The developmental expression dynamics of Drosophila melanogaster transcription factors

Abstract

Background: Site-specific transcription factors (TFs) are coordinators of developmental and physiological gene expression programs. Their binding to cis-regulatory modules of target genes mediates the precise cell- and context-specific activation and repression of genes. The expression of TFs should therefore reflect the core expression program of each cell.

Results: We studied the expression dynamics of about 750 TFs using the available genomics resources in Drosophila melanogaster. We find that 95% of these TFs are expressed at some point during embryonic development, with a peak roughly between 10 and 12 hours after egg laying, the core stages of organogenesis. We address the differential utilization of DNA-binding domains in different developmental programs systematically in a spatio-temporal context, and show that the zinc finger class of TFs is predominantly early expressed, while Homeobox TFs exhibit later expression in embryogenesis.

Conclusions: Previous work, dissecting cis-regulatory modules during Drosophila development, suggests that TFs are deployed in groups acting in a cooperative manner. In contrast, we find that there is rapid exchange of co-expressed partners amongst the fly TFs, at rates similar to the genome-wide dynamics of co-expression clusters. This suggests there may also be a high level of combinatorial complexity of TFs at cis-regulatory modules.

Figure 1

Patterns of temporal utilization of site-specific transcription factors. (a) Binary representation of TF expression according to the BDGP in situ database: black, expressed; white, not expressed. Expression behavior can be roughly categorized into six classes, no embryonic expression, diverse expression, early TFs, late TFs, those with continuous zygotic expression and continuous expression (including maternal contribution). The four largest classes are detailed in the main text. (b) Temporal utilization for different TF families. Absolute numbers of expressed TFs are provided underneath each panel. Note that roughly half of the TFs per family have been covered by the BDGP. HTH: helix-turn-helix.

Figure 2

Temporal utilization of site-specific transcription factors. (a) Percentage of the TF repertoire used during embryonic development. Samples were taken at various degrees of granularity (red, BDGP in situ database; blue, embryonic microarray gene expression time-course; green, active transcription map). Although there are differences in the absolute number of transcribed TFs, there is good agreement in the general trends. About half of the TFs are maternally contributed. The peak of TF expression is around 10 to 12 h AEL, and the number of expressed TFs declines towards the end of embryogenesis. (b) Proportion of TFs in the group of expressed genes. The ratio of TFs versus non-TFs is highest between 2 and 6 h AEL. This coincides with the functional compartmentalization of the germ layers leading to the development of various organ systems.

Figure 3

Spatio-temporal over-representation of transcription factors in embryonic development. This figure uses the slim representation of the anatomical ontology from Tomancak et al. [30] with the same color codes. The number of TFs expressed at the given spatio-temporal coordinate is provided in brackets, and the decimal number indicates the Z score for the over-representation of TFs in comparison to non-TFs. Statistically significant enrichment of TFs versus non-TFs (Z > 3) is highlighted in red.

Figure 4

Tissue specificity of site-specific transcription factors. (a) Proportion of ubiquitously expressed genes along the developmental time axis, separated into TFs and non-TFs. The expression patterns of TFs and non-TFs are significantly different between stages 4 and 12 (P < 10^-4), with less ubiquitous and more restricted expression patterns for the TFs. (b) Adult specificity. Gene expression clusters with adult tissue-specificity were queried for their specificity in late embryonic development (BDGP dataset stages 13 to 16). Shown is the expression breadth as the number of anatomical structures used in the annotation, as well as the percentage of genes with frequently encountered annotations. For example, for the 40 TFs primarily expressed in adult head, crop and gut, there is embryonic expression information for 23 of them. These genes are most frequently annotated for expression in the midgut (35%) and dorsal epidermis (35%). Interestingly, very frequently, adult tissue-specific TFs are involved in embryonic nervous system development. CNS: central nervous system; SNS: stomatogastric nervous system; VNC: ventral nerve cord.

Figure 5

Modularity of transcription factors. (a) Expression of Twi and Mef2 and their co-occurrence at genomic sites at different developmental stages. Compared are stages 5 to 9 (approximately 2-4 h AEL) and stages 9 to 11 (approximately 4-6 h AEL). In the earlier time frame, strong mesodermal Twi expression but only weak Mef2 expression can be observed. Half of the genomic sites that see occupancy by either TF at some stage during development show only Twi occupancy (43%) or Twi/Mef2 (8%) double-occupancy. By 4 to 6 h AEL, the majority of the mesoderm develops into muscle, which is characterized by strong Mef2 expression. Genomic occupancy follows this trend, showing that most sites with previous double-occupancy maintain that state while one-third of the previous Twi-only sites gain a Mef2 partner site. Overall, more than half of the Twi sites then show double-occupancy. The images showing TF expression are reproduced from [64]. ChIP data are from [22]. (b) A map of potential pairwise interactions between transcription factors. For all of the 373 TFs for which BDGP expression data are available, the frequency of co-occurrence in at least one tissue is color-coded (as fraction of sampled time frames). Most TFs are co-expressed in at least one spatio-temporal coordinate. (c) Transcription factor expression exhibits modular behavior. More modules of two (left panel) or three (right panel) transcription factors show precisely the same expression from one to the next developmental stage than is to be expected at random. Interestingly, this is an intrinsic feature of all genes and not a specific property of the transcription factors. Error bars indicate the standard deviation observed in the random experiments.