathway between genes andbehavioural symptoms. Theseendophenotypes are believed to

    face-processing strategies as onecandidate endophenotype in autism.There is mounting evidence that peoplewith autism show difficulty recognisingfacial emotions, particularly when theyare presented in subtle form. Aprevious study by the same authors[10] showed that such difficulties mightarise because people with autism takean unusual approach to telling howDepartment of Entomology, CornellUniversity, Ithaca, New York 14853, USA.E-mail:

    DOI: 10.1016/j.cub.2008.07.034

    someone is feeling. Using the Bubblestechnique, participants were asked todiscriminate between fearful andhappy faces for which only small partsof the image were revealed throughbubbles (Figure 1). The more bubblesthere were, the easier it was to seethrough them and judge the emotion.

    Importantly, the methodologygathers information about wherepeople needed the bubbles to be toidentify the emotion. For example, fortypical individuals, fear wasimmediately recognisable froma persons wide-open scared eyes.Analysis of their performance yieldedan image showing the eyes but missingmost of the other features. People withautism, however, showed a differentstrategy. They failed to make use ofinformation from the eyes, and reliedmore on cues from the mouth(Figure 1A,B).

    In their more recent paper, Adolphset al. [9] took this research oneimportant step further. Theyadministered the Bubbles task toa group of parents of typicallydeveloping children and two groups ofparents of autistic children, one ofparents who displayed a componentfeature (aloofness) of the broad autismphenotype, and another of parents whodid not. Such a design is especiallynoteworthy. Not all parents showsubclinical features of autism [7,11],and so an endophenotype, if it indeedexists, should be evident principally inthe group of parents predisposed foraloofness.

    As expected, the parents of typicalchildren showed substantial use of theeyes when judging fear or happiness.Both groups of parents of autistic

    children, however, made much less useof the eyes when making thesejudgments (Figure 1C,D). But it was thealoof parents in particular who showedthe most distinct strategy: they reliedless on information from the eyes thanthe other two parent groups andinstead gained most of their cues aboutemotions from the mouth. These latterfindings bear a striking resemblance to

  • Dispatchthe pattern shown in autistic individuals[10], although as expected, theatypicalities in parents were lesspronounced (Figure 1E,F).

    These findings present tantalizingevidence of a specific endophenotypefor one core component of autism:aloofness. Further analyses moreovershowed that the use of the eyes tomake emotion judgments decreased ina fairly linear fashion with increasinggenetic liability for autism. What isespecially remarkable here is that theface processing atypicality was alsodetected in non-aloof parents, albeit toa lesser extent. This latter findingsuggests that this endophenotypicmarker might be sufficiently sensitiveto detect risk in family members wherestandard symptom measures (tappingaloofness) cannot.

    This new study [9] suggests crucialways forward for research in this area.Two additional steps immediatelysuggest themselves. The first is toestablish whether this face processingstrategy is the sort used by parents ofautistic children during their everyday,spontaneous judgments of emotions.Previous research indicates that typicalindividuals tend to use informationfrom the whole face not just fromspecific features when judgingemotion [12]. As people generally donot see bubbled faces in the realworld, it remains possible that theBubbles task induced a bias in peoplethat they otherwise would not havehad: a bias to use particular facialfeatures rather than the overall Gestalt[13]. The presence of a distinct biasonly in autistic individuals and familymembers is intriguing. Yet this mightnot reflect unusual face processingstrategies per se, but instead couldreflect a propensity to adopt unusualstrategies given the requisite taskdemands. The use of tasks that moreclosely approximate on-lineprocessing of emotion mightdistinguish these possibilities.

    The second step is more clearly tosituate this candidate endophenotypewithin a causal story. The authorshave shown that a distinct faceprocessing strategy correlates withone core component of the autismphenotype (aloofness) at one timepoint. Now, the challenge is todemonstrate that this particularendophenotypic marker is causallyrelated to the development ofa socially-reticent disposition.Perhaps there are early-emerging

    R749determinants of this putativeendophenotype, such as activeavoidance of the eyes or reducedattention towards faces [14], whichmight better serve as anendophenotypic marker for socialaloofness. These are questions that

    a well-designed longitudinal studyshould be able to answer.

    Adolphs et al.s [9] findings pave theway for a deeper understanding ofautism. This potential endophenotypenot only should make an excellentcandidate to inform the search for

    Figure 1. The Bubbles technique for studying face processing.

    These ghostly images show the featural information required by individuals to distinguish twoemotions, happiness and fear, during the Bubbles task. Autistic individuals use informationfrom the mouth more (A), and information from the eyes less (B), than typically developing in-dividuals (see [10] for details). A similar but less pronounced set of results are found for aloofparents, who require more information from the mouth (C) and less information from the eyes(D) than non-aloof parents [9] to judge emotions. The bottom panel shows that cues from theeyes are used more with decreasing genetic risk for autism: (E) shows that the eyes are usedslightly more by aloof parents than autistic individuals; and (F) shows that the eyes are used toa greater extent by non-aloof parents than autistic individuals.

  • specific genes [3,15], in this case forsocial cognition, but also should helpto elucidate the neurobiologicalpathways involved in the liability forthe socio-cognitive characteristics ofautism.

    References1. The Autism Genome Project Consortium (2007).

    Mapping autism risk loci using genetic linkageand chromosomal rearrangements. Nat. Genet.39, 319328.

    2. Pickles, A., Bolton, P., Macdonald, H.,Bailey, A., Le Couteur, A., Sim, C.H., andRutter, M. (1995). Latent-class analysis ofrecurrence risks for complex phenotypes withselection and measurement error: A twin andfamily history study of autism. Am. J. Hum.Genet. 57, 717726.

    3. Gottesman, I.I., and Gould, T.D. (2003). Theendophenotype concept in psychiatry:etymology and strategic intentions. Am. J.Psychiatry 160, 636645.

    4. Braff, D.L., Freedman, R., Schork, N.J., andGottesman, I.I. (2007). Deconstructingschizophrenia: An overview of the use ofendophenotypes in order to understand

    a complex disorder. Schizophr. Bull. 33,2132.

    5. Folstein, S., and Rutter, M. (1977). Infantileautism: A genetic study of 21 twin pairs.J. Child Psychol. Psychiatry 18, 297321.

    6. Bishop, D.V.M., Maybery, M., Wong, D.,Maley, A., and Hallmayer, J. (2006).Characteristics of the broader phenotype inautism: A study of siblings using the ChildrensCommunication Checklist-2. Am. J. Med.Genet. Part B 141B, 117122.

    7. Losh, M., Childress, D., Lam, K., and Piven, J.(2008). Defining key features of the broadautism phenotype: A comparison acrossparents of multiple- and single-incidenceautism families. Am. J. Med. Genet. Part B147B, 424433.

    8. Ronald, A., Happe, F., and Plomin, R. (2005).The genetic relationship between individualdifferences in social and nonsocial behaviourscharacteristic of autism. Dev. Sci. 8,444458.

    9. Adolphs, R., Spezio, M., Parlier, M., andPiven, J. (2008). Distinct face processingstrategies in parents of autistic children. Curr.Biol. 18, 10901093.

    10. Spezio, M.L., Adolphs, A., Hurley, R.S.E., andPiven, J. (2007). Abnormal use of facialinformation in high-functioning autism. J.Autism Dev. Disord. 37, 929939.

    11. Bolton, P., Macdonald, H., Pickles, A., Rios, P.,Goode, S., Crowson, M., Bailey, A., andRutter, M. (1994). A case-control family historystudy of autism. J. Child Psychol. Psychiatry35, 877900.

    12. Calder, A.J., Young, A.W., Keane, J., andDean, M. (2000). Configural information in facialexpression perception. J. Exp. Psychol. Hum.Percept. Perform. 26, 527551.

    13. Murray, R.F., and Gold, J.M. (2004). Troubleswith bubbles. Vision Res. 44, 461470.

    14. Dawson, G., Webb, S.J., and McPartland, J.(2005). Understanding the nature of the faceprocessing impairment in autism: Insights frombehavioral and electrophysiological studies.Dev. Neuropsychol. 27, 403424.

    15. Piven, J. (2001). The broad autism phenotype:a complementary strategy for moleculargenetic studies of autism. Am. J. Med. Genet.105, 3435.

    Department of Experimental Psychology,University of Bristol, 12a Priory Road,Bristol BS8 1TU, UK.E-mail:

    DOI: 10.1016/j.cub.2008.07.004

    Current Biology Vol 18 No 17R750

    Autism: Face-Processing Clues to InheritanceReferences