InteractiveFly: GeneBrief

hamlet : Biological Overview | Developmental Biology | Effects of Mutation | Evolutionary Homologs | References

BIOLOGICAL OVERVIEW

The dendritic morphology of neurons determines the number and type of inputs they receive. In the Drosophila peripheral nervous system (PNS), the external sensory (ES) neurons have a single nonbranched dendrite, whereas the lineally related multidendritic (MD) neurons have extensively branched dendritic arbors. hamlet, coding for a zinc finger transcription factor, is a binary genetic switch between these contrasting morphological types. In hamlet mutants, ES neurons are converted to an MD fate, whereas ectopic hamlet expression in MD precursors results in transformation of MD neurons into ES neurons. Moreover, hamlet expression induced in MD neurons undergoing dendrite outgrowth drastically reduces arbor branching (Moore, 2002).

In Drosophila embryos, a single external sensory organ precursor (ESOP) cell gives rise to an ES neuron and an MD neuron through a stereotypical series of asymmetric cell divisions. The IIA daughter of the ESOP divides once to produce a trichogen and a tormagen, the two external support cells of the ES neuron. The IIB daughter of the ESOP generates an MD neuron and the IIIB cell, which divides to form an ES neuron and a glia (Moore, 2002).

In a genetic screen designed to identify mutants in Drosophila embryonic dendrite development, a larval lethal mutant was isolated that appeared to affect determination of cells descended from the IIB precursor of the ESOP lineage. This mutation was named hamlet (ham) after the 'To be or not to be' soliloquy in the Shakespeare play of the same name (Moore, 2002).

In embryos homozygous for the ham¹ mutation, supernumerary MD neurons are evident in each PNS cluster of the embryo, as well as a concomitant decrease in the number of ES neurons. For example, in the wild-type dorsal PNS cluster there are 13 neurons: five ES neurons, which express the pan-neural marker ELAV (Embryonic lethal, abnormal vision), and eight MD neurons, which express ELAV and the enhancer-trap E7-2-36, a pan-MD marker. In ham¹ mutants, however, the number of MD neurons is increased up to 13 and the number of ES neurons decreases (Moore, 2002).

Do these extra MD neurons arise at the expense of their sibling ES neurons? To answer this question, focus was placed on the ventral pore sensory organs (vp1 to vp4a) because their lineage is fully described and each organ develops clearly spaced from those surrounding it. These organs were labeled with an antibody to Cut, which is expressed in all cells of the lineage. In the five differentiated cells of the lineage, Cut is expressed at low levels in the ES and MD neurons and at a much higher level in the trichogen, tormagen, and glia. In addition, labeling was performed with antibodies to the MD marker E7-2-36 and Prospero (Pros), which is expressed in the IIB cell and its descendants and remains expressed in the differentiated glia. In ham¹ mutants the vp ESOP lineage division pattern appears normal and produces five cells; however, in the differentiated organ of these mutants, the ES neuron and glia are lost. A second MD neuron appears in the position normally occupied by the ES neuron, and a third external cell (trichogen) in the position normally occupied by the glia. Taken together, these observations indicate that the daughter cells of the IIIB cell are transformed to an MD neuron and a trichogen in ham¹ mutants (Moore, 2002).

To test whether the supernumerary MD neurons can extend the characteristic complex branched dendritic arbor, ham¹/ham¹ MARCM (mosaic analysis with a repressible cell marker) clones were generated in the ESOP lineage. In this analysis, clones of positively marked neurons are derived from a single mitotic recombination event within the ESOP lineage. In wild-type clones of all PNS clusters, MD neurons either label alone, representing a clone derived from recombination within the IIB cell, or in association with an ES neuron, representing recombination at the level of the ESOP cell or its precursors. In ham¹ homozygous clones of all PNS clusters, either one or two MD neurons are labeled. In the latter case, one of these neurons must represent the transformed ES cell. Both of these neurons have the arbors characteristic of MD neurons specific to the location at which they arise. Thus, in ham¹ the MD neuron transformed from an ES neuron has a full MD arbor morphology. Additionally, this analysis also shows that ham functions in a cell-autonomous manner within the ESOP lineage (Moore, 2002).

To determine whether ectopic Ham expression can alter cell fate, the full-length ham cDNA was cloned into the pUAST vector and UAS-ham transgenic flies were created. Ectopic Ham expression in the IIB precursor of the MD neuron using pros-gal4 resulted in loss of labeling for the E7-2-36 MD marker in all embryonic PNS neurons, indicating an MD-to-ES marker transformation. To test whether the dendrite morphology of these neurons also reflects an MD-to-ES transformation, elav-gal4 driving UAS-gfp was used to visualize dendrites. In the wild-type dorsal cluster of stage 17 embryos, there are several neurons with multiply branched dendrites (i.e., MDs) and two dorsal ES neurons, each with a single dorsally projecting dendrite. In embryos where both pros-gal4 and elav-gal4 drive UAS-ham and UAS-mCD8-gfp in parallel, the multiple dendritic arbor of the MD neurons has clearly been transformed; the dendrites of all neurons in the cluster are unbranched and project dorsally (Moore, 2002).

Given that Ham is expressed in the postmitotic ES neuron during dendrite extension, whether postmitotic expression of Ham in an MD neuron can alter its dendrite morphology was investigated. elav-gal4 was used to drive UAS-ham and UAS-mCD8-gfp in parallel in the embryo. Indeed, postmitotic expression of Ham in the MD neurons drastically reduces dendritic branching, leading to arbors with a structure intermediate between that of MD and ES neurons. In addition, these neurons still express the MD marker E7-2-36 at high levels, indicating that these neurons have transformed dendrite morphology but not cell fate. To investigate this effect further, the MD-specific driver 109(2)80-gal4 was used to drive UAS-gfp and UAS-ham. In these embryos 109(2)80-gal4 remains active, implying that the neurons in which it is expressed remain MD; however, the branching of dendrites in these MD neurons is clearly reduced. These two lines of evidence illustrate that postmitotic expression of Ham in MD neurons does not switch the fate of these neurons to ES but does still act to suppress the formation of complex dendritic arbors (Moore, 2002).

The PR domain and bipartite multiple ZF structure define a small family of proteins, of which Ham is the sole member described in Drosophila . The sequence of the PR domain and ZFs as well as the overall domain structure are conserved between Ham and two other proteins, human MDS1/EVI1 (myelodysplasia syndrome 1/ectopic viral integration 1) and Caenorhabditis elegans Egl43 (egg laying defective 43). Both are implicated in neural development. MDS1/EVI1 was originally isolated because ectopic expression of EVI1 alone (leading to a protein still containing the ZFs but lacking the PR domain) can cause leukemia. A partial disruption of the Mds1/Evi1 locus in mouse, however, leads to mid-gestation lethality. Among multiple defects in these embryos are regions of hypocellularity in the neuroectoderm and a failure of peripheral nerve formation. Egl43 is required for hermaphrodite-specific neuron migration and phagemid sensory neuron development. The defect in phagemid neuron development is intriguing because these neurons fail to fill with dye through normally exposed sensilla, a phenotype that could be attributable to a failure of correct dendrite formation. Thus, it will be interesting to investigate whether MDS1/EVI1 or Egl43, like ham, have a role in dendrite specification (Moore, 2002).

As far as is known, ham represents the only binary genetic switch identified that acts to repress a multiple dendritic arbor and promote single-dendrite morphology. It is expected that Ham is a transcription factor; it has a nuclear subcellular distribution, and MDS1/EVI1 can bind DNA and regulate transcription. Ham must exert its effect in a short developmental window, because its expression is initiated in the IIIB neural precursor and continues only during the initial stages of ES neuron differentiation. Given this short period of activity, it is likely that Ham induces a cascade of events. Its transcriptional targets are likely to include key players in the genetic determination of dendrite morphology that act to repress dendritic branching and promote single- over multiple-dendrite morphology (Moore, 2002).

Chromatin modification of Notch targets in olfactory receptor neuron diversification

Neuronal-class diversification is central during neurogenesis. This requirement is exemplified in the olfactory system, which utilizes a large array of olfactory receptor neuron (ORN) classes. An epigenetic mechanism was discovered in which neuron diversity is maximized via locus-specific chromatin modifications that generate context-dependent responses from a single, generally used intracellular signal. Each ORN in Drosophila acquires one of three basic identities defined by the compound outcome of three iterated Notch signaling events during neurogenesis. Hamlet, the Drosophila Evi1 and Prdm16 proto-oncogene homolog, modifies cellular responses to these iteratively used Notch signals in a context-dependent manner, and controls odorant receptor gene choice and ORN axon targeting specificity. In nascent ORNs, Hamlet erases the Notch state inherited from the parental cell, enabling a modified response in a subsequent round of Notch signaling. Hamlet directs locus-specific modifications of histone methylation and histone density and controls accessibility of the DNA-binding protein Suppressor of Hairless at the Notch target promoter (Endo, 2011).

This study analyzed the ORN lineage history that gives rise to three primary ORN identities (Naa, Nab and Nba). These three identities arise in a sensillum via iterated rounds of Notch-mediated binary cell-fate decisions. Together with previous findings, these results suggest that diversification of Drosophila ORN classes is the result of the combined output of two predominantly hardwired mechanisms; spatially localized factors determine at least 21 types of sensilla, and Notch and Ham then act in each sensillum to maximize ORN class variety (Endo, 2011).

Biochemical and molecular analyses of Ham function indicate that it can repress Notch target enhancers. In the ORN lineage, Notch signaling is used in consecutive cell fate decisions, and it was found that Ham acts to turn off Notch targets before a subsequent a round of selective reactivation. Ham is expressed specifically in pNa, the neuronal-intermediate precursor with high Notch activity, and inherited by both of the pNa daughter cells. In addition to the current findings, studies in other contexts have observed that some Notch targets require a Notch signal for their transcriptional induction, but not for maintaining their expression. These Notch targets could aberrantly persist in both pNa progeny without the intervention of a mechanism to erase the effects of the preceding Notch signal (Endo, 2011).

ham mutants showed an unusual ORN fate switch. They not only transformed ORN fate with respect to Notch state, but also altered sublineage-specific identity (low-Notch Nab to high-Notch Nba identity). This phenotype suggests that, in addition to suppressing the previous round of Notch activation, Ham may delineate the selection of the next round of targets. As Ham activity resulted in altered chromatin modifications at Notch targets, this suggests that Ham could set an epigenetic context in which the terminal round of Notch signaling occurs (Endo, 2011).

Although this was demonstrated with respect to Ham, it is suggested that this approach to modifying the transcriptional outputs of a signaling pathway may have widespread importance in other lineages that utilize iterative signals. Notch signaling iteration is a widespread phenomenon. One important example is in the maintenance of neural and other stem cells, and it is now known that some chromatin-modifying factors promote stem cell self-renewal. Notably, several Prdm factors have regionalized expression in neural precursor domains of the embryonic mouse spinal cord and could modify and diversify stem cell identity during mammalian CNS development (Endo, 2011).

In Drosophila, Notch signaling and Ham expression are transient in nascent ORNs. Thus, Notch- and Ham-mediated fate choices must be perpetuated during the later selection of alternative axon guidance factors and odorant receptors. It is possible that chromatin methylation not only sets the context of immediate Notch signaling outcomes, but also maintains initial fate choice by priming or silencing promoters for readout during differentiation. The existence of such mechanisms in neural development is now beginning to emerge. It was recently shown that, in mouse cortical precursor cells, the trithorax factor Mixed-lineage leukemia1 (Mll1) prevents epigenetic silencing of the neural differentiation gene Distal-less homeobox 2 (Dlx2), enabling it to be properly upregulated during differentiation stages. In contrast, in the mammalian olfactory system, epigenetic repression is used during the transition from multipotent precursor to immature ORN to silence all ~2,800 odorant receptor genes before subsequent de-repression of a single odorant receptor per neuron (Endo, 2011).

To determine how chromatin modifications create a context-dependent outcome from signaling and how resultant cell-fate choices are perpetuated during Drosophila ORN differentiation, it will be necessary to elucidate the components and action of the chromatin-modification complex targeted by Ham. Furthermore, genome-wide identification of the promoters targeted by Su(H) and Ham will reveal the genes regulated by these factors to confer specific ORN fates (Endo, 2011).

SWI/SNF complex prevents lineage reversion and induces temporal patterning in neural stem cells

Members of the SWI/SNF chromatin-remodeling complex are among the most frequently mutated genes in human cancer, but how they suppress tumorigenesis is currently unclear. This study used Drosophila neuroblasts to demonstrate that the SWI/SNF component Osa (ARID1) prevents tumorigenesis by ensuring correct lineage progression in stem cell lineages. Osa induces a transcriptional program in the transit-amplifying population that initiates temporal patterning, limits self-renewal, and prevents dedifferentiation. The Prdm protein Hamlet was identified as a key component of this program. Hamlet is directly induced by Osa and regulates the progression of progenitors through distinct transcriptional states to limit the number of transit-amplifying divisions. These data provide a mechanistic explanation for the widespread tumor suppressor activity of SWI/SNF. Because the Hamlet homologs Evi1 and Prdm16 are frequently mutated in cancer, this mechanism could well be conserved in human stem cell lineages (Eroglu, 2014).

The data reveal an essential function for the chromatin-remodeling SWI/SNF complex in ensuring lineage progression in stem cell lineages. When neural stem cells/NBs progress toward the transit-amplifying intermediate neural progenitor (TA/INP) fate, the SWI/SNF complex activates a transcriptional program that limits self-renewal and initiates a temporal TF cascade to confer temporal identity. Failure to do so results in lineage reversion and tumor formation. The temporal TF Dichaete (D) and the Prdm protein Ham as direct SWI/SNF targets and show that induction of Ham limits the number of TA divisions by ensuring the progression of temporal patterning. Members of the SWI/SNF complex, particularly the Osa homologs ARID1A and ARID1B, are among the most frequently mutated genes in human cancer, and the findings provide a potential mechanism for their tumor-suppressing activity (Eroglu, 2014).

A model is proposed where two distinct transcriptional programs act in concert to ensure directionality in Drosophila neural stem cell lineages. In type II NBs, a 'self-renewal' program comprising the TFs Dpn, Klu, and HLHmγ allows long-term self-renewal. Upon asymmetric division, Numb and Brat terminate this program in one of the two daughter cells, which therefore progresses toward the imINP stage. As INPs undergo maturation, Brat and Numb disappear, allowing the program to reinitiate and self-renewal to resume. The data indicate that Osa activates a second 'self-renewal restriction' program before this reinitiation occurs to ensure that INPs, unlike NBs, differentiate after around five rounds of asymmetric cell division. In osa mutants, the restriction program is not activated. The self-renewal program, however, is unaffected, and therefore, INPs regain NB-like properties resulting in unlimited self-renewal and brain tumor formation (Eroglu, 2014).

Why does Osa not activate the self-renewal restriction program in NBs? In mammalian neural stem cells, a subunit switch in the SWI/SNF complex is thought to trigger the switch from self-renewal to differentiation, but this study failed to detect a similar switch in the Drosophila larval brain. More likely, Dpn, Klu, and HLHmγ prevent Osa binding in NBs, for example by competing with SWI/SNF for binding sites. In fact, all three factors can act as transcriptional repressors, and opa (one of the SWI/SNF targets identified in this study) is actually also a direct Dpn target in the embryonic CNS (Eroglu, 2014).

The results suggest a tight functional connection between the SWI/SNF complex and the temporal TF cascade that confers temporal identity to INPs. SWI/SNF directly induces transcription of D, the first member of this cascade. In addition, it induces Ham, a chromatin regulator that can limit self-renewal capacity in INPs but also when ectopically expressed in NBs. In INPs, Ham is specifically required for the transition from Grh+, Ey+ middle-aged INPs to Grh-, Ey+ old INPs. Because transition to the terminal transcriptional state is important for timely cell-cycle exit in mINPs (Bayraktar and Doe, 2013), this explains the overproliferation phenotype observed in ham mutants (Eroglu, 2014).

How could Ham mediate temporal progression of INPs? It has been previously shown that recruitment of the earliest component of the NB 'transcriptional clock' to the nuclear periphery permanently silences its expression and limits NB competence. Evi1 and Prdm16, the mammalian homologs of Ham, have been postulated to initiate heterochromatin formation by methylating H3K9. Because H3K9 methylation is crucial for recruiting gene loci to the nuclear periphery, it is interesting to speculate that Ham acts in INPs by driving the transition to the next transcriptional state and, ultimately, to differentiation (Eroglu, 2014).

Mutations in the mammalian SWI/SNF complex subunits are potential drivers of tumorigenesis in a wide variety of tissues including the brain. The Brm homolog SMARCA4 and the Osa homologs ARID1A and ARID1B are among the chromatin modifiers that are recurrently mutated in medulloblastoma, the most common malignant childhood brain tumor. Identifying the cell of origin in brain tumors is crucial in designing effective therapeutic strategies. Stem cells could acquire oncogenic mutations that initiate tumor formation. On the other hand, tumors could also originate from more restricted progenitors that inherit these mutations and become malignant. This study offers an alternative explanation: provided that the function of SWI/SNF is conserved in humans, mutations occurring in restricted progenitors could affect lineage progression causing progenitors to revert into stem cells. In this case, the cell of origin would be a progenitor despite the fact that tumors are made up of stem cells. In fact, this possibility has been proposed for other tumor suppressors and could be tested rigorously for SWI/SNF mutations given the recent significant advances in cell lineage tracing in tumors (Eroglu, 2014).

DEVELOPMENTAL BIOLOGY

Embryonic

ham is expressed in a ventral patch in the cephalic region of the embryo from stage 5 and continues to be expressed in the cephalic region through stage 15. It is expressed in the developing PNS from stage 11 through 15 and shows transient expression in each PNS cluster. The Ham protein has the same expression profile as the mRNA and has a nuclear localization. In the ESOP lineage, it is first expressed in the IIIB cell and is inherited by both the ES neuron and glia daughter cells after IIIB division. Although Ham quickly disappears in the differentiating glia, it continues to be expressed by the ES neuron during initial dendrite extension, indicating that it may be active both pre- and postmitotically in these neurons. Ham levels become undetectable as further dendrite elaboration occurs. This expression pattern is consistent with the notion that Ham is required for proper cell fate specification of the IIIB lineage, and is required for the dendritic morphogenesis of the ES neuron (Moore, 2002).

EFFECTS OF MUTATION

For a complete and accurate sampling of sensory or presynaptic input, neuronal ensembles can organize as tiled systems, with the dendritic fields of like neurons partitioning a receptive area much like tiles covering a floor. Understanding how dendrites establish their territory is central to elucidating how neuronal circuits are built. Signaling between dendrites is thought to be important for defining their territories; however, the strategies by which different types of dendrites communicate are poorly understood. Two classes of Drosophila peripheral da sensory neurons, the class III and class IV neurons, provide complete and independent tiling of the body wall. By contrast, dendrites of class I and class II neurons do not completely tile the body wall, but they nevertheless occupy nonoverlapping territories. To address the generality of tiling and also its mechanistic basis, a strain of flies was generated with all class IV neurons marked with Enhanced Green Fluorescent Protein (EGFP) to facilitate live imaging of tiling dendritic arbors. By developing reagents to permit high-resolution studies of dendritic tiling in living animals, it was demonstrated that isoneuronal and heteroneuronal class IV dendrites engage in persistent repulsive interactions. In contrast to the extensive dendritic exclusion shown by class IV neurons, duplicated class III neurons shows repulsion only at their dendritic terminals. Supernumerary class I and class II neurons innervated completely overlapping regions of the body wall, and this finding suggests a lack of like-repels-like behavior. These data suggest that repulsive interactions operate between morphologically alike dendritic arbors in Drosophila. Further, Drosophila da sensory neurons appear to exhibit at least three different types of class-specific dendrite-dendrite interactions: persistent repulsion by all branches, repulsion only by terminal dendrites, and no repulsion (Grueber, 2003).

The growth of dendrites after cell ablation suggests that repulsive signaling is required for tiling. If these signals are communicated between dendrites, then overproduction of da neurons should lead to ectopic partitioning of territories. To test this idea, an overproduction of class IV neurons was induced by introducing mutations in the hamlet (ham) gene into a ppk-EGFP genetic background (EGFP driven by the pickpocket promoter. ham is required for the specification of external sensory (es) neurons, and loss of ham leads to es neurons adopting a multidendritic neuron fate. In a ham background, doubling of v'ada, but not of the other two abdominal class IV neurons was frequently observed. In each case (n = 11), the morphology of these two adjacent neurons was similar, and, in ten of these cases, dendrites projected to, and terminated within, distinct domains of the body wall. One case was observed in which the major trunks of the two adjacent neurons clearly projected to an overlapping region, which the finer branches partitioned into nonoverlapping subdomains. In response to the extra neuron, the dendritic field of vdaB was restricted to a more ventral region of the body wall. The ectopic partitioning of the body wall induced by overproduction of class IV neurons strongly suggests that repulsive dendritic interactions occur between dendrites of like neurons (Grueber, 2003).

The class I, II, and III neurons are organized differently from the class IV neurons; they have dendrites that are either not normally in close apposition (classes I and II) or provide a low-density coverage of the body wall (class III). It was asked whether these neurons might show exclusion if produced in excess. Supernumerary neurons were produced by making MARCM clones by using big brain (bib) or ham mutations, both of which lead to overproduction of neurons. Duplications of class I, II, and III neurons were observed in ham clones and duplication of class I neurons was observed in bib clones. In contrast to the dendritic avoidance exhibited by duplicated class IV neurons, duplicated class I neurons and class II neurons innervated overlapping regions of the body wall in each case examined. Although these dendrites did not clearly fasciculate, they intermixed very extensively, and often it was not possible to distinguish between the arbors of the two neurons. Class III neurons duplicated in ham MARCM clones behave much like the class I and II neurons in that their major trunks overlap extensively and often extend along the same direction since they cover identical territories of the epidermis. Extensive overlap was not observed, however, among the short terminal extensions of the class III neurons. It is therefore possible that these short branches contribute to the dendritic exclusion normally observed among class III neurons (Grueber, 2003).

Results from several recent studies suggest that dendritic interactions between da neurons regulate the size and shape of dendritic fields. Ablation of a cluster of da neurons results in overgrowth of dendrites from adjacent hemisegments. The neurons within these clusters are morphologically diverse and only those sharing a similar morphology consistently innervate exclusive territories. These results suggested that dendritic interactions might show type selectivity. The present results from ablation and addition-of-neuron experiments suggest that repulsive dendritic interactions indeed occur between branches of like neurons to regulate dendritic field sizes, but they also indicate that neurons can show diverse responses to like dendrites. In particular, a fairly strict dendritic exclusion is observed among the class IV neurons, whereas dendrites of supernumerary class I, II, and III neurons overlap extensively with dendrites of resident same-class neurons. Thus, like-repels-like signaling may be a property only of neurons that provide a complete coverage of a receptive territory and, even among these neurons, may be restricted to specific regions of the dendritic arbor. How are dendritic fields defined among neurons that appear not to employ homotypic repulsion? Possibly, different types of neurons have different intrinsic growth capacitites determined by their expression of particular cell identity factors. Other growth-limiting factors might include genes that, when mutated, cause overgrowth phenotypes, such as flamingo, or as yet unidentified extrinsic limitations to dendrite extension (Grueber, 2003).

The da system of Drosophila shows notable similarities to and differences from other systems, such as the vertebrate retina and leech somatosensory system, in which tiling and interbranch repulsion have been studied. Destruction of small patches of retinal ganglion cells (RGCs) causes neighboring dendrites to grow preferentially toward the depleted area. Dendritic tiling by RGCs may generally involve interactions between dendrites of adjacent cells. This exclusion appears to occur independently of afferent inputs to RGC dendrites. Thus, even though da neuron dendrites are unlike RGC dendrites in that they have no known synaptic inputs, they may prove useful for identifying conserved mechanisms of dendrodendritic interaction. In the leech also, proper innervation of peripheral targets by pressure, touch, and nociceptive neurons requires both intra- and inter-neuronal interactions between sensory fibers. Anatomical and ablation studies in leech suggest that exclusion is most rigorous between different branches of the same neuron, less strict between homologous cells, and weak or absent if cells are of a different modality. The overlap of arbors of duplicated neurons, but not of branches belonging to the same neuron, may reflect such a hierarchy in the da system. The contrasting behaviors of isoneuronal and heteroneuronal branches of class I, II, and III neurons suggest that mechanisms of avoidance may differ for 'self' and 'non-self' branches. It is also possible that isoneuronal and heteroneuronal repulsion share some common molecular mechanism, but that physical continuity of arbors contributes to a more strict avoidance by isoneuronal branches (Grueber, 2003).

Several properties of the signals regulating tiling are implied in these results. (1) The interaction between dendrites seems to be inhibitory and can result in the turning of dendrites and/or the cessation of dendrite extension. (2) It is likely that the interaction is bidirectional. In other words, it is unlikely that one dendrite is only capable of sending out the signal while the other dendrite can only receive the signal. Lastly, because interactions occur between like dendrites and are required persistently, at least some molecules are likely to show a class-specific distribution in embryonic and larval stages. With the high resolution provided by the ppk-EGFP lines, it is feasible to carry out both candidate-based approaches and unbiased genetic screens to identify molecules involved in dendrodendritic interaction and tiling (Grueber, 2003).

Conversion of neurons and glia to external-cell fates in the external sensory organs of Drosophila hamlet mutants by a cousin-cousin cell-type respecification

The Drosophila external sensory organ forms in a lineage that is elaborated from a single precursor cell via a stereotypical series of asymmetric divisions. Hamlet transcription factor expression demarcates the lineage branch that generates two internal cell types, the external sensory neuron and thecogen. In Hamlet mutant organs, these internal cells are converted to external cells via an unprecedented cousin-cousin cell-fate respecification event. Conversely, ectopic Hamlet expression in the external cell branch leads to internal cell production. The fate-determining signals Notch and Pax2 act at multiple stages of lineage elaboration and Hamlet acts to modulate their activity in a branch-specific manner (Moore, 2004).

Tissues that develop from progenitor cells, including the vertebrate hematopoietic system and the central and peripheral nervous systems, generate multiple cell types from a single precursor via iterative cell divisions. The Drosophila external sensory organ (ESO) also forms in this way. It consists of five different cell types descended from one ESO precursor cell (ESOP) via a stereotypical series of asymmetric divisions (Moore, 2004).

ESOP cell division forms the IIA and IIB cells. The IIA gives rise to the external cells, the trichogen (hair), and the tormagen (the socket) that are visible on the surface of the cuticle (external cell [E]-branch). The IIB cell divides to give rise to neuronal and glial internal cell types that lie beneath the surface of the cuticle (internal cell [I]-branch). At each stage of elaboration, each cell can be clearly visualized and distinguished; hence, the ESO is an excellent model in which to examine the elaboration of multiple cell types from one precursor at the single-cell level in vivo (Moore, 2004).

Each division of the ESO lineage is asymmetric; one of the two cells formed inherits the Numb protein, causing it to have a lower level of Notch (N) activity than its sibling. N-mediated signaling between the siblings then determines a difference in identity between them. This difference is expressed in terms of gene expression (for example, the IIB cell expresses the transcription factor Prospero [Pros], whereas the IIA does not) and in terms of cell behavior (for example, the IIB cell divides with a different plane of mitotic spindle orientation to the IIA) (Moore, 2004).

The disruption of the function of genes that generate asymmetry between siblings (e.g., N) always leads to a sibling-sibling conversion. In many cases, the disruption of transcription factors required for cell differentiation also leads to sibling-sibling conversions; for example, loss of pros activity leads to IIB-to-IIA cell conversions at low frequency, and ectopic expression of Pros in the IIA cell converts this cell into a IIB. A second type of conversion is a nephew-uncle conversion; for example, in the Drosophila embryo, the Hamlet (HAM) transcription factor is required for external sensory (ES) neuron fate, and in the absence of Ham the ES neuron is converted to the IIIBsib cell. This represents a conversion between internal cell types, with a shared ancestor (IIB cell) that is a direct parent of one of the two cell types (Moore, 2004).

To date the only type of cell-fate conversions described in the ESOP lineage are sibling-sibling or nephew-uncle conversions. This is unsurprising given the binary nature of each cell-fate decision. Loss of HAM activity in the thecogen (and also the ES neuron in the adult ESO lineage) causes a fate switch not to that of an internal cell type, but to an external cell type. This switch represents a cousin-cousin conversion, which is shown to take place by a respecification mechanism. In ham mutants, the ES neuron and thecogen first express markers associated with the I-branch of the lineage, but they fail to fully differentiate and later convert into external cell types without first reentering the cell cycle. One role of Ham in the development of the ESO is to modulate the activity of N and Pax2 signals used at multiple points in the lineage in a branch-specific manner (Moore, 2004).

In the embryonic ESO lineage, Ham is expressed in the ES neuron but not in the multidendritic (MD) neuron. Loss of Ham function converts the ES to an (uncle) MD neuron fate, and indeed Ham acts as a binary genetic switch between these two internal cell fates. Ham is also expressed in the thecogen cell, which is the sibling of the ES neuron. If the thecogen cell undergoes a cell-fate conversion in the absence of Ham, however, it could not have switched fates to a sibling or uncle cell type, as both of these are neurons, and in ham mutant ES organs there is no increase in neuron number (Moore, 2004).

To investigate the fate of the 'thecogen' in a ham mutant, wild-type and ham mutant embryos were stained with antibodies to detect expression of the transcription factors Cut, Pros, and Pax2, and the placW enhancer trap A1-2-29. Cut marks all cells descended from the ESOP. Upon terminal differentiation of the ESO, Pax2 marks the thecogen and trichogen, Pros marks only the thecogen, and A1-2-29 drives ß-galactosidase expression in the trichogen and tormagen. In the dorsal external sensory (des) and ventral pore (vp) organs of ham mutant embryos, the total number of cells remains unchanged, and the thecogen cell (expressing Cut, Pax2, and Pros) is replaced by a cell expressing Cut, Pax2, and A1-2-29; this combination of markers normally defines trichogen fate. Given that this staining pattern implies that this cell is a trichogen, the ESO structures on the cuticle surface of ham mutant second instar larvae were examined. Sixty-five percent of organs showed a wild-type one-trichogen and one-tormagen phenotype, and 34% a two-trichogen and one-tormagen phenotype (Moore, 2004).

The replacement of the thecogen by a trichogen in the ham mutant could have occurred by two different mechanisms. In the first, an alteration of the ESO lineage division pattern is responsible for the replacement of an internal cell type by an external one. In the second, the thecogen derived from the IIIB cell is converted not to another cell type derived from the I-branch of the lineage, but to a cell type that is representative of the E-branch. These two possibilities have very different implications for understanding how the ESO develops. The first is consistent with the idea that all cell-fate decisions within this lineage are essentially binary and the fates of the daughters are restricted by the identity of the precursors giving rise to them. The second, in contrast, implies there is no such restriction (Moore, 2004).

In order to distinguish between these possibilities, the adult microcheate ESO lineage, in which the elaboration of each ESO derived from a single ESOP cell by live imaging, was analyzed. The adult ESO lineage is considered analogous to that of the embryo, except that the IIIBsib in the embryo becomes an MD neuron, whereas in the adult it undergoes apoptosis (Moore, 2004).

The Ham expression pattern in the adult ESO lineage is analogous to that seen in the embryo. Pupae were dissected at a stage in which the ESO lineage was elaborating and stained with antibodies to detect Ham, Pros, and the A101 enhancer trap. A101 is expressed in all cells of the ESO lineage and Pros is expressed, at least transiently, in all cells of the I-branch. Ham is first expressed in the IIIB cell and then inherited by its daughters, the ES neuron and thecogen (Moore, 2004).

Whether the supernumerary external cell seen in embryonic ham mutant ESOs was also present in the adult was investigated. The external phenotype of ham mutant ESO clusters was examined in nota mosaic for wild-type and ham mutant tissue. ham mutant ESOs were identified by the loss of Yellow (Y) in the thecogen and tormagen cells. Seven percent of y^- sensory organs had a two-trichogen, one-tormagen phenotype similar to the ham mutant phenotype in the embryonic ESO. In addition, 42% of ham mutant ESOs had a one-trichogen, two-tormagen phenotype and 23% had a two-trichogen, two-tormagen phenotype (Moore, 2004).

To distinguish whether these supernumerary external cells have arisen from either an altered division pattern in the ESO lineage or conversion of terminal cell fates, live imaging of ESO lineage elaboration was carried in ham mutant clones between 18 and 38h after pupal formation (APF). ham mutant mosaic analysis via a repressible cell marker (MARCM). Clones were marked by Partner of Numb-GFP (PON-GFP) fusion protein expression; Pon is asymmetrically localized and inherited by only one daughter at each cell division of the ESO lineage. In ESO lineages in which the IIB cell has been converted to the IIA, for example, by ectopic expression of activated N in the IIB, the timing and orientation of the division of the cells of the I-branch are altered to resemble those of the E-branch. In contrast, the elaboration of ham mutant ESO lineages was indistinguishable from that in wild-type in the timing, orientation, and number of divisions in both the E-branch and the I-branch. Therefore, the supernumerary external cells in ham mutant clones are not due to a conversion of I-branch precursors into E-branch precursors or extra divisions within the E-branch itself; they are due to the conversion of the IIIB cell daughters to external cell fates (Moore, 2004).

The expression of I-branch specific markers was investigated in ham mutant ESO lineages during elaboration. ham mutant MARCM clones positively marked with mCD8GFP fusion protein expression were made in all ESOP-derived cells. ham mutant clones at all stages of ESO lineage elaboration were stained with antibodies that detected the following markers: Pros, Pax2, Elav (Embryonic lethal abnormal vision), which labels all differentiated neurons, and Suppressor of Hairless [Su(H)], which labels differentiated tormagen. In both wild-type and ham mutant lineages, the IIB, IIIB, and IIIBsib cells expressed Pros, thus confirming the live imaging findings that showed no differences between the elaborating I-branch in wild-type and ham mutant lineages (Moore, 2004).

The IIIB progeny in ham mutant clusters, however, shifted their patterns of differentiation over time. Shortly after division of the IIIB cell (22-24 h APF), one daughter in ham mutant clones continued to express Pros, similar to a wild-type thecogen, and the other began to express ELAV, a marker of neuron fate. By 28-30 h APF, the ham mutant ESO cell clearly differed from the wild-type control: expression of the I-branch-specific maker Pros was lost. Moreover, Elav-positive neurons were no longer present, but several ham mutant ESO showed small Elav-positive apoptotic cell fragments, indicative of the ES neuron undergoing cell death. Live confocal imaging of ham mutant clones showed the frequency of this event to be 15%. By 36-40 h, wild-type ESO clearly contained one trichogen and one thecogen (both Pax2-positive), one tormagen [Su(H)-positive], and one ES neuron (Elav-positive). In contrast, ham mutant ESOs contained no ES neurons or thecogen, two trichogen (Pax2-positive), and either one or two tormagen [Su(H)-positive]. Whereas mCD8GFP activity and Pax2 staining revealed the two-trichogen phenotype with 100% frequency, the appearance of two trichogens on the adult cuticle was less frequent, indicating that one supernumerary trichogen must have failed to grow out onto the cuticle surface. Costaining of ham mutant embryos with antibodies to detect Cut, Pros, and ELAV revealed that Pros is also transiently expressed in the thecogen cell before it undergoes conversion to a trichogen. Therefore, in both the adult and embryonic ham mutant ESO lineage, the daughters of the IIIB cell first express markers specific to internal cells, but expression of such markers ceases as these cells undergo respecification to an external cell fate (Moore, 2004).

In the ham mutant embryo, the ES neuron is transformed into a IIIBsib/MD neuron fate and the thecogen into a trichogen. On the basis of these embryo data, it is proposed that in the ham mutant adult, the thecogen is also converted into a trichogen and the ES neuron into either a IIIBsib (apoptotic cell) or a tormagen. A possible reason that the conversion properties of the ES neuron in the embryo are different from those of the adult could be that, as neuron-specific markers are already expressed in the ES neuron before it undergoes fate conversion, in the embryo, where a neuron-to-neuron respecification event can occur, it is favored over a neuron-to-nonneuron one (Moore, 2004).

If the cell-fate conversions that are proposed take place, then it would mean that the thecogen (high N) becomes a trichogen (low N) and the ES neuron (low N) can become a tormagen (high N). To confirm this, whether the high-N or low-N IIIB daughter cell becomes the supernumerary trichogen was investigated in a ham mutant ESO. ham MARCM clones were made in an N^ts (temperature sensitive) background and N was inactivated in these clones at the stage where N-mediated signaling is generating asymmetry between the IIIB daughter cells. The resulting nota were stained with antibodies to detect mCD8, Su(H), and Pax2 and the fate of the IIIB daughter cells was examined. In the ham¹, N^ts ESOs 28/28 had a one-trichogen/multiple-tormagen phenotype, whereas in the ham¹, N⁺ control 37/37 ESOs had a two-trichogen/multiple-tormagen phenotype (Moore, 2004).

This experiment clearly demonstrates that it is the high-N IIIB cell daughter that becomes the supernumerary trichogen, since reducing N activity results in a reduction in the number of trichogen to one in the ham mutant organ. The conclusion to be drawn from this experiment is, therefore, that whereas N is acting to determine the difference between daughters of an asymmetric division, of either IIA or IIIB, a second signal is acting that makes the trichogen similar to the thecogen and the tormagen similar to the ES neuron (Moore, 2004).

This study shows that in both the embryo and adult, ESO Ham is expressed solely in the IIIB cell and its daughters, the ES neuron and thecogen. Loss of Ham causes the conversion of the internal-cell-branch thecogen into an external-cell-branch trichogen by cell-fate respecification. In addition, loss of Ham in the ES neuron leads either to its conversion to an internal-cell-branch IIIBsib cell or an external-cell-branch tormagen. Therefore, Ham appears to act to determine the fate of the IIIB daughters with respect to all other terminally differentiated cell types in the ESO lineage. In other words, it acts as an intrinsic transcription factor determinant of IIIB cell-derived identity (Moore, 2004).

It was of interest to test whether Ham expression alone determines the difference between IIIB-derived and non-IIIB-derived fate. Ectopic expression of Ham in the embryonic IIB and IIIBsib cell (MD neuron) converts the IIIBsib into an ES neuron. This analysis was extended to the E-branch by using gal4^109-68 to drive UAS-ham and UAS-mCD8GFP in all cells of the adult lineage. Ham expression in all ESO lineage cells led to the loss of both the E-branch-derived trichogen and tormagen from the surface of the cuticle. Antibody staining of the ESOs showed that I-branch-specific cell types had replaced these external cells. When Ham was ectopically expressed in the entire ESO lineage, five or six cells were seen, all of which were expressing Pros or Elav or (rarely) both Pros and Elav. In contrast, in a wild-type cluster, there are four cells including only one thecogen (Pros-positive) and one ES neuron (Elav-positive). These ectopic expression experiments confirm that Ham determines IIIB-derived versus non-IIIB-derived fate (Moore, 2004).

These ectopic expression experiments confirm that Ham determines IIIB-derived versus non-IIIB-derived fate. However, why in ham mutant ESOs are the transformations that occur thecogen (high N) to tricogen (low N) and ES neuron (low N) to tormagen (high N)? Pax2 expression in the ESO highlights a connection between these pairs of cell types; the thecogen and trichogen both express Pax2, whereas the tormagen and ES neuron do not. Moreover, Pax2 itself is required for hair-shaft differentiation in the trichogen, and ectopic Pax2 expression in the tormagen leads to ectopic hair-shaft development. In the thecogen, where Pax2 and Ham are coexpressed, Pax2 expression does not lead to hair-shaft development; however, in the absence of Ham, this cell now takes on a trichogen fate including the development of a hair shaft. Therefore one role of Ham in the thecgoen cell may be to suppress the ability of Pax2 to promote hair-shaft formation (Moore, 2004).

To investigate whether Ham can repress the hair-shaft-promoting activity of Pax2, Ham or Pax2 were expressed or Ham and Pax2 were coexpressed at high levels in all cells of the ESO lineage. neu-gal4 was used to drive expression of UAS-ham and/or UAS-Pax2; however, this causes embryonic lethality. To get around this problem, ectopic gene expression was driven only in notum clones. Ectopic expression of Pax2 led to organs with multiple hair shafts, some organs with misshapen external cells, and in some cases loss of external cells. In contrast, ectopic Ham or Ham and Pax2 caused the loss of external cells in almost 100% of ESO clones and never the formation of multiple hair shafts. These experiments demonstrate that Ham has the ability to modulate a Pax2-driven differentiation program, in this case, hair-shaft growth. Therefore, loss of Ham from the "thecogen" could lead to the depression of a program at least in part controlled by Pax2, which drives this cell to a trichogen fate (Moore, 2004).

The Pax2 cell-fate-determining signal is used at multiple points during the development of the ESO lineage, as is N. The presence or absence of Ham provides a branch-specific background state against which differentiating cells of the organ can interpret these signals. It is likely that a similar modulation of iterated signals by branch-specific factors occurs in vertebrate systems. For example, N signaling occurs at multiple points during the development of the hematopoietic lineage and could be modulated by the presence of branch-specific transcription factors. It is suggested that the analysis of Ham function in ESO lineage elaboration presented in this study provides useful insight into how cell fate is determined in many invertebrate and vertebrate lineages in which a single stem/precursor cell gives rise to multiple cell types via iterative cell divisions (Moore, 2004).

EVOLUTIONARY HOMOLOGS

EGL-43, a C. elegans homolog of Hamlet

During embryonic development, the two Caenorhabditis elegans HSN motor neurons migrate from their birthplace in the tail to positions near the middle of the embryo. Of all cells that undergo long-range migrations, only the HSNs are affected in animals that lack function of the egl-43 gene. egl-43 function is required for normal development of phasmid neurons, which are sensory neurons located in the tail. The egl-43 gene encodes two proteins containing zinc finger motifs that are similar to the zinc fingers of the murine Evi-1 proto-oncoprotein. These genetic and molecular results suggest that egl-43 encodes two transcription factors and acts to control HSN migration and phasmid neuron development, presumably by regulating other genes that function directly in these processes (Garriga, 1993).

The Caenorhabditis elegans HSN motor neurons permit genetic analysis of neuronal development at single-cell resolution. The egl-5 Hox gene, which patterns the posterior of the embryo, is required for both early (embryonic) and late (larval) development of the HSN. ham-2 encodes a zinc finger protein that acts downstream of egl-5 to direct HSN cell migration, an early differentiation event. The EGL-43 zinc finger protein, also required for HSN migration, is expressed in the HSN specifically during its migration. In an egl-5 mutant background, the HSN still expresses EGL-43, but expression is no longer down-regulated at the end of the cell's migration. Finally, a new role in early HSN differentiation has been found for UNC-86, a POU homeodomain transcription factor that acts downstream of egl-5 in the regulation of late HSN differentiation. In an unc-86; ham-2 double mutant the HSNs are defective in EGL-43 down-regulation, an egl-5-like phenotype that is absent in either single mutant. Thus, in the HSN, a Hox gene, egl-5, regulates cell fate by activating the transcription of genes encoding the transcription factors Ham-2 and UNC-86 that in turn individually control some differentiation events and combinatorially affect others (Baum, 1999).

EVI1 protein structure and protein interactions

Inappropriate expression of the Evi-1 zinc finger gene is associated with myeloid leukemia and myelodysplastic syndromes in mice and humans and has been hypothesized to contribute to pathology by blocking myeloid differentiation. Evi-1 contains two domains of zinc fingers, an amino-terminal domain of seven fingers and a carboxyl domain of three fingers. The first domain binds a consensus sequence of GA(C/T)AAGATAAGATAA in binding and amplication reactions or GATA repeat containing regions of genomic DNA. The experiments described here, establish a consensus sequence for the carboxyl domain of zinc fingers consisting of GAAGATGAG. Unlike the first domain, the consensus sequence established for the carboxyl domain is identical to that which would be predicted by the current rules relating to C2H2 zinc fingers and DNA recognition. Substitution of sequences in finger 8 with those in finger 9, demonstrate that the individual fingers bind the predicted region of the consensus sequence. In an attempt to engineer binding of constructs containing the carboxyl domain, a variety of mutations were made in the middle finger that would be predicted to change the consensus sequence in specific ways. Remarkably, most of the mutations were deleterious and destroyed specific DNA binding. Although Evi-1 contains potential transcriptional activation domains, it was not able to activate gene transcription from CAT constructs containing the consensus sequence (Funabiki, 1994).

The myeloid transforming gene Evi-1 encodes a protein with two zinc finger domains, designated ZF1 and ZF2, with distinct DNA binding specificities. Evi-1 has transcriptional repressor activity which is directly proportional to the amount of Evi-1 protein in cells. Repression has been observed with two distinct promoters: the minimal HSV-1 tk promoter and a VP16 inducible adenovirus E1b minimal promoter. Optimal repression is DNA binding dependent and is mediated by either ZF1 or a heterologous GAL4 DNA binding domain (GAL4DBD) but is significantly less efficient through the ZF2 binding site. Both GAL4DBD/Evi-1 fusion and non-fusion proteins have been used to map the repressor activity to a proline-rich region located within amino acids 514-724 between the ZF1 and ZF2 domains. Constitutive expression of mutant proteins lacking the repressor domain are defective for transformation of Rat1 fibroblasts, demonstrating that this region is required for the oncogenic activity of the Evi-1 protein. These studies show that the Evi-1 gene encodes a transcriptional repressor and has important implications for the mechanism of action of the Evi-1 protein both in development and in the progression of some myeloid leukaemias (Bartholomew, 1997).

Evi-1 encodes a zinc-finger protein that may be involved in leukemic transformation of hematopoietic cells. Evi-1 has two zinc-finger domains, one with seven repeats of a zinc-finger motif and one with three repeats, and it has characteristics of a transcriptional regulator. Although Evi-1 is thought to be able to promote growth and to block differentiation in some cell types, its biological functions are poorly understood. The mechanisms that underlie oncogenesis induced by Evi-1 have been studied by investigating whether Evi-1 perturbs signalling through transforming growth factor-beta (TGF-beta), one of the most studied growth-regulatory factors, which inhibits proliferation of a wide range of cell types. Evi-1 represses TGF-beta signalling and antagonizes the growth-inhibitory effects of TGF-beta. Two separate regions of Evi-1 are responsible for this repression; one of these regions is the first zinc-finger domain. Through this domain, Evi-1 interacts with Smad3, an intracellular mediator of TGF-beta signalling, thereby suppressing the transcriptional activity of Smad3. These results define a new function of Evi-1 as a repressor of signalling through TGF-beta (Kurokawa, 1998).

The Evi-1 gene encodes a zinc finger transcriptional repressor protein that normally plays a role in development and is frequently activated in myeloid leukaemias. Evi-1 has two distinct DNA binding domains, ZF1 and ZF2, and a defined repressor domain but the function of the remainder of the molecule is unknown. The ZF1, ZF2 and repressor domains are required for transformation. An alternative splice variant of Evi-1, designated delta324, encodes a protein that lacks a portion of the ZF1 DNA binding domain and the intervening amino acids 239-514 (designated IR) located between ZF1 and the repressor domain. Delta324 can neither bind ZF1, repress transcription through this site nor transform Rat1 fibroblasts. Reconstitution studies demonstrate that the defect in delta324 is partially complemented by recreating the ZF1 DNA binding activity. However, full function also requires the IR region which has transcriptional repressor activity. This study shows therefore, that ZF1, ZF2 and repressor domains and the IR region all contribute to the transformation efficiency of the Evi-1 protein (Kilbey, 1998).

EVI-1 and its variant form, MDS1/EVI1, act in an antagonistic manner and are differentially regulated in samples from patients with acute myeloid leukaemia and rearrangements of the long arm of chromosome 3. Both EVI-1 and MDS1/EVI1 can repress transcription from a reporter construct containing EVI-1 binding sites and they interact with histone deacetylase in mammalian cells. This interaction can be recapitulated in vitro and is mediated by a previously characterized transcription repression domain, whose activity is alleviated by the histone deacetylase inhibitor trichostatin A (Vinatzer, 2001).

Ectopic production of the EVI1 transcriptional repressor zinc finger protein is seen in 4%-6% of human acute myeloid leukemias. Overexpression also transforms Rat1 fibroblasts by an unknown mechanism, which is likely to be related to its role in leukemia and which depends upon its repressor activity. Mutant murine Evi-1 proteins, lacking either the N-terminal zinc finger DNA binding domain or both DNA binding zinc finger clusters, function as dominant negative mutants by reverting the transformed phenotype of Evi-1 transformed Rat1 fibroblasts. The dominant negative activity of the non-DNA binding mutants suggests sequestration of transformation-specific cofactors and that recruitment of these cellular factors might mediate Evi-1 transforming activity. C-terminal binding protein (CtBP) co-repressor family proteins bind PLDLS-like motifs. The murine Evi-1 repressor domain has two such sites, PFDLT (site a, amino acids 553--559) and PLDLS (site b, amino acids 584--590), which independently can bind CtBP family co-repressor proteins, with site b binding with higher affinity than site a. Functional analysis of specific CtBP binding mutants shows site b is absolutely required to mediate both transformation of Rat1 fibroblasts and transcriptional repressor activity. This is the first demonstration that the biological activity of a mammalian cellular transcriptional repressor protein is mediated by CtBPs. Furthermore, it suggests that CtBP proteins are involved in the development of some acute leukemias and that blocking their ability to specifically interact with EVI1 might provide a target for the development of pharmacological therapeutic agents (Palmer, 2001).

Evi-1 is a zinc finger nuclear protein whose inappropriate expression leads to leukemic transformation of hematopoietic cells in mice and humans. Inappropriate expression blocks the antiproliferative effect of transforming growth factor beta (TGF-beta). Evi-1 represses TGF-beta signaling by direct interaction with Smad3 through its first zinc finger motif. Evi-1 represses Smad-induced transcription by recruiting C-terminal binding protein (CtBP) as a corepressor. Evi-1 associates with CtBP1 through one of the consensus binding motifs, and this association is required for efficient inhibition of TGF-beta signaling. A specific inhibitor for histone deacetylase (HDAc) alleviates Evi-1-mediated repression of TGF-beta signaling, suggesting that HDAc is involved in the transcriptional repression by Evi-1. This identifies a novel function of Evi-1 as a member of corepressor complexes and suggests that aberrant recruitment of corepressors is one of the mechanisms for Evi-1-induced leukemogenesis (Izutsu, 2001).

The leukemia-associated fusion gene AML1/MDS1/EVI1 (AME) encodes a chimeric transcription factor that results from the (3;21)(q26;q22) translocation. This translocation is observed in patients with therapy-related myelodysplastic syndrome (MDS), with chronic myelogenous leukemia during the blast crisis (CML-BC), and with de novo or therapy-related acute myeloid leukemia (AML). AME is obtained by in-frame fusion of the AML1 and MDS1/EVI1 genes. AME is a transcriptional repressor that induces leukemia in mice. In order to elucidate the role of AME in leukemic transformation, the interaction was investigated of AME with the transcription co-regulator CtBP1 and with members of the histone deacetylase (HDAC) family. AME is shown to physically interacts in vivo with CtBP1 and HDAC1; these co-repressors require distinct regions of AME for interaction. By using reporter gene assays, AME has been demonstrated to repress gene transcription by CtBP1-dependent and CtBP1-independent mechanisms. The interaction between AME and CtBP1 is shown to be biologically important and is necessary for growth upregulation and abnormal differentiation of the murine hematopoietic precursor cell line 32Dc13 and of murine bone marrow progenitors (Senyu, 2002).

Lysine acetyltransferases modulate the activity of many genes by modifying the lysine residues of both core histones and transcription-related factors. These modifications are tightly controlled in the cell because they are involved in vital processes such as cell cycle progression, differentiation, and apoptosis. Therefore, any deregulation of acetylation/deacetylation equilibrium or inappropriate modifications could lead to different diseases. Since previous studies have shown that some oncoproteins also undergo this modification, acetylation could be involved in the processes of cell transformation and oncogenesis. AML1/MDS1/EVI1 (AME), a repressor produced by the t(3;21) associated with human leukemia, physically interacts with the acetyltransferases P/CAF and GCN5. These data suggest that AME has at least two binding sites for these acetyltransferases, one of which is in the Runt domain. Both P/CAF and GCN5 efficiently acetylate AME in vivo in the central region. AME acetylation has no effect on its interaction with the co-repressor CtBP1. The co-expression of AME and either P/CAF or GCN5 abrogates the repression of an AML1-dependent reporter gene (Senyuk, 2003).

EVI1 is a very complex protein with two domains of zinc fingers and is inappropriately expressed in many types of human myeloid leukemias. EVI1 is a transcription repressor, and has been shown to interact with CtBP1. Inappropriate expression of EVI1 in murine hematopoietic precursor cells leads to their abnormal differentiation and to increased proliferation. Using biochemical assays, two groups of transcription co-regulators have been identified that associate with EVI1 presumably to regulate gene expression. One group of co-regulators includes the CtBP1 and histone deacetylase. The second group includes the two co-activators cAMP-responsive element-binding protein-binding protein (CBP) and p300/CBP-associated factor (P/CAF), both of which have histone acetyltransferase (HAT) activity. All of these proteins require separate regions of EVI1 for efficient interaction, and they divergently affect the ability of EVI1 to regulate gene transcription in reporter gene assays. Confocal microscopy analysis shows that in the majority of the cells, EVI1 is nuclear and diffused, whereas in about 10% of the cells EVI1 localizes in nuclear speckles. However, in the presence of the added exogenous co-repressors histone deacetylase or CtBP1, all of the nuclei have a diffuse EVI1 staining, and the proteins do not appear to reside together in obvious nuclear structures. In contrast, when CBP or P/CAF are added, defined speckled bodies appear in the nucleus. Analysis of the staining pattern indicates that EVI1 and CBP or EVI1 and P/CAF are contained within these structures. These nuclear structures are not observed when CBP is substituted with a point mutant HAT-inactive CBP with which EVI1 also physically interacts. The interaction of EVI1 with either CBP or P/CAF leads to acetylation of EVI1. These results suggest that the assembly of EVI1 in nuclear speckles requires the intact HAT activity of the co-activators (Chakraborty, 2001).

Intergenic splicing of EVI1

The EVI1 gene, located at chromosome band 3q26, is overexpressed in some myeloid leukemia patients with breakpoints either 5' of the gene in the t(3;3)(q21;q26) or 3' of the gene in the inv(3)(q21q26). EVI1 is also expressed as part of a fusion transcript with the transcription factor AML1 in the t(3;21)(q26;q22), associated with myeloid leukemia. In cells with t(3;21), additional fusion transcripts are AML1-MDS1 and AML1-MDS1-EVI1. MDS1 is located at 3q26 170-400 kb upstream (telomeric) of EVI1 in the chromosomal region in which some of the breakpoints 5' of EVI1 have been mapped. MDS1 has been identified as a single gene as well as a previously unreported exon(s) of EVI1. The relationship between MDS1 and EVI1 has been analyzed to determine whether they are two separate genes. In this report, evidence is presented indicating that MDS1 exists in normal tissues both as a unique transcript and as a normal fusion transcript with EVI1, with an additional 188 codons at the 5' end of the previously reported EVI1 open reading frame. This additional region has about 40% homology at the amino acid level with the PR domain of the retinoblastoma-interacting zinc-finger protein RIZ. These results are important in view of the fact that EVI1 and MDS1 are involved in leukemia associated with chromosomal translocation breakpoints in the region between these genes (Fears, 1996).

EVI1, located at chromosome band 3q26, encodes a 1051 amino acid zinc finger protein inappropriately expressed in the leukemic cells of 2%-5% of acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS) patients. The activation of EVI1 often follows a chromosomal rearrangement involving band 3q26, and the two most frequent rearrangements are the t(3;3)(q21;q26) and the inv(3)(q21q26). EVI1 exists also as a longer protein that includes 188 additional amino acids at the N-terminus, named MDS1/EVI1. Both genes are expressed at very low levels in the normal bone marrow. The genomic region between the first coding exon of MDS1/EVI1 and the first coding exon of EVI1 is 150-300 kb. The majority of the chromosomal breakpoints at the 5' end of EVI1 in the t(3;3) resulting in EVI1 activation have been mapped in this region. As a consequence of the t(3;3), the cell would be unable to express MDS1/EVI1, although it would express EVI1. The transcriptional activity of MDS1/EVI1 and EVI1 were compared; MDS1/EVI1 is a strong activator of promoters containing the AGATA motif, whereas EVI1 is a repressor. In addition, whereas EVI1 represses activation by the GATA-1 erythroid factor, MDS1/EVI1 does not, and is itself repressed by EVI1. By gene fusion to the DNA-binding domain of Gal4, it is further shown that the activation properties of MDS1/EVI1 are restricted to an acidic segment encoded by the second and third exons in the 5' untranslated region of EVI1. The relative expression of the two genes in normal bone marrow and in the bone marrow of leukemia patients with 3q26 rearrangements has also been examined. The results indicate that the rearrangements at 3q26 affect expression of EVI1, but not of MDS1/EVI1. It is proposed that rearrangements at 3q26 involving EVI1 could result in leukemia by a two-step process involving first transcriptional disruption of MDS1/EVI1, and next by inappropriately activating expression of EVI1 (Soderholm, 1997).

Expression of EVI1

An alternative form of the transcription factor EVI1, MDS1-EVI1, which previously had been believed to exist only in the context of leukemic fusion mRNAs, has recently been shown to be expressed also in normal human tissues. Moreover, it acts as an antagonist of EVI1, activating transcription of reporter constructs repressed by EVI1. The murine homolog of MDS1-EVI1 as well as mMds1 have been cloned; localization of mMds1 close to mEvi1 on chromosome 3 is demonstrated. Expression of both Evi1 forms is widespread in the adult mouse: they are upregulated during in vitro hematopoietic differentiation. These data underscore the biological importance of both EVI1 and MDS1-EVI1 and provide the basis for further studies of their function in the mouse model system (Wimmer, 1998).

EVI1 mutation

The ecotropic viral integration site-1 (Evi1) locus was initially identified as a common site of retroviral integration in myeloid tumors of the AKXD-23 recombinant inbred mouse strain. The full-length Evi1 transcript encodes a putative transcription factor, containing ten zinc finger motifs found within two domains of the protein. To determine the biological function of the Evi1 proto-oncogene, the full-length, but not an alternately spliced, transcript was disrupted using targeted mutagenesis in embryonic stem cells. Evi1 homozygous mutant embryos die at approximately 10.5 days of development. Mutants are distinguished at 10.5 days of development by widespread hypocellularity, hemorrhaging, and disruption in the development of paraxial mesenchyme. In addition, defects in the heart, somites, and cranial ganglia are detected and the peripheral nervous system fails to develop. These results correlate with whole-mount in situ hybridization analyses of embryos that show expression of the Evi1 proto-oncogene in embryonic mesoderm and neural crest-derived cells associated with the peripheral nervous system. These data suggest that Evi1 has important roles in general cell proliferation, vascularization, and cell-specific developmental signaling, at midgestation (Hoyt, 1997).

Viral insertions associated with Evi-1

Normal hematopoietic stem cells proliferate and differentiate in the presence of growth factors such as interleukin-3 (IL-3). Transformation can alter their growth factor requirements, the ability of the cells to differentiate, or both. To identify genes that are capable of transforming hematopoietic cells, IL-3-dependent cell lines, isolated from retrovirus induced myeloid leukemias, were examined for viral insertions in proto-oncogenes and in common sites of viral integration. Five of 37 cell lines contained proviruses in a common viral integration site termed the ecotropic virus integration 1 site (Evi-1). The integrations were correlated with the activation of transcription from the locus. Sequencing of cDNA clones and genomic clones have demonstrated that the integrations had occurred near or in 5' noncoding exons of a novel gene. The sequence of the cDNA clones predicts that the gene product is a 120 kd protein that contains two domains with seven and three repeats of a DNA binding consensus sequence (zinc finger) initially described in the Xenopus transcription factor III A (TFIIIA). This represents the first demonstration of the retroviral activation of a gene encoding a zinc finger protein and the first implication for a member of this gene family in the transformation of hematopoietic cells (Morishita, 1988).

Growth response of EVI1 and its leukemia promoting fusion proteins

Expression of the Evi-1 gene is frequently activated in murine myeloid leukemias by retroviral insertions immediately 5' or 90 kb 5' of the gene. The Evi-1 gene product is a nuclear, DNA-binding zinc finger protein of 145 kDa. On the basis of the properties of the myeloid cell lines in which the Evi-1 gene is activated, it has been hypothesized that its expression blocks normal differentiation. To explore this proposed role, a retrovirus vector containing the gene was constructed and its effects were examined on an interleukin-3-dependent myeloid cell line that differentiates in response to granulocyte colony-stimulating factor (G-CSF). Expression of the Evi-1 gene in these cells does not alter the normal growth factor requirements of the cells. However, expression of the Evi-1 gene blocks the ability of the cells to express myeloperoxidase and to terminally differentiate to granulocytes in response to G-CSF. This effect is not due to altered expression of the G-CSF receptor or to changes in the initial responses of the cells to G-CSF. These results support the hypothesis that the inappropriate expression of the Evi-1 gene in myeloid cells interferes with the ability of the cells to terminally differentiate (Morishita, 1992).

Inappropriate expression of the Evi-1 zinc-finger gene in hematopoietic cells has been associated with acute myelogenous leukemia and myelodysplastic syndromes in murine models and in humans. Consistent with this, aberrant expression of the Evi-1 gene in a myeloid progenitor cell line blocks granulocytic differentiation. The aberrant expression of the Evi-1 gene impairs the normal response of erythroid cells or bone-marrow progenitors to erythropoietin. Erythroid differentiation has been shown to require the GATA-1 transcription factor that binds to a sequence contained within the consensus binding sequence identified for Evi-1. Evi-1 can repress GATA-1-dependent transactivation in transient chloramphenicol acetyltransferase assays. Together the data support the hypothesis that inappropriate expression of the Evi-1 gene blocks erythropoiesis by repressing the transcription of a subset of GATA-1 target genes (Kreider, 1993).

Chromosome band 3q26 is the locus of two genes, MDS1/EVI1 and EVI1. The proteins encoded by these genes are nuclear factors each containing two separate DNA-binding zinc finger domains. The proteins are identical, aside from the N-terminal extension of MDS1/EVI1, which is missing in EVI1. However, they have opposite functions as transcription factors. In contrast to MDS1/EVI1, EVI1 is often activated inappropriately by chromosomal rearrangements at 3q26 leading to inappropriate expression of the protein in hematopoietic cells and to myeloid leukemias, which are often characterized by abnormal megakaryopoiesis. The two proteins affect replication and differentiation of progenitor hematopoietic cell lines in opposite ways: whereas EVI1 inhibits the response of 32Dc13 cells to G-CSF and TGFbeta1, MDS1/EVI1 has no effect on the G-CSF-induced differentiation of the 32Dc13 cells, and it enhances the growth-inhibitory effect of TGFbeta1. The endogenous expression of the two genes during in vitro hematopoietic differentiation of murine embryonic stem (ES) cells has been analyzed and the effects have been evaluated of their forced expression on the ability of ES cells to produce differentiated hematopoietic colonies. The expression of the two genes is found to be independently and tightly controlled during differentiation. In addition, the forced expression of EVI1 leads to a much higher rate of cell growth before and during differentiation, whereas the expression of MDS1/EVI1 represses cell growth and strongly reduces the number of differentiated hematopoietic colonies. Finally, it was found that the forced expression of EVI1 results in the differentiation of abnormally high numbers of megakaryocytic colonies, thus providing one of the first experimental models showing a clear correlation between inappropriate expression of EVI1 and abnormalities in megakaryopoiesis (Sitailo, 1999).

MDS1/EVI1, located on chromosome 3 band q26, encodes a zinc-finger DNA-binding transcription activator not detected in normal hematopoietic cells but expressed in several normal tissues. MDS1/EVI1 is inappropriately activated in myeloid leukemias following chromosomal rearrangements involving band 3q26. The rearrangements led either to gene truncation, and to expression of the transcription repressor EVI1, or to gene fusion which results in the fusion protein AML1/MDS1/EVI1. This fusion protein contains the DNA-binding domain of the transcription factor AML1 fused in-frame to the entire MDS1/EVI1 with the exclusion of its first 12 amino acids. Analysis has been made of the response of the hematopoietic precursor cell line 32Dcl3, expressing either the normal protein MDS1/EVI1 or the fusion protein AML1/MDS1/EVI1, to factors that control cell differentiation or cell replication. The 32Dcl3 cells are IL-3-dependent for growth and they differentiate into granulocytes when exposed to G-CSF. They are growth-inhibited by TGF-beta1. Whereas the expression of MDS1/EVI1 has no effect on granulocytic differentiation induced by G-CSF, expression of AML1/MDS1/EVI1 blocks differentiation resulting in cell death. This effect is similar to that for 32Dcl3 cells that express transgenic Evil. Furthermore, whereas the expression of the fusion protein AML1/MDS1/EVI1 completely abrogates the growth-inhibitory effect of TGF-beta1 and allows 32Dcl3 cells to proliferate, expression of the normal protein MDS1/EVI1 has the opposite effect, and it strengthens the response of cells to the growth-inhibitory effect of TGF-beta1. EVI1 (contained in its entirety in MDS1/EVI1 and AML1/MDS1/EVI1) is shown to physically interact with SMAD3, which is an intracellular mediator of TGF-beta1 signaling. The response of the cells to G-CSF or TGF-beta1 was corrolated with the ability of the normal and fusion proteins to activate or repress promoters which they can directly regulate by binding to the promoter site. It is proposed that mutations of MDS1/EVI1 either by gene truncation resulting in the transcription repressor EVI1 or by gene fusion to AML1 lead to an altered cellular response to growth and differentiation factors that could result in leukemic transformation. The different response of myeloid cells ectopically expressing the normal or the fusion protein to G-CSF and TGF-beta1 could depend on the different transactivation properties of these proteins resulting in divergent expression of downstream genes regulated by the two proteins (Sood, 1999).

The human t(3;21)(q26;q22) translocation is found as a secondary mutation in some cases of chronic myelogenous leukemia during the blast phase and in therapy-related myelodysplasia and acute myelogenous leukemia. One result of this translocation is a fusion between the AML1, MDS1, and EVI1 genes, that encodes a transcription factor of approximately 200 kDa. The role of the AML1/MDS1/EVI1 (AME) fusion gene in leukemogenesis is largely unknown. In this study, the effect of the AME fusion gene was analzyed in vivo by expressing it in mouse bone marrow cells via retroviral transduction. Mice transplanted with AME-transduced bone marrow cells suffer from an acute myelogenous leukemia (AML) 5-13 mo after transplantation. The disease can be readily transferred into secondary recipients with a much shorter latency. Morphological analysis of peripheral blood and bone marrow smears demonstrates the presence of myeloid blast cells and differentiated but immature cells of both myelocytic and monocytic lineages. Cytochemical and flow cytometric analysis confirms that these mice have a disease similar to the human acute myelomonocytic leukemia. This murine model for AME-induced AML will help dissect the molecular mechanism of AML and the molecular biology of the AML1, MDS1, and EVI1 genes (Cuenco, 2000).

3q21q26 syndrome, an acute myeloid leukemia (AML) or myelodysplastic syndrome (MDS) with chromosomal translocations or inversions between the bands 3q21 and 3q26, is frequently associated with dysmegakaryocytopoiesis and increased platelet counts at the initial diagnosis. Since the EVI1 gene at 3q26 is transcriptionally activated in 3q21q26 syndrome, the role of EVI1 gene expression in the abnormal megakaryocytic differentiation in 3q21q26 syndrome was assessed. RT-PCR analysis of various types of hematopoietic cells revealed that the EVI1 gene is expressed specifically in CD34(+) cells, megakaryocytes, and platelets. UT-7 is a human immature megakaryoblastic leukemia cell line with dependence for the growth on granulocyte-macrophage colony-stimulating factor (GM-CSF) (designated at UT-7/GM) and with a differentiation capacity to erythroid (UT-7/EPO) and megakaryocytic lineages (UT-7/TPO) by erythropoietin (EPO) and thrombopoietin (TPO), respectively. Among three UT-7 sublines, UT-7/GM, UT-7/EPO, and UT-7/TPO, expression of the EVI1 gene was detected at low levels in UT-7/GM and UT-7/EPO cells, but was detected at a higher level in UT-7/TPO cells. When UT-7/GM cells are cultured with TPO, the level of EVI1 expression is increased, along with increased numbers of polynuclear megakaryocytes and expression of the platelet factor 4 (PF-4) gene. Furthermore, forced expression of the EVI1 gene in UT-7/GM cells changes their morphology to polynuclear megakaryocytes, stops their growth, and induces cell death within a month. These data indicate that expression of the EVI1 gene is involved in progression of megakaryocytic differentiation and, thus, the dysmegakaryocytopoiesis in 3q21q26 syndrome could be partly due to an enhanced differentiation capacity of leukemia cells and/or megakaryocytes by constitutive expression of the EVI1 gene (Shimizu, 2002).

A single oncogenic enhancer rearrangement causes concomitant EVI1 and GATA2 deregulation in leukemia

Chromosomal rearrangements without gene fusions have been implicated in leukemogenesis by causing deregulation of proto-oncogenes via relocation of cryptic regulatory DNA elements. AML with inv(3)/t(3;3) is associated with aberrant expression of the stem-cell regulator EVI1. Applying functional genomics and genome-engineering, this study demonstrates that both 3q rearrangements reposition a distal GATA2 enhancer to ectopically activate EVI1 and simultaneously confer GATA2 functional haploinsufficiency, previously identified as the cause of sporadic familial AML/MDS and MonoMac/Emberger syndromes. Genomic excision of the ectopic enhancer restored EVI1 silencing and led to growth inhibition and differentiation of AML cells, which could be replicated by pharmacologic BET inhibition. These data show that structural rearrangements involving the chromosomal repositioning of a single enhancer can cause deregulation of two unrelated distal genes, with cancer as the outcome (Groschel, 2014).

REFERENCES

Search PubMed for articles about Drosophila hamlet

Bartholomew, C., Kilbey, A., Clark, A. M. and Walker, M. (1997). The Evi-1 proto-oncogene encodes a transcriptional repressor activity associated with transformation. Oncogene 14(5): 569-77. 9053855

Baum, P. D., et al. (1999). The Caenorhabditis elegans gene ham-2 links Hox patterning to migration of the HSN motor neuron. Genes Dev 13: 472-483. 10049362

Chakraborty, S., Senyuk, V., Sitailo, S., Chi, Y. and Nucifora, G. (2001). Interaction of EVI1 with cAMP-responsive element-binding protein-binding protein (CBP) and p300/CBP-associated factor (P/CAF) results in reversible acetylation of EVI1 and in co-localization in nuclear speckles. J Biol. Chem. 276(48): 44936-43. 11568182

Cuenco, G. M., Nucifora, G. and Ren, R. (2000). Human AML1/MDS1/EVI1 fusion protein induces an acute myelogenous leukemia (AML) in mice: a model for human AML. Proc. Natl. Acad. Sci. 97(4): 1760-5. 10677531

Endo, K,, et al. (2011). Chromatin modification of Notch targets in olfactory receptor neuron diversification. Nat. Neurosci. 15(2): 224-33. PubMed Citation: 22197833

Eroglu, E., Burkard, T. R., Jiang, Y., Saini, N., Homem, C. C., Reichert, H. and Knoblich, J. A. (2014). SWI/SNF complex prevents lineage reversion and induces temporal patterning in neural stem cells. Cell 156: 1259-1273. PubMed ID: 24630726

Fears, S., et al. (1996). Intergenic splicing of MDS1 and EVI1 occurs in normal tissues as well as in myeloid leukemia and produces a new member of the PR domain family. Proc. Natl. Acad. Sci. 93(4): 1642-7. 8643684

Funabiki, T., Kreider, B. L. and Ihle, J. N. (1994). The carboxyl domain of zinc fingers of the Evi-1 myeloid transforming gene binds a consensus sequence of GAAGATGAG. Oncogene 9(6): 1575-81. 8183551

Garriga, G., et al. (1993). Migrations of the Caenorhabditis elegans HSNs are regulated by egl-43, a gene encoding two zinc finger proteins. Genes Dev. 7: 2097-2109. 8224840

Groschel, S., et al. (2014). A single oncogenic enhancer rearrangement causes concomitant EVI1 and GATA2 deregulation in leukemia. Cell 157: 369-381. PubMed ID: 24703711

Grueber, W. B., et al. (2003). Dendrites of distinct classes of Drosophila sensory neurons show different capacities for homotypic repulsion. Curr. Biol. 13: 618-626. 12699617

Hoyt, P. R., et al. (1997). The Evi1 proto-oncogene is required at midgestation for neural, heart, and paraxial mesenchyme development. Mech. Dev. 65(1-2): 55-70. 9256345

Izutsu, K., et al. (2001). The corepressor CtBP interacts with Evi-1 to repress transforming growth factor beta signaling. Blood 97(9): 2815-22. 11313276

Kilbey, A. and Bartholomew, C. (1998). Evi-1 ZF1 DNA binding activity and a second distinct transcriptional repressor region are both required for optimal transformation of Rat1 fibroblasts. Oncogene 16(17): 2287-91. 9619838

Kreider, B. L., Orkin, S. H. and Ihle, J. N. (1993). Loss of erythropoietin responsiveness in erythroid progenitors due to expression of the Evi-1 myeloid-transforming gene. Proc. Natl. Acad. Sci. 90(14): 6454-8. 8341654

Kurokawa, M., et al. (1998). The oncoprotein Evi-1 represses TGF-beta signalling by inhibiting Smad3. Nature 394(6688): 92-6. 9665135

Moore, A. W., Jan, L. Y. and Jan, Y. N. (2002). hamlet, a binary genetic switch between single- and multiple- dendrite neuron morphology. Science 297: 1355-1358. 12193790

Moore, A. W., Roegiers, F., Jan, L. Y. and Jan, Y.-N. (2004). Conversion of neurons and glia to external-cell fates in the external sensory organs of Drosophila hamlet mutants by a cousin-cousin cell-type respecification. Genes Dev. 18: 623-628. 15075290

Morishita, K., et al. (1988). Retroviral activation of a novel gene encoding a zinc finger protein in IL-3-dependent myeloid leukemia cell lines. Cell 54(6): 831-40. 2842066

Morishita, K., Parganas, E., Matsugi, T. and Ihle, J. H. (1992). Expression of the Evi-1 zinc finger gene in 32Dc13 myeloid cells blocks granulocytic differentiation in response to granulocyte colony- stimulating factor. Mol. Cell. Biol. 12: 183-189. 1370341

Palmer, S., et al. (2001). Evi-1 transforming and repressor activities are mediated by CtBP co-repressor proteins. J. Biol. Chem. 276(28): 25834-40. 11328817

Senyuk, V., et al. (2002). The leukemia-associated transcription repressor AML1/MDS1/EVI1 requires CtBP to induce abnormal growth and differentiation of murine hematopoietic cells. Oncogene 21(20): 3232-40. 12082639

Senyuk, V., et al. (2003). P/CAF and GCN5 acetylate the AML1/MDS1/EVI1 fusion oncoprotein. Biochem. Biophys. Res. Commun. 307(4): 980-6. 12878208

Shimizu, S., et al. (2002). EVI1 is expressed in megakaryocyte cell lineage and enforced expression of EVI1 in UT-7/GM cells induces megakaryocyte differentiation. Biochem. Biophys. Res. Commun. 292(3): 609-16. 11922610

Sitailo, S., Sood, R., Barton, K. and Nucifora, G. (1999). Forced expression of the leukemia-associated gene EVI1 in ES cells: a model for myeloid leukemia with 3q26 rearrangements. Leukemia 13(11): 1639-45. 10557037

Soderholm, J., et al. (1997). The leukemia-associated gene MDS1/EVI1 is a new type of GATA-binding transactivator. Leukemia 11(3): 352-8. 9067573

Sood, R., et al. (1999). MDS1/EVI1 enhances TGF-beta1 signaling and strengthens its growth-inhibitory effect but the leukemia-associated fusion protein AML1/MDS1/EVI1, product of the t(3;21), abrogates growth-inhibition in response to TGF-beta1. Leukemia 13(3): 348-57. 10086725

Vinatzer, U., Taplick, J., Seiser, C., Fonatsch, C. and Wieser, R. (2001). The leukaemia-associated transcription factors EVI-1 and MDS1/EVI1 repress transcription and interact with histone deacetylase. Br. J. Haematol. 114(3): 566-73. 11552981

Wimmer, K., et al. (1998). Comparative expression analysis of the antagonistic transcription factors EVI1 and MDS1-EVI1 in murine tissues and during in vitro hematopoietic differentiation. Biochem. Biophys. Res. Commun. 252(3): 691-6. 9837768