© 2002 Nature Publishing Group822 | NOVEMBER 2002 | VOLUME 3 www.nature.com/reviews/genetics

H I G H L I G H T S

Although sequence similaritybetween genes is a good indicatorof their evolutionary relatedness,you wouldn’t necessarily expecttheir encoded proteins to havesimilar activities — particularly ifthey belong to different species. ButWang et al. recently found aninteresting exception: completefunctional overlap between thefruitfly atonal (ato) gene and itsmouse homologue, Math1.Although these two transcriptionfactor genes have only partialsequence similarity and havedifferent functions in the twospecies, they substitute for eachother seamlessly. This work is thefirst to find an invertebrate genesequence that can completelyrescue the mutant phenotype of itsvertebrate counterpart.

The Drosophila gene atonal hasa well-characterized function inspecifying cell fates in theperipheral nervous system (PNS)and is required for axonalbranching in the central nervoussystem (CNS). By contrast, themouse homologue Math1 specifiescell identities in both the PNS andthe CNS and, oddly enough, incertain gut cell lineages. But justhow different are these two genes?Not at all, it seems. The expressionof the Math1 coding sequenceunder the control of the atoenhancer rescues the ato−/−flyembryonic phenotype, whichincludes missing photoreceptorsand stretch receptors. The

reciprocal experiment gave anequally clear-cut result: Math1−/−

mice that expressed only one copyof ato were indistiguishable fromtheir wild-type littermates. What issurprising about this work is thatAto and Math1, although similarenough to be consideredhomologous, are only 68%identical in a crucial domain andnot at all similar elsewhere.

The differences between Ato andMath1 probably do not arise fromnew protein domains but from cis-regulatory changes that cause thetwo genes to be expressed indifferent tissues. Ato is notexpressed in the fly gut, forexample, but it substitutes perfectlyfor Math1 in this tissue in themouse. Full functionalconservation is commonplacebetween paralogous genes — thosederived from gene duplication (seethe review on p827 of this issue) —but has never before beendocumented in genes such asMath1 and ato that are relatedthrough speciation.

Tanita Casci

References and linksORIGINAL RESEARCH PAPER Wang, V. Y. et al.Drosophila atonal fully rescues the phenotype ofMath1 null mice: new functions evolve in newcellular contexts. Curr. Biol. 12, 1611–1616 (2002)FURTHER READING Prince, V. E. & Pickett, F. B.Splitting pairs: the diverging fates of duplicatedgenes. Nature Rev. Genet. 3, 827–837 (2002)WEB SITESHuda Zoghbi’s lab: http://www.hhmi.org/research/investigators/zoghbi.htmlHugo Bellen’s lab: http://www.hhmi.org/research/investigators/bellen.html

A question of control

E V O – D E V OIN BRIEF

Evolution of yellow gene regulation and pigmentationin Drosophila.Wittkopp, P. J. et al. Curr. Biol. 12, 1547–1556 (2002)

One goal of evolutionary developmental biology is tounderstand the molecular-genetic basis of phenotypicevolution. A useful system for such studies is the evolution ofthe pigmentation pattern in the Drosophila lineage. By usingtransgenic flies that carry a heterologous yellow (melanin-encoding) gene, the authors show that changes in yellowexpression in closely related species have resulted from changesin both cis-regulatory regions and trans-acting factors. Thedifferent yellow expression patterns can also be attributed toevolutionary changes in other genes.

Hedgehog-mediated patterning of the mammalianembryo requires transporter-like function ofDispatched.Ma, Y. et al. Cell 111, 63–75 (2002)

Mouse Dispatched homolog1 is required for long-range, but not juxtacrine, Hh signaling.Caspary, T. et al. Curr. Biol. 12, 1628–1632 (2002)

Signalling by the Hedgehog (Hh) family of proteins is required to pattern many vertebrate and invertebrate tissues in a concentration-dependent manner. In flies, the long-range action of Hh proteins is thought to occur through theability of Dispatched (Disp), a transmembrane protein, torelease Hh from the outer membrane of Hh-producing cells.Ma and colleagues show that this function is conserved inmouse and that DispA, a mouse disp homologue, is requiredfor all mouse Hh patterning functions. In a screen for genesrequired for the ventral patterning of the neural tube, Casparyet al. identified Disp1, a second mouse disp homologue, andhave shown that it is required for long-, but not short-, rangeHh signalling.

Gene expression during the life cycle of Drosophila melanogaster.Arbeitman, M. N. et al. Science 297, 2270–2275 (2002)

This paper reports the microarray-based transcriptionalprofiling of 4,028 wild-type Drosophila genes, assayed at 66sequential time points during the organism’s life cycle.Expression was also assayed in mutant flies that lack germlineand eye tissue to refine expression data for these tissues. Manydynamic gene-expression changes are reported that correlatewith Drosophila’s development, maturation and sex. Hierarchicalclustering analyses also provide new information on genefunction and on the components of pathways and complexes, asthese analyses clustered many genes of unknown function withgenes of known or predicted function.

F U N C T I O N A L G E N O M I C S

D E V E LO P M E N TA L B I O LO G Y

E V O – D E V O