BIOLOGICAL OVERVIEW

Although Adh transcription factor 1 (Adf-1) was first identified as a factor that bound the distal promoter of the gene for alcohol dehydrogenase (Heberlein, 1985), Adf-1 recognizes specific sites in a diverse group of promoters, including antennapedia P1 and dopa decarboxylase (England, 1990). Its ubiquitous expression (England, 1992) and its requirement for viability establish the important role of Adf-1 in Drosophila biology. The sequence of Adf-1 reveals that its presumptive DNA-binding is a distantly related member of the Myb helix-turn-helix family (England, 1992; Lipsick, 1996), whereas Adf-1 sequence reveals no similarities to known transcriptional activation domains. This suggested that Adf-1 may function through a novel type of transactivation domain (Cutler, 1998 and references therein). nalyot (nal) is a novel olfactory memory mutant, isolated as a P-element in the Adf1 gene. Following extended training sessions, Adf1 mutants show normal early memory but defective longterm memory. Adf1 shows widespread spatiotemporal expression, yet mutant alleles reveal no discernible disruptions in gross morphology of the nervous system. Nevertheless, studies at the larval neuromuscular junction reveal a role for Adf1 in the modulation of synaptic growth. It is thought that Adf1 may play an essential role in terminal stages of neuronal differentiation and function (DeZazzo, 2000).

Multiple lines of convergent evidence conclusively link the hypomorphic nal^P1 olfactory memory defect to the Adf1 gene: (1) the nal P element is inserted in a major intron of the Adf1 transcription unit, and its precise excision reverts the memory defect; (2) the nal^P1 memory defect is not complemented by the (homozygous-lethal) imprecise excision allele nal^le60, which carries a small deletion that removes only Adf1 coding sequences; (3) the nal^P1 mutation reduces levels of expression of Adf1 RNA and protein, but does not affect RNA levels of the neighboring transcription unit cn20; (4) three independent transgenic lines, each driving leaky overexpression of Adf1, show an adult memory deficit; (5) the nal^P1 memory defect is rescued with an hsp-Adf1⁺ transgene expressing the wild-type Adf1 protein (DeZazzo, 2000).

Transgenic overexpression experiments indicate that optimal memory formation requires tight developmental control of Adf1 expression. Ectopic expression of Adf1 using the UAS-GAL4 system results in lethality from late embryo to late pupa. Temperature-shift experiments of hsp-Adf1⁺ reveal a deleterious period of overexpression that appears confined to development. When raised at 25°C, olfactory memory in +; hsp-Adf1⁺ lines is lower than normal. Thus, perturbing Adf1 regulation temporally or spatially can disrupt memory formation and viability. These results are perhaps not surprising, given that Adf1 encodes a transcription factor, which likely impinges upon many downstream targets. In fact, these findings reinforce regulatory themes emerging from behavioral studies of other transcription factors. Transgenic mice expressing two additional copies of the clock transcription factor, for instance, show defective circadian function. More to the point, acute heat shock of a stonewall transgene (Clark, 1996), which also encodes a myb-related transcription factor, leads to death (DeZazzo, 2000 and references therein).

Several lines of investigation suggest a complex, multifunctional role for Adf1 in the nervous system. (1) Adf1 shows widespread expression throughout development. However, null mutants survive to the late embryonic or early larval stage, and although they exhibit a defect in synaptic structure at the neuromuscular junction, they fail to reveal any discernible disruptions of gross morphology in the nervous system. Similarly, adult-viable (hypomorphic) Adf1 mutants show normal neuropilar volume of adult mushroom bodies and the central complex, neuroanatomical centers implicated in adult olfactory memory. Nevertheless, overexpression of Adf1⁺ kills the animals, even when transgenic expression is restricted to the nervous system. Adf1 therefore may play an essential role in terminal stages of neuronal differentiation and function. (2) Leaky overexpression of hsp-Adf1⁺ during development leads to adult memory defects. Early memory is reduced in all three +; hsp-Adf1⁺ lines when they are raised at 25°C, but not at 18°C. Concomitantly, leaky hsp-Adf1⁺ expression is much greater at 25°C than at 18°C. Thus, normal adult memory formation may require tight developmental regulation of Adf1. (3) Chronic expression of Adf1⁺ in adults is sufficient to rescue the nal memory defect. Rescue in nal^P1; hsp-Adf1⁺-11 mutant transgenics was complete when these animals were raised during development at 18°C and then shifted as adults to 25°C for 3 days. (4) Genetic mutations that increase or decrease levels of Adf1 protein at the NMJ have reciprocal effects on bouton number. Once again, leaky transgene expression is capable of rescuing the defect in bouton number observed in nal^P1 mutants (DeZazzo, 2000).

The complex etiology of nal may not be unusual. The Scab/Volado locus, for example, yields at least two transcripts, Vol-l and Vol-s, each encoding the same protein. Acute induction of Vol-s (which is expressed in both head and body tissues) is sufficient to rescue fully the early memory defect of a Vol-s null mutant. The Vol-s mutant nevertheless shows normal levels of the head-specific Vol-l transcript. However, a developmental role for Vol in adult behavior cannot be excluded, since Vol-l may subserve this role. Thus, inducible transgene experiments to date have addressed whether acute expression of a transgene in adults is necessary for rescue of a memory defect, but they have not yet resolved whether such expression is sufficient (DeZazzo, 2000).

Formation of functional synapses at the larval NMJ reflects an ongoing activity-dependent process that begins during late embryogenesis and continues throughout larval development. Initial events in synapse formation, however, do not require neuronal activity. In contrast, the subsequent maturation of synaptic branches and boutons during larval development clearly is activity dependent and can be modulated by changes in neuronal excitability, in cAMP signaling, or in expression of cell adhesion molecules (DeZazzo, 2000 and references therein).

Developmental plasticity at the NMJ has been genetically dissected. Neuronal activity triggers a cAMP signaling cascade, leading to an increase in both synaptic bouton number (structure) and the quantal content of synaptic transmission (function). The structural and functional pathways appear to be independent of each other, involving the cell adhesion molecule fasciclin II (Fas II) and CREB transcription factor, respectively. Genetic mutations that decrease Fas II expression to about 50% wild-type levels increase bouton number without causing a corresponding change in synaptic strength. Similarly, overexpression of a CREB repressor (dCREB2-b) in a dunce mutant background (which constitutively elevates cAMP levels and increases the number of synaptic boutons) blocks the mutant increase in synaptic strength but not the mutant increase in synaptic structure. Conversely, overexpression of a CREB activator (dCREB2-a) has no effect at a wild-type (normal) NMJ but increases synaptic function at a mutant Fas II NMJ (DeZazzo, 2000 and references therein).

Several genes have now been implicated in growth of the Drosophila neuromuscular synapse. Fas II is expressed pre- and post-synaptically at the NMJ, where it appears to act as a cell adhesion molecule in a manner analogous to its vertebrate NCAM homolog. More recent discoveries, made with the use of genetic screens, include the identification of Highwire (Wan, 2000), a novel gene with expression localized to periactive zones of presynaptic terminals, and Futsch, encoding a MAP1B-like protein associated with the axonal, dendritic, and nerve-terminal cytoskeleton. Likewise Adf1 is involved in the structural pathway. Genetic manipulations that decrease (or increase) the amount of Adf1 give rise to a decrease (or increase) in synaptic bouton number. Thus, whereas dCREB2 affects NMJ function but not structure, Adf1 affects structure but not function. As a transcription factor, Adf1 appears not to function directly at the synapse; rather, its role reasonably may act upstream of structural effectors. Together, these observations suggest that transcriptional regulation is involved with both components of synapse maturation, and they argue that an increase in synaptic structure must precede an increase in synaptic function (DeZazzo, 2000 and references therein).

Developmental plasticity in the adult brain may share some of the cellular machinery that subserves synapse maturation at the NMJ. During metamorphosis, for instance, axonal projections from some larval mushroom body neurons first degenerate and then extend processes anew along with new MB neurons that proliferate in developing adult structures. This process of synapse formation continues for a few days after eclosion, is modulated in an experience-dependent fashion, and is aberrant in mutants with defects in cAMP signaling or in normal larvae grown in low density cultures. Positing a role for Adf1 in activity-dependent synapse formation of the maturing adult brain can explain why hypomorphic mutations produce mild defects in olfactory memory measured immediately after one training session. Several observations are consistent with this notion: (1) leaky overexpression of hs-Adf1 during development disrupts memory; (2) overexpressing the Adf1 protein in the nervous system via the UAS-GAL4 system has lethal consequences; (3) Adf1 mutants have a defect in synaptic structure at the neuromuscular junction. These observations suggest that subtle structural defects at central synapses at least partly underlie the nal memory defect (DeZazzo, 2000).

Nevertheless, the nal memory defect is not accompanied by any overt morphological changes in the embryonic or adult nervous system. In fact, normal adult morphology of the mushroom bodies and central complex is seen even in severe allelic combinations (nal^P1/nal^le60) predicted to lower Adf1 protein levels to less than 25% of wild-type. These observations suggest that defects in synaptic structure occur at very terminal stages of synaptic maturation after the growth of axons and dendrites, which represent the main contributions to planimetric measurements of neuropil. In this context, behavioral analyses reveal that assays of adult olfactory memory provides a sensitive way to detect subtle developmental abnormalities that escape morphological detection (DeZazzo, 2000).

Existence of a protein synthesis-dependent LTM appears ubiquitous in the animal kingdom and has been shown to be CREB-dependent in mammals. Long-term memory formation after Pavlovian training in Drosophila also depends on CREB-dependent gene transcription and protein synthesis. Moreover, opposite manipulations of CREB have corresponding loss- and gain-of-function effects on long-term memory formation, implying that CREB acts as a molecular switch for LTM formation -- as it does for activity dependent plasticity at the NMJ. Manipulations of CREB in cultured molluscan neurons have demonstrated activity-induced structural changes at identified sensorimotor synapses concomitant with the appearance of long-term facilitation, a cellular correlate of behavioral sensitization. Hence, the appearance of long-term memory generally may include structural, as well as functional, changes at the relevant synapses (DeZazzo, 2000 and references therein).

Long-term memory formation after spaced training is abolished in nal mutants. A trivial explanation for this result is that it derives secondarily from the milder deficit in early memory. Two observations argue against this interpretation: (1) 1 day memory after massed training is normal in nal mutants, indicating that anesthesia-resistant memory (ARM) is formed normally and, thereby, suggesting that multiple training sessions compensate for the mild memory deficit observed after one training session; (2) radish mutants have a more severe early memory deficit than nal mutants (and ARM is abolished) but nevertheless show normal LTM formation. Thus, the level of performance at earlier memory phases is not a reliable predictor of performance at later memory stages. These observations also do not readily support a general developmental etiology of the nal LTM deficit, unless early memory and LTM are anatomically distinct (DeZazzo, 2000).

Instead, observations at the NMJ suggest a possible structural role for Adf1 in adult behavioral plasticity. During the formation of LTM, the Adf1 transcriptional cascade may lead to an increase in the number of synaptic boutons, allowing the incorporation of synaptic machinery induced through the CREB transcription cascade. In nal mutants, structural changes do not occur, thereby preventing integration of the CREB-dependent increases in synaptic function, as is the case at the NMJ in the absence of Fas II- or dunce-dependent increases in synaptic boutons. Without an increase in structure and function, there is no increase in synaptic strength, and LTM does not manifest (DeZazzo, 2000).

Nevertheless, other possibilities have not yet been ruled out. As a transcription factor, for example, Adf1 may regulate the expression of functional components of neuronal activity, which are compensated for in the homeostatic context of the NMJ, but which are not compensated for in the adult CNS. Future experiments with spatially and temporally restricted transgenes likely will provide the necessary data to distinguish these possible roles for Adf1 in LTM formation (DeZazzo, 2000).

Adf1 is a complex gene. It is essential during early development and also contributes to developmental and behavioral plasticity. While these functions may seem to be disparate, a closer examination suggests a common mechanistic thread: the essential requirement for Adf1 during development does not appear to be at the level of proliferation, but instead appears to reflect a more specialized postmitotic role. In Adf1 null mutants, gross morphology is normal, the majority of embryos develop to maturity, and a significant fraction hatch. These observations suggest that Adf1 may guide the terminal stages of cellular differentiation and maintenance, as is the case for other Myb family members. In neurons, such terminal stages likely include synapse formation and refinement. In larvae, this process occurs chronically in the activity-dependent growth of neuromuscular synapses. In adults, this process may occur acutely, and in a much more spatially restricted pattern, during the formation of specific long-term memories (DeZazzo, 2000).

These various spatiotemporal roles for Adf1 support the growing notion that developmental plasticity and adult behavioral plasticity reflect similar cellular processes. The genetic approach strengthens this biological insight in two novel ways. (1) The original nal mutation was identified in a 'forward-genetic' screen for defective adult plasticity. In this manner, no assumptions were made about any a priori connections between development and adult function. (2) 'Reverse-genetic' manipulations of Adf1 yield similar effects on synapse formation at the NMJ and on adult memory, thereby suggesting a mechanistic link between these temporally (and spatially) disparate processes (DeZazzo, 2000).

What are the downstream targets of Adf1, given Adf1's role as a transcription factor? While a number of candidate targets, including Adh and Dopa decarboxylase (Ddc), have been identified in vitro, the significance of these observations in vivo is unclear. More to the point, synaptic proteins under Adf1 regulation have not yet been identified. In this regard, Fas II, Highwire, and Futsch are obvious candidates, given their established role in synaptic growth at the NMJ. The identification of Adf1 effectors promises to resolve the issue of whether Adf1 modulates synaptic morphology directly during LTM formation (DeZazzo, 2000).

GENE STRUCTURE

Mapping of more than ten Adf1 cDNAs onto the genomic region reveals a single 3' end and three introns. The distal two exons appear to be constitutively spliced from all transcripts. In contrast, a 114 bp intron, located upstream of the coding region, is sometimes retained in the mature transcript. This 5' heterogeneity potentially can give rise to two different amino-terminal sequences. The mRNA retaining the intron encodes a 253 amino acid protein, whereas removal of the intron introduces a new methionine (i2) in frame with the old one (i1), resulting in the potential addition of a nine amino acid stretch, MHTLTAAIG. The significance of this potential heterogeneity is unclear (DeZazzo, 2000). Engineered removal of 12 amino acids from the amino terminus of the smaller form abolishes DNA binding and transactivation activities in Schneider cell transfections (Cutler, 1998); the longer version described here appears to have normal DNA binding and transactivation activities (Cutler, personal communication to DeZazzo, 2000).

PROTEIN STRUCTURE

Amino Acids - 253 and 265

Structural Domains

Adf-1 is an essential Drosophila melanogaster sequence-specific transactivator that binds the promoters of a diverse group of genes. A comprehensive mapping of the functional domains of Adf-1 has been carried out to study the role of transactivators in the process of gene activation. Using a series of clustered point mutations and small deletions, regions of Adf-1 required for DNA binding, dimerization, and activation have been identified. In contrast to most enhancer-binding factors, the Adf-1 activation regions are nonmodular and depend for activity on an intact protein, including the Adf-1 DNA-binding domain. Like many transcriptional activators, Adf-1 contains a TFIID-binding domain that can interact with specific TAF subunits. Although TAFs are required for Adf-1-directed activation, TAF binding is not sufficient, suggesting that Adf-1 may direct multiple essential steps during activation. Interestingly, both the TAF-binding domain and the DNA-binding domain contain sequences homologous to those of the Myb family of DNA-binding domains. Thus, Adf-1 has evolved an unusual structure containing two versions of the Myb motif, one that binds DNA and one that binds proteins (Cutler, 1998).

A BLAST search for sequences similar to that of the TAF-binding domain of Adf-1 has revealed a sequence similarity with a region of the S. cerevisiae transcriptional adapter protein ADA2. In fact, the number of residues at the carboxy terminus of Adf-1 (AdfC) that are identical or homologous to those of the yeast ADA2 is almost as high as that between the human and yeast ADA2 proteins over this region. The region of highest homology is in the region between amino acids 223 and 247 of Adf-1. The corresponding region of ADA2 has been shown to be essential for its biological function and is required for binding another transcriptional adapter protein, GCN5. Visual examination of the sequence of Adf-1 also suggests that this same region has homology with the Adf-1 amino-terminal Myb domain (AdfN). Interestingly, it has been reported that the corresponding region of ADA2 also contains a Myb domain homology. ADA2 is more similar to Adf-1 than to Myb (52% similarity between Adf-1 and yADA2 versus 43% similarity between mouse Myb repeat 2 and yADA2), suggesting that Adf-1 and ADA2 may contain domains belonging to a novel subfamily of Myb sequences (Cutler, 1998).

Unlike the AdfC and ADA2 sequences, previously characterized Myb domains have only been shown to bind DNA. The high-resolution structures of several Myb domains have been solved, revealing that they are variants of the helix-turn-helix motif. An alignment was performed of AdfC to yADA2 and two DNA-binding Myb domains (AdfN and the second Myb repeat of the mouse Myb protein). The level of similarity between AdfC and Myb is similar to that between Myb and the two other sequences which are already recognized as being Myb homologs, AdfN and yADA2. Interestingly, the region of greatest similarity between AdfC and yADA2 is also the region of least similarity between those two proteins and Myb and AdfN. This region corresponds to Myb helix 3, the DNA-recognition helix of DNA-binding Myb domains. The mouse Myb structure reveals that seven residues in helix 3, including those that specifically recognize DNA, project out from the protein. While all seven of these residues are hydrophilic in the mouse Myb sequence and six are hydrophilic in the Adf-1 DBD, only three are hydrophilic in the Adf-1 carboxy-terminal region. Thus, this region of Adf-1, which may correspond to the DNA-recognition surface of typical Myb domains, may be much more hydrophobic than that of any of the DBDs. Fewer hydrophobic substitutions are seen in this region with ADA2, but there are several large hydrophobic residues in both ADA2 and AdfC immediately carboxy terminal to the helix 3 region that may further contribute to a potential helical hydrophobic surface. This structural arrangement of the Adf-1 carboxy-terminal Myb-like domain is consistent with a role for this domain in protein-protein rather than protein-DNA interactions (Cutler, 1998).

date revised: 2 September 2022

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.