Are you looking for a refreshing new nuclear organization assay this summer? Well look no further than SPRITE, it will quench your thirst for a better DNA conformation assay. There are a lot of methods for assessing higher-order DNA structure, but they mostly fall into two categories: proximity ligation and in situ imaging. Proximity ligation (ex. Hi-C) ligates spatially close genomic regions, which are then sequenced. Hi-C shows us that most DNA-DNA interactions are intra-chromosomal. On the contrary, in situ imaging of DNA, RNA, and protein in the nucleus show many inter-chromosomal interactions organized around nuclear bodies, ex. rRNA genes into the nucleolus. These disparate results suggest that proximity ligation is missing many inter-chromosomal interactions.
To address this problem, the laboratory of Mitchell Guttman at Caltech developed Split-Pool Recognition of Interactions by Tag Extension (SPRITE). The process involves:
- Crosslinking of DNA, RNA, and protein followed by nuclei isolation and chromatin fragmentation.
- Barcoding interacting molecules using a split-and-pool approach, meaning the DNA is split into 96 wells, tagged, then pooled together. This is repeated iteratively to generate unique barcodes.
- Interactions are then identified by sequencing and matching reads with identical barcodes. By using different bin sizes, the resolution can be honed from fine- to large-scale.
Since SPRITE doesn’t use proximity ligation, it can detect wider spaced DNA-DNA interactions. Furthermore, since it doesn’t rely on whole-genome amplification, it can also identify DNA-RNA interactions. In their new publication, the SPRITEly team report their new method and test it on mammalian cell types previously examined by Hi-C and in situ imaging.
Here’s what they found:
- In both mouse embryonic stem cells and human lymphoblastoid cells (GM12878), fine-resolution SPRITE and Hi-C are similar, where they preferably identify intra-chromosome interactions
- SPRITE identifies known Hi-C structures including chromosome territories, compartments, topologically associated domains, and loop structures
- By simultaneously examining DNA and RNA interactions, they found two core nuclear organizing hubs: the nucleolus (rRNA production) and nuclear speckles (splicing and RNA metabolism)
- Gene-dense regions that are highly transcribed are organized around nuclear speckles
- Gene-poor regions near centromeres are organized around the nucleolus.
- A large portion of the genome is preferentially positioned relative to these nuclear bodies
- Regions close to the nucleolus are linearly close to their centromere
- Regions that are closer to nuclear speckles strongly associate with high levels of active Pol II transcription
Taken together, the application of this exciting new technique shows that nuclear bodies shape the overall organization of DNA in the nucleus; rRNA genes associate with the nucleolus while transcribed Pol II genes associate with nuclear speckles. It is still unclear whether spatial organization around nuclear bodies directly impacts transcription or whether it is a consequence of Pol II occupancy within a genomic region. These results demonstrate the broad applicability of SPRITE for mapping local to large-scale genomic organization. Given that it is faster, easier, and more accurate than similar methods, it may prove very appealing in nuclear architecture research. So, if you’re looking for a refreshing new take on interaction mapping, grab a SPRITE and head for the bench!
Get all the refreshing details over at Cell, July 2018.