Altering DNA packaging to regulate gene transcription
Oliver Hendy, a PhD student in the lab of Alex Stark, is in the last stages of his doctoral studies. For his thesis, he studied how a group of proteins called ‘chromatin remodelers’ change the way DNA is packaged, and how this modulates gene transcription in fruit fly cells. In his latest work, now published in the journal Molecular Cell, he found that two major sets of genes require the help of distinct chromatin remodelers to be transcribed.
What is chromatin and why does it need to be ‘remodelled’?
DNA is often represented as a stretched-out double helix, but in reality, it’s very neatly packaged in the cell nucleus in a form called chromatin. The DNA molecule is repeatedly wrapped around protein complexes, a bit like a rope along which you would tie knots at regular intervals. These ‘knots’ are called nucleosomes and can make it difficult for enzymes to access the DNA sequence and transcribe it into messenger RNA. For gene transcription to work properly, nucleosomes have to be moved around to expose the right regions – that is one way that chromatin can be remodelled. Chromatin remodelers are proteins that change the way DNA is packaged when it’s needed. In my project, we focused on the way they influence gene transcription.
What did we know about chromatin remodelers and transcription when you started your PhD?
In eukaryotes, there are four families of chromatin remodelers: SWI/SNF, Iswi, Ino80 and Chd. There have been many studies on the roles that chromatin remodelers play in transcriptional activation, particularly in yeast, where each remodeler plays a distinct role in nucleosome organisation. But the yeast genome is relatively simple compared to that of a multicellular organism, like an insect, or a mouse, and until recently, the roles they played there were not clear. We wanted to take a systematic approach and study the direct regulatory roles of remodelers in a single cell type in the fruit fly, in a highly temporal manner. We perturbed the remodelers in living cells and measured the direct effects on transcription about six hours later.
How did you do that?
We used a method that was inspired from plant physiology. Plant cells have evolved a system in which they can degrade some unwanted proteins on demand, when a specific hormone called auxin is nearby. The enzyme that degrades these proteins only recognises them if they are labelled with a specific protein sequence, a sort of “hit me” tag, that we can genetically attach to any protein of interest. So what we did is that we tagged chromatin remodelers and added auxin to our cells just before measuring the impact it would have on ongoing transcription. This system hasn’t been around for long, it brought new opportunities for that kind of project.
What type of fruit fly genes did you focus on?
We looked at the transcription of two major, distinct groups of genes: ‘housekeeping’ genes, which are transcribed in all cells, and ‘developmental’ genes, which are only active in certain tissues at certain times. These two groups have different kinds of promoters – the region of DNA where the gene’s transcription begins. They also have different types of enhancers – the distal regions of DNA that activate transcription. The organisation of nucleosomes around the promoters is also distinct, which gave us a clue that there might be differences in their interactions with chromatin remodelers.
What did you find out about chromatin remodelers?
We discovered that the transcription of housekeeping genes and developmental genes depends on separate sets of chromatin remodelers. The promoters and enhancers of housekeeping genes are mostly unpacked, readily accessible by default at all times. Two remodelers – called Iswi and Ino80 – fine-tune the transcription of these genes by moving nucleosomes downstream of their promoters.
What we saw with developmental genes was completely different, though. While the loss of Iswi and Ino80 didn’t affect their transcription, SWI/SNF perturbation caused most developmental genes to shut down. The enhancers of developmental genes tend to bind nucleosomes easily if they aren't kept in check, so when SWI/SNF was absent, these enhancers became crowded with nucleosomes and were therefore inaccessible for the transcription machinery. Surprisingly, when we depleted SWI/SNF, it had no effect on housekeeping gene transcription or accessibility, which shows that these programs have different remodeler requirements.
Where do you see this project going next?
Within the families of chromatin remodelers we studied, there are subtypes that we could look into. It would be interesting to follow up and see if there is any further division of labour between those: what they do, which genes they affect, and whether this is conserved in other species.
My colleagues in the lab of Alexander Stark have been of huge help in this project, and I think generally inducible depletion will continue to be a powerful approach to decipher the roles of the many activators in transcription.
Original publication
Further reading
Milestone in understanding regulation of gene transcription
Harnessing artificial intelligence to predict and control gene regulation