Decoding the grammar of gene regulation
Animal genomes harbour thousands of genes. Within the genome, DNA sequences called ‘enhancers’ activate gene expression in a tissue-specific manner. The lab of Alexander Stark has made it their mission to investigate the ‘grammar rules’ that link an enhancer’s sequence and its activity. In a study now published in the journal Genome Research, they show that enhancer activity is encoded through a combination of their sequence motifs’ identity and context.
Our genes provide the instructions to produce all the proteins that our bodies need to function. While all tissues in one body carry the same genes, they do not use them uniformly. Genes in different tissues are activated on-demand thanks to regulatory DNA sequences called ‘enhancers’, the “on” switches of gene transcription.
Enhancers rely on a series of short sequences called ‘motifs’ that are recognised and bound by regulatory proteins to modulate gene transcription. In a study now published in the journal Genome Research, scientists in the lab of Alexander Stark have investigated how these motifs contribute to enhancer activity.
Like any other DNA sequence, enhancer motifs are made of the four-letter code of DNA: A, T, G, and C. Within an enhancer, motifs are comparable to words arranged in a sentence. “We wanted to know how much of a motif we can change without affecting enhancer activity – this is one more step towards understanding the grammar of enhancer sequences,” explains co-first author Franziska Reiter, a student in the Vienna BioCenter PhD Program. “We created a comprehensive screen in which we randomised eight letter-long motifs in seven enhancer positions and observed the consequences on enhancer function. How many letters in a word can you change without altering the meaning of the sentence?”
The scientists found that altering the composition of motifs can affect their power over enhancer activity in three ways. Some letter combinations decreased enhancer activity while others boosted them, just as a single word can modulate the power of a sentence: turning “I’m hungry” into “I’m starving”, or the other way around. A third type of motifs acted like synonyms and only maintained enhancer activity, suggesting there is some flexibility to motif content.
Once the team had studied the importance of motif sequences under fixed conditions, they turned the problem around: they copied and pasted eight motifs in 500 different locations in enhancers to evaluate whether the motifs’ context influences their function.
Circumstances did matter: while some motifs were strong activators in most contexts – suggesting they might be general activators – others failed to work in a foreign context, or when they were too close to other specific motifs.
“We tested a few motifs in many contexts, in fruit fly cells and in human cells, to test if the rules we found were applicable across species,” says Bernardo Almeida, co-first author and also a student in the Vienna BioCenter PhD Program. “Although some rules are similar, others differed between species or even between motif types. Our findings show that the context of a motif plays an important role in its function.”
The study demonstrates that enhancer activity is encoded through a complex interdependence between motifs and context. The underlying experimental approach complements DeepSTARR, a deep learning algorithm developed by the Stark lab that predicts enhancer activity from their DNA sequence with high accuracy.
“Mutations in enhancer sequences can only be understood in the context of the grammatical features that we identified experimentally and with DeepSTARR,” says Alexander Stark. “My lab combines experimental and artificial intelligence approaches to decipher the rules of gene regulation, and this is an important step in this direction.”
Original publication
Franziska Reiter*, Bernardo P. de Almeida*, Alexander Stark: “Enhancers display constrained sequence flexibility and context-specific modulation of motif function”. Genome Research (2023), DOI: 10.1101/gr.277246.122.
*These authors contributed equally to this study.
Further reading
Harnessing artificial intelligence to predict and control gene regulation
About the Vienna BioCenter PhD Program
The work underlying this publication was done by doctoral students of the Vienna BioCenter PhD Program. Are you interested in a world-class career in molecular biology? Find out more: https://training.vbc.ac.at/phd-program/