![]() |
Background Research Publications Current Lab |
| Sir Philip Cohen |
| E: p.cohen@dundee.ac.uk |
| T: 44 1382 384238 |
| F: 44 1382 223778 |
| Michelle Mulligan PA to Sir Philip Cohen |
| E: m.z.mulligan@dundee.ac.uk |
| T: 44 1382 384238 |
| F: 44 1382 223778 |
Sir Philip Cohen
Research
The Interplay Between Protein Phosphorylation and Protein Ubiquitylation in Regulating Innate Immunity
The goal of my research is to elucidate the signaling pathways that become activated during infection by bacteria and viruses, and to discover how they trigger the production of inflammatory mediators and interferons to combat and destroy these pathogens. Understanding this system is critical, not only because of its importance in defence against infection, but also because the uncontrolled production of inflammatory mediators is a major cause of many diseases, including rheumatoid arthritis, psoriasis, lupus, septic shock and some forms of leukaemia. Drugs that target particular components of these signaling pathways may therefore be of great benefit in treating a number of diseases.
Pathogen-derived molecules activate intracellular signalling pathways when they interact with Toll-like receptors (TLRs) in the plasma membranes and endosomes of immune cells. Signaling “downstream” of TLRs 1, 2, and 5-9, as well as the interleukin-1 (IL-1) receptor, requires the adaptor protein MyD88 and protein kinases of the IRAK family and leads to the activation of the E3 ubiquitin ligase TRAF6. TRAF6 then generates Lys63-linked polyubiquitin (K63-pUb) chains, which bind to the regulatory (TAB2 and TAB3) subunits of a “master” protein kinase TAK1, inducing conformational changes that activate TAK1. The K63-pUb chains also bind to the regulatory (NEMO) subunit of IKKα and IKKβ (termed the canonical IKK complex) inducing conformational changes that facilitate their activation by TAK1 (Fig 1). The canonical IKKs in turn activate the transcription factor NFκB and the protein kinase Tpl2, leading to the production of many inflammatory mediators.
Recently, we discovered that the canonical IKK complex also participates in the activation of the IKK-related kinases, TBK1 (TANK-binding kinase-1) and IKKε. We further showed that, once activated, the IKK-related kinases negatively regulate the canonical IKK-complex, preventing the hyper-phosphorylation of its substrates [1]. We also found that this "cross-talk" between the canonical IKK complex and the IKK-related kinases is lost in cells that do not express the adaptor protein TANK [2], which may explain why TANK-/- mice spontaneously develop autoimmune disease as they age.
We have obtained preliminary evidence that the IKK-related kinases may limit the strength of signaling in the innate immune system in other ways and the discovery of new substrates and physiological roles of these protein kinases is one focus of our current research. These studies have and are benefiting from a collaboration with MRC Technology, which has led to the development of the first potent and relatively specific inhibitor of the IKK-related kinases (MRT 67307) [1].

Figure 1. Model for how the K63-pUb chains produced by TRAF6 activate TAK1, as well as activation and crosstalk between the members of the IKK subfamily.
Mutations in the polyubiquitin-binding domain of NEMO that disrupt binding to polyubiquitin chains, cause a severe human immunodeficiency disease and greatly increased susceptibility to infection by tuberculosis-causing bacteria. Intriguingly, similar polyubiquitin-binding domains are present in several other proteins, termed ABIN1 (A20-binding inhibitor of NFκB), ABIN2 and Optineurin (OPTN) (Fig 2). We have established that, like NEMO, these proteins bind to K63-pUb chains and linear-pUb chains and that polyubiquitin-binding is prevented by mutations equivalent to those that prevent the interaction of NEMO with pUb chains. We have therefore generated knock-in mice in which each protein is replaced with a mutant in which polyubiquitin binding has been disabled by mutating the key aspartic acid residue (highlighted in red in Fig 2) to asparagine. We found that the MyD88-dependent signaling pathways activated by TLR ligands are hyper-activated, and pro-inflammatory cytokines overproduced in immune cells from the ABIN1[D485N] mice. This caused the ABIN1[D485N] mice to develop a lupus-like autoimmune disease as they aged, which could be prevented by crossing them to MyD88-/- mice [3]. The ABIN1[D485N] mice we have generated appear to be a good model for some types of human lupus, because polymorphisms in human ABIN1 have recently been reported to pre-dispose to lupus, psoriasis and other autoimmune diseases in five different human populations.

Figure 2. The Lys63-linked polyubiquitin-binding domain of NEMO is also found in several other human proteins. Mutation of the Asp residue highlighted in red to Asn prevents these proteins from binding to Lys63-linked or linear polyubiquitin chains.
Toll-like receptor 3 (TLR3), which is activated by viral double-stranded (ds) RNA, does not signal via the adaptor MyD88 or the IRAKs, but via a distinct adaptor, called TRIF, while TLR4, which responds to bacterial lipopolysaccharide (LPS), is unique among the TLRs in signaling via TRIF as well as MyD88. The early events in the TRIF-dependent signalling pathway are not well understood but one branch of this pathway culminates in the activation of the IKK-related kinases (TBK1 and IKKε). The IKK-related kinases then phosphorylate and activate the transcription factor interferon regulatory factor 3 (IRF3), stimulating transcription of the gene encoding interferon β. We have identified OPTN (Fig 2) as a binding partner of TBK1 [4] and found that the LPS or dsRNA-stimulated activation of TBK1 and IRF3, as well as the secretion of interferon β, is reduced in macrophages from mice expressing the polyubiquitin binding-deficient mutant OPTN[D477N] [5]. This suggests that the binding of K63-pUb or linear-pUb chains to OPTN is important for the activation of one or more forms of TBK1.
Recently, we discovered that the IKK-related kinases phosphorylate and activate the E3 ubiquitin ligase Pellino 1 in response to dsRNA or LPS [6]. We also found that the TRIF-dependent signaling pathway stimulates transcription of the gene encoding Pellino 1 in response to dsRNA or LPS, leading to greatly increased expression of the Pellino 1 protein [6]. We have generated knock-in mice in which wild type Pellino 1 is replaced by an E3 ligase-defective mutant to study the physiological roles of Pellino 1 in this pathway. We have also generated knock-in mice expressing equivalent mutants of the other Pellino isoforms, Pellino 2 and Pellino 3.

Figure 3. The signaling pathway that leads to the production of Interferon β (IFNβ) and the activation and expression of the E3 ubiquitin ligase Pellino 1 in response to TLR3 and TLR4 ligands that signal via TRIF. Protein kinases are shown in red, components of the Ubiquitin system in green, and receptors and adaptors in blue.
Interest in targeting components of the ubiquitin system to develop drugs to treat disease is increasing [7] and components of the ubiquitin system that play critical roles in regulating innate immune signaling pathways may become an important area of drug discovery in the years to come.
An overview of my contributions to the field of cell signaling before I started to work on the interplay between protein phosphorylation and protein ubiquitylation in regulating innate immunity has been published [8].
Relevant recent references
[1] Clark, K., Peggie, M., Plater, L., Sorcek, R.J., Young, E.R.R., Madwed, J.B., Hough, J., McIver, E.G. and Cohen, P. (2011) Biochem. J. 434, 93-104. "Novel cross-talk within the IKK family controls innate immunity."
[2] Clark, K., Takeuchi, O., Akira, S. and Cohen, P. (2011) Proc. Natl. Acad. Sci. (USA) published online. "TANK facilitates crosstalk within the IKK family during TLR signaling" http://www.pnas.org/content/early/2011/09/22/1114194108.full.pdf+html
[3] Nanda, S.K., Venigalla, R.K.C., Ordureau, A., Patterson-Kane, J.C., Powell, D.W., Toth, R.J., Arthur, J.C.C. and Cohen, P. (2011) J. Exp. Med. 208, 1215-1228. “Polyubiquitin binding to ABIN1 is required to prevent autoimmunity.”
[4] Morton, S., Hesson, L., Peggie, M. and Cohen, P. (2008) FEBS Lett. 582, 997-1002. “Enhanced binding of TBK1 to an optineurin mutant that causes a familial form of primary open angle glaucoma.”
[5] Gleason, C.E., Ordureau, A., Gourlay, R., Arthur, J.S.C. and Cohen , P. (2011)
J. Biol. Chem. published online. "Polyubiquitin binding to optineurin is required for optimal activation of TANK-binding kinase 1 and the production of interferon β.” http://www.ncbi.nlm.nih.gov/pubmed/21862579
[6] Smith, H., Liu, X-Y., Dai, L., Goh, E.T.H., Chan, A-T., Xi, J., She, C-C., Qureshi, I.A., Ruedl, C., Gourlay, R., Morton, S., Hough, J., McIver, E.G., Cohen, P. and Cheung, P.C.F. (2011) Biochem. J. 434, 537-548. “The role of TBK1 and IKKε in the expression and activation of Pellino 1."
[7] Cohen, P. and Tcherpakov, M. (2010) Cell 143, 686-93. “Will the ubiquitin system furnish as many drug targets as protein kinases?”
[8] Cohen, P. (2009) J. Biol. Chem. 284, 23891-23901. “Keep nibbling at the edges.”
