Copyright 0 1994 by the Genetics Society of America Perspectives

Anecdotal, Historical and Critical Commentaries on Genetics Edited by James F. Crow and William F. Dove

Discovery and Genetic Definitionof the Complex

Rob Denell Division of Biology, Kansas State University, Manhattan, Kansas 66506-4901

NE day in 1976 THOMKAUFMAN called me in great gued fordecades that they must play key regulatory roles 0 excitement. KAUFMANand I shared an interest in in assigning developmental commitments realized Drosophila homeotic located near the third chre through theaction of downstream genes. Moreover, the mosome centromere. He hadrecently arrivedat Indiana apparent clustering of genes of similar function in a University, and the resultsof his first set of crosses higher was potentially very interesting. convinced him that these genes were part of a complex However, uncertainties as to allelic relationships and important to anterior developmental commitments (KAuF- mapping difficulties due to centromeric inhibition of MAN et aL 1980),just as ED LEWIS’S bithorax complex con- recombination in this region (HANNAH-ALAVA 1969) left trolled fate in more posterior regions. He invited me to thenumber of homeoticgenes and their spatial Bloomington the next summer to join him in a genetic relationships in considerable doubt. study of this Antennapedia complex (ANTC). All known homeotic loci in that region had been rec- In the late 196Os, KAUFMANand I had been graduate ognized by adult transformations, and included Extra students together in BURKEJUDD’S lab at the University of sex combs ( SCX), proboscipedia ( pb),Antennapedia Texas. Althoughmy interest in segregation distortion was (Antp),Polycomb (PC), Multiplesex combs (Msc),and Ear from the focus of the lab, KAUFMANworked at its very Nasobemia (Ns)(Figure 1A). [ThatDeformed (Dfd)and center: a mutational analysis of the restewhi& region. JUDD Humeral (Hu)were homeotic was not then was and remains interested in organization, and recognized.] Many Antennapedia mutant shared in this pre-molecularera he wished to ascertain allzygoti- a common dominant antenna +. leg transformation cally active functions in this Xchromosome interval nec- , rearrangement breakpoint, andrecessive le- essary for normal morphology and/or viability and to ex- thality. WALTERGEHRING (1966) had discovered a muta- amine the distribution of these loci along the polytene tion associated with a similar (albeit more extreme) chromosome map (JUDD et al. 1972). In addition to the transformation that was free of recessive lethality and genetic aspects of this“saturation mutagenesis” effort, gross chromosomal rearrangement.He believed this KAUFMANjoined MARY SHANNONin a study of the develop variant to identify a locus separate from Antennapedia mental consequences of the recessive-lethal variants iso- and named it Nasobemia after MORGENSTERN’Smytho- lated (SHANNONet al. 1972). logical creature that walked on its nose. ELIEZERLIF- I left Austin in 1970 for postdoctoral studies with DAN SCHYTZ, then also in LINDSLEY‘Slab, had recently shown LINDSLEYat UCSD and arrived there ata particularly ex- that the dominant phenotype of a gain-of-function al- citing time. LINDSLEYand long-time collaborator LARRY lele could be “reverted” by loss-of-function mutations SANDERwere also interested in functional aspects of isolated at rates characteristic of forward genetic organization, and I joined their massive effort (LIFSCHWZand FALK1969). LINDSLEYsuggested that I use to systematically generate and study small deficiencies LIFSCHYTZ’Sapproach for the sex-transforming mutation and duplications distributed throughout the genome then known as transformer-dominant (which proved to (LINDSLEYet a,!. 1972).LINDSLEY has served as a repository be doublesex-dominant instead), andI also applied itto and chronicler of Drosophila genetic knowledge and Nasobemia. Most revertants were associated with reces- takes seriously his responsibility to impartthis lore tohis sive lethality that failed to complement the lethality of students. He pointedout to me an apparent enrichment Antennapedia mutations, showing that Nasobemia was of homeotic and sex-transforming genes near the third an Antennapedia and that thephenotype that was chromosome centromere. These geneswere fascinating its namesake was neomorphic (DENELL 1973).Extra sex from a developmental standpoint, and it had been ar- combs appeared to be anAntennapedia allele as well. A

Genetics 138: 549-552 (November, 1994) 550 R. Denell very similar reversion analysis was done independently BXC. He envisaged the middle thoracic segment as a by KAUFMAN and IAN DUNCAN, then an undergraduateat ground state, such that the loss-of-function phenotype of the University of British Columbia (DUNCANand KAuF- Sex combs reduced caused posteriorlydirected changes ofthe MAN 1975). [In this issue, TALBERTand GARBER(1994) first thoracic leg, and dominant gain-of-function variants discuss the molecular nature of Nasobemia revertants.] Multiple sex mbs and Extra sexcombs caused opposing These results clarified the picture a little, but the ques- changes (KAUFMAN et al. 1980). tion of relationships in the genome and their potential How could the structure and function of this complex significance remained unsolved. be elucidated? KAUFMAN recognized that the available col- At this time ED LEWISwas performing his well known lection of variants, isolatedon the basis of adult homeotic and elegant work at Caltech on other homeotic muta- , probablyhad been strongly influenced by as- tions. In this case, alarge number of recessiveand domi- certainment bias. Except for some poboscipedia alleles, all nant variants causing homeotic transformations of the available visible mutations were neomorphs, and null al- adult thorax and abdomen were unambiguously very leles were needed to assess their normal developmental tightly clustered within the bithorax complex (BX-C), significance.Moreover, he anticipated that additional which mapped more distally on the right arm of chro- functionally related genes existed withinthe complex that mosome 3. LEWIS(1978) arguedthat loss-of-function could be recognized by the embryonic or larval phencl variants caused anteriorly directed transformations, types of recessive lethal alleles. Thus, he proposed that we whereas those associated with gain-of-function variants applyJuDD’s concept of saturation mutagenesis to identify were posteriorly directed. Moreover, themutations and characterize all loci in this regionthat mutate to yield mapped along the chromosome in an order colinear recessive lethal or visible adult phenotypes. WhenI arrived with their mutant effects along the anterior-posterior in Bloomington in the summer of 1976, KAUFMANand body axis. This distribution led Lewis to the hypothesis graduate student RICKILEWIS had set up an experiment to that the numberof genes corresponded to the number isolate X-ray- and EMSinduced mutations that failed of thoracic and abdominal segments. He suggested a to complement Df!3R)Ns’””. Later, graduate student model inwhich themiddle thoracic segment was a BARBMWAKIMOTO screened for additionalmutations “ground state”characteristic of no BX-C activity in the using Df(3R)Sm Both sets of variants were subjected to context of an otherwise normal , and the ex- complementation and recombinational analysis, and the pression of a progressively greater number ofBX-C map shownin Figure 1Bwas generated (LEWIS et aZ., genes definedsuccessively more posterior segments. He 1980a,b).As predicted, point mutations associated with a interpreted gain-of-function mutations as being due to dominant adult T1 + T2 transformation were isolated the expression of a in an inappropriate domain. at the Sex combs reduced locus. The results also suggested For me at least, this idea provided one satisfymg expla- that the EbRll complementation group included the nation for thedifficult question of howneomorphic mu- Deformed mutation, and preliminary observations showed tations could acquire “a new function” (MULLER1932). that the lethal syndromes associated with the EbRll and Noting that Japanese geneticists studying the silk moth Ep36 complementation groups included larval head equivalent of the BX-C had placed emphasis on larval abnormalities. mutant phenotypes (TAZIMA1964), LEWISalso stressed The description of complementation groups with arbi- theimportance of examining the terminal lethal trary names generated little widespread excitement. The syndromes of dying at preadult stages. real payoff began when WAKIMOTOand KAUFMAN(1981) KAUFMAN’S first crossesat Indiana used a small deficiency and WAKIMOTOet al. (1984) described the phenotypes of isolated as a revertant of Nasobemia (Df{3R)Ns+R17),and lethal larvae and adult clones homozygous for recessive an overlapping deletion isolated at the University of British lethal alleles (seeUUFMAN etal. 1990). KAUFMAN (1978) had Columbia by DON SINCWR calledDf(3R)Sm The latter was earlier made the unexpected observation that proboscipedia associated with areduction in the size of the sex combsof function is dispensable for normal embryonic develop the male prolegshared by Multiple sex combs, and KAUFMAN ment, so it was satisfying that the null phenotypes of An- postulated that deletion of a hapleiinsufficient Sex combs tennapedia and Sex mbsreduced were associated withlarval reduced gene caused a partial transformationof first to sec- homeotic transformations. The Ep14 complementation ond leg. Complementation analysis usingthe deficiencies group was associated withaberrant embryos missing half of indicated that polytene chromosome region 84A-B in- the normal number of segments. Wishingto give this gene cluded all of the homeotic loci known in the proximal a descriptiveJapanese name, WAKIMOTOand KAUFMANcon- third chromosome region except Polycomb, which by then sulted colleagueBOB TOGASAKIas well as WAKIMOTO’Sfather had been shown to reside on the other side of the and settled on @hi tarazu (ft.), a term that roughly trans- centromere (P~Roand NKREN1975). This exciting de- lates to describe a shortening of bamboo by 10s of seg- finitive evidenceof clustering led Kaufr-nan to propose that ments. This segmentationgene has been among the most these genes were part of a complex controlling anterior intensively studied in Drosophila. They alsofound that the determinative decisions,and to predict similarities to the Ern36 complementation group caused abnormalities in 55 1

A. Recombinational map positions, from LINDSLEY and GRELL(1967)

SIX Dfd pb AntpMsc PC Ns Hu 47 47.5 47.7 48 48.0 48.0 48-54

B. Complementation groups, after LEWIS et ul. (1980a)

".EbRl1 Sa EM4 I Antp I Dfd? Msc Ns SCX Hu

C. Post-molecular organization, after KAUFMAN ef a/.(1990)

cuticle lab genes pb 22 Zen bod arna Md Sa ffz Antp FIGURE1.-Conceptual evolution of the Drosophila Antennapediacomplex. (A) Map positions of the homeotic mutations (as well as Deformed and Humeral) recognized in 1967 in the proximal region of chromosome 3, with the centromere position indicated by a circle. Becauseof centromeric inhibitionof recombination, the region fromScx to Ns potentially spanned salivary chromosome map region 77-84.All recognized Antp alleles were associatedwith rearrangements sharing breakpoints in proximal 84, and its relative positionwas assigned on that basis. Likewise, Humeral is associated with an inversion of salivary chromosome map region 84-86 and was placed at recombinationalmap position 48-54. Later studies would show thatPC is in the leftarm and all others in the rightarm at salivary map position 84B1,Z.(B) Complementation groups recognizedby LEWIS et al. (1980a). They have been aligned with the current version of the map (based on molecular well as genetic studies) shown in (C), although 1. as LEWISet al. &owed them in a slightly different order. morphogenetic rearrangements associated with gastrula- integrated manner is an ancient feature of the Metazoa. In tion, and renamed the locus zerknullt (m)(German for 1987my colleague DICKBEEMAN mapped the homeotic mu- crumpled). Later work hasdefined the importance of this tations then extant in the red flour beetle, Tribolium m- gene in dorsal-ventral patterning. Studies of the EbRll tuneum. He found that apparent homologs of both ANT-C complementation group (later to be DefMmed) detected and BXC genes were tightlylinked, suggesting that a single anterior abnormalitiesthatwere not obviously homeotic in homeotic complex (HOM-C) was the ancestral organiza- nature, and later work on the Ep9 or labial (lab) group tion among . As recently related by MCGINNIS also demonstrated similar nonhomeotic head defects. (1994), the facilitated comparative molecular Which of these loci associated with diversemutant phe- studies that showed that this complex arosevery early in notypes potentiallybelonged to a complexof functionally evolution and that (with the single recognized ex- related genes? KAUFMAN argued that all of the genes nec- ception of Drosophila) its integrityhad been maintained essary for early embryonicdevelopment (as well as pobosci- over hundreds of millions of years. Until recently, ho- @diu) were members of the ANT-C. He believed that the mologs of the nonhomeotic members of the complex had lack of overt embryonichomeotic phenotypes of labiul and not been detected outside of the Diptera, suggesting that Defonnedwas a consequence of the highly derivednature of these genes originated as a concomitant of evolutionarily the maggot head; labial and Deformed mutations werelater advanced aspects of early development in higher flies. shown to causehead transformation in homozygousadult However, a homolog offiLshi tarcuu has now been recog- clones. He further argued that the interspersion offiLshi nized in Tribolium(BROWN et al. 1994), and probable he tarazu and zerknullt among the homeotic genes could not mologs havebeen described in the brine shrimp and grass- be merely fortuitous. hopper (AVEROFandh 1993; DAWetal. 1994),leading Off the record, membership of the complex (or indeed to the hypothesis that this and other nonhomeotic mem- the idea that there was more than a chance proximity of bers of the complex show a rate of sequence divergence developmentally significant genes) was sometimes ques- much higher than do homeotic genes. Thus, the question tioned, but additional work proved KAUFMAN correct. Mo- of their origin remains unresolved. lecular analysis showed that the homeotic genes (and in- The kind ofsaturation mutagenesis effort devotedto the deed all other interspersed nonhomeotic genes except the Antennapedia complex could not detect all of the func- cuticle clustermentioned below) shared en- tions included (Figure 1C). Screening for zygotic effects coding sequence-specific DNA-binding domains impor- failed to identify the maternal effect gene bicoid (but) tant to the function of their proteins as transcription fac- (FROHNHOFER and NUSSLEIN-VOW 1987), which plays a tors. Further, it was realized that a cluster of homeobox critical role in the establishment of embryonic anterior/ genes regulating developmental commitments in an posterior polarity. The gene amalgam (am) and the 552 R. Denell mkniillt paralog 22 were discovered molecularly,but mu- (Ns) von Drosophilamelanogaster. Arch. Julius Klaus-Stift. tant effects have not yet been described. Perhaps most Vererbungsforsch. Sozialanthropol. Rassenhyg. 41: 44-54. HANNAH-ALAVA,A,, 1969 Localization of PC and SCX.Drosophila Inf. enigmatic is a molecularly ascertained clusterof genes pu- Ser. 44: 75-76. tatively atlecting cuticle synthesis and structurally unre- HAYES, P. H., T. SATO and R. E. DENELL,1984 Homoeosis in lated to any other members of the complex. Drosophila: the larval syndrome. Proc. Natl. Acad. Sci. USA 81: 545-549. Saturation mutagenesis also proved very important to JuDn, B. H., M. W. SHENand T. C. KAUFMAN,1972 The anatomy and our understanding of the BX-C. Independent work by function of a segment of the X chromosome of Drosophila mela- three groups detected only three lethal complementa- nogaster. Genetics 71: 139-156. KARCH, F., B. WEEFENBA~H,M. PElFER, W. BENDER,I. DUNCAN et uL, 1985 The tion groups (SANCHEZ-HERREROet al. 1985; TIONGet al. abdominal region of the bithorax complex. Cell 43: 81-96. 1985; KARCH et al. 1985). Molecular studies showed that KAUFMAN, T. C., 1978 Cytogenetic analysis of chromosome 1 in Dro- they correspond to the only protein-coding transcrip- sophila melanogustel: isolation and characterization of four new alleles of the proboscipedia ( pb) locus. Genetics 90: 579-596. tion units in the complex and that many mutations giv- KAUFMAN, T. C., R. LEWISand B. WAKIMOTO,1980 Cytogenetic analysis ing adult transformations affect complex cis-regulatory of chromosome 3 in Drosophilamelanogaster. The homeotic regions. Earlier views that the BX-C (and some ANT-C) gene complex in polytene interval 84A-B. Genetics 94: 115-133. KAUFMAN, T. C., M.A. SEECERand G. OLSON,1990 Molecular and genes function within segmental domains werealso genetic organization of theAntennapedia gene complex of modified by the discovery that mutant phenotypes cor- Drosophila melanogaster. Adv. Genet. 27: 309-362. respond to metameric units offset from segments (HAYES LEWIS,E., 1978 A gene complex controlling segmentation in Dro- sophila. Nature 276: 141-152. et al. 1984; STRUHL1984) and later called parasegments LEWIS,R. A., T. C. KAUFMAN,R. E. DENELLand P. TALLERICO,1980a Ge- (MARTINEZ-ARIASLAWRENCE and 1985). netic analysis of the Antennapedia gene complex (ANT-C) and The facility with which homeo genes can be molecu- adjacent chromosomalregions of Drosophilamelanogaster. I. Polytene chromosome segments 84B-D. Genetics 95: 367-381. larly cloned and studied (see MCGINNIS1994) has al- LEWIS,R. A., B. T. WAKIMOTO,R. E. DENELLand T. C. KAUFMAN, lowed an investigation of the roles of these developmen- 1980b Genetic analysis of theAntennapedia gene complex tally significant genes inphylogenetically diverse (ANT-C) and adjacent chromosomal regions of Drosophila mela- nogaster. 11. Polytene chromosome segments 8&-84B1,2. Genet- , as well as insights into the evolution of tran- ics 95 383-397. scriptional regulatory mechanisms. In the past, some of LIFSCHITZ,E., and R. FAW(,1969 A system for screening of rare events my colleagues have questioned how the arcane results in genes of Drosophila melanogaster. Genetics 63: 343-352. LINDSLEY,D. L., and E. H. GRELL,1968 GeneticVariations of from Drosophila studies could possibly be relevant to Drosophila melanogaster. Carnegie Inst. Wash. Publ. 627. more important matters such as mammalian develop- LINDSLEY,D. L., L. SANDLER,B. S. BAKER, A.T. C. CARPENTER, E. R. DENELL ment. Thus,it is especially satisfylngto contemplatehow et al., 1972 Segmental aneuploidy and the genetic gross struc- the genetic and molecular characterization of these bi- ture of the Drosophila genome. Genetics 71: 157-184. MARTINEZ-hAS, A., and P. A. LAWRENCE,1985 Parasegments and com- zarre mutations has led to the discovery of a gene com- partments in the Drosophila embryo. Nature 313 639-642. plex that is providing ever increasing insight into mam- MCGINNIS,W., 1994 A centuryof homeosis, adecade of homeoboxes. Genetics 137: 607-611. malian regulatory mechanisms and incidentally has MULLER,H. J., 1932 Further studies on the nature andcauses of gene been proposed as the criterion thatdefines members of mutations, pp. 213-255 in Proceedings of the Sixth International the animal kingdom (SLACKet al. 1993). Congress of Genetics, edited byD. F. JONES. Brooklin Botanic Gardins, Menasha, Wisc. Many thanks to THOM KAUFMAN, BARB^ WAKIMOTOand RICH LEWIS PURO,J., and T. NYGREN,1975 Mode of action of a homoeotic gene for helpful suggestions, as well as to MONICAJUSTICE and MARNETTE in Drosophilamelanogaster. Localization and dosage effect of DENELLfor comments on the manuscript. Polycomb. Hereditas 81: 237-248. SANCHEZ-HERRERO,E., I. VERNOS,R. MARCO and G. MORATA,1985 Ge- netic organization of the Drosophila bithorax complex. Nature LITERATURE CITED 313 108-113. SHANNON,M. P., T. C. KAUFMAN, M. W. SHENand B. H. Jum,1972 Le- AVEROF, M.,and M. AKAM,1993 HOM/Hox genes of Artemia: impli- thality patterns andmorphology of selected lethaland semi-lethal cations for the origin of and crustacean body plans. Curr. mutations in the zeste-white region of Drosophila melanogaster. Biol. 3: 73-78. Genetics 72: 615-638. BEEMAN,R.W., 1987 A homoeotic genecluster in the redflour beetle. SLACK,J. M. W., P.W. H. HOLLANDand C. F. GRAHAM,1993 The Nature 327: 247-249. zootype and the phylogenetic stage. Nature 361: 490-493. BROWN,S. J., R. B.HILCENFELII and R. E. DENELL,1994 The beetle, STRUHL,G., 1984 Splitting thebithorax complex of Drosophila. Tribolium castaneum, has a fushi tarazu homolog expressed in Nature 308: 454-457. stripes during segmentation.Proc. Natl. Acad. Sci. USA 91: TALBERT,P. B., and R. L. GARBER,1994 The Drosophila homeotic (in press). mutation nasobemia (AntpNs)and its revertants: an analysis of DAWS,R., I. DAWN, F. FALCIANI, G.TEAR and M. AKAM, 1994 Dux, mutational reversion. Genetics 138: 000-000. a locust relatedfushi-tararu to but showing no pair-rule TAZIMA,Y., 1964 E-group as a tool of developmental genetics, pp. expression. Development 120: 1561-1572. 6075 in The Genetics of the Silkworm, edited by Y. TAZIMA.Logo DENELL,R. E., 1973 Homeosis in Drosophila. I. Complementation Press, London. studies with revertants of Nasobemia. Genetics 75: 279-297. TIONC,S., L. M. BONEand J. R. S. WHITI'LE,1985 Recessive lethal DUNCAN,I. W., and T.C. KAUFMAN,1975 Cytogenetic analysis of chro- mutations within the bithorax complex in Drosophilamelano- mosome 3 in Drosophila melanogustel: mapping of the proximal gaster. Mol. Gen. Genet. 200: 335-342. portion of the right arm. Genetics 80: 733-752. WAKIMOTO,B. T., and T.C. KAUFMAN, 1981 Analysis of larval segmen- FROHNHOFER,H. G., and C. N~SSLEIN-VOLHARD,1987 Maternal genes tation in lethal genotypesassociated with the Antennapedia gene required for the anterior localization of bicoid activity in the em- complex in Drosophila melanogaster. Dev. Biol. 81: 51-64. bryo of Drosophila. Genes Dev. 1: 880-890. WAKIMOTO,B. T., F. R. TURNERand T. C. GUFMAN, 1984 Defects in GEHRING,W., 1966 Bildung eines vollstiindigen Mittelbeines mit Ster- embryogenesis in mutants associatedwith the Antennapediagene nopleura in der Antennenregion bei der Mutante Nasobemia complex of Drosophila melanogaster. Dev. Biol. 102: 147-172.