Targeted DamID (TaDa)
Targeted DamID (TaDa) (Southall, et al., 2013) enables cell-type specific profiling of DNA-binding proteins with fine spatial and temporal resolution. TaDa is based on DamID (van Steensel and Henikoff, 2000), a technique which takes advantage of the lack of significant adenine methylation in eukaryotes. By fusing a bacterial DNA adenine methylase (Dam) to any protein of interest, only the regions bound by the protein will be methylated at the sequence GATC. Methylated GATC fragments are then selectively amplified via methylation-sensitive digestion and ligation-mediated PCR. By sequencing these fragments, the binding profile of the protein of interest can be obtained.
TaDa combines DamID with the GAL4 system in Drosophila, allowing Dam-fusion proteins to be expressed only in specific cell-types at precise temporal windows. TaDa is simple, robust, sensitive and highly reproducible, and works over a wide variety of different cell types and conditions. A further advantage of TaDa is that Dam-fusion proteins are expressed at almost undetectable levels such that cell fate remains unaffected.
(Note: this protocol now uses Zymo and Machery-Nagel kits. The old protocol using Qiagen kits is still available; Qiagen kits work perfectly well, but are more costly.)
Our current lab protocol for TaDa (version 2017-Oct-09). This protocol is based on published work in Southall et al., Dev cell, 2013; Marshall and Brand, Bioinformatics, 2015; Marshall et al., Nature Protocols, 2016.
The protocol is regularly updated, so please check for updates.
The original TaDa publications used mCherry as the primary ORF in the TaDa bicistronic expression system. We have now created new vectors (pTaDaG and pTaDaG2) that use membrane-bound GFP as the primary ORF, and as such clearly label the driving pattern within any TaDa experiment. The TaDaG2 vector is freely available upon request, and we are happy to send the TaDaG2-Dam control flies.
The original mCherry-based pUAST-attB-LT3-Dam vector sequence is available through GenBank.
Some notes on data representation
DamID datasets are somewhat different to ChIP datasets. In general, we recommend that data is presented:
- As a log2(Dam-fusion/Dam) ratio
- At GATC fragment resolution
- Without obscuring or removing negative peaks (unbelievably, people actually do this; more unbelievably, this sometimes makes it through peer-review. Journal editors, please note!)
If datasets are compared between different cell types (and/or different conditions), scaling of the data (by dividing each dataset by its standard deviation) is appropriate before comparison. In general, it’s worth noting that DamID data is at most semi-quantitative. Direct quantitative comparisons are best made when using identical drivers and conditions.