Graphic of networks and cells
Categories: Sanger Science30 November 20124.4 min read

Making messengers work hard for their money

30 November 2012

By Marija Buljan

Even though all cells in the human body have the same genome, there are several hundred different cell types. Each cell type is specialized for distinct functions and belongs to a different tissue type. The attributes that define cells and tissues are largely built upon the activity – the expression – of specific genes.

Cartoon: differently spliced forms can lead to different protein interactions

Tissue-specific splicing can produce isoforms that have different interacting protein segments, which can result in the rewiring of an interaction network in the respective tissues.

Genes are copied (transcribed) into ephemeral messengers called transcripts, which direct the production of proteins. But each gene might produce several overlapping transcripts that contain different but related sequences. The individual genome sequence segments that are transcribed and joined together through splicing in the final messenger transcript are called exons. Following this complex processing, transcripts are translated into proteins – complex molecules that do most of the work in cells.

Thanks to advances in sequencing technologies, we can now examine the transcripts in different tissue types across the entire genome – thus examining all transcripts (the transcriptome).

As a result, we are starting to appreciate that not only the expression of specific genes but also the expression of differently spliced  transcripts plays a role in the development and definition of a tissue. Thanks to splicing, certain exons can be included in some tissues, but skipped in others. However, our understanding of the functional impact of the observed splice events is still limited.

In our study, we addressed this by using the available transcriptome data for several different tissues and investigating how alternative splicing shapes protein function in different tissues.

We first used descriptions of gene structures from gene annotations in the Ensembl database to identify tissue-specific exons that were likely to encode functional information corresponding to a protein segment when translated. We then used other publicly available data sets to look at the structural, functional and evolutionary features of these tissue-specific protein segments alone. We also extracted data on interaction properties and functional characteristics of the whole genes that encoded such segments from established data sets. Together, these results allowed us to analyse the likely impact of the tissue-specific changes in the produced protein sequences.

When the same gene is programmed, via splicing, to produce different protein products in different tissue types, the general expectation is that these will be functionally different proteins. Most proteins are composed of structurally and functionally independent units – protein domains, so we searched whether inclusion of a tissue-specific exon possibly added a new domain to the protein.

In contrast to those exons that are always present in final transcripts, we found that exons that are included in the tissue-specific way very rarely added a new domain and thus in general did not tend to dramatically affect protein structure. In fact, these exons frequently encoded protein regions that are intrinsically disordered.

However, even though they themselves are without a structure, those tissue-specific protein segments still  encompassed conserved binding motifs for other proteins and sites of regulation – posttranslational modifications – that are both important in activity. Equally interestingly, the tissue-specific exons were highly conserved between mouse and human, suggesting selective pressure has favoured their conservation.

We also found that the genes that encoded such exons are frequently central in protein interaction networks and many of them are linked to roles in signalling or in cancer. The protein partners for their interactions tended to be different in different tissues, thus suggesting that inclusion of a specific segment can decide on the partners of choice.

Taken together, these results suggest that inclusion of different, tissue-specific protein segments can modulate the binding properties of a protein and affect the choice of interaction partners, depending on the tissue of gene expression. This can thus result in the tissue-specific rewiring of interaction networks.

These results unravel the molecular principles by which the function of a single gene can be adapted to different contexts. Moreover, they show how even seemingly small changes in protein sequences can be carefully regulated and crucial for the proper functioning of cellular networks.

Our work calls for more studies in greater detail of the individual versions of the same gene in a tissue-, developmentally- or response-specific manner. Studies of the isoforms of medically relevant genes will be particularly interesting. Individual isoforms can have functions adapted to specific cell-types or environmental responses. Thus, mutations in the tissue-specific segments or in the mechanisms that regulate their inclusion can underlie tissue-specific diseases.  However, this also opens hope of the development of drugs specific to each gene variant that would have fewer side effects compared to the drugs that target constitutively spliced regions.

Marija Buljan did her PhD at the Sanger Institute with Dr Alex Bateman and she now works as a postdoctoral researcher in Madan Babu’s group more...

Research paper

Marija Buljan, Guilhem Chalancon, Sebastian Eustermann, Gunter P. Wagner, Monika Fuxreiter, Alex Bateman, M. Madan Babu (2012) Tissue-Specific Splicing of Disordered Segments that Embed Binding Motifs Rewires Protein Interaction Networks. Molecular Cell 46: 871–83

Related Links