The strategy of this protocol is to coat dissected Drosophila imaginal discs or tissues with ConA magnetic beads, and perform all washes and buffer changes by magnetic capture of the sample. This method uses small numbers of imaginal discs, and uses digitonin to gently permeabilize the unfixed tissues. Finally, the protocol uses Ampure beads to efficiently recover DNA, avoiding the risk of contamination by phenol or chloroform. Bottles where larvae are not too dense provide well-fed crawling larvae that are easiest to dissect and provide the most tissue.
The protocol workflow is as follows:
Day 1: Larvae to primary antibody incubation
Preparing working solutions and conA beads (Steps 1-4, 15')
Dissecting larvae (Steps 5-6, 15')
Binding tissues to beads (Step 7, 10')
Binding primary antibody (Steps 8-10, O/N)
Day 2: primary antibody incubation to DNA recovery
Binding primary antibody (Steps 11-12, 10')
Optional: binding secondary antibody (Steps 13-16, 1 hr 15')
nuclease tethering (Steps 17-20, 1 hr 20')
DNA cleavage (Steps 21-25, 1 hr 5')
DNA recovery (Steps 26-36, 3 hrs 30')
Preparing Libraries, sequencing, data processing and analysis is performed as perviously described for CUT&RUN protocols (eg. Skene et al (2018). Targeted in situ genome-wide profiling with high efficiency for low cell numbers. Nature Protocols 13:1006-1019. doi: 10.1038/nprot.2018.015.
The success of CUT&RUN depends on the affinity of an antibody for its target and its specificity under the conditions used for binding. Because antibodies bind to their epitopes in the solid state using CUT&RUN, we expect that antibodies successfully tested for specificity by immunofluorescence (IF) will be likely to work with CUT&RUN, with the caveat that IF generally involves fixation, whereas formaldehyde fixation decreases the efficiency of CUT&RUN.
One of the limitations of working with small amounts of tissues is that the amount of DNA recovered can be very low, such that analysis even by sensitive capillary electrophoresis or picogreen assays (e.g. Agilent Tapestation and Qubit) are problematic. In addition, high resolution mapping techniques that cleave a minimal footprint are not suitable to PCR-based analysis of known binding loci, as it is not commonly possible to design ~50 bp PCR amplicons. As such, we recommend using a positive control antibody that targets an abundant epitope and therefore the DNA can be readily detected. We have successfully used a rabbit monoclonal antibody raised against the H3K27me3 histone modification, with capillary electrophoresis showing with the amount of cleaved fragments being proportional to the number of starting cells. A nucleosomal ladder is expected by Tapestation or other sensitive electrophoretic analysis method, and the use of a monoclonal antibody avoids potential lot-to-lot variation that can complicate troubleshooting. For less abundant epitopes such as trnascription factors, we often do not detect any cleaved fragments by Tapestation, yet successful libraries can be prepared from these reactions. As a negative control, we recommend the use of a non-specific rabbit IgG antibody that will randomly coat the chromatin at low density without sequence bias. We do not recommend a no-antibody control, as the lack of tethering increases the possibility that any slight carry-over of pA-MN will result in preferential fragmentation of hyper-accessible DNA. We recommend activating the nuclease for 30 minutes as a starting point that can be tailored based upon epitope abundance and antibody concentration.
In the standard CUT&RUN protocol we recommend allowing the cleaved chromatin complexes to diffuse out of the nuclei, thereby permitting simple isolation of the cut DNA from the supernatant fraction with the undigested genome retained in the intact nuclei. In such cases, DNA extraction and size-selection of fragments below ~700 bp can be used for profiling.