grainy head: Biological Overview | Evolutionary Homologs | Regulation | Developmental Biology | Effects of Mutation | References

Gene name - grainy head

Synonyms - Ntf1 and Elf1

Cytological map position - 54F1-2

Function - transcription factor

Keyword(s) - ectodermal transcriptional cofactor

Symbol - grh

FlyBase ID: FBgn0259211

Genetic map position - 2-86

Classification - bHLH

Cellular location - nuclear

NCBI links: Precomputed BLAST | Entrez Gene
Recent literature
Nevil, M., Bondra, E. R., Schulz, K. N., Kaplan, T. and Harrison, M. M. (2016). Stable binding of the conserved transcription factor Grainy head to Its target genes throughout Drosophila melanogaster development. Genetics [Epub ahead of print]. PubMed ID: 28007888
It has been suggested that transcription factor binding is temporally dynamic and that changes in binding determine transcriptional output. Nonetheless, this model is based on relatively few examples in which transcription factor binding has been assayed at multiple developmental stages. The essential transcription factor Grainy head is conserved from fungi to humans and controls epithelial development and barrier formation in numerous tissues. Drosophila melanogaster, which possess a single grainy head gene, provide an excellent system to study this conserved factor. To determine whether temporally distinct binding events allow Grainy head to control cell fate specification in different tissue types, a combination of ChIP-seq and RNA-seq was used to elucidate the gene regulatory network controlled by Grainy head during four stages of embryonic development (spanning stages 5 - 17) and in larval tissue. Contrary to expectations, Grainy head was found to remain bound to at least 1146 genomic loci over days of development. In contrast to this stable DNA occupancy, the subset of genes whose expression is regulated by Grainy head varies. Grainy head transitions from functioning primarily as a transcriptional repressor early in development to functioning predominantly as an activator later. The data reveal that Grainy head binds to target genes well before the Grainy head-dependent transcriptional program commences, suggesting it sets the stage for subsequent recruitment of additional factors that execute stage-specific Grainy head functions.
Yao, L., Wang, S., Orzechowski-Westholm, J., Dai, Q., Matsuda, R., Hosono, C., Bray, S., Lai, E. C. and Samakovlis, C. (2017). Genome-wide identification of Grainy head targets in Drosophila reveals regulatory interactions with the POU-domain transcription factor, Vvl. Development [Epub ahead of print]. PubMed ID: 28760809
Grainy head (Grh) is a conserved transcription factor (TF) controlling epithelial differentiation and regeneration. To elucidate Grh functions, embryonic Grh targets were identified by ChIP-seq and gene expression analysis. Grh was shown to control hundreds of target genes. Repression or activation correlates with the distance of Grh binding sites to the transcription start sites of its targets. Analysis of 54 Grh-responsive enhancers during development and upon wounding suggests cooperation with distinct TFs in different contexts. In the airways, Grh repressed genes encode key TFs involved in branching and cell differentiation. Reduction of the POU-domain TF, Vvl, (ventral veins lacking) largely ameliorates the airway morphogenesis defects of grh mutants. Vvl and Grh proteins additionally interact with each other and regulate a set of common enhancers during epithelial morphogenesis. It is concluded that Grh and Vvl participate in a regulatory network controlling epithelial maturation.

Interest in grainy head function will vary, depending on one's developmental perspective. For the biologists who take a classical approach, studying morphological or phenotypic results caused by mutation, grainy head is not the most exciting gene in the lab. But over the past half dozen years, a second developmental perspective has gained adherents, and made meticulous, if understated, contributions to an understanding of developmental processes by revealing the inner workings of gene regulation. Grainy head has been used as a tool to reveal the structure of the general transcription apparatus involved in gene regulation, one cog in a vast machine used to transcribe genes into messenger RNA, into proteins, and ultimately, into the mature organism. The difference in perspective is the difference between wanting to know how the engine operates versus where the engine will carry the organism. Both are valid. morphology: reads like shopping list of disparate activities Grainy head is synthesized during oogenesis and deposited in the developing oocyte to await the call to action. Grainy head maternal mRNA contributes to ventral repression during early embryogenesis.

The earliest involvement of GRH in development must be in the regulation of tailless downstream of Torso. Here GRH acts as a cofactor with GAGA (Liaw, 1995). GRH also acts as a cofactor of the transcription factor Dorsal in the repression of dpp and zerknüllt as mentioned above. As a repressor GRH acts as a downstream target of both Torso and Toll receptor tyrosine kinase pathways. Activation of either pathway results in a phosphorylation cascade; MAP kinase is responsible for attaching the phosphates to GRH. Said another way, GRH is a substrate of MAP kinase. (Liaw, 1995 and Huang, 1995).

The role of GRH in tailless regulation is both critical and complex. grainy head mutants show an enlargment of the tailless range of transcription. GRH and GAGA/Trithorax-like act at the Torso response element of tailless. GAGA is a transcriptional activator, but the GRH-GAGA interaction produces repression. It is concluded that GAGA is involved in relief of GRH-caused repression. Signaling through Torso results in the phosphorylation of GRH and its consequent inactivation (Liaw, 1995).

Overexpression of grainy head in the postblastoderm embryo results in a phenotype consistent with its role in the repression of dpp and zerknüllt later in embryogenesis (Huang, 1995). In spite of all these observed regulatory interactions, mutants of grainy head have only a minimal effect on fly morphology, in particular, an effect on cuticular morphology.

In contrast to its role as a repressor, as originally characterized, GRH is also a transcriptional activator: Grainy head regulates genes involved in epidermal development, including Ultrabithorax, engrailed, and fushi tarazu. An incredible amount of work has been expended in understanding the mechanism by which GRH acts as a transcriptional activator. This work has paid off grandly in an understanding of the basic apparatus for transcriptional activation. Studies of the role of GRH in transcriptional activation have resulted in the isolation of coactivators associated with the TATA-binding protein that mediates transcriptional activation (Dynlacht, 1991).

The activities of Grainyhead and other members of the family appear to be modulated so that they can participate in different developmental processes. The structure and function of mRNAs from the Drosophila grainyhead gene have been examined. Alternate splicing is responsible for generating a neuroblast-specific isoform of the protein. The N transcript of grh is expressed in larval epidermis, trachea, and foregut as well as in imaginal discs and optic lobes. However the mRNA detected by the O-specific fragment is expressed in the CNS, where the O-specific fragment is found at high levels in neuroblasts but not in the developing optic lobes. Similar results have been achieved in embryos, where O mRNA is predominantly detected in the neuroblasts of the CNS (Uv, 1997).

A mutation which abolishes the O isoform results in pupal and adult lethality and in the absence of Grh protein from neuroblasts. In other places, such as the epidermis, head skeleton, trachea, anterior spiracles, discs, and the foregut, Grh levels appear normal. The specific mutant has been localized to a base pair change in exon 5. Reporter genes containing different Grainyhead binding sites exhibit tissue-specific patterns of expression that correlate with the Grainyhead isoforms, suggesting that the alternate splicing serves to alter the repertoire of target genes controlled in the neuroblasts. Two Grh binding elements (be1 and be2) derived from the Dopa decarboxylase gene direct grh expression to different cells. be1 activity is detected in the optic lobes, neuroblasts and tracheal cells, whereas reporter be2 is detected only in the neuroblasts. The expression of both reporters is dependent on Grh function, implying that Grh can activate expression of reporter be2 in a tissue-specific manner: this correlates with the expression pattern of the O isoform of GRH. Since the be1 and be2 sites differ (be1 is palindromic and be2 is nonpalindromic), it is suggested that be1 and be2 represent targets for homo- and heterodimers of Grh, respectively, and that the O isoform preferentially forms heterodimers with an unknown partner. Similarly, regulation of globin genes by mammalian Grh homolog CP2 requires complex with a second, tissue-specific factor, which is as yet not cloned but which confers tissue and stage specificity on CP2 function (Uv, 1997).

Drosophila Grainyhead specifies late programmes of neural proliferation by regulating the mitotic activity and Hox-dependent apoptosis of neuroblasts

The Drosophila central nervous system is generated by stem-cell-like progenitors called neuroblasts. Early in development, neuroblasts switch through a temporal series of transcription factors modulating neuronal fate according to the time of birth. At later stages, it is known that neuroblasts switch on expression of Grainyhead (Grh) and maintain it through many subsequent divisions. The function of this conserved transcription factor is to specify the regionalised patterns of neurogenesis that are characteristic of postembryonic stages. In the thorax, Grh prolongs neural proliferation by maintaining a mitotically active neuroblast. In the abdomen, Grh terminates neural proliferation by regulating the competence of neuroblasts to undergo apoptosis in response to Abdominal-A expression. This study shows how a factor specific to late-stage neural progenitors can regulate the time at which neural proliferation stops, and identifies mechanisms linking it to the Hox axial patterning system (Cenci, 2005).

Thoracic neuroblasts normally continue dividing into pupal stages, stopping at ~120 hours, by which time ~100 adult-specific neurons have been generated. By compromising grh function, it was observed that neurogenesis ceases two days prematurely, at ~72 hours. This limits the average size of neuroblast clones to ~30 cells, indicating that Grh is required to generate 70% of all adult-specific neurons in the thorax (Cenci, 2005).

Four lines of evidence are provided suggesting that the underlying basis for premature cessation of thoracic proliferation in grh mutant clones is reduced mitotic activity of the neuroblast, most probably followed by Hox-independent apoptosis. (1) Although grh mutant neuroblasts are present at 72 hours they are mitotically inactive; (2) by 96 hours, no recognisable grh mutant neuroblasts remain; (3) inhibiting cell-death effector caspases by misexpressing P35 rescues the loss of grh mutant neuroblasts; (4) although misexpression of Hox proteins in thoracic neuroblasts induces apoptosis, Ubx, the resident Hox protein of the posterior thorax, remains excluded from grh mutant neuroblasts at 72 hours. Importantly, the role of Grh in maintaining mitotically-active neuroblasts is not a general 'housekeeping' function but is specific for their age. Thus, wild-type neuroblasts in the early embryo are Grh-negative yet viable and actively dividing. This observation suggests that the late switch to Grh-dependency involves additional factors. These could be intrinsic to the neuroblast or provided by a glial-cell niche. Consistent with the niche idea, neuroblast divisions within the postembryonic brain require DE-cadherin-dependent interactions between glia and neural cells (Cenci, 2005).

In the central abdomen, it has been found that, at 72 hours, many neuroblasts downregulate Grh and become TUNEL positive. When the neuroblast death pathway is blocked in H99 clones, Grh expression continues in mitotically active neuroblasts long after the 72-hour stage. This indicates that abdominal neuroblasts remain in Grh-positive mode during their final division and that Grh is only downregulated after the onset of apoptosis. Moreover, loss of Grh activity leads to the failure of neuroblasts to undergo apoptosis. As these persistent neuroblasts not only survive but also remain actively engaged in the cell cycle, they generate a 3.5-fold excess of cells within each abdominal neuroblast lineage. Together, these findings identify Grh as a terminal neuroblast factor that is an essential component of the abdomen-specific 'stop' programme (Cenci, 2005).

Two different interactions with the Hox gene AbdA underlie the dramatic reversal of Grh function from pro-proliferative in the thorax to anti-proliferative in the abdomen. (1) Grh acts upstream of AbdA to maintain its late phase of expression, and (2) it functions in parallel with AbdA to activate apoptosis. Although the functional significance of grh-dependent AbdA maintenance is not clear, it may be that efficient neuroblast apoptosis requires AbdA levels to remain high for a significant proportion of the interval separating initial AbdA upregulation and the TUNEL-positive stage. More definitively, epistasis tests were used to show that Grh, acting in parallel with AbdA activity, is essential for abdominal neuroblast apoptosis. Thus, when the AbdA-maintenance deficit is rescued using hs-AbdA, neuroblast death remains blocked. Since AbdA is not required to activate neuroblast Grh expression, Grh and AbdA must work in parallel to activate apoptosis. Together with the finding that AbdA is required to activate H99 gene activity, this study demonstrates that inputs from Grh and AbdA are both essential to activate proapoptotic genes and thus trigger neuroblast apoptosis. Whereas the late upregulation of AbdA provides a timing cue to schedule the onset of apoptosis, the much broader phase of Grh expression defines the period of neuroblast competence to respond appropriately to it (Cenci, 2005).

The restricted temporal pattern of Grh expression ensures that competence to undergo AbdA-dependent apoptosis, rather than some other AbdA-dependent output, is only installed at late stages. Consistent with this, neuroblasts in the early embryo that are AbdA positive but Grh negative go on to generate substantial embryonic lineages. Low levels of expression from UAS-grh transgenes make it difficult to test whether Grh is sufficient to confer apoptotic competence to these early embryonic neuroblasts. In the late embryo, however, neuroblasts have already switched on Grh, and, within the central abdomen, all but three undergo abdA-dependent death The observation that reduced neural grh function leads to supernumerary postembryonic neuroblasts positioned outside the vm, vl and dl rows, raises the possibility that Grh is required for all developmentally programmed neuroblast apoptosis (Cenci, 2005).

Cooperation between Su(H) and Grainyhead

Cell-cell signaling mediated by Notch is critical during many different developmental processes for the specification or restriction of cell fates. Currently, the only known transduction pathway involves a DNA binding protein, Suppressor of Hairless [Su(H)] in Drosophila and CBF1 in mammals, and results in the direct activation of target genes. It has been proposed that in the absence of Notch, Su(H)/CBF1 acts as a repressor and is converted into an activator through interactions with the Notch intracellular domain (Morel, 2000). It has also been suggested that the activation of specific target genes requires synergy between Su(H) and other transcriptional activators. An assay has been designed that allows a direct test of these hypotheses in vivo. The results clearly demonstrate that Su(H) is able to function as the core of a molecular switch, repressing transcription in the absence of Notch and activating transcription in the presence of Notch. In its capacity as an activator, Su(H) can cooperate synergistically with a DNA-bound transcription factor, Grainyhead. These interactions indicate a simple model for Notch target-gene regulation that could explain the precision of gene activation elicited by Notch signaling in different developmental fate decisions (Furriols, 2001).

Activation of Notch by its ligands promotes proteolytic processing, releasing an intracellular fragment (Nicd) that embodies most functions of the activated receptor. There is substantial in vivo and in vitro evidence demonstrating that Su(H) and its homologs in other species are required for the activation of Notch target genes, such as the Enhancer of split/HES genes, and it is proposed that Su(H) DNA binding proteins cooperate with Nicd to promote transcription. However, Su(H) and Nicd are relatively ineffectual at activating Enhancer of split [E(spl)] genes in ectopic locations and it appears that their capacity to promote transcription of specific target genes requires synergistic interactions with other enhancer-specific factors. Cell transfection assays have also revealed a potential repressive role for the mammalian homolog of Su(H), CBF1, and indicate that in the absence of Nicd, Su(H)/CBF1 could recruit a corepressor complex to shut off target genes. Recent work in Drosophila has supported this model through the analysis of single-minded, one target gene whose expression is derepressed in animals that lack Su(H) function (Furriols, 2001).

An assay has been designed that allows investigation of whether this is a general mechanism by first testing whether Su(H) can mediate repression of a heterologous activator, and second, whether it can synergize with the same activator in the presence of Nicd to promote transcription (Furriols, 2001).

In order to assess whether Su(H) is able to function as a repressor as well as an activator, it was necessary to target it to a well-defined enhancer that independently confers widespread expression. Through work on the Grainyhead (Grh) transcription factor, a palindromic binding site (Gbe) has been defined that, when combined in three copies with a minimal promoter, confers expression throughout the imaginal discs, epidermis, and trachea of the Drosophila larvae. Since Su(H) is expressed ubiquitously, it was anticipated that when Su(H) sites are combined with Gbe, Su(H) would cooperate with Grh to yield high levels of expression in the cells where Notch is active (e.g., dorsal/ventral boundary, interveins in the wing imaginal disc) and would prevent Grh-mediated activation in cells where Notch is inactive (e.g., larval epidemis, where there is no evidence for Notch activity based on expression patterns of known Notch target genes (Furriols, 2001).

The Su(H) binding sites used were the paired sites derived from the regulatory region of the Enhancer of split m8 gene, which is primarily expressed in association with proneural clusters in the imaginal discs. On their own, two pairs of Su(H)m8 sites only give extremely limited activity; patchy expression was detected at the wing disc dorsal/ventral boundary and the tracheal branchpoints [Su(H)m8]. In contrast, the Su(H)m8 sites have a dramatic effect when combined with three copies of Gbe [Gbe+ Su(H)m8]. In the imaginal discs, strong activation is detected in a pattern reminiscent of the most widely expressed Notch target gene, Enhancer of split mß [E(spl)mß]. Expression also occurs at tracheal branchpoints in a similar manner to E(spl)mß suggesting that this, too, is a site of Notch activity (Furriols, 2001).

The activation was coupled with apparent inhibition of Gbe-driven expression in some patches in the discs that correspond to the places where E(spl)mß is also silent. More definitive, however, is the effect in the epidermis and the trachea. The widespread expression throughout these tissues that is normally elicited by Gbe is shut off, while the activation at the tracheal branchpoints is enhanced. Similar results were obtained using a single copy of the paired Su(H)m8 site. This construct has virtually no expression on its own but give an E(spl)mß-like pattern in the discs with Gbe and in two out of six lines represses Gbe-derived expression in the epidermis and trachea. Overall, the patterns obtained with Gbe+ Su(H)m8 indicate first that Grh and Su(H) can cooperate synergistically to confer high levels of transcription in places known to have Notch activity and second that Su(H) is able to repress the Grh activation function in regions without Notch activity. Intriguingly, the resulting pattern strongly resembles that of E(spl)mß, although, since no evidence is as yet available that Grh normally confers this expression, Su(H) may synergize with a different activator on this E(spl)mß enhancer. It is important to note, however, that neither the synergy nor the repressive effects imply direct interactions between Su(H) and the DNA-bound activators. Based on the experiments with CBF-1, it is likely that Su(H) exerts its effects through the recruitment of cofactors, which probably include chromatin-modifying enzymes such as histone deacetylases (Furriols, 2001).

To confirm that the effects of adding the paired Su(H)m8 sites to Gbe are due to the activity and not simply to the length of the Su(H) sequences inserted, a similar construct was generated in which the Su(H)m8 sites had been mutated by substituting critical bases in the recognition sequence. The resulting transgene [Gbe+ Su(H)MUT] has an expression pattern similar to the parental Gbe sites alone, although the levels of expression are reduced. Since these mutations restore the widespread activity of the enhancer, it must be the Su(H) sequence per se that confers the activation and repression detected with Gbe+ Su(H)m8 (Furriols, 2001).

If this interpretation is correct and the E(spl)mß -like expression from Gbe+ Su(H)m8 reflects a synergistic interaction between Su(H)/Nicd and Grh, this expression should be dependent on Notch and Su(H). Reducing Su(H) activity [Su(H)SF8] in clones of cells in the wing disc leads to an autonomous loss of the high levels of expression from the mutant cells. Likewise, reducing Notch activity using a temperature sensitive combination (Nts1/N55e11) at the nonpermissive temperature also eliminates the intervein pattern and reduces the dorsal/ventral boundary expression of Gbe+ Su(H)m8. Thus, the strong disc expression requires Notch and Su(H) and, as it is not seen with the Su(H)m8 sites alone, must involve cooperation with Gbe-bound protein (Furriols, 2001).

Similar experiments were carried out to assess whether repression depends on Su(H). In this case, it would be anticipated that mutations in Su(H) should cause derepression, restoring Gbe-mediated epidermal expression, whereas mutations in Notch should not. In Su(H) mutant animals [Su(H)SF8/Su(H)AR9], there is widespread expression from Gbe+ Su(H)m8 throughout the epidermis and tracheal cells, and the strong activation at tracheal branchpoints is lost. In contrast, there is no expression in the epidermal cells or most tracheal cells when Notch function is reduced. In the latter case, the activity at the branchpoints is reduced, as in Su(H) mutants, consistent with this being Notch-dependent activation, but there is no derepression in the other tracheal cells. Clearly, the transgene can be expressed in a similar pattern to the parental Gbe when there is little or no Su(H) protein present, confirming, therefore, that Su(H) is critical for the repression. In contrast, reducing Notch activity has no effect on repression. This differential highlights the fact that mutations in Notch and Su(H) are unlikely to have the same consequences on many target genes, as shown recently for singleminded. Since nonconsonance in phenotypes has been taken to indicate that certain Notch functions are independent of Su(H), it will be important to reevaluate these phenotypes, taking into consideration the possibility that Su(H) mutations can lead to derepression of target genes (Furriols, 2001).

If Su(H) has the ability to function as a molecular switch, the silencing of Gbe+ Su(H)m8 expression in the epidermis should be alleviated by ectopic activation of Notch in this tissue. To test this, hsNicd flies, which have the intracellular domain of Notch (Nicd) under the control of the heat-shock promoter, were used. Exposure to 37°C induces ubiquitous expression of Nicd, which is a constitutively active fragment of Notch, and under these conditions Gbe+ Su(H)m8 confers expression throughout the epidermis and the trachea. In the presence of Nicd, therefore, the silencing is alleviated, and the transgene becomes activated in all the places where Grh is present (Furriols, 2001).

These data indicate that in the absence of Notch activation, Su(H) is capable of binding to its cognate sites and repressing transcription. Notch activation can alleviate the repression so that Su(H) is able to cooperate with other DNA-bound activators, like Grh, to promote transcription. These results are in agreement with recent models and strongly suggest that this is a general mechanism through which Su(H) acts at native targets. Thus, Su(H) is capable of acting as the pivot in a sensitive switch that would ensure that Notch target genes can be poised but silent until Notch is activated. For example, the E(spl) genes, which mediate the inhibitory effect of Notch during lateral inhibition, appear to be targets of proneural proteins. However, E(spl) genes are not expressed in the cells that are selected to be neural, even though proneural proteins accumulate at highest levels in these cells. According to the model, Su(H) would be able to suppress activators like the proneural proteins until Notch is activated. As soon as levels of Notch are sufficient to overcome Su(H)-mediated repression, the synergistic interactions with activators would lead to a sharp transition in the expression of E(spl) genes. The potent effect of combining Su(H) and Grh also gives a precedent for the way that individual target genes might respond to Notch in specific contexts, if each involves a different transregulator cooperating with Notch. This demonstrates the potential for designing specific molecular assays for Notch activity in different cellular contexts. By replacing the Gbe sites with elements that respond to other activators, it should be possible to generate a transcriptional readout for Notch activity in any cell type (Furriols, 2001)


Four distinct classes of c DNAs have been described (N, N', O, and O'). There are two sites of variation. One is a small intron within exon 12 which remains unspliced in some cDNA variants (e.g., N and O) and codes for an extra 30 amino acids. The other consists of exons 4 and 5, which are present in the rarer large cDNAs (O and O') and lead to an inserion of 810 bp. It appears that exons 4 and 5 are always inserted together (Uv, 1997).
Genomic length - 37 kb (Uv, 1997)

cDNA clone length - Three cDNAs representing alternative splicing are present (Bray, 1989).

Bases in 5' UTR -930 for N transcript

Exons -16

Bases in 3' UTR - 729 for N transcript


Amino Acids The N transcript of Grainy head has 1063 amino acids (Bray, 1989).

Structural Domains

Grainyhead belongs to a recently identified group of transcription factors that share a 250-amino-acid domain required for binding to DNA and a carboxy-terminal dimerization domain. grainy head has a basic helix-loop-helix motif (Bray, 1989), and a novel isoleucine-rich activation motif, distinct from the glutamine-rich activation domain of Sp1 (Attardi, 1993a). NTF-1 has an unusually large, unique DNA-binding and dimerization domain, as well as a novel, isoleucine-rich activation domain. This 56-amino-acid activation region fails to interact with the putative Sp1 coactivator, dTAFII110, and thus appears to use a mechanism distinct from the glutamine-rich activation domain of Sp1 (Attardi, 1993b).

The regions in the Drosophila tissue-specific transcription factor Grainyhead have been mapped that are required for DNA binding and dimerization. These functional domains correspond to regions conserved between Grainyhead and the vertebrate transcription factor CP2, which has similar activities. The identified DNA-binding domain is large (263 amino acids) but contains a smaller core that is able to interact with DNA at approximately 400-fold lower affinity. The major dimerization domain is located in a separate region of the protein and is required to stabilize the interactions with DNA. The data also suggest that Grainyhead activity can be modulated by an N-terminal inhibitory domain (Uv, 1994; full text of article).

grainy head: Evolutionary Homologs | Regulation | Developmental Biology | Effects of Mutation | References

date revised: 1 February 98

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.