back to timetable

Hopeful Monsters ....

HOMEOTIC TRANSFORMATIONS involving the Antennapedia Complex (ANTC)

Adult fly heads: wild type, antennapedia (antennae transformed to 2nd thoracic legs) and a double mutant proboscipedia/ antennapedia (antennae to 2nd thoracic and proboscis to 1st thoracic legs). Note how "clean" the transformations are ie transformation of one head structure does not much affect adjacent structures. This illustrates how each structure is developmentally isolated from its neighbours - the fly is put together like a construction toy with interchangeable parts.

Note that the adult body parts develop from groups of precursor cells that are laid down during embryogenesis but which lie semi-dormant during larval life until their development is triggered again during pupation - when the fly undergoes metamorphosis. These groups of precursor cells are called imaginal discs (the imago is the final stage of an insect e.g. butterfly or adult fruit fly). There is an imaginal disc for each adult structure, e.g. leg disc, wing disc, genital disc, eye/antennal disc, etc. (see Wolpert, chapter 2, p47 or Gilbert, chapter 20, p729).

Mutant flies like those above were noticed as long ago as the 1890s. Most of the early mutations were dominant - ie only one copy of the mutant gene is required to affect a transformation (e.g. Antennapedia: antenna-to-leg). Later, recessive mutations Antennapedia were discovered, which cause the reverse transformation (leg-to-antenna). Many more analogous mutants were isolated that cause transformations of abdominal as well as thoracic and head segments. The mutations were genetically mapped and shown to fall in two gene complexes, the Antennapedia Complex (ANTC) and the Bithorax Complex (BX-C) on separate chromosomes. ANT-C controls the identites of segments anterior to T3, while BX-C controls T3-A8. In the 1970s Ed Lewis undertook a detailed genetic analysis of the BX-C for which he gained a Nobel Prize (with Christiane Nusslein-Volhard and Eric Wieschaus), nearly twenty years later in 1995.

Lewis found that, in general, dominant mutations in the BX-C changed segment identity to that of the next more posterior segment while recessive mutations caused transformations in the opposite direction. For example, the recessive gene bithoraxoid (bxd) causes the first abdominal segment (A1) to take on the character of the third thoracic segment (T3). Since T3 normally carries a pair of legs as well as halteres (balancers), bxd mutants have four pair of legs instead of the normal three, as well as two pairs of halteres. Other recessive BX-C genes are bithorax (bx), mutations of which convert the anterior half of T3 to the anterior half of T2 (i.e. T3a -> T2a), postbithorax (pbx) which converts the posterior half of T3 to the posterior half of T2 (T3p ->T2p). Thus, the double recessive mutant (bx, pbx) carries two pairs of wings*.

In addition, Lewis identified recessive mutations infra-abdominal-2 (iab2), that changed A2->A1, iab3 (A3->A2), iab4 (A4->A3) and so on up to iab8 (A8->A7). Dominant mutations in the BX-C include Contrabithorax (Cbx: T2->T3), Hyper-abdominal (Hab: A1->A2), Ultra-abdominal (Uab: A2->A3) and Miscadestral pigmentation (Mcp: A4->A5).

Lewis also noticed that large deletions that took out the entire BX-C converted all the thoracic and abdominal segments to T2, suggesting that T2 represents the "ground state" that is generated when no BX-C genes are active, and that fates more posterior to that require the activation of one or more BX-C genes. Lewis proposed a model whereby an additional BX-C gene is activated in each more posterior segment; no genes are active in T2, bx+pbx are active in T3, bx+pbx+bxd are active in A1, bx+pbx+bxd+iab2 are active in A2 and so on until all of the BX-C genes are active in A8. In order to achieve this stepwise activation of BX-C genes, Lewis postulated that there was a gradient of a repressor molecule, highest in T2 (where no BX-C genes are expressed) and lowest in A8 (where all are expressed). Further, he postulated that the promoter region of each gene in the BX-C had a different affinity for the repressor, with iab8 having the highest affinity (and thus remaining repressed at lower concentrations of the putative repressor than any other BX-C gene). This model was consistent with the finding that mutations in extra sex combs (esc), a regulatory gene that lies outside the BX-C locus, can apparently activate all of the BX-C genes in all segments T2-A8, generating a mutant fly in which all segments resemble A8.

This model, based purely on genetics, was a remarkable achievement and contains the essence of the truth. However, when the BX-C was cloned and analyzed it was found to be wrong in one important respect: there are only three transcription units (protein coding units) not nine or ten as Lewis had proposed. These units are named Ubx, abd-A and Abd-B, in that order along the chromosome. They are expressed in broad overlapping domains in the embryo, Ubx from T3 to A8, abd-A from A2 to A8 and Abd-B from A5 to A8. Note that they are expressed in progressively more posterior domains that correspond to their relative positions along the chromosome. If the Ubx protein coding sequence is deleted, segments A1 and T3 develop as T2 (i.e. the sum of Lewis's mutations bx, pbx and bxd). If abd-A is deleted, then T2-A1 develop more-or less normally (judging by the bristles on the larval cuticle) but A2-A4 develop as A1 (the sum of iab2, 3 and 4). If Abd-B is deleted, T2-A4 develop normally but A5-A8 develop as A4 (the sum of iab5, 6,7 and 8) (see below).

[red text/digits means deleted or abnormal, black means wild-type. The segment designations beneath the larvae refer to the bristle pattern on the cuticle. From "The making of a fly" by Peter A Lawrence (Blackwell Scientific Publications, 1992) Fig 5.3 p117.]

How do Ubx, abd-A and Abd-B relate to Lewis's mutations? It turns out that most of the mutations Lewis identified do not fall within protein coding sequences, but in regulatory elements that control the expression of Ubx, abd-A and Abd-B. For example, mutation of the cis-acting regulatory control element bxd prevents expression of Ubx in segment A1, iab2 prevents expression of abd-A in A2, and mutation of iab6 abolishes expression of Abd-B in A6 [regulatory mutations].

[*Since more "primitive" flies like dragonflies normally carry two pairs of wings, while in dipteran flies like Drosophila the second pair is stunted to form halteres , it seems likely that mutations of homeotic genes might have played an important role during evolution.]


1. Wolpert Chapter 5 pp161-167 (not a very full treatment)

2. Gilbert Chapter 5 pp552-562

[link to flybase - HOM genes]

back to timetable