Bioinformatics and Network Biology for Unravelling Regulation of T Helper Cell Differentiation

Over the past ten years or so, the Teichmann group and others have shown that the ensemble of transcriptional regulatory interactions in a cell or organism forms a network, which can be described and analysed using bioinformatics, graph theory and mathematical modelling to reveal profound new insights (e.g. Teichmann & Babu, Nature Genet., 2004; Luscombe et al., Nature, 2004). Transcriptional regulatory networks control development and differentiation (e.g. Soneji et al., Ann. N.Y. Acad. Sci., 2007), and here we propose to study the differentiation of T helper cell types of the immune system. The T helper (Th) cell system, one terminal branch of haematopoiesis (Reiner, Cell, 2007), is an experimentally amenable model of differentiation with the added advantage that the various Th cell types can be obtained in large amounts as homogeneous populations. It is also of fundamental importance to human health. The Th cell system regulates immune responses, and impairment of the T helper cell compartment leads to dramatic immune deficiency as seen in late-phase HIV infections (7). Misbalanced differentiation from naïve T helper cells to one of the currently known subtypes, Th1, 2 and 17 and iTregs, is causally involved in diseases like autoimmunity and allergy (Reiner, Cell, 2007). The whole Th differentiation process from naïve Th cells to the three mature subtypes (Th1, 2 and 17) as well as iTregs, can be followed in large numbers of primary cells, and can be accurately simulated in vitro from primary T cells (Reiner, Cell, 2007). The signalling within and between cells in tissues that takes place in most animal developmental and differentiation process is limited in Th differentiation, as the cells are practically not matrix-associated, and their inter-cell interactions are well characterised and can be simulated in vitro (Reiner, Cell, 2007). Therefore, the Th system provides an ideal model to study the transcriptional changes ensuing upon differentiation. There are several fundamental questions about the role of transcriptional regulation in differentiation that can be addressed in this system. First, only a handful of transcription factors (TFs) have been identified for each subtype. Therefore, we propose to identify new candidate transcription factors for each Th subtype by deep sequencing (RNA-seq) of each subtype in a timecourse, and subsequent bioinformatics analysis of the data. This will provide insight into the complexity of the network underlying differentiation into each subtype. Secondly, we will look for binding sites of these transcription factors using both computational screens and Chromatin Immunoprecipitation followed by sequencing (ChIP-seq). It will be crucial to have ChIP-grade antibodies for these transcription factors. We know already from published microarray data that there are will be a number of transcription factors in Th cells that are poorly characterised, with either no antibody or no ChIP-grade antibody commercially available. This is where antibody development is essential, and where Abcam will be instrumental in driving the project forward. These ChIP-seq and computational analyses will shed light on the molecules involved and the topology of their interactions. Thirdly, by superimposing ChIP-seq data of histone modifications on these transcriptional regulatory interactions, we will evaluate the role of chromatin-level regulation in the differentiation process.