How is magnet based bead capture used for indexing-first ChIP-seq?

Question

The context for this question can be found in this paper by Lara-Astiaso et al. Here, to study H3K4 trimethylation in various cell types, they

1. sort cells
2. sonicate chromatin
3. immobilize chromatin on Dynabeads with anti-H3 antibody
4. index chromatin (barcode)
5. pool samples
6. perform H3K4me3 ChIP
7. sequence

I am interested in the purpose of step 3 and how that contributes to step 4. What could be the purpose of essentially adding another IP step?

The Supplementary Materials (where all the methodology is described) mentions that "magnet based bead capture was used to efficiently add, wash and remove the different master mixes used in the indexing process." I'm somewhat confused as to how this adds efficiency, as I thought barcoding prior to sequencing is typically performed as a ligation step without selection (they perform the H3K4me3 ChIP after the indexing step anyway, so why select for histone H3 chromatin first?).

Answer 1

The anti-H3 antibody will precipitate all chromatin. They could have used an antibody against any histone monomer; that they initially used anti-H3 and then later used anti-H3K4me3 is coincidental. The idea behind the first immunoprecipitation is to immobilize the chromatin on the beads to allow easy recovery of pure chromatin after each step of the indexing process. These steps are seen in the supplementary material (4. Chromatin Indexing). For example, after immobilization, they perform end repair using specific buffers and T4 polymerase and polynucleotide kinase. These enzymes must be removed and the buffer exchanged before the subsequent indexing. The beads facilitate this process.