BIOLOGICAL OVERVIEW

The response kinetics of known and putative target genes of Ftz has been examined in order to distinguish between direct and indirect Ftz targets. This work also has bearing on the classification of primary or secondary pair rule gene. Kinetic analysis was achieved by providing a brief pulse of Ftz expression and measuring the time required for genes to respond. The time required for Ftz to bind and regulate its own enhancer, a well-documented interaction, is used as a standard for other direct interactions. Surprisingly, both positively and negatively regulated target genes respond to Ftz with the same kinetics as autoregulation. The rate-limiting step between successive interactions (<10 minutes) is the time required for regulatory proteins to either enter or be cleared from the nucleus, indicating that protein synthesis and degradation rates are closely matched for all of the proteins studied. The matching of these two processes is likely to be important for the rapid and synchronous progression from one class of segmentation genes to the next. In total, 11 putative Ftz target genes have been analyzed, and the data provide a substantially revised view of Ftz roles and activities within the segmentation hierarchy (Nasiadka, 1999).

Ftz is known to directly bind and regulate the expression of its own promoter. Hence, the temporal response of this interaction was measured and used as a basis of comparison for other possible direct interactions. Endogenous ftz stripes respond to ectopic Ftz expression by limited expansion: each stripe expands anteriorly by about one cell. To determine the kinetics of this response, HSFtz embryos were heat treated for 8 minutes and aliquots were fixed at 5 minute intervals. The fraction of embryos exhibiting broadened Ftz stripes was determined. Embryos were scored as positive when the regions of broadening could be visually determined with certainty. Autoregulation, resulting in ftz stripe widening, was shown to occur between 15 and 20 minutes post-heat shock. After this time, a plateau is reached with about 65% of embryos having responded. Since the HSFtz line used in this experiment is heterozygous, a maximum of 75% of the embryos examined would be expected to exhibit this response. Failure to reach this theoretical maximum at most time points is probably due to the presence of embryos that were either inaccurately staged, unfertilized or inefficiently heat shocked. One exception was the value determined for the 30 minute time point. This was a highly reproducible 72%, after which the number of embryos with widened stripes appeared to steadily decrease. While this suggests the possibility of complex influences on the ability of Ftz to autoregulate in these cells, the authors were reluctant to make conclusions based on a single time point (Nasiadka, 1999).

The second best-characterized target of Ftz is the segment polarity gene engrailed. Stripes of en initiate at the anterior edge of each parasegment. Every second stripe overlaps with the anterior portion of each ftz stripe and is lost in ftz mutant embryos. Ubiquitous expression of Ftz causes a broadening of expression of these ftz-dependent en stripes, making them 2-3 cells wide instead of 1-2 cells wide. This expansion occurs in the same cells as endogenous ftz stripe expansion. The kinetics of this response were determined in the same way as for endogenous ftz, using embryos from the same set of collections. The curve generated for en overlaps very closely that of the endogenous ftz response, suggesting that both genes are direct targets of Ftz. If en were regulated indirectly, a delay would be expected reflecting the time required for intermediary gene products to be expressed, to accumulate and to elicit a response (Nasiadka, 1999).

In contrast to en, the segment polarity gene wingless (wg) has been identified genetically as a negative target of Ftz. This negative interaction has also been demonstrated in HSFtz embryos. Although all wg stripes are repressed in these embryos, the predominant effect is on odd-numbered stripes, which are completely repressed following Ftz induction. Repression of even-numbered stripes is much less efficient. To assess whether this repression is direct, the kinetics of repression were examined. The differential repression of odd- versus even-numbered stripes of wg was a helpful tool and an internal control for recognizing affected embryos. This curve follows very closely that of ftz autoregulation, with the midpoint of both curves occuring at 18 minutes post-heat shock. This indicates that repression of wg by Ftz is also likely to be direct, and that Ftz can act as both an activator and repressor of transcription. Repression is not the only response exhibited by wg in HSFtz embryos. Weak activation within most of each odd-numbered parasegment is also detected. The kinetic curve of this activation response is considerably delayed relative to the kinetics of the other three responses measured thus far. This suggests that wg activation results from an indirect genetic interaction. A likely intermediary gene in this response is the paired gene. prd is genetically required for the proper initiation of all 14 wg stripes; all 14 wg stripes expand rapidly in HSPrd embryos (Nasiadka, 1999).

To test whether the prd gene acts as an intermediate in the positive response of wg to ectopic Ftz, the spatial and temporal responses of prd were determined in HSFtz embryos. If prd does function as an intermediary gene, its expression should be induced in odd-numbered parasegments where wg activation is later observed. Moreover, the induction of prd transcripts should occur with the same rapid kinetics as the ftz, en and early wg responses. The prd expression pattern was examined 20 minutes after ectopic expression of Ftz. Stripes are significantly wider than those in similarly staged wild-type embryos. Using the most posterior stripe of prd as a landmark, it was seen that each of the expanded stripes had broadened at its anterior edge. These regions of expansion comprise most of each odd-numbered parasegment, which is exactly where ectopic expression of wg occurs. The time course of prd mRNA induction was assessed as described for endogenous ftz, en and wg. Although the slopes of the prd and ftz activation curves differ, the initial responses occur at about the same time, suggesting that the interaction between Ftz and prd is also direct. The differences in the slopes of the two curves are likely due to the autoregulatory nature of the ftz response: for a short time, Ftz is expressed from both heat shock and endogenous promoters, and then maximal expression is sustained via autoregulation at the endogenous locus. In contrast, prd activation takes place in regions of the embryo where neither ftz nor prd autoregulates. Hence, prd transcripts do not accumulate as quickly as those of ftz and soon disappear due to degradation of the ectopically expressed Ftz activator (Nasiadka, 1999).

The role of prd as an intermediary factor in wg activation was tested further by examining wg expression in a HSFtz;prd minus background. In the absence of prd, wg stripes should no longer expand. Ectopic Ftz does indeed fail to activate ectopic wg in the absence of prd. The expression pattern in HSFtz;prd - minus embryos is essentially identical to the pattern of wg expressed in prd - embryos: odd-numbered stripes are weak and even-numbered stripes are essentially absent. This result is consistent with the proposed role of prd as a direct activator of wg and as a genetic intermediate between Ftz and wg during ectopic stripe broadening (Nasiadka, 1999).

To verify that the prd protein (Prd) is a direct activator of wg, the nature of the temporal delay between prd and wg activation was analysed. Specifically, the temporal accumulation of Prd protein with respect to prd and wg transcripts was examined. If the interaction between prd and wg is direct, then one would expect that much of the interval between accumulation of the two transcripts would be occupied by synthesis and nuclear transport of the Prd protein. The kinetic curve for ectopic Prd induction closely resembles that of WG mRNA activation except that it is shifted by 1-2 minutes to the left (earlier). This indicates that most of the delay observed between the accumulation of prd and wg transcripts (about 8 minutes) is consumed by the synthesis and localization of Prd protein. The time required (~6-7 minutes) may be fairly typical of other segmentation proteins expressed at this stage. Indeed, the delay between detection of en transcript and protein responses is also 6-7 minutes, with curves that are virtually identical to those of prd transcripts and protein. These data do not exclude the possibility that there are genes in addition to prd that are required for ectopic activation of wg. However, if such gene products are required, their rates of synthesis or removal do not appear to supercede the temporal limitations imposed by the synthesis of Prd (Nasiadka, 1999).

The pair-rule genes hairy (h), runt (run) and even-skipped (eve) are designated as primary pair-rule genes, because they have been considered to regulate, and not to be regulated by, the other pair-rule loci. However, the results of more recent genetic analyses suggest that eve and run may be targets of Ftz. Given the conflicting nature of these results, expression patterns of the primary pair-rule genes were examined in HSFtz embryos to clarify their regulatory relationships. Since primary and secondary responses that peak respectively 20 and 35 minutes post-heat shock have been consistently observed, the primary pair-rule gene expression patterns were examined at these time points. The more efficient Ftz-expressing line, hsf2, was used due to the sensitivity of eve and run to longer (>4 minutes) heat shocks. Twenty minutes after induction of ectopic Ftz there no obvious differences in the observed pattern from the wild-type pattern. Likewise, expression of h and run are also unaffected 20 minutes after induction of Ftz. This was true for all stages (5-7) examined. Thus, no evidence has been found to support the possibility that these genes are direct targets of Ftz. Somewhat surprisingly, no differences in primary pair-rule gene expression patterns were apparent at the later recovery time either, indicating a lack of both direct and indirect responses. Similarly, no primary or secondary responses are observed for another pair-rule gene: odd-paired. Thus, only a subset of genes investigated in this study respond to ectopic Ftz, further demonstrating the specificity of this assay and Ftz function (Nasiadka, 1999).

To determine the regulatory relationships between Ftz and the other non-primary pair-rule genes, the expression of odd-skipped (odd) and sloppy-paired (slp) was examined in HSFtz embryos fixed 20 and 35 minutes post Ftz induction. Expression patterns of these genes were also examined in ftz mutant embryos to obtain genetic confirmation of the interactions observed. Like ftz, en and prd, odd appears to be directly activated by Ftz. In stage 5 embryos, ectopic expression of Ftz causes rapid expansion of odd, from its initiating pattern of six stripes, to near homogeneous expression across the germband. In stage 6 embryos, odd is normally expressed in 14 evenly expressed stripes. Ectopic Ftz causes an intensification of the primary odd stripes at this stage. These stripes are derived from the original 7 stripes that overlap ftz stripes. In stage 7 embryos, these primary stripes are not only intensified, but expand anteriorly as well (from about 1 cell wide to 2 cells wide). The percentage of embryos responding to ectopic Ftz, at all stages tested, is about the same as the percentage of embryos that show ftz autoregulation, en and prd activation and wg repression. Thus, Ftz appears to be an activator of odd at all stages tested. This positive relationship between Ftz and odd is consistent with the differences in odd expression observed in ftz mutant embryos. Stripes of odd appear to be diminished in intensity in stage 5 embryos, and primary stripes are weak or missing in stage 6 and 7 embryos (Nasiadka, 1999).

Unlike prd and odd, the pair-rule gene slp is negatively regulated by Ftz: ectopic expression of Ftz results in the differential repression of secondary slp stripes. Again, the penetrance of repression at the 20 minute recovery time was about 60%, as has been measured for the other genes exhibiting direct responses. As might be expected, slp stripes expand in ftz mutant embryos, filling the regions where Ftz is normally expressed. Thus, as with wg, Ftz appears to act as a direct repressor of slp. This effect is likely exerted through the response elements or trans-acting factors that regulate secondary stripe expression (Nasiadka, 1999).

Genetic studies have suggested that the segment polarity gene gooseberry (gsb), like wg, may be repressed by Ftz. In ftz mutant embryos, gsb stripes expand into the regions where Ftz is normally expressed, fusing to form seven wide stripes. To test whether this interaction is direct, gsb expression was examined in hsf2 embryos fixed 20 and 35 minutes after a 4 minute heat shock. No response is observed 20 minutes post heat shock. However, changes are detected in embryos fixed 35 minutes post heat shock; stripes expand into the ventral regions of odd-numbered parasegments. Interestingly, this response is a positive one, in contrast to the negative response predicted from expression in ftz minus embryos. This apparent contradiction can be reconciled by postulating that gsb is indirectly regulated by Ftz and that different intermediary factors are involved in each case. The delayed nature of the response in HSFtz embryos is consistent with this interpretation, and a likely intermediary factor is Prd, since gsb, like wg, appears to be activated by Prd. In ftz minus embryos, a likely intermediary activator is wg, since wg also expands in ftz - embryos and appears to function as an activator of gsb. An alternative explanation is that Ftz has the ability to repress gsb directly, but that this effect is spatially limited to regions where ftz is normally expressed. An exception would have to be made, however, in the anterior-most cell of each ftz stripe, where Ftz and gsb normally overlap (Nasiadka, 1999).

A major goal of the post genome sequencing era will be to organize all newly identified genes into comprehensive genetic circuits. Powerful new methods will be required to perform this task efficiently and accurately. One new approach that shows a great deal of potential is the use of high-density oligonucleotide arrays. Like the kinetic approach described here, this is a functional assay in that it has the potential to identify genes that respond directly to specific cellular stimuli or programmed steps in differentiation. As currently employed, however, changes in gene expression may not represent direct interactions. For example, a recent analysis of genes affected by mutations in components of the general transcription machinery used a 45 minute recovery time following inactivation of temperature-sensitive alleles. Based on the results presented here, 45 minutes is enough time for several successive gene interactions to take place. To identify direct targets, a kinetic approach, much as described here, would have to be employed. Genes could be turned on or off using a variety of techniques, and then the responses monitored at intervals thereafter. Genome-wide analyses of this type can then be used to comprehensively solve complex hierarchies such as the segmentation gene hierarchy studied here. Once direct circuitry is established, the more time-consuming business of establishing the relevance and molecular mechanisms underlying each interaction can follow (Nasiadka, 1999 and references).

Ftz modulates Runt-dependent activation and repression of segment-polarity gene transcription

A crucial step in generating the segmented body plan in Drosophila is establishing stripes of expression of several key segment-polarity genes, one stripe for each parasegment, in the blastoderm stage embryo. It is well established that these patterns are generated in response to regulation by the transcription factors encoded by the pair-rule segmentation genes. However, the full set of positional cues that drive expression in either the odd- or even-numbered parasegments has not been defined for any of the segment-polarity genes. Among the complications for dissecting the pair-rule to segment-polarity transition are the regulatory interactions between the different pair-rule genes. An ectopic expression system that allows for quantitative manipulation of expression levels was used to probe the role of the primary pair-rule transcription factor Runt in segment-polarity gene regulation. These experiments identify sloppy paired 1 (slp1), most appropriately classified as segment polarity genes, as a gene that is activated and repressed by Runt in a simple combinatorial parasegment-dependent manner. The combination of Runt and Odd-paired (Opa) is both necessary and sufficient for slp1 activation in all somatic blastoderm nuclei that do not express the Fushi tarazu (Ftz) transcription factor. By contrast, the specific combination of Runt + Ftz is sufficient for slp1 repression in all blastoderm nuclei. Furthermore Ftz is found to modulate the Runt-dependent regulation of the segment-polarity genes wingless (wg) and engrailed (en). However, in the case of en the combination of Runt + Ftz gives activation. The contrasting responses of different downstream targets to Runt in the presence or absence of Ftz is thus central to the combinatorial logic of the pair-rule to segment-polarity transition. The unique and simple rules for slp1 regulation make this an attractive target for dissecting the molecular mechanisms of Runt-dependent regulation (Swantek, 2004).

The role of Runt as a primary pair-rule gene complicates interpreting the alterations in segment-polarity gene expression that are observed in run mutants. Recent experiments utilizing a GAL4-based NGT-expression system [the transgene construct used to express GAL4 maternally contains the nanos promoter and the 3' untranslated region of an alpha-tubulin mRNA and is thus referred to as an NGT transgene (nanos-GAL4-tubulin)] to manipulate expression in the blastoderm embryo have demonstrated that low levels of Runt repress en in odd-numbered parasegments without altering expression of the pair-rule genes eve and ftz. This observation suggested that this approach might provide a useful tool for defining the role of Runt in regulating other segment-polarity genes. A systematic survey was undertaken of the response of other segmentation genes to increasing levels of NGT-driven Runt expression. These experiments revealed significant differences in sensitivity as well as interesting differences in the nature of the response of different genes to ectopic Runt. The odd-numbered en stripes are repressed at both intermediate and high levels of ectopic runt. After en, the second most sensitive target is slp1. This gene shows a partially penetrant and subtle defect in the spacing of the segmentally repeated stripes in embryos with low levels of NGT-driven Runt. A more pronounced alteration is obtained in embryos with intermediate levels of Runt. In these embryos the slp1 pattern is converted from a segment-polarity-like, 14-stripe pattern to a pair-rule-like, seven-stripe pattern. At this level, expression of other segmentation genes is normal although there are subtle changes in the spacing of the wg stripes and a partial loss of the odd-numbered hh stripes. All three of these genes show clearer alterations at higher levels of NGT-driven Runt, with wg responding in a manner similar to slp1 and hh responding in a manner similar to en. High Runt levels also produce spacing defects in the expression of odd and gsb, as well as a more subtle effect on prd. Several of the changes observed at high levels of ectopic Runt are likely to be indirect and due to alterations in the expression of eve, ftz and hairy. The response of slp1 to ectopic Runt is notable both because of its sensitivity and apparent simplicity, thus suggesting that Runt plays a pivotal role in regulating slp1 transcription (Swantek, 2004).

The differential combinatorial effects of Runt and Ftz on segment-polarity gene regulation emerged as a result of analyzing the sensitive and relatively simple response of slp1 to ectopic Runt. The slp1 transcription unit is one of two redundant genes that comprise the slp locus. This locus was initially characterized as having a pair-rule function in the segmentation gene hierarchy based on a weak pair-rule phenotype associated with loss of slp1 function. The slp1 and slp2 genes are expressed in similar patterns during early embryogenesis. Embryos deficient for both slp1 and slp2 have an unsegmented lawn cuticle phenotype similar to that produced by wg mutations. This raises the question of whether it is most appropriate to consider slp as a pair-rule or segment-polarity locus. In the most straightforward interpretation of the segmentation hierarchy, the role of the pair-rule genes is to establish the initial metameric expression patterns of the segment-polarity genes. The initial expression of the key segment polarity genes en and wg is normal in gastrula stage embryos that are deleted for both slp1 and slp2. The expression of wg begins to become abnormal and is lost during early germband extension. These observations are consistent with the proposal that slp expression identifies cells that are competent to maintain wg expression subsequent to the blastoderm stage. Based on these observations, it is concluded that slp1 and slp2 are most appropriately classified as segment polarity genes, not pair-rule genes (Swantek, 2004).

The expression of slp1 (and slp2) differs from several other segment-polarity genes in that the metameric pattern is comprised of two-cell wide, rather than single-cell wide stripes. These two cell-wide stripes comprise the posterior half of each parasegment. slp1 activation in odd-numbered parasegments requires the cooperative action of Runt and Opa, whereas in even-numbered parasegments Runt works together with Ftz to repress slp1 expression. The simple rules involving these three factors fully account for slp1 regulation in all of the Runt-expressing cells in the blastoderm embryo but also raise a question regarding the positional cues that regulate slp1 expression in cells that do not express Runt (Swantek, 2004).

There are four other pair-rule transcription factors that could be involved in slp1 regulation: Eve, Hairy, Odd and Prd. Expression of both Odd and Prd overlaps the slp1 stripes in a manner that suggests that neither of these factors provides positional information crucial for slp1 regulation. Consistent with this, there are no substantial changes in the early 14-striped slp1 pattern in embryos mutant for either odd or prd. By contrast, elimination of either Eve or Hairy leads to changes in both the number and spacing of the slp1 stripes. However, as these are both primary pair-rule genes some of these changes are certainly indirect and due to alterations in Runt and Ftz expression (Swantek, 2004).

Several lines of evidence indicate that Eve has a direct role in slp1 repression. Experiments with the temperature-sensitive eve[ID19] mutation indicate that transient elimination of Eve at the cellular blastoderm stage leads to expanded six cell-wide slp1 stripes because of de-repression in the anterior two cells of each odd-numbered parasegment. These two are the cells with the highest level of Eve, indicating that the primary role of Eve at this stage is to repress slp1 expression. Complementary experiments with an inducible hs-Eve transgene reveal that ectopic Eve blocks slp1 activation in both odd- and even-numbered parasegments. This result not only confirms Eve's role as a repressor, but also reveals a crucial difference between Eve and Ftz-dependent repression. Ftz-dependent repression is restricted to odd-numbered parasegments unless Runt is also ectopically expressed. This same restriction is observed in experiments with hs-Ftz transgenes, indicating that the difference between Eve and Ftz is not due to the mode of ectopic expression. Taken altogether these results indicate that Eve and Ftz normally have comparable roles in repressing slp1 transcription in the anterior half of the odd- and even-numbered parasegments, respectively, in late blastoderm stage embryos. The key distinction in the regulation of slp1 by these two homeodomain transcription factors is the critical role that Runt plays in Ftz-dependent repression (Swantek, 2004).

One aspect of slp1 expression not accounted for by the above rules is the factor (or combination of factors), referred to here as factor X, that is responsible for slp1 activation in the posterior half of the even-numbered parasegments. Activation in these cells is blocked either by the combination of Runt+Ftz or by ectopic Eve. Runt and Ftz are co-expressed anterior to these even-numbered stripes and presumably both play a role in defining the anterior margin of these stripes. Conversely, Eve is expressed posterior to these cells and probably has a role in defining the posterior margins of these stripes. The sole pair-rule transcription factor that remains as a candidate for Factor X is Hairy, which is expressed in the posterior half of even-numbered parasegments. However, it is not thought that factor X is Hairy for several reasons. All of the evidence to date indicates that Hairy functions as a repressor. Furthermore, NGT-driven expression of Hairy does not lead to slp1 activation in anterior blastoderm cells similar to that produced by the co-expression of Runt and Opa. Identification of factor X is clearly important for a complete understanding of slp1 regulation (Swantek, 2004).

Previous studies have indicated that Runt has roles in both activating and repressing transcription of different target genes in the Drosophila. The current results provide additional compelling evidence for this dual activity and also provide insight on factors that contribute to this context-dependent regulation. The dramatic effects of Ftz on Runt-dependent slp1 regulation clearly demonstrate that one important component of context is the specific combination of other transcription factors that are present in a cell. Indeed, the unique and relatively simple rules for slp1 regulation make this an especially attractive target for dissecting the molecular mechanisms whereby Ftz converts Runt from an activator to a repressor of transcription. It seems likely that the rules governing the Runt-dependent regulation of slp1 will provide a foundation for understanding the regulation of wg and gsb, two segment-polarity genes that are expressed in a subset of slp-expressing cells and that respond to Runt in a manner similar, but not identical to slp1 (Swantek, 2004).

The results also point to a second important component of context-dependent regulation by Runt. The specific combination of Runt + Ftz, which represses slp1, does not always give repression, since these same two factors work together to activate en in some of these same cells at the same stage of development. Thus, cellular context alone cannot fully account for the regulatory differences and there must be a target-gene specific component of context-dependent regulation. A similar gene-specific example of context-dependent regulation has recently been described for the Runx protein Lozenge. In this case, the presence of binding sites for the Cut homeodomain protein helps to stabilize a complex that leads to repression of deadpan transcription in the same cells in which Lozenge is responsible for activation of Drosophila Pax2. In a strict parallel of this model, it would be speculated that the slp1 regulatory region contains binding sites for some factor that helps to stabilize a repressor complex that includes the Runt and Ftz proteins. In a reciprocal, and not mutually exclusive model, perhaps there are binding sites for a factor in the en regulatory region that helps to stabilize a Runt- and Ftz-dependent transcriptional activation complex. Further studies on the en and slp1 cis-regulatory regions are needed in order to address these questions at the molecular level. This future work is crucial for understanding the context-dependent activity of Runt and thus the molecular logic of the control system that underlies the pair-rule to segment-polarity transition in Drosophila segmentation (Swantek, 2004).

Visualizing the role of boundary elements in enhancer-promoter communication

Formation of self-associating loop domains is a fundamental organizational feature of metazoan genomes. This study employed quantitative live-imaging methods to visualize impacts of higher-order chromosome topology on enhancer-promoter communication in developing Drosophila embryos. Evidence is provided that distal enhancers effectively produce transcriptional bursting from target promoters over distances when they are flanked with boundary elements. Importantly, neither inversion nor deletion of a boundary element abrogates this 'enhancer-assisting activity,' suggesting that they can facilitate intra-domain enhancer-promoter interaction and production of transcriptional bursting independently of topologically associating domain (TAD) formation. In contrast, domain-skipping activity of distal enhancers was lost after disruption of topological domains. This observation raises a possibility that intra-domain and inter-domain enhancer-promoter interactions are differentially regulated by chromosome topology (Yokoshi, 2020).

This study presents evidence that boundary elements help distal enhancers to induce transcriptional bursting from their target promoters. The data suggest that TAD formation is not a prerequisite for facilitating intra-domain enhancer-promoter interaction because inversion or deletion of Homie did not abrogate this 'enhancer-assisting activity.' Supporting this view, deletion of SF1 boundary element from the endogenous locus did not eliminate ftz stripe formation. In contrast, analysis of the neighboring Hox gene Scr revealed that domain-skipping activity of distal T1 enhancer relies on pairing between SF1 and SF2 boundary elements, suggesting that intra-domain and inter-domain enhancer-promoter interactions have differential dependence on TAD formation (Yokoshi, 2020).

It is well established that TAD formation is mainly regulated by a Zn-finger DNA-binding protein CTCF and the ring-shaped SMC complex cohesin in vertebrates. When CTCF or Rad21, a kleisin subunit of cohesin, is acutely removed by auxin-inducible degron, essentially all TADs are lost in mammalian cultured cells. Consistent with this, knockout of cohesin-loading factor Nipbl leads to loss of TADs in mouse liver. Importantly, these studies agree well in a point that elimination of TADs has a modest influence on overall transcription activities. Thus, topological boundaries and resulting TAD structures have been implicated to play an auxiliary role in enhancer-promoter communication and gene expression. This study combined quantitative live imaging and genome-editing methods to directly visualize impacts of higher-order genome topology on enhancer-promoter interaction. The data suggest that TAD formation itself has minor influence on intra-domain enhancer-promoter interaction and spatial patterning of gene expression because only ~20% reduction was seen in the total output after Homie inversion (Nhomie/Homie versus Nhomie/Homieinv) and ~10%-13% reduction after SF1 deletion. Consistent with this, recent whole-genome sequencing study of balancer chromosomes that contain massive rearrangements of genome topology revealed that alternation of TAD structures has minor impacts on overall level and specificity of gene expression in Drosophila. Similarly, in mammalian system, genome-editing analysis of Sox9 locus showed that TAD disruption results in only ~10%-15% reduction of Sox9 expression and no detectable phenotypes in developing mouse limb buds, again suggesting non-essential role of TADs in gene expression. Supporting this view, it was recently reported that Drosophila CTCF-null mutant can progress through embryogenesis. Moreover, it is known that C. elegans and Arabidopsis lack CTCF and TAD-like structures, yet they achieve productive gene expression during development. It is speculated that TAD-mediated intra-domain interactions confer robustness of transcriptional program under environmental fluctuations, such as temperature change and nutrient availability, thereby providing selective advantages in a population. Contrary to intra-domain regulatory interactions, the data clearly showed that inter-domain T1-Scr interaction was compromised in ΔSF1 embryos, suggesting that pairing of boundary elements is required for gene-skipping activity of distal enhancers. Given an essential role of SF1 in correct Scr expression, it might be possible that ftz TAD is formed as a consequence of bringing distal T1 enhancer and Scr promoter together by looping out an intervening ftz transcription unit. Importantly, ΔSF1 mutant embryos developed normally even without early Scr expression, suggesting that enhancer redundancy acts as an alternative mechanism to achieve reliable gene expression under various genetic perturbations, including misregulation of TAD formation (Yokoshi, 2020).

The data suggest that boundary elements are capable of facilitating enhancer-promoter communication independently of TAD formation because the level of gene activities remained to be substantially higher even after Homie inversion or deletion. As in mammalian genome, TAD boundaries are enriched for CTCF and Rad21 in Drosophila. In past studies, much effort has been focused on functional analysis of these proteins because CTCF and cohesin have been long known to play a fundamentally important role in genome organization. However, these are not the only proteins enriched at boundaries. Whole-genome ChIP assays revealed that so-called insulator proteins, such as Su(Hw), Mod(mdg4), CP190, and BEAF-32, colocalize with CTCF-binding sites in the Drosophila genome. Moreover, other regulatory proteins, such as Fs(1)h, DREF, Chromator, L3mbt, Z4 (also known as Putzig), TFIIIC, and condensin subunits (CAP-H2 and Barren), are shown to be a major constituent of the complex at TAD boundaries. Indeed, Nhomie and Homie used in this study are highly enriched for these components. Among these, Fs(1)h is known as a Drosophila homolog of bromodomain-containing protein Brd4. Recent studies suggested that Brd4 undergoes dynamic liquid-liquid phase separation through the conserved intrinsically disordered region at C-terminal domain, producing transcriptional condensates or transcription hubs that colocalize with the Mediator and Pol II clusters. In addition, Brd4 is known to interact with the positive transcription elongation factor P-TEFb, a complex of Cdk9 and its regulatory subunit cyclin T1. Intriguingly, P-TEFb is also suggested to undergo liquid-liquid phase separation though the conserved histidine-rich domain of cyclin T1 to increase processivity of Ser2 phosphorylation at Pol II CTD. Thus, it might be possible that boundary-enriched Fs(1)h acts as a scaffold to increase local concentration of transcription factors, co-activators, elongation factors, and Pol II complexes, thereby affecting the timing and the size of transcriptional bursting independently of TAD formation. Importantly, other boundary-enriched proteins DREF, Chromator, and Z4 are also suggested to be involved in recruitment of Pol II into a subset of genes, giving rise to a possibility that these proteins further contribute to assembly of transcription hubs at nearby enhancers. It appears that this mechanism is particularly important for distal enhancers to overcome large genomic separation because extension of enhancer-promoter distance from 6.5 kb to 9 kb leads to significant (more than 50%) diminishment in the level of total output when boundary elements are not present. It is speculated that this is a common regulatory mechanism in the Drosophila genome because typical enhancer-promoter separation is estimated to be ~10 kb. Future functional studies should elucidate the molecular mechanism underlying 'enhancer-assisting activity' of boundary elements (Yokoshi, 2020).

A bistable autoregulatory module in the developing embryo commits cells to binary expression fates

Bistable autoactivation has been proposed as a mechanism for cells to adopt binary fates during embryonic development. However, it is unclear whether the autoactivating modules found within developmental gene regulatory networks are bistable, unless their parameters are quantitatively determined. This study combined in vivo live imaging with mathematical modeling to dissect the binary cell fate dynamics of the fruit fly pair-rule gene fushi tarazu (ftz), which is regulated by two known enhancers: the early (non-autoregulating) element and the autoregulatory element. Live imaging of transcription and protein concentration in the blastoderm revealed that binary Ftz fates are achieved as Ftz expression rapidly transitions from being dictated by the early element to the autoregulatory element. Moreover, this study discovered that Ftz concentration alone is insufficient to activate the autoregulatory element, and that this element only becomes responsive to Ftz at a prescribed developmental time. Based on these observations, a dynamical systems model was developed, and its kinetic parameters were quantitated directly from experimental measurements. This model demonstrated that the ftz autoregulatory module is indeed bistable and that the early element transiently establishes the content of the binary cell fate decision to which the autoregulatory module then commits. Further in silico analysis revealed that the autoregulatory element locks the Ftz fate quickly, within 35 min of exposure to the transient signal of the early element. Overall, this work confirms the widely held hypothesis that autoregulation can establish developmental fates through bistability and, most importantly, provides a framework for the quantitative dissection of cellular decision-making (Zhao, 2023).

One of the central questions in developmental biology concerns how cells precisely and irreversibly adopt distinct cellular fates. It has been argued that cells assume their unique gene expression profiles through a sequence of decisions among branching paths, famously encapsulated by 'Waddington's landscape' of peaks and valleys delineating the possible trajectories that a cell can follow. Genetic networks that lock a cell into one of these trajectories may be thought of as 'memory modules' that guide cells through valleys in the landscape to their ultimate fates. In the simplest case, where a decision is made between two alternative developmental fates, the memory module is binary and referred to as a switch. The state of the switch is set by the action of transient upstream regulatory signals. Several genetic motifs, including autoactivation and mutual repression, are capable of maintaining binary cell fates. However, the mere presence of these motifs is insufficient to guarantee that a network can remember its expression state once upstream regulators have degraded. The ability to lock onto high or low expression levels results from bistability, a systems-level property that depends upon the quantitative details of the kinetics of the involved chemical reactions (Zhao, 2023).

Though bistability is widely invoked to explain cell fate decisions, relatively little quantitative data exist to confirm bistability in gene expression modules within developing embryos. Previous studies in cell culture and fixed embryos have provided evidence for bistability in hematopoietic differentiation, the Shh network, the vertebrate hindbrain, between the BMP and FGF morphogens, and within the Notch-Delta signaling system. Quantitative evidence for multistability in fruit fly embryos has also been derived from fitting the parameters of high-dimensional network models to fixed tissue measurements While these models are capable of reproducing the observed phenomenology, there is no guarantee that the optimal set of inferred parameter values reflects actual biophysical quantities (Zhao, 2023).

Thus, it is important to verify that the conclusions drawn from computational modeling and in vitro experiments apply to developmental systems in vivo in the context of models that quantitatively capture the molecular interactions that underlie cellular decision-making. Evidence for the bistability of a genetic module based on these molecular interactions in an intact multicellular organism has not yet been demonstrated (Zhao, 2023).

The early development of the fruit fly Drosophila melanogaster is an ideal model system for studying binary cell fate determination, due to the presence of pair-rule genes such as fushi tarazu (ftz) that form discrete stripes at the cellular blastoderm stage prior to gastrulation (2.5–3.5 h after fertilization. The expression of ftz is regulated by two main enhancers: the early, or zebra, element and the autoregulatory element. The early element responds to upstream transcription factors such as the gap genes to establish the initial expression pattern of seven stripes. This element is functionally distinct from the autoregulatory element, which contains multiple Ftz binding sites that allow Ftz to activate its own expression. This autoactivation network motif is theoretically capable of exhibiting bistability and has therefore been hypothesized to act as a binary memory module (Zhao, 2023).

Whether a cell possesses a memory module determines whether observed states of gene expression are transient in the absence of continued external signaling, or whether these states can be locked into permanent cell fates that can be maintained without further intervention. Specifically, if the autoregulatory module is bistable, then it maintains high ftz expression driven by the transient presence of upstream factors, even once those factors degrade (or until further regulatory mechanisms intervene). If, instead, the autoregulatory element is monostable, then the observed separation of Ftz concentration into high and low levels persists only as long as upstream factors are present to regulate expression. In their absence, Ftz expression would revert to a single fate for all cells. It is important to note, however, that in this case, the transiently high or low trajectory of Ftz concentration could still be instructive for regulating downstream genes (Zhao, 2023).

This study characterize the ftz autoregulatory module in vivo through quantitative real-time measurements in living fruit fly embryos. Focusing on stripe 4 expression, Ftz expression was observed to separate into discrete high and low levels at the blastoderm stage during the 20 min prior to gastrulation, concurrent with a transition in regulatory control from the early to the autoregulatory element. It was discovered that autoregulation is triggered at a specific time point in development—presumably through the action of 'timer genes'—rather than through a readout of Ftz concentration alone. Based on these observations, a dynamical systems model was developed and its parameters were quantitated from simultaneous real-time measurements of ftz transcription and Ftz protein dynamics in single cells of living embryos. This model predicts binary Ftz expression levels at gastrulation with high accuracy and demonstrates that, indeed, the ftz autoregulatory module is bistable. It is concluded that the ftz autoregulatory element acts as a memory module to commit cells to binary fates that are otherwise transiently defined by the early element, thereby validating a long-standing hypothesis in developmental and systems biology that bistability underlies cell fate determination. Simulations further make it possible to quantitatively define a developmental commitment window, which shows that the autoregulatory module requires about half an hour to establish a memory of the transient signal from the early module. Thus, yhid work provides a framework for the dissection of other regulatory modules in the gene regulatory networks that dictate development based on this interplay between dynamical systems models and real-time experiments (Zhao, 2023).

For decades, developmental biologists have used Waddington's landscape to conceptualize cellular decision-making. Under this framework, cells roll down valleys in a predetermined landscape to adopt their ultimate fates. This framework has been repeatedly mathematicized using dynamical systems theory. Many of these studies have hypothesized that autoactivation helps establish and maintain binary cell fates through bistability, which can be thought of as introducing forks in Waddington's landscape. Though experiments in cell culture and fixed tissue have provided evidence for the bistability of various autoregulatory modules found within gene regulatory networks, until now, these results have not been confirmed by direct examination of dynamics in intact, living embryos (Zhao, 2023).

This work utilized live imaging to quantitatively characterize the dynamics of the fruit fly ftz regulatory system in vivo. Tight temporal coordination was elucidated between the two enhancer elements that regulate ftz expression and combined dynamical systems modeling with biophysical measurements to show that the bistability of the autoregulatory module can maintain otherwise transient expression levels driven by upstream factors. Based on the prevalence of autoregulatory motifs in nature, it is speculated that the approach employed by the Ftz system to decide cell fate is not limited to fruit flies but might also be widely adopted during development in other organisms (Zhao, 2023).

One of the central discoveries of this study is that ftz autoregulation is triggered at a specific developmental time rather than being triggered when Ftz reaches a certain threshold concentration. Recent work has suggested candidates for 'timer genes' that are expressed at distinct developmental time points and appear to facilitate the expression of other genes. It is speculated that timer genes might also bind the ftz autoregulatory element to trigger its responsiveness to Ftz (Zhao, 2023).

This study relied on quantitative modeling with no free parameters to provide strong evidence demonstrating that the Ftz autoregulatory module is bistable. Dynamical systems models such as the one employed in this study are advantageous for this approach since the parameters are biophysically interpretable, which is not true in, for example, more coarse-grained Boolean models of genetic networks. Although the model in the main text is deterministic, in preliminary work a stochastic chemical reaction network simulated by the Gillespie algorithm, was considered as well as a stochastic differential equation model. Aside from introducing noise to Ftz expression levels, and in contrast to other patterning networks where noise appears crucial to drive fate determination, these models did not exhibit any unique behaviors that better accounted for the experimental observations compared to the deterministic model (Zhao, 2023).

The results presented in this work were derived from experiments paired with mathematical analysis of the theoretical model. The conclusions could be further supported—or challenged—by experiments where Ftz initial conditions are altered through, for example, heat shock or optogenetic approaches. Future research on Ftz autoregulation would also benefit from technological advancements that extend quantitative live imaging capabilities beyond gastrulation in order to monitor Ftz concentrations over extended timescales, from the onset of autoregulation to the final adoption of cellular fate (Zhao, 2023).

A basic assumption of the approach employed in this study is that the behavior of the whole network can be predicted from the behavior of the parts (modules) in isolation. In this model, the early and late populations of Ftz were divided into two separate modules. The early module produces the early protein, which acts as an input to the autoregulatory module responsible for producing the late protein. Positive feedback arises because ftz also activates its own expression. This is not the only way to define the module; for example, if it were known which regulatory factors rendered the autoregulatory element responsive, then those as inputs could be included instead of describing their activity implicitly through the parameter. Similarly, if the dynamics of the upstream regulatory factors for the early element were known, β, the phenomenological temporal decay in the early transcription rate, could be substituted with a mathematical expression relating the concentrations of these regulatory factor inputs to the early Ftz transcription rate (Zhao, 2023).

The representation of the autoregulatory module can predict the fate of ftz expression from arbitrary trajectories of early Ftz. As a result, it was possible to predict the effect of modifications to upstream signaling on the resulting gene expression patterns. This leads to asking what forms of input are appropriate to achieve particular patterning outcomes. Such ability to reverse engineer the process of cellular decision-making could facilitate designing perturbations to manipulate the system, identifying constraints placed on upstream modules by the needs of downstream modules, and analyzing whether biologically evolved signals match those that are mathematically 'optimal' for such needs as patterning speed or information transmission.70 Different methods of generating predictions may be appropriate depending on the types of inputs under consideration. In this paper, the fact that increasing any one of the parameters that define the early Ftz input, increases total Ftz concentration at all points in time (a property known as monotonicity) made it possible to analyze the model using a switching separatrix. However, this may not be true for other regulatory systems, as in the case where a gene within a module represses its own production (Zhao, 2023).

Throughout developmental biology, the concept of a commitment window has been repeatedly utilized to describe the amount of time cells need to be exposed to upstream signals in order to decide their developmental fates. This quantitative dynamical systems model enabled conducting of a detailed examination of this commitment window and to identify what fraction of cells adopt certain fates as developmental timing is varied. From an engineering perspective, a gene expression pattern may be considered as an objective that must be achieved with a prescribed level of precision (i.e., as a design specification) and work backward to see what inputs satisfy this requirement. This approach complements existing work on precision that emphasizes how tightly protein concentrations are controlled and how accurately cells can locate their position by reading out concentrations of upstream factors. In particular, the latter approaches indicate what level of precision is actually achieved by a patterning network, while the framing of this study focuses rather on what range of parameters allows a system to attain a predefined level of precision. A combination of the two perspectives could help elucidate what biophysical and evolutionary factors influence stochastic variation in phenotypes, including how precise expression patterns must actually be to produce functional, healthy organisms (Zhao, 2023).

In summary, by turning widespread schematic models of autoactivation modules into precise mathematical statements and experimentally testing the resulting predictions, this study has provided support for a widely held hypothesis about how developmental fates are established in embryos. In the future, combining quantitative measurements with precise spatiotemporal perturbations and synthetic reconstitution methods promises to enable yet another iteration of the dialogue between theory and experiment that constitutes the basis of this work, ultimately leading to a predictive understanding of function in developmental networks and the myriad forms and fates to which they give rise (Zhao, 2023).

Earlier Analysis

Any discussion of the function of fushi tarazu means giving equal time to its apparent complement, even-skipped . Both are pair rule genes; both are expressed in vertical stripes very early in development. At this early stage, the embryo looks like a belted bean. Seven stripes of ftz are interspersed with seven stripes of eve , forming a total of fourteen evenly spaced alternating bands that coincide with what will be the boundaries between future body segments in the adult fly. even-skipped is transcribed in odd numbered embryonic segments (called parasegments) and fushi tarazu is transcribed in even numbered segments. EVE is a transcriptional repressor, and FTZ is a gene activator. These two complementary functions are crucial for the activation of segment polarity genes, the downstream targets of EVE and FTZ. The segment polarity genes are responsible for proper differentiation within each segment, based on the earlier successful activity of ftz and eve.

A current source of controversy illustrates a fundamental difference between ftz and eve, and has to do with the way they become activated. eve is termed a primary pair rule gene, since its activation depends on maternal and gap transcription factors. Each stripe of eve transcription is independently activated by separate, stripe specific enhancers. A decade of work has not yet identified multiple stripe specific enhancers in ftz. Does ftz depend on eve and hairy and other so-called "primary" pair rule genes for its activation? If it does, then ftz is a "secondary" pair rule gene. If not, then ftz is a primary pair rule gene.

Certainly Hairy regulates ftz, as sites that bind Hairy have been identified in the ftz promoter (Tsai, 1995). Nevertheless, some evidence refutes the "secondary" status. ftz can be activated in a pair rule pattern in mutants of eve and hairy. The 'zebra" enhancer site can by itself target ftz to the seven wild type stripes. And herein lies the mystery. Drosophila biologists are accustomed to the notion of position specific enhancers. How then can a single enhancer sequence target ftz to seven different stripes? Where do the cues come from for stripe specific transcription, as is the case with ftz? Those biologists who think ftz is a "secondary" pair rule gene raise these questions, and answer them by suggesting that the cues come from "primary" pair rule genes.

Over time, with ever mounting information, this argument has lost ground (Yu, 1995). Transcription factors known to bind to the zebra stripe enhancer include FTZ-F1, an orphan nuclear receptor responsive to ecdysone. FTZ-F1 and Tramtrack (FTZ-F2) function in the stripe specific activation of FTZ (Lavorgna, 1991 and Han, 1992). It may be that this debate is rather narrowly focused. There is good evidence that mRNA stability is a prime source of the seven striped pattern. Thus the concern with transcriptional activation is only half the story as far as generation of the pair rule pattern is concerned (Riedl, 1996).

Whatever the primary-secondary pair-rule resolution for ftz, there will be a fascinating story to tell about ftz regulation during molting, and its role in structuring the adult, but at this time, this too has yet to be completely explained. After more than a decade, ftz retains its mystery.

GENE STRUCTURE

fushi tarazu is located within the Antennapedia-Complex (ANTP-C) between Antennapedia and Sex combs reduced. Molecular mapping of Scr breakpoint lesions has defined a segment of greater than 70 kb of DNA necessary for proper Scr gene function. This region is split by the fushi tarazu gene, with lesions affecting embryonic Scr function molecularly mapping to the region proximal (5') to ftz and those exhibiting polyphasic semilethality predominantly mapping distal (3') to ftz (Pattatucci, 1990a and b).

Bases in 5' UTR - 121

Exons - two

Bases in 3' UTR - 426

PROTEIN STRUCTURE

Amino Acids - 398

Structural Domains

FTZ has a PEST domain and a homeodomain, the latter being situated between the center of the sequence and the C-terminal end (Laughon, 1984) . The homeodomain is closely related to that of ANTP. The protein is heavily phosphorylated (Krause, 1989).

EVOLUTIONARY HOMOLOGS

Although fushi tarazu in Drosophila is flanked by homeotic selector genes conserved throughout the metazoa, there is no evidence that ftz was part of the ancestral homeotic complex; it has been unclear when the gene arose and acquired a role in segmentation. The beetle Tribolium castaneum has a ftz homolog located in its Homeotic complex and expressed in a pair-rule fashion, albeit in a register differing from that of the fly gene. These and other observations demonstrate that a ftz gene preexisted the radiation of holometabolous insects and suggest that ftz has a role in beetle embryogenesis which differs somewhat from that described in flies (Brown, 1994).

In short germ insects, the procephalon and presumptive anterior segments comprise most of the embryonic rudiment which lengthens as posterior segments are added during development (see Tribolium early embryonic development). The expression pattern of a grasshopper ortholog of the primary pair-rule gene even-skipped suggests that it is not relevant to segmentation in this short germ insect. However in Drosophila, a long germ insect that forms all segments simultaneously, eve plays a vital role in segment formation. The eve ortholog of the beetle Tribolium castaneum (Tc eve) has been cloned. Tribolium has been considered either a short germ insect (because the abdominal segments are added after formation of the germ rudiment) or intermediate germ (because the procephalon, gnathals and thorax are defined in the germ rudiment) but never a long germ insect like Drosophila. The homeodomain sequence is highly conserved between beetle, fly, and grasshopper eve orthologs. Tc eve is expressed in stripes during segmentation, but in a pattern differing in some details from that of the fly gene. In both insects eve is expressed in the anterior part of each parasegment. However, in Tribolium this expression appears equivalent in both even and odd numbered parasegments. In contrast, Drosophila expression in the even numbered parasegments arises later in development, after expression has taken place in odd numbered parasegments. This pattern is coincident with that detected with a cross-reacting antibody. Tribolium has a fushi tarazu ortholog in its Homeotic complex, but embryos deficient for this region do not have a pair-rule phenotype. Perhaps in the evolution of Drosophila ftz assumed a function in even numbered parasegments that was performed by an ancestral eve ortholog. Thus, an ancestral even-skipped gene appears to have evolved a role in segmentation in a common ancestor of flies and beetles. Central nervous system expression of even-skipped orthologs is of ancient origin. CNS expression preceded the separation of the protostomes and deuterostomes. Unlike vertebrate orthologs but similar to eve, Tc eve is not linked to the homeotic complex. In humans and mice, eve orthologs have been mapped to the 5' end of the homeotic complex consistent with the colinearity between chromosome location and the area of function (the anterior domain). Since Tc eve and Drosophila eve are not linked to the HOM-C complexes of the two species, it appears that eve was lost from the complex before the divergence of beetles and flies, and assumed a new role in segmentation found in Drosophila and Tribolium (Brown, 1997).

The DNA-binding homeobox motif was first identified in several Drosophila homeotic genes but also in fushi tarazu, a gene found in the Hox cluster yet involved in segmentation, not anteroposterior patterning. Homeotic transformations are not seen in insect ftz mutants, and insect ftz genes do not have Hox-like expression except within the nervous system. Insect ftz homeobox sequences link them to the Antp-class genes and Tribolium and Schistocerca orthologs have Antp-class YPWM motifs, amino-terminal to the homeobox. Orthologs of ftz cloned from a centipede and an onychophoran show that the origin of ftz predates the emergence of the arthropods, but the inability to pinpoint non-arthropodan orthologs has suggested that ftz is the product of a Hox gene duplication in the arthropod ancestor. ftz orthologs have been isolated from a mite and a tardigrade, arthropod outgroups of the insects. Mite ftz is expressed in a Hox-like pattern, confirming its ancestral role in anteroposterior patterning. Phylogenetic analyses indicate that arthropod ftz genes are orthologous to the Lox5 genes of lophotrochozoans (a group that includes molluscs) and, possibly, with the Mab-5 genes of nematodes and Hox6 genes of deuterostomes, and would therefore have been present in the triploblast ancestor (Telford, 2000).

Phylogenetic analyses of the homeodomain securely unite the mite (Archegozetes longisetosus; Alftz) and tardigrade (Milnesium tardigradum, Mtftz) ftz genes with the ftz genes of insects as well as those of a crustacean, a centipede and an onychophoran. All arthropod and onychophoran ftz sequences are grouped with bootstrap supports of 74% (neighbor joining; NJ) and 51% (maximum parsimony; MP) and 66% support from Puzzle Maximum Likelihood (PML). Striking amino acid identities are seen in the hexapeptide consensus F(F/Y)PWM(K/R)SYTD (in the single-letter amino acid code) in onychophora, tardigrades, mites and centipedes. In further support of this orthologous relationship, Alftz was found adjacent to, upstream of and in the same orientation as the Archegozetes Sex-combs-reduced ortholog (A/Scr) in genomic library lambda clones in precisely the relative position and orientation inferred for the ancestral insect ftz homolog (Telford, 2000).

The insect and crustacean ftz genes are particularly divergent, as can be inferred through outgroup comparison. Chelicerates and centipedes are phylogenetically closer to the insects than to the onychophoran, yet their ftz homeodomain is considerably closer in sequence to onychophoran and tardigrade ftz genes. This demonstrates that the onychophoran, tardigrade, chelicerate and centipede genes are closest to the ancestral arthropod ftz sequence and that the differences between them and the crustacean/insect lineage are due to changes in the latter. All but two of the residues of the inferred ancestral homeodomain sequence are seen in at least one of the crustacean/insect sequences. In subsequent phylogenetic analyses, the complete chelicerate ftz ortholog is used as representative of arthropod ftz genes in general (Telford, 2000).

Earliest ftz expression in insects is in a broad band throughout the blastoderm. The anterior-most early ftz expression in the beetle Tribolium lies at the front of the maxillary segment, just behind the mandibular engrailed stripe (in register with the segments rather than the parasegments), but later fades as expression resolves into a pair-rule pattern. Drosophila ftz is also expressed early, in a broad domain in the syncytial blastoderm, again resolving into seven stripes marking the front of even-numbered parasegments. Unlike in Tribolium, the early anterior boundary of Drosophila ftz is in the posterior of the maxillary segment [the parasegment 1/2 (PS1/2) boundary]; however, it seems that there would be an exact correspondence between Tribolium and Drosophila anterior boundaries of ftz expression at the front of the maxillary segment if Drosophila ftz were not repressed in the front of the maxilla by evenskipped (eve) expression. Drosophila ftz expands anteriorly into this domain in eve^- embryos. This repression by eve of Drosophila ftz is likely to be a derived feature, because eve and ftz are co-expressed in Schistocerca and to some degree in Tribolium. There are no segmental markers to identify the earliest anterior boundary of expression of Sgdax, which has been suggested to be a Schistocerca ftz ortholog. In all of the insects, the broad early domain disappears very early and long before there is any evidence of appendages (Telford, 2000).

It seems clear that insect ftz is derived from a homeotic gene: it is suggested that the earliest expression of ftz in insects indicates its primitive expression domain with a boundary at the anterior of the maxillary segment. This suggestion is supported by in situ hybridizations using Alftz probes on Archegozetes embryos that show that Alftz is expressed in a typically Hox-like pattern in all stages examined, with a sharp anterior boundary at the front of the second leg segment. This segment has been shown to be homologous with the insect maxilla. The posterior boundary is at the rear of the fourth leg bud. Unlike the early broad insect ftz domain, expression in the insect nervous system is slightly more anterior and in parasegmental register (its anterior border is at the PS0/1 boundary) in the posterior of the mandibular segment. This discrete parasegmental anterior boundary is similar to the nervous system expression of the canonical Hox genes and is conserved in all insects studied. Nervous system expression has not been studied in Archegozetes (Telford, 2000).

Alftz expression overlaps almost exactly with AlScr expression in the fourth appendage-bearing segment (the mite second leg, which is equivalent to the insect/crustacean second maxilla). This anterior boundary of Scr expression seems to be primitive for the arthropods, since it is seen almost identically in the isopod crustacean Porcellio, although it is expressed only in the posterior of the maxilla in insects. It seems plausible that ftz has lost its role as a homeotic gene in the ancestor of the insects as a result of redundancy of function following overlap of its expression domain with that of Scr. Coincidence of anterior boundaries also correlates with a loss of homeotic function of the Hox3/zen homolog in arthropods. Identical with Hox3/zen in insects, the loss of AP patterning function might have released the homeodomain from stabilizing selection and led to the rapid sequence divergence seen in the insect ftz genes; indeed in its new role as a pair-rule gene, Drosophila ftz can function even with its homeodomain almost entirely deleted. Sequence analyses suggest that ftz orthologs are present in all protostomes and possibly all triploblasts (Telford, 2000).

In further phylogenetic analyses, arthropod Ftz homeodomain amino acid sequences have been compared with others of the closely related central-group Hox genes (those related to Antp, Ubx and abdA) from a range of other animals, as well as with Hox4/Dfd and Hox5/Scr orthologs in order to root the trees. Recent phylogenetic analyses of Hox genes have implied that arthropod ftz and Antp genes arose from one duplication and Lox5-like genes and the genes related to nemertean LsHox7 arose from another separate duplication, such that there is no ortholog of ftz outside the Arthropoda and Onychophora (Telford, 2000).

Phylogenetic analyses using the 60 amino acids of the homeodomain show that, in fact, arthropod ftz genes cluster with the Lox5 genes of lophotrochozoa (molluscs, annelids, and relatives. Furthermore the ecdysozoan Antp genes (except AlAntp and the priapulid Antp orthologs) cluster with the lophotrochozoan LsHox7 and LaHB1 genes. There is no bootstrap support for this latter clade but there is support separating Antp+LsHox7+LaHB1 plus UbdA+Lox2+Lox4 from ftz+Lox5 and the outgroup as well as support separating UbdA+Lox2+Lox4 from Antp+LsHox7+LaHB1 (UbdA refers to Ubx and abdA genes). A number of further similarities outside the homeobox also support the orthology of ftz and Lox5. This scenario requires the minimum number of gene duplications (two) and losses (none) to produce UbdA+Lox2+Lox4, Antp+Lshox7 and ftz+Lox5 representatives in the ecdysozoa and lophotrochozoa. As such, this is a parsimonious explanation of the data. All alternative trees were rejected either as significantly less parsimonious or requiring four or more gene duplication/loss events (Telford, 2000).

In further support of the proposed orthology relationship between ftz and Lox5, the anterior expression boundary of leech Lox5 is immediately posterior to that of the leech Scr ortholog (Lox20), as would be expected of a ftz ortholog according to the rule of colinearity between chromosomal position and anterior boundary of expression domain. Genes from two other ecdysozoans (Mab-5 genes from two nematodes) and from a deuterostome (the amphioxus Hox-6) also group with ftz/Lox5 in further analyses, and these share with ftz/Lox5 one or the other of the two amino acids typical of ftz/Lox5. These genes also have chromosomal positions (amphioxus) or anterior boundaries of expression (nematodes) adjacent to the respective Scr orthologs, lending further credence to these orthology assignments (Telford, 2000).

In conclusion, the demonstration that ftz is derived from a Hox gene and has lost its anteroposterior patterning role in the insects means that the number of HOM/Hox genes involved in anteroposterior patterning in the arthropod/onychophoran ancestor must have been ten (eight Hox genes plus zen/Hox3 and ftz). Furthermore, the most parsimonious conclusion from this sequence analysis shows that a ftz ortholog was present in the protostome ancestor and possibly in all bilaterians. Since there is no convincing evidence for grouping Ubx and abdA with each other rather than each being an ortholog of either Lox2 or Lox4 of lophotrochozoans, the assumption of two independent duplications can be avoided and it is suggested that Ubx and abdA orthologs were also present in the protostome common ancestor. This would mean that the number of Hox genes in the Precambrian common ancestor of ecdysozoans and lophotrochozoans must certainly have been nine and probably all ten of the Hox genes that are found in extant arthropods (Telford, 2000).

The expression patterns of Hox genes have not previously been comprehensively analyzed in a myriapod. The expression patterns are presented of the ten Hox genes in a centipede, Lithobius atkinsoni, and these results are compared to those from studies in other arthropods. Three major findings are reported. (1) It has been found that Hox gene expression is remarkably dynamic across the arthropods. The expression patterns of the Hox genes in the centipede are in many cases intermediate between those of the chelicerates (spiders) and those of the insects and crustaceans, consistent with the proposed intermediate phylogenetic position of the Myriapoda. (2) Two 'extra' Hox genes were found in the centipede compared with those in Drosophila. Based on its pattern of expression, Hox3 appears to have a typical Hox-like role in the centipede, suggesting that the novel functions of the Hox3 homologs zen and bicoid were adopted somewhere in the crustacean-insect clade. In the centipede, the expression of the gene fushi tarazu suggests that it has both a Hox-like role (as in the mite), as well as a role in segmentation (as in insects). This suggests that this dramatic change in function was achieved via a multifunctional intermediate, a condition maintained in the centipede. (3) It was found that Hox expression correlates with tagmatic boundaries, consistent with the theory that changes in Hox genes had a major role in evolution of the arthropod body plan (Hughes, 2002).

The expression of the Hox genes corresponds roughly with the tagmatic divisions in the centipede. The expression of the genes lab, pb, Hox3 and Dfd is confined to the head, while the trunk is apparently under the control of Antp, Ubx, abd-A and Abd-B. Interestingly, the maxilliped segment has expression of three genes that extend both into the head (Scr and ftz) and into the trunk (Antp). The maxilliped segment is thought to be homologous to the first trunk or thoracic segment of other mandibulate arthropods. The appendages of this segment in the centipede, however, have been highly modified. While their leg-like structure is still evident, they develop to become short and broad fangs, complete with a poison gland. Thus, the first legs of the centipede are modified to become more mouthpart-like, and are used for prey capture and manipulation. This mixed head/trunk identity of the segment seems to be reflected in the Hox code found there. While the segment itself has only a 'trunk' Hox gene (Antp), the appendages have expression of Antp as well as the 'head' genes Scr and ftz, which are also expressed in the maxillary II segment. It remains to be determined how these genes contribute to the development of the centipede fangs. It would also be interesting to know whether the evolution of this novel appendage is correlated with a shift in the expression of these genes. Further studies of Hox expression in other myriapods such as a millipede, or functional studies in the centipede, would be very interesting regarding these issues (Hughes, 2002).

With regard to fushi tarazu, a process of Hox gene change may be revealed in this study. Although ftz has a role in segmentation in Drosophila, ancestrally in the arthropods it seems to have been a more typical Hox gene. The transition between a Hox-like role and a role in segmentation may have occurred via an intermediate state in which the gene played multiple roles in development, and that this transition state was maintained in the centipede lineage (Hughes, 2002).

The results of this study suggest that a role for ftz in the process of segmentation may have an ancient origin, and may be conserved across the mandibulate arthropods (myriapods, crustaceans and insects). In early centipede embryos, the pattern of expression in the posterior growth zone plus stripes in new segments (not unlike that of even-skipped) suggests a role in segment formation. But in later embryos, a clear Hox-like domain in the maxillary II and maxilliped segments emerges. Thus, it is suggested that fushi tarazu made its evolutionary transition from a Hox-like role to a role in segmentation via an intermediate stage that is retained in the centipede. Based on its combined domains of expression, it would appear that ftz may be able to play multiple roles in the same embryo, one of which was lost in the insects (perhaps owing to redundancy with Scr). Further studies of ftz homologs in the crustaceans and insects should clarify where in arthropod evolution the Hox role was lost. The results reported in this study suggest that the complex, dynamic expression domains in the centipede reflect multiple roles for the centipede ftz gene. The observed expression domains of this gene in the centipede suggest that major transitions in the function of a developmentally important gene may happen gradually via a multifunctional intermediate, and not necessarily only by duplication and divergence of two copies of a gene (Hughes, 2002).

Drosophila fushi tarazu (Dm-ftz) is first expressed in seven stripes defining alternate parasegments of the embryo; this is a 'pair-rule' segmentation function. It is then expressed in specific neural precursor cells in the central nervous system and finally in the developing hindgut. An Orthopteran ortholog of ftz (Sg-ftz, formally Dax) has been isolated from the grasshopper Schistocerca gregaria. The pattern of Sg-ftz expression in Schistocerca embryos suggests that some developmental roles of the ftz gene are likely to be conserved between these two species (e.g., CNS functions) while others may have diverged (e.g., segmentation functions). To test whether the function of the Ftz protein itself differs between these two species, the functions of Sg-Ftz and Dm-Ftz proteins were compared by expressing both in Drosophila embryos. Sg-ftz mimics only poorly several segmentation roles of Dm-ftz (engrailed activation, wingless repression, and embryonic cuticle transformation). However, the two proteins are similarly active in the rescue of a CNS-specific ftz mutant. These findings argue that this ftz CNS function is mediated by conserved parts of the protein, while efficient pair-rule function requires sequences present specifically in the Drosophila protein (Alonso, 2001).

The ftz gene of Drosophila was first identified as a 'pair-rule' segmentation gene. In its absence, the embryo forms only half the normal number of segments. Dm-ftz is transiently expressed in alternate parasegments of the blastoderm stage embryo where it is required for the regulation of engrailed and other genes that define and maintain segment boundaries. Dm-ftz gene is located within the Drosophila Hox cluster between the genes Sex combs reduced and Antennapedia, but the sequence of Dm-ftz is quite divergent and does not immediately betray its ancestry. However, orthologs of ftz have been isolated from a number of other arthropods. Most of these retain sequence motifs that are characteristic of the Hox genes, but which have been lost from the Dm-ftz gene (e.g., the YPWM motif) (Alonso, 2001).

Pair-rule patterns of ftz expression similar to that seen in Drosophila have been observed in other holometabolous insects, though even within this group not all the developmental roles of the ftz genes are conserved. In the beetle Tribolium, ftz is expressed in pair-rule stripes, but ftz mutations do not prevent segment formation. In Schistocerca, a representative of a more basal order of insects, the ftz ortholog is transiently expressed throughout the trunk primordium during early embryogenesis, but it never shows a spatially regulated pattern of expression analogous to that of the Drosophila pair-rule genes. However, during the subsequent development of the nervous system, the Drosophila and Schistocerca ftz genes are expressed in strikingly similar patterns, labeling specific cells that are clearly homologous between these two species (Alonso, 2001).

These observations suggest that the ftz family of genes has acquired new developmental roles in the lineage leading to Drosophila , and that these new roles may be mediated in part by changes in the structure of the Ftz protein. It is therefore relevant to ask to what extent an Orthopteran Ftz protein can substitute for the endogenous Ftz protein of Drosophila . To compare the functions of these proteins, transgenic Drosophila was used in which either the Drosophila orSchistocerca ftz coding sequences were expressed ectopically at different times during development, under heat shock regulation. Transgenic lines expressing the complete coding sequence of Schistocerca Ftz under control of a heat shock promoter (HS-Sg-ftz) were generated and compared with established HS-Dm-ftz lines. Brief heat shock treatments (36.5°C, 8 min) were used to induce both gene products. After a recovery period at 25°C (30 min), embryos were examined for cuticle defects (Alonso, 2001).

One of the best characterized targets of Ftz is the segment polarity gene engrailed (en). Ubiquitous Ftz expression expands the boundary of even-numbered en stripes anteriorly, causing the stripes to appear in closely spaced pairs, rather than evenly spaced. Under the conditions of these experiments, ectopic expression of Dm-ftz leads to the pairing of Engrailed stripes in 31% of treated embryos. The partial penetrance of this and other phenotypes in these experiments is due in part to the range of ages in the treated population. Eggs were collected over a period of 1 hr, but segmentation is maximally sensitive to ectopic Ftz expression for only a few minutes (Alonso, 2001).

Overexpression of Sg-ftz does not induce a similar phenotype under these conditions, though it mildly affects the width of en stripes in a small fraction of the population (stripe width changing from 2-3 to 3-4 cells in 5% of treated embryos. Neither of these effects is seen in control embryos heat shocked in parallel (Alonso, 2001).

Another target of Ftz activity is the segment polarity gene wingless (wg). Ubiquitous expression of Ftz represses wg transcription in odd-numbered parasegments. Kinetic experiments show that this is a direct interaction. Dm-ftz-mediated repression of wg was observed in 42% of the treated embryos. In contrast, wg expression in Sg-ftz-expressing embryos was indistinguishable from that seen in controls (Alonso, 2001).

The effects of ectopic Ftz expression on the larval cuticle pattern, secreted at the end of embryogenesis, was studied. Ectopic expression of Drosophila Ftz protein leads to a strong 'anti-ftz' phenotype characterized by the elimination of odd-numbered parasegments -- a pair-rule phenotype. The Schistocerca protein was also able to produce pair-rule cuticular phenotypes, but only in a much smaller proportion of the embryos (4% for HS-Sg-ftz compared to 36% in the case of HS-Dm-ftz (Alonso, 2001).

To study the activities of Dm-ftz and Sg-ftz in the Drosophila CNS, the same constructs were used to express these two Ftz proteins in embryos carrying a mutation that specifically disables at least one CNS function of ftz but has little effect on segmentation. This mutation, ftz^11.3, replaces an arginine at position 309 in the Ftz homeodomain with a histidine. Homozygous ftz^11.3 embryos have the normal number of segments, but lack the characteristic Ftz-dependent expression of Even-skipped (Eve) protein in RP2 neurons (Alonso, 2001).

The ubiquitous expression of Ftz proteins was induced in this mutant background by heat shocking embryos at 4-5 hr after egg laying, around the time that Ftz is normally first expressed in the developing CNS. After a recovery time of 6 hr at 25°C, embryos were fixed, stained with antibody to Eve protein, and scored for the rescue of Eve-staining neurons in the normal position of RP2 cells. In wild-type embryos, RP2 neurons always lie anterior to the aCC/pCC neurons, which also express Eve. Under these conditions, the two Ftz proteins activate eve in neurons at comparable, though low, frequencies. However, while Sg-Ftz activates eve in symmetrically located neurons at the position expected for RP2 neurons, Dm-Ftz expression results in a more extensive and complex pattern of eve activation. An analogous unrestricted expression of eve in CNS cells has been previously observed in certain Polycomb (Pc) group mutants, revealing that most CNS cells can be competent to activate eve expression under specific regulatory circumstances (Alonso, 2001).

At present the Ftz-induced eve-expressing neurons are only provisionally identified as induced RP2 cells. However, the fact that both Drosophila and Schistocerca proteins induce eve⁺ cells in a comparable proportion of embryos, and that the Schistocerca protein specifically induces cells in the location of normal RP2 neurons, suggests that the Ftz protein of Schistocerca is able to replace the Drosophila protein for this role in CNS development. These results also suggest that the failure of the HS-Sg-ftz construct to activate the normal targets of Dm-ftz during Drosophila segmentation results from a difference in the intrinsic function of the two proteins and not simply from the inability of this construct to provide sufficient levels of protein (Alonso, 2001).

Although the homeodomain (HD) of Ftz is the only part of the protein that is well conserved between Drosophila and Schistocerca, it has previously been shown to be dispensable for segmentation in Drosophila . Therefore, what sequences outside the HD might be responsible of the differential behavior of Dm-Ftz and Sg-Ftz proteins in segmentation? An alignment of these two protein sequences highlights two particular modules that may confer functional specificity. The first is a Ftz-F1 interaction domain present only in the Drosophila protein. Ftz-F1 is a transcription factor from the nuclear receptor family and an obligatory cofactor for Dm-ftz in segmentation functions. Careful deletion mapping of Dm-ftz protein has identified two interaction domains that mediate the contact with Ftz-F1. Interestingly, one of these domains maps outside the HD and matches the LXXLL consensus for a Nuclear Receptor Box (NRB) motif found in most coactivators of nuclear receptors. This NRB/Ftz-F1 interaction domain is absent from the Sg-ftz protein (Alonso, 2001).

The second module is a hexapeptide motif present only in the locust protein. This element is also involved in interactions with cofactors. Sg-Ftz and some other insect Ftz proteins contain a short sequence generally known as the 'Hexapeptide' or 'YPWM motif' located N-terminal to the HD. This motif is a prefolded domain present in Hox proteins and responsible for their interaction with cofactors of the TALE family such as Extradenticle/Pbx proteins. The Drosophila Ftz protein lacks the YPWM motif (Alonso, 2001).

The presence of the Ftz-F1 interaction domain exclusively in the Drosophila protein and of the Hexapeptide/YPWM motif exclusively in the Schistocerca protein is likely to modulate the activity of these two transcriptional regulators differentially in different developmental contexts. Other, less well-defined regions that could also contribute to functional specificity are a paired interaction domain (aa 100-150) and a region of a putative transactivation domain of the proline-rich class (aa 79-90), both only present in the Drosophila protein (Alonso, 2001).

In summary, these results imply that one conserved role of Ftz in CNS development -- eve activation -- is mediated by the HD or other conserved sequences, while the pair-rule function of Ftz in Drosophila segmentation is mainly mediated by the divergent regions of the protein outside of the HD, probably through protein-protein interactions (Alonso, 2001).

Hox genes specify cell fate and regional identity during animal development. These genes are present in evolutionarily conserved clusters thought to have arisen by gene duplication and divergence. Most members of the Drosophila Hox complex (HOM-C) have homeotic functions. These genes, by definition, have the ability to transform the characteristics of one body part into those of another body part. However, a small number of HOM-C genes, such as the segmentation gene fushi tarazu (ftz), have nonhomeotic functions. If these genes arose from a homeotic ancestor, their functional properties must have changed significantly during the evolution of modern Drosophila. In this study it was asked how Drosophila ftz evolved from an ancestral homeotic gene to obtain a novel function in segmentation. Ftz proteins at various developmental stages were examined to assess their potential to regulate segmentation and to generate homeotic transformations. Drosophila Ftz protein has lost the inherent ability to mediate homeosis and functions exclusively in segmentation pathways. In contrast, Ftz from the primitive insect Tribolium (Tc-Ftz) has retained homeotic potential, generating homeotic transformations in larvae and adults and retaining the ability to repress homothorax, a hallmark of homeotic genes. Similarly, Schistocerca Ftz (Sg-Ftz) causes homeotic transformations of antenna toward leg. Primitive Ftz orthologs have moderate segmentation potential, reflected by weak interactions with the segmentation-specific cofactor Ftz-F1. Thus, Ftz orthologs represent evolutionary intermediates that have weak segmentation potential but retain the ability to act as homeotic genes. It is concluded that ftz evolved from an ancestral homeotic gene as a result of changes in both regulation of expression and specific alterations in the protein-coding region. Studies of ftz orthologs from primitive insects have provided a 'snap-shot' view of the progressive evolution of a Hox protein as it took on segmentation function and lost homeotic potential. It is proposed that the specialization of Drosophila Ftz for segmentation results from loss and gain of specific domains that mediate interactions with distinct cofactors (Löhr, 2001).

ftz is a rapidly evolving member of the Hox gene family and thus provides an opportunity to study the mechanisms underlying the functional evolution of Hox proteins. Based on the sequence similarity between Ftz and Antp in the N-terminal arm of the HD it has been suggested that ftz arose as a duplication of an ancestral Antp gene. The experimental data supports this hypothesis, considering the similarity of phenotypes induced by ancestral ftz genes and Dm-Antp. It is further suggested that redundancy between ftz and Antp relieved constraints on the primordial ftz gene, resulting in changes within both its regulatory and protein-coding regions (Löhr, 2001).

The expression of ftz has evolved from that which is typical of Hox genes found in mites to a Hox-like pattern in grasshoppers and a striped expression in beetles, which is typical for segmentation genes. In addition, two major changes occurred in the protein-coding region of ftz genes to facilitate the functional evolution from a homeotic to a segmentation protein. (1) Ftz proteins acquired segmentation functions not present in the homeotic Hox proteins. This change appears to have happened gradually, because Sg-Ftz displays very weak segmentation potential, Tc-Ftz has moderate segmentation potential, and a major function of Dm-Ftz is to promote segmentation. This change correlates with the ability to interact with the cofactor Ftz-F1. Sg-Ftz is able to interact weakly with Ftz-F1, and this interaction is enhanced by the addition of an LXXLL motif in the Tc-Ftz and Dm-Ftz proteins. This motif has been shown to strengthen the interactions of coactivators with nuclear hormone receptors, to which family Ftz-F1 belongs. The LXXLL motif was presumably acquired after the divergence of grasshoppers and beetles, although the possible loss of a more ancient LXXLL motif in the grasshopper lineage cannot be ruled out. (2) Ftz proteins lost their ancestral homeotic properties. Homeotic functions of Hox proteins are dependent on the presence of a YPWM domain upstream of the HD, which mediates the interactions with a cofactor, the HD protein Extradenticle (Exd) in vivo and in vitro. The loss of homeotic potential of Ftz correlates with the loss of the YPWM motif: Sg-Ftz and Tc-Ftz retain a YPWM motif and homeotic potential, whereas Dm-Ftz has lost the YPWM motif and the ability to mediate homeotic transformations. Dm-ftz gain-of-function mutants in which Ftz protein is abnormally stable display unique properties suggested to be homeotic. Whether this mutation has revealed residual homeotic activity of Dm-Ftz remains to be determined (Löhr, 2001).

Ernst Mayr suggested that "if, in the course of evolution, some of the proteins in an organism undergo evolutionary changes ... this might create a selection pressure in favor of remodeling other proteins in order to improve interaction. " It is suggested that the capability of ancestral Ftz to interact with Ftz-F1 was selected for, leading to the acquisition of a protein domain that stabilizes this interaction and the loss of a domain that could cause competition with another cofactor. This competition was observed in experiments with Tc-Ftz, which contains interaction motifs for both Exd and Ftz-F1. Ectopic expression of Dm-ftz produces anti-ftz phenotypes when induced over a broad time range (2.25 and 3.25 hr AEL). Ectopic expression of Tc-ftz causes anti-ftz phenotypes in less than half of the embryos scored when induced early (2.25 and 2.75 hr AEL). Slightly later, when Exd has become nuclear, Tc-Ftz induces only homeotic transformations and head involution defects. Since Ftz-F1 is nuclear throughout embryogenesis, but Exd is only nuclear at approximately 3 hr AEL, these observations suggest that Tc-Ftz preferentially interacts with Exd when it is present, and, thus, its segmentation functions are inhibited. In Drosophilids, loss of the YPWM motif abolished this competition for cofactors, allowing for exclusive interaction of Ftz with Ftz-F1 such that the function of Dm-Ftz was devoted entirely to segmentation. Both Hox cofactors, Exd and Ftz-F1, influence Hox protein function by cooperative binding to composite binding sites that contain DNA recognition sequences for both the Hox protein and the partner. The different partner pairs recognize qualitatively different sets of DNA sequences in the genome and thus regulate different sets of target genes. Homeotic genes function to determine regional identity, while Dm-Ftz, in conjunction with Ftz-F1, has evolved to regulate a unique set of target genes, presumably involved in promoting cell survival and morphogenesis during segmentation (Löhr, 2001).

The ability of Ftz proteins to regulate segmentation in Drosophila correlates with the germ band length of its native organism: Sg-ftz, from an extreme short germ band insect in which only the anterior-most segments are established during blastoderm and all other segments are established during gastrulation, shows the weakest ability to function as a segmentation gene. In the case of Tc-Ftz, which is native to an intermediate germ band insect, in which head and trunk segments are established during blastoderm and remaining segments during gastrulation, segmentation properties were more pronounced. Dm-Ftz is entirely devoted to segmentation in the long germ band insect Drosophila, in which all segments are established at the blastoderm stage. It is possible that the segmentation functions of Ftz evolved to fit the needs of more complex forms of embryogenesis, such as that found in Drosophila. It remains to be seen if ftz orthologs from ancestral insect species with a long germ band mode of embryogenesis have segmentation functions (Löhr, 2001).

Elegant studies have highlighted the importance of changes in cis-regulatory regions that drive the functional evolution of homeobox genes. The results presented here suggest that the functional evolution of the Hox gene ftz is dependent not only upon such regulatory changes, but also upon changes in protein sequence that modified interactions with specific protein partners. Studies of these molecular modulations in rapidly evolving control genes, such as ftz, provide insight into the developmental and evolutionary mechanisms that promote changes in the body plan and ultimately the diversity of species throughout the animal kingdom (Löhr, 2001).

date revised: 18 February 2024 

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.