Role of STAT1 Cooperative DNA Binding in Interferon Signalling

This proposal aims to investigate how cells manage to respond quickly to sudden changes in their environment. An example of the need for rapid responsiveness is the cells of the body's immune system, which protect us against microbes, i.e. viruses and bacteria, that attempt to establish infections. Once the immune cells have sensed the presence of intruding microbes they release signalling molecules with diverse roles. We are concerned with signalling molecules that interfere with the ability of viruses and bacteria to multiply and therefore are called interferons (IFNs). Since their discovery 50 years ago, the IFNs have come to be regarded as the model signalling molecules with therapeutic efficacy not just against microbial infections, but moreover in the treatment of cancers and neurodegeneration. However, their very potency can also cause problems manifesting in autoimmune disorders or severe depression. It is thus not surprising that the interferons and their activities are still intensely studied not least to realize fully their clinical potential. A lot has been learned though about how the interferons function. One important aspect is that IFNs exert their influence on cellular behaviour by regulating the activity of genes. Yet interferons do not enter the cells but attach to receptors in the cell membrane. Here, they activate a protein called STAT1, which then carries the signal to the genes in the cell nucleus. STAT1 is a transcription factor. Transcription factors bind to DNA -the material containing the genes- in order to regulate the expression of genes. Transcription factors can be regarded as the keys that allow extracellular signalling molecules, like the IFNs, access to their target genes. How STAT1, and for that matter transcription factors in general, find specific sites that are few in number and hidden in a massive excess of unspecific sites, is a formidable, incompletely understood, problem in cell signalling. A way forward was suggested when we discovered that STAT1 can bind to DNA not only in single copies, but that it can also polymerise into larger structures. These polymers are believed to then access specific sets of genes within the IFN target gene repertoire. It was suggested that polymerization is widely important for transcription factors to steer the behaviour of individual cells and to orchestrate complex phenomena like immunity, but direct experimental evidence is scarce.
To address this lack of knowledge we have expanded upon a number of crucial discoveries in the PI's lab regarding STAT1 DNA binding and polymerization, and have created a genetically modified mouse strain with normal STAT1 function, except that it is specifically deficient in polymerization. It is thus a promising model for studying the effects of polymerization on gene regulation and cellular behaviour. On the molecular level we hope to identify the target genes of STAT1 single molecules and polymers, and the "DNA code" by which STAT1 discriminates between them. As such the proposed research aims at solving a fundamental biological problem. Mice moreover are a tested model for the functioning of the human immune system. Our mutant STAT1 mice thus enable us to explore the effects of polymerization on protection against bacterial and viral infections, central tenets of immunity. As interferons are effective in diverse clinical settings, an improved understanding of how they achieve target gene specificity is critical to better tailor their use according to clinical need. The possibility that STAT1 inhibitors may limit excessive inflammation is another potential clinical application. Current pharmaceuticals inhibit STAT1 indiscriminately. Our pilot work indicates that targeting STAT1 polymerization, in contrast, might hold the prospect for novel STAT1 inhibitors with higher specificity and fewer side effects. The proposed experiments will help to decide whether these intriguing possibilities are in fact realistic.

The proposed research seeks to answer the question whether cooperative DNA binding of the transcription factor STAT1 is critical for discriminating signalling of type I IFN from that of type II IFN. Given the often antagonistic activities of both types and their demonstrated different clinical effectiveness against viruses, malignant cell growth, and Multiple Sclerosis, the answer is of importance both for cytokine biology and interferon therapy. Our studies depend upon a novel knockin mouse strain generated in our laboratory. These mice express point-mutated STAT1, whose ability to interact with DNA is unchanged, but which cannot sustain homotypic protein interactions, so that DNA binding cooperativity is lost. This study is thus distinguished by the fact that a transcription factor deficient in cooperative DNA binding is being tested only for the second time in a mammalian system for its ability to carry out its normal in vivo functions. Extensive transcriptome analyses of type I and type II IFN stimulated macrophages expressing WT or the mutated STAT1 will be done using RNA-sequencing. These studies are complemented with STAT1 chromatin binding data from ChIP-sequencing in order to identify the sequence motifs that distinguish the targets for STAT1 dimers and multimers both in type I and type II IFN signalling. To examine the contribution of DNA affinity differences to target gene selectivity we propose to compare the DNA-binding profiles of WT and mutant STAT1. This is done with total internal reflectance fluorescence protein binding microarrays (TIRF-PBM) containing 1 kB-long promoter regions from 200 different IFN-regulated genes. In order to assess the role of STAT1 cooperativity in antimicrobial immunity, and possible differences between type I and type II IFNs in this regard, we propose to study the progression of infections with both bacterial and viral pathogens.

The proposed research is addressing an unsolved problem of basic biology and its manifestations concerning interferon signalling. Hence it is foremost academic impact that this work will generate. Based on our results this work is likely to change current scientific thinking regarding fundamental control mechanisms in gene expression. It is presently accepted that cooperative DNA binding, with its resultant switch-like "on-off" transitions in promoter occupancy, is suitable for the generation of sharp transcriptional demarcations typical of developmental decisions. Responses to extracellular stimuli such as interferons, in contrast, are assumed to require transcription factors that function in a non-cooperative, i.e. "graded" manner to deliver graded transcriptional outputs commensurate with the varying strengths of the input signals. However, our data indicate that cooperative DNA binding is indispensable for promoter occupancy of STAT1 homodimers in type II IFN signalling. In contrast, STAT1 cooperativity is not required for the functioning of type I IFN-induced ISGF3, which contains STAT1 heterodimers. Yet another variation of cooperativity is manifest for STAT5 homodimers, which require cooperativity for only a very small subset of target genes. Thus, not only do transcription factors that seemingly function very similarly in fact differ starkly in their dependence on cooperativity, but even the same transcription factor can function very differently in the context of different stimulation conditions. It is furthermore becoming clear that the need for cooperativity cannot be inferred from the presence of tandem canonical recognition sites in the gene promoters. Together, these discoveries will undoubtedly stimulate research into the role of DNA binding cooperativity for extracellular signal-regulated transcription factors. With regard specifically to interferon signalling, the novel mutant STAT1 mouse strain that we have generated will allow more precisely to distinguish type I from type II interferon activities. To unravel the shared and distinctive properties of the interferons requires in depth characterization of intracellular signalling. We think that the animal model that has been developed is a very useful tool to achieve this. We are using clinically relevant pathogens in our experiments and will study influenza-induced alveolar pathogenesis, which is directly relevant to human health. It should also be remembered that unlike other animal models the interferons themselves are already used in humans. So at the very least translation of our data into better understanding of the human field will have an impact. The proposed work continues existing and initiates new collaborations between leading labs working on STATs, transcription regulation, infectious diseases, and bioinformatics. These partnerships will benefit not least two UK Universities and thus contribute to raising the profile of UK science further.

A potential medium-term beneficiary of the proposed research is the Pharmaceutical Industry. Immunotherapy approaches concerned with modulating STAT1 activity are already actively pursued to treat viral infections, cancers, and auto-immune diseases such as Multiple Sclerosis to name but a few. However, the presently available STAT1 inhibitors inactivate STAT1 indiscriminately and hence can compromise anti-viral protection, which might be avoided if cooperativity was targeted. Moreover, the efficacy of current activation immunotherapies such as IFN-alpha may be reduced by the concurrent activation of STAT1 homodimers, as suggested by the often adverse effects associated with IFN-gamma. Inhibitors of STAT1 cooperativity may function as an adjuvant to specifically suppress STAT homodimer activity during IFN-alpha therapy. The proposed research would thus help clarify whether targeting STAT1 cooperativity can lead to the development of more selective STAT1 inhibitors with fewer side effects.