p Interactive Fly, Drosophila

broad


DEVELOPMENTAL BIOLOGY

Larval and Pupal Stages

The broad primary response gene is composed of several distinct genetic functions and encodes a family of related transcription factor isoforms. Are BR-C isoforms components of the primary ecdysone response in all tissues and is tissue-specific isoform expression associated with tissue-specific metamorphic outcomes? To study protein expression patterns during the initial stages of metamorphosis, specific antibody reagents have been used that recognize and distinguish among the Z1, Z2 and Z3 BR-C protein isoforms. BR-C isoforms are induced at the onset of metamorphosis, each with unique kinetics of induction and repression. The BR-C proteins accumulate in the nuclei of all larval and imaginal tissues; this indicates that the BR-C is induced as a primary response in many tissues. Several tissues express different levels and combinations of the BR-C isoforms, suggesting that the BR-C is important in determining the tissue-specific outcome of many parallel ecdysone response cascades. For example, prepupal salivary glands (destined for histolysis during metamorphosis) express Z1 isoforms, while imaginal discs (destined for cell differentiation and morphogenesis) shift from the synthesis of Z2 isoforms to the synthesis of Z1 isoforms. The prepupal central nervous system (destined for tissue remodeling) expresses all isoforms, with Z3 predominating. Salivary gland chromosome immunostaining indicates that BR-C proteins interact directly with numerous loci in the polytene genome. Distinct BR-C genetic functions can be correlated with single and specific BR-C protein isoforms (Emery, 1994).

In Drosophila, secretion of the steroid hormone ecdysone from the prothoracic ring gland coordinates and triggers events such as molting and metamorphosis. In the developing Drosophila compound eye, pattern formation and cell-type specification initiate at a moving boundary known as the morphogenetic furrow. The role of ecdysone has been investigated in eye development and the ecdysone signaling pathway has been found to be required for progression of the morphogenetic furrow in the eye imaginal disc of Drosophila. A temperature-sensitive mutation in ecdysoneless (ecd1) reduces ecdysone titer in vivo up to 20-fold, but does not eliminate it completely. Genetic mosaic analysis has shown that ecdysoneless is required in the ring gland (and ovary) and the mutation is thus likely to affect either ecdysone synthesis or release. Genetic disruption, either of the ecdysone signal in vivo with the ecd1 mutant, or of ecdysone response with a Broad-Complex mutant, results in disruption of morphogenetic furrow progression (Brennan, 1998).

To investigate how the ecdysone signal might be transduced in the eye disc, a study was carried out of the expression and role of Broad complex, an early ecdysone response gene complex known to play an important role in metamorphic responses to ecdysone. With an antibody specific to the Z1 finger-containing forms of this protein (the isoform expressed more strongly in differentiating imaginal tissues), this protein is found localized near the furrow. Z1-containing isoforms of this protein begin to be expressed just anterior to the furrow and reach maximal levels posterior to the furrow, following in time the activation of EcRE:lacZ, as visualized by beta-galactosidase staining just anterior to the furrow. This suggests that although BR-C is immediately downstream of the Ecdysone Receptor, there is a delay in its maximal expression as compared with the expression of EcRE:lacZ, which may reflect autoregulation or post-transcriptional control of expression. Both cross-sections and whole mount stainings show that, in addition to its furrow domain of expression, Br-C Z1 is also expressed ubiquitously in the peripodial membrane. Br-C Z1 expression both near the furrow and in the peripodial membrane is greatly reduced when ecdysone titer is reduced in ecd-ts flies. Males hemizygous for npr-1 (an allele null for all Broad-Complex functions), exhibit failures of furrow progression and photoreceptor recruitment. The correlation between effects of removal of ecdysone on furrow progression and Br-C expression and the effects of direct removal of Br-C function suggest that Br-C may mediate a subset of the ecdysone effect in the eye, but doesn't preclude the possibility that Br-C may also have independent functions, as does Usp (Brennan, 1998).

Insect metamorphosis provides a valuable model for studying mechanisms of steroid hormone action on the nervous system during a dynamic phase of functional remodeling. The Drosophila Broad Complex (BRC) holds a pivotal position in the gene expression cascade triggered by the molting hormone 20-hydroxyecdysone (20E) at the onset of metamorphosis. BrC is essential for transducing 20E signals into the morphogenetic movements and cellular assembly that alter the CNS from juvenile to adult form and function. The relationship of BR-C to two other genes was examined: Ecdysone-inducible gene E1 (IMP-E1), coding for an EGF-like domain cysteine pattern protein, and Deformed (Dfd), involved in the metamorphic transition of the CNS. Both BR-C and Dfd are required for maturation of the subeosphageal ganglion. Representatives of the whole family of BrC transcript isoforms accumulate in the CNS during the larval-to-pupal transition and respond directly to 20E in vitro. IMP-E1 is also directly regulated by 20E, but its induction is independent of BR-C, revealing that 20E works through at least two pathways in the CNS. Dfd expression is also independent of BR-C function. Full induction of a number of other primary response genes (e. g., E74 and E75) requires BR-C function. Surprisingly, BR-C and Dfd proteins are expressed in distinct, nonoverlapping subsets of neuronal nuclei of the subesophageal ganglion (SEG) even though both are required for SEG migration into the head capsule. The midline of the ventral ganglion contains many BR-C expressing cells, which may be neurons or glia. Expression in the optic lobes, which undergo BR-C dependent developmental events, is most extensive, with the vast majority of cells staining from BR-C. In contrast, the brain, SEG, and thoracic ganglia show clusters of stained cells, often surrounding a central zone devoid of staining. The pattern of "holes" is reminiscent of the positions of proliferating neuroblasts. This suggests that the segment identity (represented by Dfd expression) and ecdysone cascades operate in separate but parallel pathways to control region-specific reorganization during metamorphosis (Restifo, 1998).

An understanding of the molecular basis of the endocrine control of insect metamorphosis has been hampered by the profound differences in the responses of the Lepidoptera and the Diptera to juvenile hormone (JH). In the presence of JH, there is no change in form; in the absence of JH, ecdysone causes the switching in gene expression necessary for metamorphosis, first to the pupa, then to the adult. JH therefore prevents this switching action of ecdysone and thus maintains the 'status quo' during a molt. In the Coleoptera and in Lepidoptera such as the tobacco hornworm, Manduca sexta, where the epidermis sequentially makes several larval cuticles, the pupal cuticle and finally the adult cuticle, JH prevents each of the metamorphic transitions. By contrast, in Drosophila and the other higher flies, the pupal epidermis, except for the abdomen, is derived from imaginal discs, and exogenous JH does not prevent the larval-pupal transformation, even when given throughout larval life. Nor does JH have any effect on the subsequent external adult differentiation of the head and thorax, although JH disrupts metamorphosis of the nervous and muscular systems when given during the prepupal period. However, JH application during the final larval instar or during the prepupal period prevents the normal adult differentiation of the abdomen, whose cells arise from proliferation of the histoblasts during the prepupal period (Zhou, 2002 and references therein).

In both Manduca and Drosophila, the broad (br) gene is expressed in the epidermis during the formation of the pupa, but not during adult differentiation. Misexpression of Br-Z1 during either a larval or an adult molt of Drosophila suppresses stage-specific cuticle genes and activates pupal cuticle genes, showing that br is a major specifier of the pupal stage. Treatment with a JH mimic at the onset of the adult molt causes br re-expression and the formation of a second pupal cuticle in Manduca, but only in the abdomen of Drosophila. Expression of the Br isoforms during adult development of Drosophila suppresses bristle and hair formation when induced early or redirects cuticle production toward the pupal program when induced late. Expression of Br-Z1 at both of these times mimics the effect of JH application but, unlike JH, it causes production of a new pupal cuticle on the head and thorax as well as on the abdomen. Consequently, the 'status quo' action of JH on the pupal-adult transformation is mediated by the JH-induced re-expression of Br (Zhou, 2002).

Br has long been known to be required for the onset of metamorphosis of Drosophila because the nonpupariating (npr) alleles lack all Br proteins and remain as final instar larvae. In both Drosophila and Manduca, Br transcripts and proteins are expressed prominently during the larval-pupal transformation with different isoforms showing different temporal and tissue specificities through this period and causing either activation or suppression of specific genes. For example, in the Drosophila salivary gland, the induction of Sgs-4 and L71 and the suppression of Pig-1 during the mid and late third instar require the Z1 isoform, while the later suppression of Sgs-4 at puparium formation is due to the downregulation of another transcription factor Forkhead (Fkh) by the Z3 isoform. By contrast, the Z3 isoform activates the expression of Fbp1 in larval fat bodies during the second half of the third instar, while Z2 may play a role in repressing its premature expression. Br proteins also may play a role in the regulation of chromatin structure, since they are found in over 300 sites on the salivary gland chromosomes including sites in the interband regions and in the heterochromatin (Zhou, 2002).

Br-Z1 is the predominant isoform during the time of pupal cuticle formation in Drosophila. Whenever Br-Z1 is expressed during an ecdysone-induced molt, it can direct the epidermis into a program of pupal cuticle production. For example, the molt to the third larval instar in Drosophila begins with the rise of the ecdysteroid that peaks about 12 hours after ecdysis to the second instar. Shortly thereafter, mRNAs for larval cuticular proteins are upregulated. Expression of Br-Z1 during this time suppresses the larval cuticle gene Lcp65A-b and prematurely activates the pupal cuticle gene Edg78E. The ability of Br to be a pupal specifier is also evident during an adult molt. This molt begins about 24 hour APF with the rise of the ecdysteroid titer, and adult procuticle deposition begins about 53 hours APF during the decline of the ecdysteroid titer. Br-Z1 is most effective in activating pupal cuticle genes and suppressing an adult cuticle gene when expressed just before the normal onset of adult procuticle gene expression. This temporal restriction suggests that although Br selects which cuticle genes will be expressed, it can only do so within the confines of an ecdysone-induced program that determines the timing of cuticle gene expression at every molt. Therefore, in either a larval or an adult molt, the expression of Br-Z1 is sufficient to redirect that molt towards the pupal program (Zhou, 2002).

Adult differentiation of the epidermis can be divided into two developmental phases: cellular morphogenesis followed by cuticle deposition. Morphogenesis of the epidermis begins with the formation and outgrowth of the bristles (macrochaetes, microchaetes) between 30 and 45 hours APF, first in the head and thorax, then in the abdomen. Trichomes (hairs) are then formed by most of the general epidermal cells, beginning on the wing at 33 hours APF and on the abdomen about 48 hours APF. The general epidermis deposits three cuticular layers: cuticulin, epicuticle and procuticle. The bristle and hair shafts lack the procuticle layer. Cuticulin formation begins in patches on the wings and legs at 35-36 hours and on the abdomen at 40-45 hours APF, followed by synthesis of a continuous epicuticle once morphogenesis is complete. Adult procuticle synthesis occurs primarily between 53 and 90 hours APF. The expression of the adult cuticle gene Acp65A is restricted to flexible cuticle regions of the abdomen, the wing hinges, leg joints and the ptilinum and begins about 55-60 hours APF (Zhou, 2002).

Br disappears before the onset of adult differentiation in both Manduca and Drosophila. This disappearance is crucial for normal adult development since the misexpression of Br in Drosophila can affect both adult morphogenesis and adult cuticle production. When expressed between 30 and 40 hours APF, Br causes truncation of the bristles with early times affecting the bristles of the head and thorax and slightly later times affecting those of the abdomen. This timing corresponds to the onset of bristle outgrowth in the different regions. Suppression of bristle outgrowth occurs with misexpression of each of the Br isoforms, although the Z1 isoform has the strongest effects because the truncation is seen with expression of only two copies as well as with four copies of Br-Z1. Bristle outgrowth occurs by extension of the longitudinal actin microfilament arrays that surround the microtubular core. These actin filaments are bundled together, then crosslinked to support the elongating bristles, using sequentially the product of the forked gene and fascin. Although an occasional forked bristle is seen after misexpression of Br, the primary effect is truncation similar to that seen after exposure to inhibitors of microfilament elongation, indicating Br may be able to interfere with this process, either directly or indirectly (Zhou, 2002).

Trichome production in the abdominal epidermis is suppressed by Br-Z1 expression between 36 and 39 hours APF. Since nearly every epidermal cell normally produces a trichome, this result shows that early Br expression also suppresses morphogenesis of the general epidermis. In this case, the effective time is about 10-12 hours before abdominal trichome production. By 42 hours APF bristle and trichome morphogenesis is no longer affected by expression of Br. Between this time and 60 hours APF, the effects are primarily on the types of cuticle proteins produced. Br-Z1 is most effective in suppressing adult cuticle gene expression and causing re-induction of pupal cuticle gene expression with the resultant formation of a thin, transparent, pupal-like cuticle by the general epidermis. None of the other isoforms have such a dramatic effect on the external appearance of the cuticle, although Br-Z2 causes re-expression of the two pupal cuticle genes, and Br-Z3 causes re-expression of one pupal cuticle gene and suppression of the adult cuticle gene studied, indicating that they normally play a role in production of pupal cuticle. Cuticle is composed of many proteins, so a predominance of adult cuticle proteins could maintain the cuticular morphology despite the presence of some pupal cuticle proteins or the absence of specific adult cuticle proteins. Further study is required to resolve this issue (Zhou, 2002).

Although bristle morphogenesis is unaffected by expression of Br during the onset of cuticle formation, bristle pigmentation and sclerotization are subsequently inhibited. Whether this suppression is due to the type of epicuticle deposited or to an inhibitory action of Br on the melanization and sclerotization pathways themselves is unclear. In the case of Br-Z1, this effect is most pronounced when expression is either between 43 and 48 hours APF or later during 54-60 hours APF. Although the pupal cuticle genes used in this study all encode proteins found in the pupal exocuticle (the outer procuticle), Br-Z1 probably also directs pupal epicuticle production. If so, the earlier expression of Br-Z1 may be suppressing the deposition of the proenzymes necessary for tanning and melanization that are normally associated with adult cuticle. Such a suppression would not be unexpected, since normal pupal cuticle does not tan or melanize. These proenzymes are often laid down very early in formation of the new cuticle. Br-Z3 or Br-Z4 also suppresses bristle pigmentation but only when expressed late between 52 and 60 hours APF. This effect of later expression of any of these three isoforms is probably due to an interference with the production or deposition of the substrates for these enzymes, which normally appear in the cuticle shortly before the proenzymes are activated. However, an effect on the pigmentation process itself that occurs later cannot be ruled out (Zhou, 2002).

These different effects of Br misexpression depending on its timing indicate that Br and/or the unknown proteins whose expression Br regulates must be present to direct the pupal program. Once they disappear, the cells can revert back to the expression of the adult program. In these experiments, Br transcripts disappear by 6 hours after the heat shock, but the proteins are present until at least 9 hours. Thus, in order to obtain a second pupal cuticle that lacks bristles and trichomes, one must express Br-Z1 during both the initiation of bristle outgrowth and the onset of procuticle formation (Zhou, 2002).

An important finding of these studies is the fact that the presence of Br-Z1 at the time of cuticle formation is sufficient to redirect the program of cuticle gene expression into a pupal mode in cells that have completed their adult morphogenesis. This is most clearly seen after expression of Br at 48 or 52 hours APF. The cells of the general abdominal epidermis make the adult hairs but then deposit procuticle that includes pupal cuticle proteins. Thus, cells already committed to and expressing aspects of adult differentiation are plastic and can be caused to re-express pupal products when given the proper transcription factor. Clearly the suppression of br through the duration of adult development is essential for the normal completion of metamorphosis (Zhou, 2002).

JH has long been known to prevent metamorphosis without interfering with the molting process itself. In both Manduca and Drosophila abdomens, JH causes the formation of a second pupal cuticle only when given before the onset of the adult molt. These studies have revealed that this re-expression of the pupal program in both species is associated with the re-induction and maintenance of Br expression during the molt. This renewed Br expression appears to be sufficient to mediate the 'status quo' action of JH, since Br can both activate pupal genes and suppress adult genes. Thus, during the crucial period of adult commitment, ecdysone in the absence of JH must switch off Br so that the adult-specific program of differentiation can occur (Zhou, 2002).

In Drosophila the JH-sensitive period of the abdomen is during the prepupal period with the highest sensitivity being at the time of pupariation and loss of sensitivity after head eversion at 12 hours APF. During this time the histoblasts are proliferating rapidly. After this JH-sensitive period is over, beginning about 15 hours APF, these imaginal cells spread over the pupal abdomen and replace most of the larval cells by about 28 hours APF. Throughout this period, both types of cells express Br. JH given at pupariation has no apparent effect on the proliferation or spreading of these cells or on their replacement of the larval epidermis. Nor does it interfere with their normal Br expression during this period. Its effect is only to cause renewed and sustained expression of Br in the imaginal cells during the adult molt up through 72 hours APF (Zhou, 2002).

JH at pupariation has no apparent effect on the adult development of the Drosophila head and thoracic structures that are derived from the imaginal discs. This study shows that the refractoriness of the head and thorax to the JH treatment is due to the inability of JH to cause Br re-expression in these regions during the adult molt. Yet appropriate misexpression of Br during adult differentiation results in pupal cuticle formation in both the head and thorax, showing that Br retains its pupal-specifying function in these regions. Hence, the refractoriness to JH of the head and thorax must be due to a lesion in the pathway from the JH receptor to Br re-induction, possibly to the loss of the receptor itself (Zhou, 2002).

In all insects including Drosophila, JH is present during the larval molts, then declines during the last larval instar. In both Manduca and Drosophila epidermis and imaginal discs, Br is not expressed during the larval molt. Pupal commitment of the polymorphic epidermis of Manduca by 20E at the end of the larval feeding period is correlated with the appearance of Br, and both can be prevented by JH. By contrast, in Manduca wing imaginal discs, Br appears earlier in the final larval instar as the discs become competent to metamorphose, and JH cannot prevent this appearance but only delays it. In Drosophila and the higher flies, the pupa is derived from imaginal discs except for the abdominal cuticle that is produced by the persisting larval epidermal cells and the histoblasts. Although the effect of JH on the appearance of Br in Drosophila discs and larval epidermis has not been directly studied, dietary JH throughout larval life delays the onset of metamorphosis but does not prevent pupation, indicating that these tissues can turn on Br despite the presence of JH. Thus, the derivation of the Drosophila pupa from primarily imaginal discs probably accounts for the inability of JH to prevent the larval-pupal transformation, although the lack of effect of JH on the abdominal epidermis in its switch to pupal cuticle production remains unexplained. The mechanism whereby JH prevents the switching-on of Br by ecdysone during a larval molt and also prevents its switching-off by ecdysone at adult commitment is still unclear (Zhou, 2002).

These studies demonstrate for the first time that by the misexpression of a single transcription factor of the ecdysone cascade, the Br-Z1 isoform, one can redirect cells undergoing either larval or adult differentiation into a pupal developmental program. They also provide the first molecular basis for the 'status quo' action of JH on the pupal-adult transformation, by showing that JH causes the re-induction of Br expression and consequently re-expression of the pupal program during the molt (Zhou, 2002).

Anciently duplicated Broad Complex exons have distinct temporal functions during tissue morphogenesis

Broad Complex (BRC) is an essential ecdysone-pathway gene required for entry into and progression through metamorphosis in Drosophila. Mutations of three BRC complementation groups cause numerous phenotypes, including a common suite of morphogenesis defects involving central nervous system (CNS), adult salivary glands (aSG), and male genitalia. These defects are phenocopied by the juvenile hormone mimic methoprene. Four BRC isoforms are produced by alternative splicing of a protein-binding BTB/POZ-encoding exon (BTBBRC) to one of four tandemly duplicated, DNA-binding zinc-finger-encoding exons (Z1BRC, Z2BRC, Z3BRC, Z4BRC). Highly conserved orthologs of BTBBRC and all four ZBRC were found among published cDNA sequences or genome databases from Diptera, Lepidoptera, Hymenoptera, and Coleoptera, indicating that BRC arose and underwent internal exon duplication before the split of holometabolous orders. Tramtrack subfamily members, abrupt, tramtrack, fruitless, longitudinals lacking (lola), and CG31666 were characterized throughout Holometabola and used to root phylogenetic analyses of ZBRC exons, which revealed that the ZBRC clade includes Zabrupt. All four ZBRC domains, including Z4BRC, which has no known essential function, are evolving in a manner consistent with selective constraint. Transgenic rescue was used to explore how different BRC isoforms contribute to shared tissue-morphogenesis functions. As predicted from earlier studies, the common CNS and aSG phenotypes were rescued by BRC-Z1 in rbp mutants, BRC-Z2 in br mutants, and BRC-Z3 in 2Bc mutants. However, the isoforms are required at two different developmental stages, with BRC-Z2 and -Z3 required earlier than BRC-Z1. The sequential action of BRC isoforms indicates subfunctionalization of duplicated ZBRC exons even when they contribute to common developmental processes (Spokony, 2007).

Temporal patterns of broad isoform expression during the development of neuronal lineages in Drosophila

During the development of the CNS of Drosophila, neuronal stem cells, the neuroblasts (NBs), first generate a set of highly diverse neurons, the primary neurons that mature to control larval behavior, and then more homogeneous sets of neurons that show delayed maturation and are primarily used in the adult. These latter, 'secondary' neurons show a complex pattern of expression of broad, which encodes a transcription factor usually associated with metamorphosis, where it acts as a key regulator in the transitions from larva and pupa. The Broad-Z3 (Br-Z3) isoform appears transiently in most central neurons during embryogenesis, but persists in a subset of these cells through most of larval growth. Some of the latter are embryonic-born secondary neurons, whose development is arrested until the start of metamorphosis. However, the vast bulk of the secondary neurons are generated during larval growth and bromodeoxyuridine incorporation shows that they begin expressing Br-Z3 about 7 hours after their birth, approximately the time that they have finished outgrowth to their initial targets. By the start of metamorphosis, the oldest secondary neurons have turned off Br-Z3 expression, while the remainder, with the exception of the very youngest, maintain Br-Z3 while they are interacting with potential partners in preparation for neurite elaboration. That Br-Z3 may be involved in early sprouting is suggested by ectopically expressing this isoform in remodeling primary neurons, which do not normally express Br-Z3. These cells now sprout into ectopic locations. The expression of Br-Z3 is transient and seen in all interneurons, but two other isoforms, Br-Z4 and Br-Z1, show a more selective expression. Analysis of MARCM clones shows that the Br-Z4 isoform is expressed by neurons in virtually all lineages, but only in those cells born during a window during the transition from the second to the third larval instar. Br-Z4 expression is then maintained in this temporal cohort of cells into the adult. These data show the potential for diverse functions of Broad within the developing CNS. The Br-Z3 isoform appears in all interneurons, but not motoneurons, when they first begin to interact with potential targets. Its function during this early sorting phase needs to be defined. Two other Broad isoforms, by contrast, are stably expressed in cohorts of neurons in all lineages and are the first examples of persisting molecular 'time-stamps' for Drosophila postembryonic neurons (Zhou, 2009).

In Drosophila and other insects with complete metamorphosis, broad expression is associated with the initial phase of metamorphosis - the transition of the larva to the pupa. The broad isoforms play critical roles in this transition and are needed for the proper 'read out' of ecdysone-coordinated molecular events. In broad null mutants (for example, brnpr3) the premetamorphic growth and development of larval and imaginal tissues appear normal, but metamorphosis is blocked in all tissues before puparium formation. In brnpr3 mutants, the larval CNS appears to be normal before the wandering stage, although its subsequent metamorphosis is disrupted (Zhou, 2009).

While the classic role for broad is in the regulation of metamorphosis, the gene is known to have a wider role in development. It is involved in chorion formation during oogenesis in adult Drosophila, and this study shows that Br-Z3 expression is initiated in the CNS of the late embryo and that Br-Z1 and Br-Z4 continue to be expressed in the adult. The origins of the four isoforms of broad are ancient, extending back into the Crustacea (Spokony, 2007), much earlier than the evolution of complete metamorphosis in insects. Functions typified by its role in the nervous system may represent an ancestral function for broad, and its role in metamorphosis a more recently derived function for this gene (Zhou, 2009).

In the developing vertebrate CNS, and especially the spinal cord, the initial set of neurons that set-up the basic architecture of the CNS are often called primary neurons. They are typically large and readily identifiable and may be later replaced by a larger flood of secondary neurons. In metamorphic species, like frogs, the primary neurons are ascribed to function in the larval tadpole while secondary neurons are generated in preparation for metamorphosis to the adult frog. An analogous concept of primary and secondary neurons is becoming established for Drosophila development. Similar to that seen in metamorphic vertebrates, the term primary neuron is typically associated with cells born during the first phase of neurogenesis in the embryo and secondary neuron for the cells born during the later, larval phase. The larval-born cells arrest their development soon after their birth and wait until metamorphosis before maturing into functional neurons. However, BrdU feeding experiments show that some embryonic-born neurons also arrest without showing any branching, a condition identical to that seen in the neurons that are born in the larva. Moreover, it appears that these arrested, embryonic neurons show a sustained expression of Br-Z3, a feature characteristic of the secondary neurons born during the larval phase of neurogenesis (Zhou, 2009).

In light of the latter observations, it is worth contemplating the definition of primary and secondary neurons. There are three potential uses of this designation/terminology. One would be to use it to denote embryonic versus postembryonic born neurons. A second would be to use it to refer to the functional status of a neuron, where embryonic neurons that are fully differentiated, functional components of a network could be termed primary neurons and neurons that had initiated neurite growth but 'stalled' their development prior to establishing pre- and postsynaptic specializations would be secondary neurons. A third use would be to refer to the developmental transition within lineages where a neural precursor switches from producing diverse progeny at each division to generating homogenous 'blocks' of cell types (Zhou, 2009).

There is a striking difference between the mechanisms of specification of neuronal identities of early and late-born neurons in insects. For a given NB, the characteristics of the first few progeny can be strikingly diverse and are based on birth-order of the parent GMC. The determination of birth-order is based on unique molecular identities that are bestowed on successive GMCs through an intrinsic, temporal program of transcription factor production by the NB and the stable passage of that expression to the GMC during cell division. A series of loss-of-function and gain-of-function studies confirmed that the presence of these transcription factors is responsible for the diversity of neuronal phenotypes within a given lineage and underlies the great diversity of early neuronal types seen across the lineages. This intrinsic progression, however, appears to stop after the expression of castor, and remaining neurons, both embryonic and postembryonic, express the transcription factor gene. During the postembryonic neurogenic period, the NBs typically produce groups of neurons with similar characteristics. This is most evident during the postembryonic phase of neurogenesis in moths and flies, where neurons are produced in discrete blocks of similar cells, rather than as unique individuals. The expression of Chinmo is required for establishing the identities of an early-born block of cells in the mushroom bodies and antennal lobes, among other lineages. This study reports that the expression of Broad-Z4 apparently provides a molecular marker for the next block of postembryonic neurons. Unlike the early embryonic identity factors, though, Broad is not expressed first in the NB and then passed into this cohort of cells. Rather, it appears in neurons some time after their birth, likely in response to hormonal conditions experienced by them or their parent GMC (Zhou, 2009).

Clonal studies on Drosophila embryos suggest that the generation of neuron classes does not abruptly begin with the postembryonic neurons. Although the embryonic lineages are remarkable for the diversity of their early-born neurons, the later-born cells (farther from the neuropil) in many lineages show clustered cell bodies and similar neurite trajectories, indicating that they may constitute a discrete neuronal type. This study proposes that these later cells, along with the post-embryonic neurons, should be called the secondary neurons and the term primary neurons should be confined to the initial neurons that are produced during embryogenesis during the progression from Hunchback to Castor. Typically, the first four GMCs produced by a given NB would account for its primary neurons, although, in some lineages, like NB 7-1, the expression of a factor may be prolonged to give two GMCs that produce similar properties. In the ventral CNS, the primary neurons would account for the majority of the neurons present at hatching. Most of the early-born secondary neurons are also functional and contribute to larval behavior, while others show the arrested development described in this study (Zhou, 2009).

Uncoupling of the terms primary and secondary from embryonic versus postembryonic origins circumvents problems with using these terms in a comparative context. In more basal insects, like grasshoppers, that have direct development, all of the neurons in the ventral CNS are made during embryogenesis, and, hence, would be considered as primary neurons if time of birth was the deciding criterion. Similarly, within the insects with complete metamorphosis, neurogenic arrest occurs at different points in the lineage depending on whether you are making a reduced nervous system for a relatively simple larva like a fly maggot, or a complex CNS that is a feature of the hunting larvae of ground beetles. Basing the designation of primary and secondary on the mechanisms that generate cellular diversity, rather than on the stage at their birth or their functional status within the network allows for homologues to have the same designation across these diverse taxa (Zhou, 2009).

In Drosophila, most primary neurons express low to moderate levels of Br-Z3 during late embryogenesis. Their embryonic expression was not characterized in detail, and these neurons do not then re-express Br-Z3 (or any other Broad isoform) when they undergo remodeling at the end of larval life. The secondary interneurons also transiently express Br-Z3 during their early development. The results from feeding larvae on a BrdU diet from hatching indicate that these embryonic secondary neurons account for many of the strong Br-Z3+ cells seen at hatching (Zhou, 2009).

The main focus of the paper has been on the expression of Broad in the neurons born during the larval phase of neurogenesis. BrdU incorporation was used to measure the latency to the onset of Br-Z3 expression in the secondary neurons. For larvae at 96 h AEL, the BrdU pulse length needed to see labeling in Br-Z3+ neurons was 11 h, and defines the latency from the end of DNA synthesis in a GMC to the appearance of Br-Z3 in its daughters. At 96 h AEL (72 h post-hatching), the average lifetime of a GMC is about 7 h and its G2 phase lasts about 4 h. Hence, a young neuron starts expressing Br-Z3 about 7 h after its birth, which is about 14 h after the birth of its parent GMC. The arrested embryonic NBs resume dividing early in the second instar larva, and GMCs are seen associated with them by 60 h AEL (36 h post-hatching). In this study, Br-Z3 started to appear in the clusters of neurons by about 72 h AEL, a timing consistent with a 14 h latency between the birth of a GMC and the subsequent appearance of Br-Z3 in its daughter cells. Consequently, it is assumed that this latency is likely similar throughout the period of postembryonic neurogenesis (Zhou, 2009).

Although the secondary neurons may be relatively uniform in the timing of the onset of Br-Z3 expression, they differed in how long this expression was maintained. One pattern was shown by the embryonic and oldest postembryonic secondary cells. These showed initial Br-Z3 expression but then lost it by the start of wandering. They correspond to those cells characterized as being Broad negative and Chinmo+ at pupariation. The other pattern was shown by later born neurons (whose GMCs are born after about 72 h AEL) and these show the persistence of Br-Z3 through pupariation but then lose it early in adult differentiation (Zhou, 2009).

The distinction between the Chinmo+ and Broad+ groups of secondary neurons is set up by the postembryonic actions of Seven-up and Castor. For many lineages, removal of Seven-up resulted in the loss of Broad+ neurons in favor of the larger Chinmo+ cells. A similar result was seen in some of the lineages by the prolonged expression of Castor. It is interesting that both Seven-up and Castor have roles in temporal identity in the embryo. It could be that Br-Z3 expression in the embryonic secondary neurons may likewise be the result of the embryonic interplay of these two genes (Zhou, 2009).

The appearance of Br-Z3 in thoracic interneurons about 7 h after their birth coincides with the loss of expression of Notch in the membrane of these young neurons. Notch is prominent in the membranes of newborn neurons during pathfinding but then rapidly disappears. Consequently, the appearance of Br-Z3 is correlated with a new neuron arriving at its initial target. The best data on the behavior of growth cones in this post-pathfinding stage are for in-growing retinal axons in the first optic neuropil in Drosophila. After arriving at their initial location these axons undergo a 'sorting phase', during which their growth cones spread filopodia laterally to select their cartridge partners. Interestingly, photoreceptors are the only sensory neurons that express Br-Z3 and they do so during this early sorting phase (Zhou, 2009).

Expression of Br-Z3 in the developing medulla interneurons may also be correlated with the sorting of potential partners. When Br-Z3 first appears in the medulla in the mid third instar, it is confined to the most medial columns of neurons. Through time, more lateral columns progressively acquire Br-Z3 expression, and all outer medulla neurons are expressing this isoform by 14 h APF. The neurons of the most medial columns interact with the first photoreceptors growing in from the posterior border of the eye, and as axons grow in from the subsequent rows of ommatidia over the next 2 days, these interact with the neurons of progressively more lateral columns. The wave of Br-Z3 expression across the medulla therefore matches the wave of afferent in-growth into this structure. Br-Z3 expression in the medulla remains prominent through 27 h APF after which it fades. Throughout this time the axon terminals of photoreceptors R7 and R8 are refining their initial positions in the layers of the medulla (Zhou, 2009).

The Br-Z3 expression in thoracic interneurons suggests that there may be a similar 'sorting phase' within the developing central neuropils. The exact role of Br-Z3 in this sorting process is still being explored, but the ectopic expression of Br-Z3 in the LNv neurons shows that the presence of Br-Z3 during initial outgrowth results in excessive sprouting. If the expression of Br-Z3 in the secondary neurons marks a sorting phase, then groups of these neurons may vary in the time when the sorting is completed. Notably, the early, Chinmo+ secondary neurons lose Br-Z3 expression by the start of wandering. This early loss may mean an early commitment to initial targets, and they may serve a pioneer function in organizing the structure of the nascent adult neuropils. The bulk of secondary neurons, however, retain BR-Z3 expression until neurogenesis is almost complete in the central brain and the thorax. The last-born cells in these regions show only a very brief Br-Z3 expression as they are inserting into a neuropil that is already highly sorted (Zhou, 2009).

Br-Z3 expression in the larval γ neurons of the mushroom body is notable because it is the rare case of functioning neurons that continue to express Br-Z3. The significance of this expression is not known. This may represent a unique function of Br-Z3 in larval mushroom body neurons. Alternatively, it may indicate that a 'sorting phase' may continue even in mature neurons. γ neurons are continually added through the early larval stages and the older γ neurons may have to continue to sort and rewire as the γ class swells in number. Besides the mushroom body γ neurons, it was found that a few mature-looking neurons in the brain and thorax of newly hatched larvae were also BR-Z3 positive. The identity of these neurons is not known, but as proposed for the mushroom body γ neurons, they may also have some unique plasticity demands as the larva grows (Zhou, 2009).

A final puzzle involving Br-Z3 is its absence from motoneurons. In the growing larva, the axons from the lineage 15 motoneurons extend to the leg imaginal disc where they end without obvious elaborations. The subsequent interactions of their sprouting axon terminals with nascent muscles may require a very different set of processes from those used by interneurons as they sort out central targets (Zhou, 2009).

The BTB domain transcription factor gene chinmo was the first temporal specifying gene found for the postembryonic neurogenic period. Its presence in early-born neurons in both mushroom body and antennal lobe projection neuron lineages establishes early neuronal phenotypes in these lineages. Just prior to metamorphosis, the secondary neuron lineages can be divided into older, Chinmo+ and younger, Broad+ neurons, the latter being due to the Br-Z3 isoform. Both Chinmo and Br-Z3 disappear early in metamorphosis, though, and it is not known how their transient expression is then turned into persisting markers of temporal identity. Other Broad isoforms, notably Br-Z1 and Br-Z4, were found to provide permanent molecular time-stamps of birthdates (Zhou, 2009).

Prior to metamorphosis, the Br-Z4+ neurons were always found clustered within a given lineage, a pattern suggesting that they arose from consecutively born GMCs. Br-Z4+ neurons were found in all of the thoracic lineages, except for the very small motor lineage generated by NB 24. Although brain lineages were not examined in detail, the broad pattern of Br-Z4 expression shows that this isoform is also expressed in most brain lineages. Thoracic lineages that made two classes of interneurons had roughly twice the number of Br-Z4+ neurons as lineages that produced only one class of interneuron. The secondary neurons generated by a NB can be divided into two hemilineages, each composed of the collective Notch+ or Notch- daughters arising from the division of the series of GMCs. Studies blocking programmed cell death showed that in lineages with one bundle, one of the daughters consistently dies and only one hemilineage survives. The lineages with both hemilineages had roughly twice the number of neurons compared with those that retain a single hemilineage. This pattern is consistent with a thoracic NB typically generating 13 to 16 GMCs whose daughters will express Br-Z4, but then cell death reducing the number by half in those that have a single hemilineage. Some lineages may deviate from this pattern; for example, lineage 0 has more Br-Z4+ neurons than would be expected for a lineage with a single surviving daughter, and atypically small lineages, such as lineages 2 and 15, have reduced numbers of Br-Z4+ cells (Zhou, 2009).

MARCM clones induced at 24 and 54 h AEL contained their complete complement of Br-Z4+ cells. Those induced at 76 h contained only a few Br-Z4+ neurons, and most Br-Z4+ cells were outside of the clone boundaries. The presence of about six Br-Z4+ neurons within the latter clones suggests that only three to six GMCs with the Br-Z4 fate were left to be born at the time of the clone induction. Given that a NB produces about one GMC per hour at this time in larval life, it is estimated that the GMCs that generate the Br-Z4 cells are born from about 66 to about 80 h AEL (that is, from about 18 h into the second instar until about 8 h of the third larval stage). The analysis of MARCM clones in the adult shows that neurons born during this period in the larva continue to express Br-Z4 after metamorphosis. Br-Z4 expression in the adult therefore provides a molecular marker for the neurons whose GMCs were born during this window of larval life. Br-Z1 is co-localized in a subgroup of the Br-Z4 neurons, but it is not known if this expression defines a specific temporal subset of the Br-Z4 neurons. Within the postembryonic lineages, neuronal fate is clearly related to the time of birth. How Br-Z4 and Br-Z1 participate in these fate decisions remains to be examined (Zhou, 2009).

Broad differs from the embryonic temporal specification genes in that neither Br-Z4 nor Br-Z1 is expressed in the NBs or parent GMCs. Br-Z4 expression only becomes evident towards the start of metamorphosis, well after the neurons are born. This suggests a mechanistic difference in the way in which primary versus secondary neurons are 'time-stamped'. The cells that will express Br-Z4 at the outset of metamorphosis are included in the Br-Z3+ group and not in the Chinmo+ group. Based on its timing, the Chinmo/Broad-Z3 boundary likely marks the start of production of the Br-Z4 neurons. Interestingly, Chinmo and Br-Z3 then disappear early in metamorphosis, whereas Br-Z4 persists as a permanent temporal marker for these neurons. This is the first reported example of a marker for temporal identity that persists into the adult (Zhou, 2009).

A second potential difference from the embryonic pattern is in the signals that control expression of the timing genes. In the embryo, the switch from one timing gene to the next is intrinsically controlled within each lineage. In the postembryonic phase of neurogenesis, by contrast, extrinsic factors appear to play a role in determining the timing of switching from one cell type to another. The switching on and off of GMCs that will make Br-Z4+ neurons coincides with two 'milestones' in Drosophila larval development. The GMCs with the Br-Z4 fate start to be made late in the second instar, around the time larvae achieve the threshold size for metamorphosis and become committed to entering the terminal larval stage. They stop being made at the time in the third instar when the larva achieves its 'critical weight' sufficient to carry it through to pupariation and metamorphosis. Both of these milestones represent endocrine-generated decisions. These extrinsic signals may be responsible for the expression of Br-Z4 and the specification of a distinct subtype of neuron, but this remains to be examined (Zhou, 2009).

Oogenesis

Translating available food into the number of eggs laid by Drosophila: Differential expression of BR-C isoforms plays a key role in controlling whether the fate of the egg chamber is to develop or undergo apoptosis

In Drosophila and other insects egg production is related to the nutrients available. Somehow the nutritional status of the environment is translated into hormonal signs that can be 'read' by each individual egg chamber, influencing the decision to either develop into an egg or die. BR-C is a control gene during oogenesis and the differential expression of BR-C isoforms plays a key role in controlling whether the fate of the egg chamber is to develop or undergo apoptosis (Terashima, 2004).

In Drosophila oogenesis, the nutritional status of the female's environment affects YP synthesis and egg production. Stage 8 and 9 egg chambers in starved flies are degenerated by apoptosis, leading to a reduced number of egg chambers beyond stage 8. Hence fewer eggs are laid. During nutritional shortage, the nutrients must be used to develop a smaller number of normal eggs. To achieve this, it is essential to select which egg chambers will develop to mature eggs and which will die (Terashima, 2004).

When the ovary accumulates stage 14 oocytes in the sexually mature virgin female, the production of new eggs is prevented by resorption of egg chambers at stage 9, mediated by the high level of 20E in the hemolymph. It seems likely that the resorption of stage 9 egg chambers under nutritional stress may well occur by a similar mechanism, since reduced food induces increased ecdysone concentration in the flies, and 20E in turn induces apoptosis of the egg chambers at stages 8 and 9. While ecdysone induces a reduction in the number of egg chambers, JH acts to suppress this reduction. Females homozygous for ap56f produce low levels of JH but are fertile although vitellogenesis is subject to a delay. The ap56f ovaries are capable of ultimately producing greater than wild-type amounts of ecdysteroids. It is possible that the ecdysone concentration in the flies under starvation is suppressed when JHA is applied to the abdomen of the flies and thus JHA suppresses the ecdysone-induced apoptosis in egg chambers at stage 8/9. It is likely that both ecdysone and JH levels are affected by the nutritional status of Drosophila. Ecdysone and JH concentration could be modulated by synthesis of hormones or by inducing metabolic pathways that degrade the hormones. Using this mechanism a female can relate the number of eggs produced to the food available (Terashima, 2004).

BR-C isoforms and BR-C mRNA are expressed differentially in a number of tissues during metamorphosis and control subsequent cell differentiation. Late in Drosophila oogenesis, BR-C regulates the dorsal-ventral signaling pathway and leads to prolonged endoreplication of follicle cell DNA and to additional amplification of specific genes. The BR-C expression pattern differs between starved and fed flies earlier in oogenesis at stages 5, 6, and 8. BR-C expression in oogenesis is first detected in the egg chamber at stage 5 under normal nutritional conditions, but under nutritional shortage it is detected later, at stage 6. This shows that nutritional shortage is detected in the follicle cells of the egg chamber in previtellogenic stages. During metamorphosis of holometabolous insects, the developmental program is switched from 'larval' to 'pupal' during the final larval instar, a switch that is referred to as pupal commitment. To execute apoptosis, tissues and glands must first become committed to apoptosis and then execute the decision and die. It is proposed that the altered timing of BR-C expression affects this commitment stage in the egg chamber. By stage 8, when the egg chamber must develop or die, the different isoforms of BR-C, specifically Z2 and Z3, differ between starved and fed flies. Z2 and Z3 are not expressed in egg chambers at stage 8 under normal nutritional conditions, but they are expressed in egg chambers at stage 8 under nutritional shortage or if 20E is injected into fed flies. This means that Z2 and Z3 are expressed under apoptotic conditions. Furthermore, Z2 and Z3 overexpression induces apoptosis at stages 8 and 9 (Terashima, 2004).

In the ovary of starved flies, Z2 and Z3 expression is independent, but in fed flies Z3 overexpression induces Z2 expression. 20E injection overrides the nutritional signals and induces apoptosis in the egg chambers at stages 8 and 9 in fed flies probably by activating Z3 that in turn induces Z2 expression. Z3 therefore acts as key transcript for inducing apoptosis. On the basis of these results, BR-C acts as a selector gene for progressing oogenesis or inducing apoptosis at stages 8 and 9 (Terashima, 2004).

A model is presented for the control of oogenesis in relation to the nutritional status. Each Drosophila egg chamber selects between two fates during development. These fates are triggered by the hormonal conditions that are in turn regulated by the nutritional environment. The higher ecdysone concentration in starved flies induces BR-C Z2 and Z3 expression at stage 8, and these genes activate the pathway of apoptosis in the egg chambers at stages 8 and 9. In fed flies, the ecdysone concentration is lower and hence Z2 and Z3 are not expressed in the egg chambers at stage 8. The JH analog, methoprene, suppresses apoptosis at stages 8 and 9, suggesting that JH acts as a suppressor of apoptosis and an inducer of normal development. The normal developmental pathway requires that yp genes are expressed by stage 8; however, Z3 suppresses yp expression. Thus yp expression is suppressed by nutritional shortage and by 20E injection in fed flies. The expression of Z3 that is induced by nutritional shortage or 20E injection suppresses yp expression at stage 8 in the follicle cells and hence prevents normal development from continuing (Terashima, 2004).

In summary, the environmental status of the fly is sensed by each individual egg chamber and that some develop and some die in proportion to the food available. The nutritional status of the environment controls the developmental fate of the egg chamber in Drosophila by adjusting the hormonal balance in the fly. The key gene in follicle cells that is controlled by the hormone balance is BR-C; a the BR-C Z3 isoform is particularly important for egg-chamber development in Drosophila in relation to the nutrition available (Terashima, 2004).

Nutritional status affects 20-hydroxyecdysone concentration and progression of oogenesis in Drosophila: Ecdysone activates the apoptosis pathway, including BR-C Z2, Z3 and E75A expression, in follicle cells

Drosophila egg production depends upon the nutrition available to females. When food is in short supply, oogenesis is arrested and apoptosis of the nurse cells is induced at mid-oogenesis via a mechanism that is probably controlled by ecdysteroid hormone. Expression of some ecdysone-response genes is correlated with apoptosis of egg chambers. Moreover, ecdysteroid injection and application of juvenile hormone induces and suppresses the apoptosis, respectively. In this study, an investigation was carried out to see which tissues show increases in the concentration of ecdysteroids under nutritional shortage to begin to link together nutrient intake, hormone regulation and the choice between egg development or apoptosis made within egg chambers. Ecdysteroid levels in the whole body, ovaries and haemolymph samples were measured by RIA, and it was found that the concentration of ecdysteroid increased in all samples. This contributes to the idea that nutritional shortage leads to a rapid high ecdysteroid concentration within the fly and that the high concentration induces apoptosis. Low concentrations of ecdysteroid are essential for normal oogenesis. It is suggested there is threshold concentration in the egg chambers and that apoptosis at mid-oogenesis is induced when the ecdysteroid levels exceed the threshold. Starvation causes the ovary to retain the ecdysteroid it produces, thus enabling individual egg chambers to undergo apoptosis and thus control the number of eggs produced in relation to food intake (Terashima, 2005).

The prothoracic glands, which are the principal source of ecdysone in the immature stages, are no longer present in adults. The egg chambers produce ecdysone, which, at least in some insects, accumulates in the oocyte. In the fat body, ecdysone is converted to 20E, the active hormone, and shade, which encodes 20-hydroxylase for converting ecdysone to 20E, is expressed in nurse cells and follicle cells in the ovary and fat body (Terashima, 2005).

Ecdysteroid synthesis is affected by the nutritional status of the female, and ecdysteroids affect oogenesis in many insects. Egg production in mosquitoes is triggered by a blood meal. The digested products of the blood meal stimulate the brain to secrete egg development neurosecretory hormone (EDNH), which is also known as ovarian ecdysteroidogenic hormone (OEH). EDNH stimulates the ovary to synthesize ecdysteroids, which instruct the fat body cells to make vitellogenin for the oocytes. Vitellogenin is critical for egg production, thus without the blood meal there is no vitellogenin and no eggs, so to produce mature eggs ecdysteroids are essential. In contrast, nutritional shortage induces an increase in ecdysteroid concentration in Drosophila females, ecdysteroid concentration increases in Drosophila whole body, haemolymph and ovaries during starvation. Feeding suppresses the high ecdysteroid concentration that is induced by nutritional shortage (Terashima, 2005).

Under starvation, apoptosis of nurse cells in stage-8 and -9 egg chambers is induced; 20E injection into the females under adequate nutrition also induces the apoptosis and JHA treatment of females under nutritional shortage suppresses this apoptosis. Presumably high ecdysteroid concentrations in the haemolymph and/or the ovary, which are induced by starvation, may induce the apoptosis of nurse cells in stage-8 and -9 egg chambers. However, ecdysteroid is indispensable to produce mature eggs in Drosophila. Oogenesis in ecd-1 mutants is arrested at mid-oogenesis, and germline clones of EcR mutations led to developmental arrest; egg chambers degenerated during mid-oogenesis in Drosophila. Presumably, there is an ecdysteroid threshold for inducing apoptosis of nurse cells at stages 8 and 9 and ecdysteroids induce normal development when below the threshold concentration and induce apoptosis of nurse cells at stages 8 and 9 when over the threshold. Starvation induces an increase in ecdysteroid concentration to above the threshold level in the haemolymph and the ovary through activation of the ecdysone synthesis pathway in the egg chamber. Ecdysteroid secretion from the ovary decreased following nutritional shortage. Thus, ecdysteroid secretion from the fat body or other ecdysteroid-synthesizing tissues must be stimulated to induce the high ecdysteroid concentration observed in haemolymph (Terashima, 2005).

JHA suppresses the high ecdysteroid concentration that is induced by starvation. JH and JHA suppress ecdysone synthesis/secretion from the prothoracic glands in larvae of Maduca sexta. It is likely that JHA suppression decreases the high ecdysteroid concentration in the ovary that induces apoptosis of nurse cells in stage-8 and -9 egg chambers under starvation, and therefore JHA treatment retains minimal ecdysteroid levels needed for inducing normal oogenesis (Terashima, 2005).

There is a developmental checkpoint at stage 8 of oogenesis. YP synthesis commences at stage 8 and YP is accumulated during development into mature eggs. Drosophila egg chambers normally transit through stages 8 and 9 during a 6-h period, but starvation induced an accumulation of stage-8 and -9 egg chambers in Drosophila oogenesis. The number of stage-8 egg chambers is increased during a 5-12-h period after starvation starts, but the number of stage-9 egg chambers does not increase for 0-12 h after starvation started. This means that oogenesis progresses from stage 7 to 8, but does not progress from stage 8 to 9 and then to 10 under nutritional shortage. When 20E is injected into the fed flies, the accumulation of stage-8 chambers is not seen; therefore this arrest of oogenesis at stage 8 was not caused by the increasing 20E concentration in haemolymph and ovary. Perhaps starvation signals induce the arrest of oogenesis at stages 8 and 9 directly, or they could inhibit YP uptake. Some nutrient- and stress-response genes exhibit different expression patterns in the ovaries of females under adequate nutrition and starvation. It is suggested that the genes which respond directly to stress and nutrients interact with the ecdysone-synthesis pathway, resulting in the induction of apoptosis of nurse cells in stage-8 and -9 egg chambers through activation of BR-C Z2, Z3 and E75A expression in the follicle cells. Other genes could have altered their expression levels, so as to arrest oogenesis at stages 8 and 9 and to check the developmental status of the egg chamber. As a result, the decision is made to develop into a mature egg or undergo apoptosis at stages 8 and 9. The arrest in the progression of oogenesis at stages 8 and 9 is independent of increasing ecdysteroid levels (Terashima, 2005).

Starvation signals are needed to activate a number of pathways to adjust the rate of egg production in Drosophila. These pathways could be classified into two groups: one to stimulate ecdysone synthesis in the follicle cells and/or nurse cells to activate the apoptosis pathway, including BR-C Z2, Z3 and E75A expression in the follicle cells, and another one to interact with and participate in the developmental checkpoint, giving rise to an arrest in oogenesis at stage 8 under nutritional shortage. A possible scheme is presented for the regulation of oogenesis related to nutrition in Drosophila. It is likely that starvation signals from the gut activate ecdysteroid synthesis in the ovary in Drosophila under starvation. Ecdysteroid is then accumulated in the egg chamber by decreasing 20E secretion from the ovary, and the fat body secretes 20E to haemolymph. It is suggested that there are two thresholds of 20E concentration in Drosophila ovary -- one is the concentration for normal oogenesis and the other is the concentration for inducing apoptosis -- and that starvation elevates the ecdysone levels in some egg chambers over the threshold that leads to apoptosis (Terashima, 2005).

A combinatorial code for pattern formation in Drosophila oogenesis

Two-dimensional patterning of the follicular epithelium in Drosophila oogenesis is required for the formation of three-dimensional eggshell structures. Analysis of a large number of published gene expression patterns in the follicle cells suggests that they follow a simple combinatorial code based on six spatial building blocks and the operations of union, difference, intersection, and addition. The building blocks are related to the distribution of inductive signals, provided by the highly conserved epidermal growth factor receptor and bone morphogenetic protein signaling pathways. The validity of the code is demonstrated by testing it against a set of patterns obtained in a large-scale transcriptional profiling experiment. Using the proposed code, 36 distinct patterns were distinguished for 81 genes expressed in the follicular epithelium, and their joint dynamics were characterize over four stages of oogenesis. The proposed combinatorial framework allows systematic analysis of the diversity and dynamics of two-dimensional transcriptional patterns and guides future studies of gene regulation (Yakoby, 2008b).

Drosophila eggshell is a highly patterned three-dimensional structure that is derived from the follicular epithelium in the developing egg chamber. The dorsal-anterior structures of the eggshell, including the dorsal appendages and operculum, are formed by the region of the follicular epithelium, which is patterned by the highly conserved epidermal growth factor receptor (EGFR) and bone morphogenetic protein (BMP) signaling pathways. The EGFR pathway is activated by Gurken (GRK), a transforming growth factor α-like ligand secreted by the oocyte. The BMP pathway is activated by Decapentaplegic (DPP), a BMP2/4-type ligand secreted by the follicle cells stretched over the nurse cells (Yakoby, 2008b).

Acting through their uniformly expressed receptors, these ligands establish the dorsoventral and anteroposterior gradients of EGFR and DPP signaling and control the expression of multiple genes in the follicular epithelium. Under their action, the expression of a Zn finger transcription factor, Broad (BR), evolves into a pattern with two patches on either side of the dorsal midline. The BR-expressing cells form the roof (upper part) of the dorsal appendages. Adjacent to the BR-expressing cells are two stripes of cells that express rhomboid (rho), a gene that is directly repressed by BR and encodes ligand-processing protease in the EGFR pathway. These cells form the floor (lower part) of the appendages (Yakoby, 2008b).

The patterns of genes expressed during the stages of egg development that correspond to appendage morphogenesis are very diverse. At the same time, inspection of a large number of published patterns suggests that they can be 'constructed' from a small number of building blocks. For instance, the T-shaped pattern of CG3074 is similar to the domain 'missing' in the early pattern of br, while the two patches in the late pattern of br appear to correspond to the two 'holes' in the expression of 18w. Based on a number of similar observations, it was hypothesized that all of the published patterns could be constructed from just six basic shapes, or primitives, which reflect the anatomy of the egg chamber and the spatial structure of the patterning signals (Yakoby, 2008b).

In computer graphics, representation of geometrical objects in terms of a small number of building blocks is known under the name of constructive solid geometry, which provides a way to describe complex shapes in terms of just a few parameters -- the types of the building blocks, such as cylinders, spheres, and cubes, their sizes, and operations, such as difference, union, and intersection. Thus, information about a large number of structures can be stored in a compact form of statements that contain information about the types of the building blocks and the operations from which these structures were assembled. This study describes a similar approach for two-dimensional patterns and demonstrate how it enables the synthesis, comparison, and analysis of gene expression at the tissue scale (Yakoby, 2008b).

The six building blocks used in the annotation system can be related to the structure of the egg chamber and the spatial distribution of the EGFR and DPP signals. The first primitive, M (for 'midline'), is related to the EGFR signal. It reflects high levels of EGFR activation and has a concave boundary, which can be related to the spatial pattern of GRK secretion from the oocyte. The second primitive, denoted by D (for 'dorsal'), reflects the intermediate levels of EGFR signaling during the early phase of EGFR activation by GRK, and is defined as a region of the follicular epithelium that is bounded by a level set (line of constant value) of the dorsoventral (DV) profile of EGFR activation. The boundary of this shape is convex and can be extracted from the experimentally validated computational model of the GRK gradient. The third primitive, denoted by A (for 'anterior'), is an anterior stripe which is obtained from a level set of the early pattern of DPP signaling in the follicular epithelium. This pattern is uniform along the DV axis, as visualized by the spatial pattern of phosphorylated MAD (P-MAD). Thus, the D, M, and A primitives represent the spatial distribution of the inductive signals at the stage of eggshell patterning when the EGFR and DPP pathways act as independent AP and DV gradients (Yakoby, 2008b).

Each of the next two primitives, denoted by R (for 'roof') and F (for 'floor'), is composed of two identical regions, shaped as the respective expression domains of br and rho, and reflect spatial and temporal integration of the EGFR and DPP pathways in later stages of eggshell patterning. The mechanisms responsible for the emergence of the F and R domains are not fully understood. It has been shown that the R domain is established as a result of sequential action of the feedforward and feedback loops within the EGFR and DPP pathways. The formation of the F domain requires the activating EGFR signal and repressive BR signal, expressed in the R domain. Thus, at the current level of understanding, the R and F domains should be viewed as just two of the shapes that are commonly seen in the two-dimensional expression patterns in the follicular epithelium. The sixth primitive, U (for 'uniform'), is spatially uniform and will be used in combination with other primitives to generate more complex patterns (Yakoby, 2008b).

While a number of patterns, such as those of jar and Dad, can be described with just a single primitive, more complex patterns are constructed combinatorially, using the operations of intersection (∩), difference ( ), and union (∪) For example, the dorsal anterior stripe of argos expression is obtained as an intersection of the A and D primitives (A∩D). The ventral pattern of pip is obtained as a difference of the U and D primitives (U D). The pattern of 18w is constructed from the A, D, and R primitives, joined by the operations of union and difference (A∪D R). For a small number of published patterns, the annotations reflect the experimentally demonstrated regulatory connections. For example, the U D annotation for pip reflects that actual repression of pip by the dorsal gradient of EGFR activation. For a majority of genes, the annotations should be viewed as a way to schematically represent a two-dimensional pattern and as a hypothetical description of regulation (Yakoby, 2008b).

The geometric operations of intersection, difference, and union can be implemented by the Boolean operations performed at the regulatory regions of individual genes. Boolean operations evaluate expression at each point and assign a value of 0 (off) or 1 (on). As an example, consider a regulatory module, hypothesized for argos, that performs a logical AND operation on two inputs: the output of the module is 1 only when both inputs are present. When both of the inputs are spatially distributed, the output is nonzero only in those regions of space where both inputs are present, leading to an output that corresponds to the intersection of the two inputs. Similarly, a spatial difference of the two inputs can be realized by a regulatory module that performs the ANDN (ANDNOT) operation. This is the case for pip, repressed by the DV gradient of GRK signaling and activated by a still unknown uniform signal. Finally, a regulatory module that performs an OR operation is nonzero when at least one of the inputs is nonzero. When the inputs are spatially distributed, the output is their spatial union (Yakoby, 2008b).

Boolean operations on primitives lead to patterns with just two levels of expression (the gene is either expressed or not). In addition to Boolean logic, developmental cis-regulatory modules and systems for posttranscriptional control of gene expression can perform analog operations, leading to multiple nonzero levels of output. Consider a module that adds the two binary inputs, shaped as the primitives. The output is nonzero in the domain shaped as the union of the two primitives, but is characterized by two nonzero levels of expression. This type of annotation is reserved only for those cases where the application of Boolean operations would lead to a loss of the spatial structure of the pattern (such as the A + U expression pattern of mia at stage 11 of oogenesis. For example, the union of the A and U primitives is a U primitive, whereas the sum of these primitives is an anterior band superimposed on top of a spatially uniform background (Yakoby, 2008b).

Signaling pathways guide organogenesis through the spatial and temporal control of gene expression. While the identities of genes controlled by any given signal can be identified using a combination of genetic and transcriptional profiling techniques, systematic analysis of the diversity of induced patterns requires a formal approach for pattern quantification, categorization, and comparison. Multiplex detection of gene expression, which has a potential to convert images of the spatial distribution of transcripts into a vector format preferred by a majority of statistical methods, is currently feasible only for a small number of genes and systems with simple anatomies. This paper presents an alternative approach based on the combinatorial construction of patterns from simple building blocks (Yakoby, 2008b).

In general, the building blocks can be identified as shapes that are overrepresented in a large set of experimentally collected gene expression patterns. This approach can be potentially pursued in systems where mechanisms of pattern formation are yet to be explored. At the same time, in well-studied systems, the building blocks can be linked to identified patterning mechanisms. This study chose six primitives based on the features that are commonly observed in real patterns and related to the structure of the tissue as well as the spatial distribution of the inductive signals. A similar approach will be useful whenever a two-dimensional cellular layer is patterned by a small number of signals, when cells can convert smoothly varying signals into spatial patterns with sharp boundaries, and when the regulatory regions of target genes have the ability to combinatorially process the inductive signals. One system in which this approach could be feasible is the wing imaginal disc, which is patterned by the spatially orthogonal wingless and DPP morphogens (Yakoby, 2008b).

The six primitives are sufficient to describe the experimentally observed patterns during stages 10-12 of oogenesis. A natural question is whether it is possible to accomplish this with a smaller number of primitives. Two of the primitives, R and F, could be potentially constructed from the D, M, and A primitives, which are related to the patterns EGFR and DPP activation during the earlier stages of eggshell patterning. Specifically, recent studies of br regulation suggest that the R domain is formed as a difference of the D, A, and M patterns (Yakoby, 2008a). Furthermore, the formation of the F domain requires repressive action in the adjacent R domain. With the R and F domains related to the other four primitives, the size of the spatial alphabet will be reduced even further (from six to four), but at the expense of increasing the complexity of the expressions used to describe various spatial patterns (Yakoby, 2008b).

Previously, the question of the diversity of the spatial patterns has been addressed only in one-dimensional systems. For example, transcriptional responses to the Dorsal morphogen gradient in the early Drosophila embryo give rise to three types of patterns in the form of the dorsal, lateral, and ventral bands. This work provides an attempt to characterize the diversity and dynamics of two-dimensional patterns. Thirty-six qualitatively different patterns were constructed, and it is proposed that each of them can be constructed using a compact combinatorial code. The sizes of the data sets from the literature and from transcriptional profiling experiments are approximately the same (117 and 96 patterns, respectively. Based on this observation, it is expected that discovered patterns will be readily described using this annotation system (Yakoby, 2008b).

A gene expressed in more than one stage of oogenesis is more likely to appear in different patterns, and it was found that groups of genes sharing the same pattern at one time point are more likely to scatter in the future than to stay together. More detailed understanding of the dynamics of the spatial patterns of the EGFR and DPP pathway activation is crucial for explaining these trends and the two observed scenarios for the emergence of complex patterns. A gene that makes its first appearance as a complex pattern, such as the A∩D pattern of argos at stage 10B, can be a direct target of the EGFR and DPP signal integration. In contrast, a gene such as Cct1, which changes from the A to the R pattern, can be a dedicated target of DPP signaling alone, and changes as a consequence of change in the spatial pattern of DPP signaling. Future tests of such hypotheses require analysis of cis-regulatory modules responsible for gene regulation in the follicular epithelium. While only a few enhancers have been identified at this time, this categorization of patterns should accelerate the identification of enhancers for a large number of genes (Yakoby, 2008b).

Proposed for the spatial patterns of transcripts, these annotations can also describe patterns of protein expression, modification, and subcellular localization. For example, the stage 10A patterns of MAD phosphorylation and Capicua nuclear localization can be accurately described using the A and U D annotations, respectively. The ultimate challenge is to use the information about the patterning of the follicular epithelium to explore how it is transformed into the three-dimensional eggshell. A number of genes in the assembled database encode cytoskeleton and cell adhesion molecules, suggesting that they provide a link between patterning and morphogenesis. It is hypothesized that the highly correlated expression patterns of these genes give rise to the spatial patterns of force generation and mechanical properties of cells that eventually transform the follicular epithelium into a three-dimensional eggshell (Yakoby, 2008b).


broad: Biological Overview | Evolutionary Homologs | Regulation | Targets of Activity | Effects of Mutation | References

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.