Skip to main content
NCBI home page
UKPMC Funders Author Manuscripts logoLink to UKPMC Funders Author Manuscripts
. Author manuscript; available in PMC: 2012 Feb 27.
Published in final edited form as: Mech Dev. 2007 Jul 21;124(9-10):647–656. doi: 10.1016/j.mod.2007.07.006

Functional distinctness of closely related transcription factors: a comparison of the Atonal and Amos proneural factors

Sam M T W Maung 1,*, Andrew P Jarman 1,**
PMCID: PMC3287285  EMSID: UKMS43596  PMID: 17709231

Abstract

Using the well-characterised paradigm of Drosophila sensory nervous system development, we examine the functional distinctness of the Amos and Atonal (Ato) proneural transcription factors, which have different mutant phenotypes but share very high similarity in their signature bHLH domains. Using misexpression and mutant rescue assays, we show that Ato and Amos proteins have abundantly distinct intrinsic proneural capabilities in much of the ectoderm. The eye, however, is an exception: here both proteins share the capability to direct the R8 photoreceptor fate choice. Therefore, functional distinctness between these closely related transcription factors varies with developmental context, indicating different molecular mechanisms of specificity in different contexts. Consistent with this, the structural basis for their distinctness also varies depending upon the function in question. In previous studies of neural bHLH factors, specificity invariably mapped to the bHLH domain sequence. Similarly, and despite their high similarity, much of the Amos’ specificity relative to Ato maps to Amos-specific residues in its bHLH domain. For Ato-specific functions, however, the Amos bHLH domain can substitute for that of Ato. Consequently, Ato’s specificity relative to Amos requires the non-bHLH portion of the Ato protein. Ato provides a powerful precedence for a role of non-bHLH sequences in modulating bHLH functional specificity. This has implications for structural and functional comparisons of other closely related transcription factors, and for understanding the molecular basis of specificity.

Keywords: Drosophila, neurogenesis, proneural, transcription factor

INTRODUCTION

A major question of transcription factor biology is how structurally related factors attain their functional distinctness in vivo, particularly as their cognate DNA binding sites often display loose and ill-differentiated consensus sequences. This is particularly true of the wide range of basic-helix-loop-helix (bHLH) transcription factors that are central to neurogenesis in all metazoans studied (Bertrand et al., 2002). In the Drosophila paradigm, bHLH factors encoded by the proneural genes (genes of the achaete-scute complex (AS-C), ato, and amos) are required for formation of the progenitor cells of the peripheral nervous system (PNS), known as sense organ precursors (SOPs). Loss- and gain-of-function experiments have shown that these proneural genes endow ectodermal cells with the capacity to become SOPs. For instance, each of the proneural factors can induce ectopic neurogenesis when misexpressed in Drosophila (Goulding et al., 2000; Huang et al., 2000; Jarman et al., 1993; Rodríguez et al., 1990), whereas bHLH factors related to vertebrate neurogenin cannot (Ledent et al., 1998; Quan et al., 2004). It is thought that proneural factors activate a core SOP developmental programme, and a number of ‘pan-neural’ target genes have been identified that may form part of this shared programme.

Loss- and gain-of-function experiments show also that the different proneural factors are required for distinct subsets of SOPs. The AS-C (particularly scute (sc)) are required for external sense organs, including sensory bristles (Campuzano and Modolell, 1992). ato specifies chordotonal organs (internal stretch receptors) (Jarman et al., 1993; Jarman and Jan, 1995), R8 photoreceptor precursors (Jarman et al., 1994), and a subset of olfactory sense organs (sensilla coeloconica) (Gupta and Rodrigues, 1997). amos specifies the two remaining subsets of olfactory sensillum (sensilla basiconica and trichodea) and some larval multiple dendritic neurons (Goulding et al., 2000; Huang et al., 2000). This suggests that each proneural factor differentially regulates subtype-specific target genes that modify the core SOP developmental programme. Understanding how they achieve these different functions is an important goal.

To understand the protein properties that underlie their neural subtype specificity, one approach is to map the specificity determinants within the proneural proteins. These factors are Class II bHLH proteins, as are other neural factors (such as vertebrate neurogenin, neuroD), MyoD (myogenic factor) and SCL (haematopoietic factor) (Massari and Murre, 2000). They share a related bHLH domain but otherwise have little or no similarity. The bHLH domains of diverse members of the class share a number of residues that are concerned with the folding of the domain, its dimerisation with a bHLH partner protein (in this case, the Class I E-protein, Daughterless), and its contact with the generic DNA binding site known as the E box. Despite these commonalities, structure-function studies on a variety of bHLH factors have concluded that sequence differences within the bHLH domain determine the functional specificity of different family members (Chien et al., 1996; Davis and Weintraub, 1992; Talikka et al., 2002). For example, the proneural capability of Ato relative to neurogenin has been mapped in chimeric protein experiments to three residues in the basic region (Quan et al., 2004). These residues are shared between Ato, Sc and Amos, thereby defining general determinants of proneural capability in Drosophila, which are presumably related to the activation of the core SOP developmental pathway. At a finer level of specificity, the difference in subtype specificity between Ato and Sc was also mapped to their bHLH domains (Chien et al., 1996). Much of this was due to residues that differed between the Ato and Sc basic regions. It is not clear how these ‘specificity residues’ affect protein function. It is commonly proposed that they promote differential interaction with protein ‘specificity cofactors’ rather than directly affect DNA binding properties, since the residues concerned are not among those predicted to contact DNA (Chien et al., 1996; Quan et al., 2004). Nevertheless, differences in in vivo DNA binding site utilisation have been established for Ato and Sc (Powell et al., 2004).

The Ato bHLH domain is quite divergent from that of Sc (~42% identity). In contrast, Ato and Amos have high similarity in their bHLH sequences (73% identity), and, strikingly, their basic regions are identical. Yet this is not the result of a recent gene duplication: there is no similarity between these proteins outwith their bHLH domains, and distinct Amos and Ato orthologues are maintained in the genomes of Anopheles and Apis (unpublished observations). This strongly suggests that they perform distinct functions that cannot be achieved by a single Ato-like protein (Ato and Amos are jointly referred to hereafter as the ‘Ato-like proteins’). Misexpression data only partly supports this at present: some specificity is shown by Amos, which alone promotes formation of ectopic olfactory sensilla (Goulding et al., 2000), but specificity has not been explored fully in the context of the antenna – a location in which the proteins function in close proximity. Outside the antenna, no apparent specificity was reported in chordotonal sensillum formation: both Ato-like proteins can promote large numbers of ectopic chordotonal sensilla. The reason for this lack of specificity is not clear. One possibility is that their close bHLH sequence relationship results in overlap of their functional properties. Similar questions of specificity could be asked of many other bHLH factors that share high similarity, such as Math1 and 5, and neurogenin1 and 2.

Here we show that Amos and Ato have fully distinct functional capabilities in vivo in chordotonal and olfactory sensillum development. This was not previously observed because Amos can ectopically activate Ato when expressed outside the antenna, leading to an apparent chordotonal specifying capability. A notable exception to this distinctness occurs during eye development, where Amos can apparently replace Ato function effectively. In general, however, Ato and Amos are specialised to perform unique functions. Using chimeric protein constructs, we show that the structural basis of specialisation differs for different functions. Relative to Ato, Amos specificity can be inferred to map partly to a few residues in its HLH region. In contrast, Ato functions require bHLH sequences that appear to be shared between Ato and Amos. In this case, specificity is achieved through the modulation of bHLH activity by sequences outwith Ato’s bHLH domain. We suggest that non-bHLH sequences may generally play an important and unrecognised role in modulating bHLH specificity.

RESULTS

amos and ato have intrinsic abilities to specify distinct olfactory sensillum fates on the funiculus

On the antennal funiculus (essentially the third antennal segment), both ato-like genes are required for olfactory sensilla (Gupta and Rodrigues, 1997; zur Lage et al., 2003). To determine the specificity of the genes for these fate choices, we investigated their ability to rescue the olfactory sensillum loss in amos mutant flies. Flies null for amos gene function completely lack sensilla basiconica and trichodea, whereas the ato-dependent sensilla coeloconica are unaffected (Fig. 1C,D) (zur Lage et al., 2003). Moreover, ectopic sensory bristles are formed owing to derepression of ac/sc. In such flies, misexpression of UAS-amos using the proneural driver, 109-68Gal4, rescues significant numbers of these sensilla, and also suppresses the ectopic bristles (Fig. 1A(i),E). Rescue is not complete, probably owing to suboptimal patterning or timing of misexpression. Whereas amos-dependent sensilla are rescued, misexpression does not increase the number of sensilla coeloconica. Instead there is a 27.9% decrease, suggesting that UAS-amos suppresses endogenous ato function (Fig. 1B(i)). In contrast, identical misexpression of ato failed to rescue significant numbers of sensilla basiconica or trichodea (Fig. 1A(i)) and failed to suppress ectopic bristles (Fig. 1F), even though such misexpression resulted in 30.2% more sensilla coeloconica (Fig. 1B(i)). Thus, in the funiculus the Ato-like proneural factors exhibit strongly distinct specificities in directly controlling the fate choice between different olfactory sensilla.

Figure 1. Abilities of Ato-like proteins in rescuing olfactory sensillum loss of amos1 mutant flies.

Figure 1

Open in a new tab

(A,B) Graphs of olfactory sensilla numbers on funiculus. Bars indicating standard deviations are shown on all graphs. All figures are scores of antennal sensilla from amos1 mutant flies in which the indicated proneural protein has been expressed using 109-68Gal driver. (A) Graph showing rescue of sensilla trichodea and basiconica by misexpression of proneural proteins in amos1 mutant antenna. These sensilla are normally absent in amos1 (not shown). (A(i)) Misexpression of Ato and Amos. (A(ii)) Misexpression of Ato-Amos chimeras. (B) Graph showing numbers of sensilla coeloconica after proneural protein misexpression. These are not significantly affected by the amos1 mutation itself, and the horizontal dotted line indicates the baseline number of sensilla in such mutant flies. (B(i)) Misexpression of Ato and Amos. (B(ii)) Misexpression of Ato-Amos chimeras. (C-H) Examples of adult funiculi. Arrows indicate examples of sensilla: closed arrow, sensillum basiconicum; closed arrowhead, sensillum trichodeum; open arrow, ectopic mechanosensory bristle; open arrowhead, sensillum coeloconicum; bracket: clusters of sensilla basiconica. (C) Wildtype. (D) amos1 mutant. (E) 109-68Gal4 UAS-amos; amos1. (F) 109-68Gal4 UAS-ato; amos1. (G) 109-68Gal4 UAS-ato-bHLH(amos); amos1. (H) 109-68Gal4 UAS-amos-bHLH(ato); amos1.

amos largely lacks the intrinsic ability to specify chordotonal organs

In most of the larval ectoderm except the funiculus and the eye, ato is required for chordotonal sensillum formation whereas amos has no function as it is not expressed. When misexpressed, however, both ato and amos are able to promote ectopic chordotonal sensillum formation (Goulding et al., 2000; Huang et al., 2000; Jarman et al., 1993), suggesting a lack of specificity for this neural fate. We confirmed this observation using a GFP-nompA fusion protein as a sensory neuron marker (Chung et al., 2001)(Fig. 2A-C). When misexpressed using 109-68Gal4, UAS-amos is very efficient at promoting chordotonal organs (Fig. 2B,C(i)), which recalls previous suggestions that Amos appears to be a particularly powerful proneural protein (Villa Cuesta et al., 2003). We considered two possible explanations for amos’s lack of specificity in this context. Firstly, Amos protein may share the protein sequence features of Ato that are required to activate chordotonal target genes. Secondly, Amos may inappropriately cross-activate Ato. These possibilities were tested by examining Ato expression in wing imaginal discs from 109-68Gal4, UAS-amos larvae. In this case, Ato shows strong ectopic activation within the areas of ectopic amos expression (Fig. 2D,E). Activation of Ato is also observed when amos is misexpressed in embryos (hairy-Gal4 UAS-amos) (Fig. 2F,G). In contrast ato does not cross-activate amos when misexpressed (data not shown).

Figure 2. Chordotonal organ formation by Ato-like proteins.

Figure 2

Open in a new tab

(A,B) GFP-nompA marker of chordotonal dendrites. (A) wildtype femoral chordotonal organs. GFP expression in ES organs (barely visible) is punctate whilst expression in chordotonal organs is spear-shaped (bracket). (B) Ectopic chordotonal organs induced in the scutellum by misexpression of 109-68Gal4 UAS-ato. By contrast, in the wild-type scutellum there are no chordotonal sensilla present (not shown). (C) Graph of chordotonal organ numbers in scutellum after proneural misexpression in wild type or ato1 mutant background. (C(i)) Misexpression of Ato and Amos. (C(ii)) Misexpression of Ato-Amos chimeras. (D,E) Ato expression in wing imaginal discs. (D) Wild type; note that Amos is not expressed in the wild-type wing disc. (E) 109-68Gal4, UAS-amos. Bracket indicates the region of the third wing vein. (F,G) Ato expression in stage 10 embryos. (F) Wild type; (I) hairyGal4 UAS-amos.

To test whether this cross-activation of Ato contributes to Amos’s promotion of chordotonal sensilla, we examined the effect of ato mutation on the amos misexpression phenotype. Significantly, UAS-amos promoted 10-fold fewer chordotonal sensilla in the absense of ato function (Fig. 2C(i)). In contrast, chordotonal production by misexpression of Ato was only slightly reduced in ato1 flies (Fig. 2C(i)). To a large extent, therefore, amos is unable to promote chordotonal specification in the absence of ato function. Contrary to previous findings, we conclude that Amos does not exhibit Ato’s intrinsic specificity in promoting chordotonal organs.

ato antagonises amos’s olfactory promoting ability

Outwith the funiculus, misexpression of amos, but not ato, also results in a small number of ectopic olfactory-like sensilla, particularly on the second antennal segment, along the third wing vein, and on the dorsal thorax (Fig. 3B) (Goulding et al., 2000). Interestingly, the number of ectopic olfactory sensilla induced by UAS-amos increases when endogenous ato gene function is reduced by one or two copies (Fig. 3A). It seems, therefore, that inappropriate cross-activation of ato masks amos’s intrinsic specificity in two ways: it causes ectopic chordotonal sensillum formation, and it also antagonises amos’s olfactory sensillum promoting ability.

Figure 3. ato interferes with ectopic olfactory organ specification by amos misexpression.

Figure 3

Open in a new tab

(A) Graph showing number of olfactory-like sensilla observed on second antennal segment after misexpression of UAS-amos. In wildtype, there are no olfactory sensilla on this segment. (B) Example of second antennal segment from 109-68Gal4, UAS-amos fly. The ectopic olfactory organs are usually found in clusters. Sensilla resembling different subtypes are produced but cannot be assigned with certainty. Indicated are examples of trichoid-like sensilla (arrows), basiconica-like sensilla (open arrow) and stunted olfactory-like sensilla (bracket). Occasional forked sensilla are produced (arrowhead).

amos can specify R8 photoreceptors independently of ato

A third context for ato function is the developing eye. Here, ato promotes the R8 photoreceptor fate (Jarman et al., 1994). To investigate whether this ability is confined to ato, we tested whether misexpression can rescue the loss of R8 formation in ato mutant flies. The wild-type compound eye comprises some 750 ommatidia, each of which develops around a founding R8 photoreceptor. Ommatidia are completely lacking in ato mutant flies owing to lack of R8 specification (Jarman et al., 1994) (Fig. 4A,B). Using 109-68Gal4, misexpression of ato results in small rescue of ommatidial numbers (Fig. 4C). Under similar conditions, misexpression of amos results in a substantial rescue of eye development (Fig. 4D). Moreover, this correlates with rescued formation of R8 photoreceptor precursors, as assayed by the expression of the Senseless in developing eye-antennal imaginal discs (Fig. 4E,F). This is consistent with the finding that ectopic amos expression resulting from the amosRoi mutation also promotes R8 rescue in ato mutants (Chanut et al., 2002). Therefore, in the context of the eye the ato-like genes cannot specify alternative fate choices; either gene is sufficient for R8 specification. In this respect, Amos and Ato share an ability that Sc does not have (Sun et al., 2000). Unlike other ato fate choices, the structural features required for this function must be shared by the two proneural proteins (but not by Sc).

Figure 4. Both Ato-like proteins can rescue R8 formation in ato mutant flies.

Figure 4

Open in a new tab

(A–D) Adult compound eyes. (A) Wild type. (B) ato1 mutant. (C) 109-68Gal4,UAS-ato; ato1, showing weak rescue of ommatidia. (D) 109-68Gal4,UAS-amos; ato1, showing substantial rescue of ommatidia. (E,F) Eye antennal discs. (E) Wild type, showing expression of Ato (green) and Sens (magenta) in the morphogenetic furrow (arrow), ocelli (arrowhead), and antenna (bracket). (F) 109-68Gal4,UAS-amos; ato, showing expression of Amos (green) and Sens (magenta).

Amos specificity relative to Sc maps to its bHLH domain

Studies on the structural basis of functional diversity of bHLH factors have overwhelmingly indicated that a key determinant of specificity is the sequence of the bHLH domain. To assess the contribution of the bHLH domain to the specificity of Amos, hybrid proneural genes were constructed that allow the expression of chimeric bHLH proteins. In the first instance, domain swaps were carried out between Amos and Sc. Thus, UAS-amos-bHLH(sc) produces an Amos protein in which the bHLH domain is exchanged for that of Sc. UAS-sc-bHLH(amos) produces a Sc protein in which the bHLH domain has been exchanged for that of Amos.

When misexpressed using 109-68Gal4, UAS-sc promotes ectopic sensory bristles, but cannot efficiently specify chordotonal sensilla (Fig. 5A,D) (Jarman and Ahmed, 1998). Similarly, UAS-sc promotes few ectopic olfactory sensilla on the second antennal segment (Fig. 5A). In contrast, UAS-amos promotes ectopic chordotonal and olfactory sensillum formation but not sensory bristles (Fig. 5A,C). For the chimeric proteins, the activity of UAS-sc-bHLH(amos) closely resembled that of UAS-amos, whereas UAS-amos-bHLH(sc) resembled that of UAS-sc (Fig. 5). Thus, in a comparison between Amos and Sc, functional specificity is mediated by their bHLH sequences. This mirrors the findings of a similar comparison between Ato and Sc (Chien et al., 1996).

Figure 5. Amos specificity relative to Sc maps to its bHLH domain.

Figure 5

Open in a new tab

(A) Graph of sense organ numbers after misexpression of Amos, Sc, or Amos-Sc chimeras. Chordotonal (Ch, divided by ten) and bristle (ES) numbers are scores from the scutellum; Olfactory numbers (Olf) are from the second antennal segment. (B–F) Scutella (rear dorsal thoraces) from adult flies showing the effect of proneural factor expression on macrochaete bristle formation in this region. (B) Wild type, showing the four scutellar macrochaetae; (C) UAS-amos, showing suppression of macrochaetae. (D) UAS-sc, showing many ectopic macrochaetae. (E) UAS-sc-bHLH(amos), showing very few ectopic macrochaetae. This is from a weak line, stronger lines show bristle suppression. (F) UAS-amos-bHLH(sc), showing many ectopic bristles.

Non-bHLH sequences define Ato’s specificity relative to Amos

The above result suggests that the bHLH domains of Amos, Ato and Sc completely define their distinctive functional specificities. Yet the bHLH domains of Amos and Ato are highly similar (Goulding et al., 2000). Moreover, Ato’s chordotonal specificity relative to Sc has been mapped largely to the sequence of its basic domain (Chien et al., 1996), yet this sequence is identical in Amos. To investigate the structural basis of specificity of the Ato-like proteins, hybrid genes were constructed that encode the chimeric proteins Ato-bHLH(Amos) and Amos-bHLH(Ato). Initially, the Ato-specific characteristics of these proteins were determined. When misexpressed in wild-type larvae using 109-68Gal4, both Amos/Ato chimeric proteins promoted efficient ectopic chordotonal sensillum formation (Fig. 2C(ii)), and therefore share these characteristics with both Ato and Amos. This suggests that each chimeric gene produces a functional Ato-like protein. To assess specificity, chordotonal sensillum production was assayed in ato mutant flies (Fig. 2C(ii)). In such flies, UAS-ato-bHLH(amos) was still able to promote substantial numbers of chordotonal sensilla formation (albeit 40% reduced compared with wild type), thus more closely resembling UAS-ato. In contrast, the ability of UAS-amos-bHLH(ato) to promote chordotonal sensilla was almost abolished (96% reduction in sensilla produced compared with wild type), thus resembling UAS-amos. The former result suggests that both bHLH domains are capable of supporting chordotonal development and therefore share most of the sequences required for this function. The latter result suggests that this bHLH capability is modified by sequences within the non-bHLH portions of one or both of these proneural proteins. A similar conclusion is reached when assaying Ato-specific function in the funiculus (i.e sensillum coeloconicum formation). Here, UAS-ato-bHLH(amos) promoted increased numbers of sensilla coeloconica (like UAS-ato), whereas UAS-amos-bHLH(ato) did not (like UAS-amos) (Fig. 1B(ii)).

Two possible mechanisms for the effect of non-bHLH sequences on Ato specificity were considered. Firstly, Ato’s non-bHLH sequence may confer chordotonal specificity on the Ato (or Amos) bHLH domain. Alternatively, Amos’s non-bHLH sequence may inhibit the chordotonal capability of the Amos (or Ato) bHLH domain. As a test of this, the Ato specificity of Amos-Sc chimeric proteins was examined: we reasoned that the non-bHLH sequence of Sc would be ‘neutral’ with regard to Ato specificity since the Sc bHLH domain does not encode any potential for Ato specificity. As expected from the capabilities of Amos and Sc proteins, UAS-sc-bHLH(amos) induces large numbers of ectopic chordotonal sensilla in the scutellum and wings; but UAS-amos-bHLH(sc) could not produce this phenotype (Fig. 5). However, this ability of UAS-sc-bHLH(amos) was greatly reduced (by 81%) in ato1 mutant larvae (Fig. 2C(ii)). A similar finding is observed for sensillum coeloconicum formation: unlike UAS-ato-bHLH(amos), UAS-sc-bHLH(amos) does not increase sensillum coeloconicum numbers (Fig. 1B(ii)). These results lead to two conclusions. Firstly, chordotonal/coeloconicum sensillum formation requires an Ato/Amos bHLH domain: Ato specificity therefore requires bHLH features shared by Ato and Amos but not Sc. Secondly, an Ato/Amos bHLH domain is not sufficient for Ato specificity: the non-bHLH portion of Ato is positively required.

Amos’s specificity depends on sequences both within and outwith its bHLH domain

The Amos-like properties of Ato-Amos chimeric proteins were assayed in amos1 rescue experiments. Interestingly, both UAS-amos-bHLH(ato) and UAS-ato-bHLH(amos) were able to rescue the formation of sensilla basiconica and trichodea to a greater degree than UAS-ato (Fig. 1A(ii),G,H). For UAS-ato-bHLH(amos), this rescue was significantly stronger and sensilla were often clumped as observed for UAS-amos (Fig. 1E,G). This chimera also suppresses ectopic bristle formation more strongly. Taken together with the finding that UAS-sc-bHLH(amos) suppresses sensillum coeloconicum formation (Fig. 1B(ii)), these observations suggest that the Amos bHLH domain is somewhat specialised in conferring Amos specificity. This in turn implies that, unlike Ato, Amos specificity depends more strongly on bHLH sequences unique to Amos. This specialisation is not complete, however, and Ato’s bHLH domain can partially substitute, with the help of Amos’s non-bHLH sequence. An interesting consequence of this is that Ato-bHLH(Amos) combines to some degree the functions of both Amos and Ato.

DISCUSSION

In this study, we have analysed the functional specificities of two proneural proteins, Ato and Amos, as examples of how highly related transcription factors achieve functional distinctness. This analysis is complex, partly because each proneural protein is multifunctional in neural development, each being required for different neural fates in different developmental contexts. Nevertheless, our results point to two general conclusions. Firstly, closely related transcription factors can exhibit high functional distinctness, but this distinctness can vary strongly with developmental context. Secondly, the function of the bHLH domain itself can depend on the protein context: most notably, chimeric proteins reveal that the Amos bHLH domain can largely support Ato-specific functions when in the context of the Ato protein. Unravelling the molecular and structural bases underpinning these observations is likely to be complex, but an important conclusion is that, for a given factor, different functional specificities will have different protein structural requirements, pointing to different underlying molecular mechanisms.

Context-dependent neuronal specificity of the closely related ato-like proneural genes

Despite their relatedness, ato and amos show different functional capacities in misexpression experiments. Within the antennal funiculus the proteins have very distinctive specificities for different olfactory sensillum subtypes, which appear further sharpened by mutual antagonism. Outside the funiculus they also show specificity: amos does not share ato’s intrinsic ability to promote chordotonal specification, but this specificity is masked because amos is able to activate the endogenous ato gene. It is likely that this activation is an aberrant reflection of a true physiological regulatory relationship between the two genes, since recent evidence suggests that ato is regulated by amos in some embryonic locations (Grillenzoni et al., 2007; Holohan et al., 2006). Despite this, their olfactory specificity suggests that amos does not cross-activate ato when misexpressed in the antennal disc, and immunohistochemistry confirms this (unpublished observations). This suggests that some mechanism prevents cross-activation in the antenna — the only larval location in which the two genes normally function in close proximity to specify different fates. Our finding that amos is poorly able to promote chordotonal specification in the absence of ato function contrasts with a previous report that used an embryonic assay (Chien et al., 1996; Huang et al., 2000). We suggest that this reflects a difference between the embryo and the imaginal discs: in embryos, there is evidence that some chordotonal organs can be specified in the absence of ato function (Jarman et al., 1995), and that specificity may be provided in this case by the non-proneural ato-like gene, cato (P. zur Lage and A.P.J., in prep.).

A notable exception to the distinctness of Ato and Amos occurs in the eye. Here, both proteins appear intrinsically capable of specifying R8 formation (unlike Sc (Sun et al., 2000)). Thus, for R8 formation the difference in wild-type function of the two genes is due to differences in expression pattern rather than intrinsic protein capability. This reinforces the danger in inferring transcription factor specificities from loss-of-function mutant phenotypes alone.

The comparison of Ato and Amos provides a useful precedent for other pairs/groups of closely related bHLH factors, such as Math1/Math5, Mash1/Mash2, Ngn1/Ngn2. Very little is known of the functional distinctness of these, but it may be supposed that similar functional studies may also reveal complex patterns of specificity. In such studies, it is important to bear in mind that an apparent lack of specificity may simply indicate that the appropriate context or cell fate has not been assayed. Moreover, although misexpression phenotypes provide useful specificity assays, mutant rescue may be a more discerning assay for closely related proteins.

Identifying structural requirements for specificity: specificity determinants within and outwith the bHLH domains

The bHLH domain sequence is clearly important for specificity, and this is reflected in its strong sequence conservation in orthologues across species (Fig. 6A). For example, the sequence identity in the bHLH domain between orthologues from D. melanogaster and D. virilis is 100% for Ato and 97% for Amos. This strongly implies that most bHLH residues have functional significance, either for general bHLH properties or for specificity. A proportion of bHLH residues (19/60) are highly conserved across class II bHLH domains because they function in dimerisation or in contact with the generic E box sequence (Fig. 6A and (Ellenberger et al., 1994; Ma et al., 1994)). The potential importance of the remaining conserved (i.e. consistently different) residues can be inferred by correlating functional properties with cross-species patterns of sequence similarity.

Figure 6. Sequence comparisons and cartoon summary of chimera results.

Figure 6

Open in a new tab

(A) Alignment of bHLH domains of Ato and Amos orthologues from insect species: Drosophila melanogaster (Dm), D. pseudoobscura (Dps), D. virilis (Dv), Anopheles gambiae (Ag) and Apis mellifera (Am). Coloured arrowheads indicate the pattern of conservation of specific residues and their implied functions. Residues implicated in dimerisation and DNA contact are based on crystal structures of MyoD and E47 (Ellenberger et al., 1994; Ma et al., 1994). Residues shared by the Drosophila proneural genes but not neurogenin have been implicated in core proneural functions (Quan et al., 2004). Residues shared between Ato and Amos but not Sc are implicated in shared Ato/Amos functions (and also for Ato-specific functions for which the Amos bHLH domain can substitute for that of Ato). Differences between Amos and Ato can be divided into two types. Residues that differ consistently between all Amos and Ato orthologues are particularly implied in Amos-specific functions, for which the Ato bHLH domain cannot effectively substitute. Some residues are unique to Drosophila orthologues of Ato or Amos, implying that they have appeared later in evolution, perhaps underlying ongoing functional divergence. (B) Cartoon summary of pattern of conservation within the full length Ato and Amos proteins. Percentages refer to amino acid sequence identity between D. melanogaster and D. virilis.

Distinct bHLH specificity of Amos and Ato

Seventeen amino acid residues differ between Amos and Ato; most of them consistently differ in orthologues. Our experiments suggest that these differences may be more important for the specificity of Amos than for Ato, since the Amos bHLH domain can influence the olfactory specificity of the Ato-bHLH(Amos) chimera. Interestingly, seven residues are unique to Drosophila orthologues of Amos, consistent with some functional specialisation relative to a putative Ato/Amos ancestor. Only four residues are unique to Drosophila Ato.

Shared functions may have constrained Ato-Amos bHLH divergence

The functional distinctness of the Ato-like proteins provides an explanation for why highly conserved Amos and Ato orthologues have been maintained in the genomes of diverse insects. Yet Amos and Ato retain high bHLH sequence similarity. Twelve amino acid residues are shared between Amos and Ato (but not Sc) (Fig. 6A). This is most striking in the almost complete identity of their basic regions. The retention of these residues in both proteins implies a common functional constraint that has prevented greater bHLH divergence. Indeed the two proteins may have several shared functions. One is the ability to repress endogenous bristle formation (Goulding et al., 2000; Jarman and Ahmed, 1998). We also speculate that shared features might be responsible for common aspects of the function of both proteins in olfactory sensillum development (albeit of different subtypes).

Whilst these functions are physiologically relevant, the constraint they provide may also have had two ‘accidental’ consequences. The shared bHLH residues may be important in R8 determination, thereby explaining why Amos (but not Sc) inappropriately shares Ato’s ability to drive R8 formation. Perhaps this has not been selected against because Amos is not expressed in the context in which this R8 ability is manifest (the eye). Another accidental consequence would be that the Amos bHLH domain shares the capability of Ato for driving chordotonal formation. However this is not seen in the Amos protein as a whole because chordotonal capability also requires other non-bHLH Ato sequences – i.e. a specific protein context.

Distinct Ato specificity – non-bHLH sequence modulates bHLH function

Ato-specific functions can be performed by an Ato gene with a bHLH domain from either Ato or Amos, but not Sc. Thus, an Ato-like bHLH domain is necessary for Ato specificity, but it is not sufficient. Ato sequences outwith the bHLH domain are required to modulate its function, enabling Ato to attain its specificity relative to Amos. Comparison of the orthologous Ato proteins from three Drosophila species shows that they share regions of strong amino acid sequence conservation outside the bHLH domain (summarised in Fig. 6B). Overall there is 59% identity between the three, with many areas of greater than 70% identity. The N terminus shows particularly high identity over 99 amino acids (84%). Interestingly, the non-bHLH sequences of the three orthologous Amos proteins share much less similarity. Overall identity is only 29%, with the highest identity of 62.5% found in a region of 32 amino acids. This supports a scenario in which the non-bHLH sequence of Ato has a strong role in promoting Ato specificity, whereas Amos’s non-bHLH sequence appears to have a more accessory role for Amos specificity.

These findings contrast strongly with all previous studies on neural bHLH proteins in both Drosophila and vertebrates, which have overwhelmingly highlighted the bHLH domain as the sole determinant of protein specificity (Chien et al., 1996; Jarman and Ahmed, 1998; Nakada et al., 2004; Quan et al., 2004; Quan and Hassan, 2005). For Ato, whilst the bHLH domain is clearly important for distinguishing Ato from Sc, it is the non-bHLH sequences that distinguish Ato from Amos. Similar structure-function comparisons between other closely related pairs of proteins may also reveal the importance of non-bHLH sequences. However, such comparisons would require a much greater understanding of the functional distinctness of such closely related factors than is currently known.

In summary, proneural specificity in different contexts or for different neural fates requires different underlying amino-acid sequence elements. Moreover, the pattern of specificity is complicated by the simultaneous need for shared and diverged functions. Strong constraint for shared functions may have left little room for functional specialisation via divergence of the bHLH domain. To some extent, bHLH specialisation was achieved during Amos’s evolution, but for Ato, functional divergence required the acquisition of a modulatory role by its non-bHLH sequences.

Molecular basis of specificity

Proneural specificity must entail differences in target gene regulation that result from differences in interaction specificity either with DNA binding sites or with cofactor proteins (or both). Differential interaction with cofactor proteins is widely favoured as an explanation for specificity (Chien et al., 1996; Jarman and Ahmed, 1998; Quan et al., 2004; Quan and Hassan, 2005), partly because context dependence is most easily explained by interaction with different spatially restricted cofactors. This would be consistent with the conclusion that the different sequence motifs mediate different aspects of specificity, since they might provide different protein interaction interfaces. Despite the attractiveness of cofactor interactions as the major basis of specificity, it has recently been demonstrated that Ato-specific and Sc-specific target genes have E box binding sites that conform to distinct (Ato- and Sc-specific) consensus sequences (Powell et al., 2004). One possibility is that Ato and Sc have intrinsic abilities to distinguish between different E boxes, and it is certainly possible that amino acid differences in their basic regions may subtly alter their DNA binding properties, even though the amino acids involved are not predicted to contact DNA directly. A possible alternative is that cofactor interactions might enforce differences in DNA binding site specificity through conformational effects. It is open to speculation whether the functional differences between Ato and Amos have a similar molecular basis. Since Amos’s basic region is identical to that of Ato, it seems possible that Ato and Amos utilise similar DNA binding sites. The in vivo binding site preference of Amos, however, is currently unknown as there are no known Amos-specific target genes. It remains an important priority in the study of this group of transcription factors to determine the molecular basis of their functional specificity.

EXPERIMENTAL PROCEDURES

Fly stocks

Fly stocks used were Oregon R and w1118 (both wild type), ato1 (Jarman et al., 1994), amos1 (zur Lage et al., 2003), UAS-ato.1, UAS-amos.3, UAS-GFP-nompA, 109(2)68Gal4, hairy-Gal4, sca-Gal4.

Chimeric gene constructs

Chimeric proneural genes were constructed by overlap extension PCR (Horton et al., 1990). In each case, the two gene fragments to be combined were amplified separately from genomic DNA (primers 1+2 and 3+4). The two products were combined by denaturation/reannealing, extended to give a single template, and then further amplification was performed using the two ‘outside’ primers only (primers 1+4). The hybrid product was then cloned into the EcoRI site of pUAST and sequenced. For sc-bHLH(amos), in which the bHLH domain is in the middle of the ORF, a third gene fragment is amplified from the sc gene (primers 5+6) and this was combined with the hybrid product of the first two fragments, the final hybrid product being amplified with primers 1+6. The primers used were as follows:

amos-bHLH(ato):

1 5′-CAGGAATTCGTTGTCCCTTGAGCCTGC

2 5′-CTTCACGACTTCGCCACCGAATCCCGC

3 5′-GGTGGCGAAGTCGTGAAGAGGAAGCGT

4 5′-CGGGAATTCAGCGCAGCAGATCCCC

ato-bHLH(amos):

1 5′-CGCGAATTCGTATCGTTTCATCCAG

2 5′-TTTGAGGACCACGACGGGTGTGATCTG

3 5′-CCCGTCGTGGTCCTCAAAAAACGGCGA

4 5′-GCAGAATTCATCGCCCACACTGGCTAG

sc-bHLH(amos):

1 5′-CAGGAATTCGTTGATCGTTATCCGGAAAGTG

2 5′-TTTGAGGACCTGGGATTGGTCTACATT

3 5′-CAATCCCAGGTCCTCAAAAAACGGCGA

4 5′-TAGGTCATCGTAGTCTCTGGACAGCAA

5 5′-AGAGACTACGATGACCTAAATGGGGGC

6 5′-CCTGAATTCTCACTGCTCCTGCCATAG

amos-bHLH(sc):

1 5′-CAGGAATTCGTTGTCCCTTGAGCCTGC

2 5′-TTGGACCGATTCGCCACCGAATCCCGC

3 5′-GGTGGCGAATCGGTCCAAAGGCGCAAT

4 5′-GGAGAATTCCTACACCAGATCCTGAAGGCT

Germline transformation

Transformant fly stocks containing the chimeric gene constructs were made by DNA microinjection into syncytial embryos. At least two independent insertions were used for the majority of experiments. The lines used were UAS-amos-bHLH(ato): 4.6C, also 4.6E; UAS-ato-bHLH(amos): 14a; UAS-sc-bHLH(amos): 18.1.1, also 18.1.6; UAS-amos-bHLH(sc): 4.16.

Immunohistochemistry

Immunohistochemistry of imaginal discs and embryos were carried out by standard procedures (zur Lage et al., 2003). Primary antibodies used were anti-Ato (1:5000) (Jarman et al., 1995), anti-Amos (1:1000) (zur Lage et al., 2003), anti-Sens (1:6250) (Nolo et al., 2000), anti-Futsch (1:200, obtained from the Developmental Biology Hybridoma Bank, Iowa City, Iowa). Secondary antibodies (1:1000) were from Molecular Probes. Confocal microscopy analysis was carried out using a Leica TCS SP2 or a Zeiss PASCAL microscope.

Assaying adult sense organ phenotypes

Adult chordotonal organs were assayed as described in Jarman et al. (Jarman et al.), except that in most cases, chordotonal organs were identified by virtue of their characteristic localisation of the dendrite marker, GFP-nompA (Chung et al., 2001). Adult olfactory sensilla were scored as described in zur Lage (zur Lage et al.). In general at least five individuals were assayed.

ACKNOWLEDGEMENTS

We thank Maurice Kernan for the GFP-nompA flies and David Prentice for technical assistance. SMTWM was the recipient of a research studentship from the Medical Research Council of Great Britain. This work is supported by the Wellcome Trust (042182 and 077266).

REFERENCES

  1. Bertrand N, Castro DS, Guillemot F. Proneural genes and the specification of neural cell types. Nature Rev. Neuro. 2002;3:517–530. doi: 10.1038/nrn874. [DOI] [PubMed] [Google Scholar]
  2. Campuzano S, Modolell J. Patterning of the Drosophila nervous system: the achaete-scute gene complex. Trends in Genetics. 1992;8:202–208. doi: 10.1016/0168-9525(92)90234-u. [DOI] [PubMed] [Google Scholar]
  3. Chanut F, Woo K, Pereira S, Donohoe TJ, Chang S-Y, Laverty T, Jarman AP, Heberlein U. Rough eye is a gain-of-function allele of amos that disrupts regulation of the proneural gene atonal during Drosophila retinal differentiation. Genetics. 2002;160:623–635. doi: 10.1093/genetics/160.2.623. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Chien C-T, Hsiao C-D, Jan LY, Jan YN. Neuronal type information encoded in the basic-helix-loop-helix domain of proneural genes. Proceedings of the National Academy of Sciences, USA. 1996;93:13239–13244. doi: 10.1073/pnas.93.23.13239. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chung YD, Zhu J, Han Y, Kernan MJ. nompA encodes a PNS-specific, ZP domain protein required to connect mechanosensory dendrites to sensory structures. Neuron. 2001;29:415–428. doi: 10.1016/s0896-6273(01)00215-x. [DOI] [PubMed] [Google Scholar]
  6. Davis RL, Weintraub H. Acquisition of myogenic specificity by replacement of three amino acid residues from MyoD into E12. Science. 1992;256:1027–1030. doi: 10.1126/science.1317057. [DOI] [PubMed] [Google Scholar]
  7. Ellenberger T, Fass D, Arnaud M, Harrison SC. Crystal structure of transcription factor E47: E-box recognition by a basic region helix-loop-helix dimer. Genes and Development. 1994;8:970–980. doi: 10.1101/gad.8.8.970. [DOI] [PubMed] [Google Scholar]
  8. Goulding SE, zur Lage P, Jarman AP. amos, a proneural gene for Drosophila olfactory sense organs that is regulated by lozenge. Neuron. 2000;25:69–78. doi: 10.1016/s0896-6273(00)80872-7. [DOI] [PubMed] [Google Scholar]
  9. Grillenzoni N, de Vaux V, Meuwly J, Vuichard S, Jarman A, Holohan E, Gendre N, R.F. S. Role of proneural genes in the formation of the larval olfactory organ of Drosophila. Dev Genes Evol. 2007 doi: 10.1007/s00427-007-0135-6. [DOI] [PubMed] [Google Scholar]
  10. Gupta BP, Rodrigues V. Atonal is a proneural gene for a subset of olfactory sense organs in Drosophila. Genes Cells. 1997;2:225–33. doi: 10.1046/j.1365-2443.1997.d01-312.x. [DOI] [PubMed] [Google Scholar]
  11. Holohan EE, zur Lage PI, Jarman AP. Multiple enhancers contribute to spatial but not temporal complexity in the expression of the proneural gene, amos. BMC Developmental Biology. 2006;6:53. doi: 10.1186/1471-213X-6-53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Horton RM, Cai ZL, Ho SN, Pease LR. Gene splicing by overlap extension: tailor-made genes using the polymerase chain reaction. Biotechniques. 1990;8:528–535. [PubMed] [Google Scholar]
  13. Huang ML, Hsu CH, Chien CT. The proneural gene amos promotes multiple dendritic neuron formation in the Drosophila peripheral nervous system. Neuron. 2000;25:57–67. doi: 10.1016/s0896-6273(00)80871-5. [DOI] [PubMed] [Google Scholar]
  14. Jarman AP, Ahmed I. The specificity of proneural genes in determining Drosophila sense organ identity. Mech Dev. 1998;76:117–125. doi: 10.1016/s0925-4773(98)00116-6. [DOI] [PubMed] [Google Scholar]
  15. Jarman AP, Grau Y, Jan LY, Jan YN. atonal is a proneural gene that directs chordotonal organ formation in the Drosophila peripheral nervous system. Cell. 1993;73:1307–1321. doi: 10.1016/0092-8674(93)90358-w. [DOI] [PubMed] [Google Scholar]
  16. Jarman AP, Grell EH, Ackerman L, Jan LY, Jan YN. atonal is the proneural gene for Drosophila photoreceptors. Nature. 1994;369:398–400. doi: 10.1038/369398a0. [DOI] [PubMed] [Google Scholar]
  17. Jarman AP, Jan YN. Multiple roles for proneural genes in Drosophila neurogenesis. In: Juurlink BHJ, et al., editors. Neural Cell Specification: Molecular Mechanisms and Neurotherapeutic Implications. Plenum Press; New York: 1995. pp. 97–104. (Altschul symposia series Vol. 3). [Google Scholar]
  18. Jarman AP, Sun Y, Jan LY, Jan YN. Role of the proneural gene, atonal, in formation of Drosophila chordotonal organs and photoreceptors. Development. 1995;121:2019–2030. doi: 10.1242/dev.121.7.2019. [DOI] [PubMed] [Google Scholar]
  19. Ledent V, Gaillard F, Gautier P, Ghysen A, Dambly-Chaudière C. Expression and function of tap in the gustatory and olfactory organs of Drosophila. International Journal of Developmental Biology. 1998;42:747–750. [PubMed] [Google Scholar]
  20. Ma PCM, Rould MA, Weintraub H, Pabo CO. Crystal structure of MyoD bHLH domain-DNA complex: perspectives on DNA recognition and implications for transcriptional activation. Cell. 1994;77:451–459. doi: 10.1016/0092-8674(94)90159-7. [DOI] [PubMed] [Google Scholar]
  21. Massari ME, Murre C. Helix-loop-helix proteins: regulators of transcription in eukaryotic organisms. Mol. Cell. Biol. 2000;20:429–440. doi: 10.1128/mcb.20.2.429-440.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Nakada Y, Hunsaker TL, Henke RM, Johnson JE. Distinct domains within Mash1 and Math1 are required for function in neuronal differentiation versus neuronal cell-type specification. Development. 2004;131:1319–1330. doi: 10.1242/dev.01008. [DOI] [PubMed] [Google Scholar]
  23. Nolo R, Abbott LA, Bellen HJ. Senseless, a Zn finger transcription factor, is necessary and sufficient for sensory organ development in Drosophila. Cell. 2000;102:349–62. doi: 10.1016/s0092-8674(00)00040-4. [DOI] [PubMed] [Google Scholar]
  24. Powell LM, zur Lage PI, Prentice DRA, Senthinathan B, Jarman AP. The proneural proteins Atonal and Scute regulate neural target genes through different E-box binding sites. Mol. Cell. Biol. 2004;24:9517–9526. doi: 10.1128/MCB.24.21.9517-9526.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Quan X-J, Denayer T, Yan J, Jafar-Nejad H, Philippi A, Lichtarge O, Vieminckx K, Hassan BA. Evolution of neural precursor selection: functional divergence of proneural proteins. Development. 2004;131:1679–1689. doi: 10.1242/dev.01055. [DOI] [PubMed] [Google Scholar]
  26. Quan X-J, Hassan BA. From skin to nerve: flies, vertebrates and the first helix. Cell. Mol. Life Sci. 2005;62:2036–2049. doi: 10.1007/s00018-005-5124-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Rodríguez I, Hernández R, Modolell J, Ruiz-Gómez M. Competence to develop sensory organs is temporally and spatially regulated in Drosophila epidermal primordia. EMBO J. 1990;9:3583–3592. doi: 10.1002/j.1460-2075.1990.tb07569.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Sun Y, Jan LY, Jan YN. Ectopic scute induces Drosophila ommatidia development without R8 founder photoreceptors. Proceedings of the National Academy of Sciences, USA. 2000;97:6815–6819. doi: 10.1073/pnas.110154497. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Talikka M, Perez SE, Zimmerman K. Distinct patterns of downstream target activation are specified by the helix-loop-helix domain of proneural basic helix-loop-helix transcription factors. Dev Biol. 2002;247:137–148. doi: 10.1006/dbio.2002.0677. [DOI] [PubMed] [Google Scholar]
  30. Villa Cuesta E, de Navascues J, Ruiz Gomez M, del Corral RD, Dominguez M, de Celis JF, Modolell J. Tufted is a gain-of-function allele that promotes ectopic expression of the proneural gene amos in Drosophila. Genetics. 2003;163:1403–12. doi: 10.1093/genetics/163.4.1403. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. zur Lage PI, Prentice DRA, Holohan EE, Jarman AP. The Drosophila proneural gene amos promotes olfactory sensillum formation and suppresses bristle formation. Development. 2003;130:4683–4689. doi: 10.1242/dev.00680. [DOI] [PubMed] [Google Scholar]

RESOURCES