University of Dundee

Professor Helen Walden

Specificity and disease in the ubiquitin system
Professor of Protein Structure & Function and Programme Leader
School of Life Sciences, University of Dundee, Dundee
Full Telephone: 
+44 (0) 1382 384109, int ext 84109


The main focus of my lab is understanding how ubiquitin is conjugated to a target. This particular signal underpins every aspect of eukaryotic life, from cell-cycle control to the wound response in plants. The best understood role of ubiquitin conjugation is to steer the targeted substrate to the proteasome for degradation. However, there are myriad non-degradative ubiquitin signals, including a key role in the DNA damage response.  There is a range of different ubiquitin signals that can be produced.  A substrate can be monoubiquitinated, or multiply-monoubiquitinated, where a single ubiquitin is conjugated to one or more specific residues.  Alternatively, a substrate may be modified by a chain of ubiquitins conjugated to themselves through one of 7 lysine (K) residues and the N-terminal methionine.  In addition to homotypic chains, branched chains of ubiquitins conjugated to different lysines can be formed.  With such a complex process, it is no surprise that there are hundreds of enzymes involved in the ubiquitin system, for adding ubiquitin, removing ubiquitin, and for relaying the ubiquitin signal. Fundamentally, ubiquitin conjugation requires an E1, E2 and E3 enzyme cascade to relay an activated ubiquitin moiety onto a target lysine of a substrate (Figure 1).


Figure 1 A schematic of the ubiquitin (Ub) pathway and the members involved at each step; activation (E1), conjugation (E2) and ligation (E3). Ubiquitin activation is catalysed by the E1 in an energy-consuming step. The ubiquitin thioester conjugate is then passed onto the catalytic cysteine site on the E2 and finally ligated onto a target lysine of a substrate, an event mediated by E3 ligases. Also indicated is the numerical hierarchy of the ubiquitin pathway in humans. (Figure from Chaugule and Walden, 2016).

There are many E2s and E3 enzymes, found in different abundancies and in different cellular localization. Given the importance of ubiquitination in biology, it is unsurprising that mutations and dysfunction in ubiquitin system components are found in many human diseases. Although the identity of the enzymes required for ubiquitination are well known, there is still much to understand about how a particular signal is achieved. The key questions my lab aims  to answer are how specific signals are attached to individual lysines on particular targets, and how specificity within given pathways is achieved. To address these questions, we use two model systems that are important both for these fundamental questions, and for the role they play in human disease: 1) FANCL - an exquisitely specific E3 ligase that targets one lysine on a substrate for modification, functions with limited auxiliary enzymes including E2s, and performs only one type of ubiquitination (Figure 2A). 2) Parkin - a broad-spectrum, promiscuous ligase that can cooperate with multiple E2s, to effect multiple types of ubiquitin signal (Figure 2B).


Figure 2 A) Schematic of the FANCL and Ube2T mediated specific substrate monoubiquitination. B) Schematic of the variety of substrate ubiquitination events mediated by Parkin through several E2 enzymes. (Figure from Chaugule and Walden, 2016).

FANCL – a specific ligase required for DNA damage repair

Fanconi Anemia (FA) is a devastating childhood disease caused by chromosomal instability. The FA pathway is required to fix interstrand crosslinks that tie two strands of DNA together. The key signal to repair the damage is a single ubiquitin attached to a single lysine on the substrate protein FANCD2. The goal in our lab is to understand, in atomic detail, how that signal is attached to the protein, how it is recognised by repair proteins, and how it is removed (Figure 3).

Figure 3 The ubiquitination and deubiquitination cycle in the regulation of DNA interstrand crosslink repair. FANCI/FANCD2 are shown in cyan and green respectively, with buried ubiquitination sites highlighted (pdb code 3S4W – structure from Pavletich lab, Joo et al., 2011, Science). Core complex components are labeled and shown in yellow, with Ube2T in magenta pdb codes 3K1L and 4CCG, Cole et al., 2010; Hodson et al., 2014). The deubiquitinase complex is in teal/brown, and unknown interactors of the monoubiquitinated proteins are shown in orange and red shapes. The ICL is depicted by a red line.

Since ubiquitin is an optimal acceptor of other ubiquitins, a major unanswered question in ubiquitin biology is how such a signal is restricted to a single ubiquitin. The Fanconi Anemia (FA) DNA repair pathway comprises at least 19 proteins, one of which (FANCL) is an E3 ligase. In 2010, we determined the crystal structure of full length FANCL from Drosophila, providing the correct domain architecture, and identifying domains required for activity, and substrate binding (Cole et al., 2010). An intriguing feature of the pathway is that it apparently utilises only one E2, Ube2T, in cells. In 2014, we solved the 2.2 Å structure of human FANCL RING domain in complex with Ube2T (Figure 4), and worked out what the molecular determinants of the exquisite specificity of FANCL for Ube2T are, and vice versa (Hodson et al. 2014 ). Subsequently, FA patients with mutations in Ube2T have been identified, and this is now called FANCT. 

Figure 4 Structure of FANCL RING (magenta) bound to Ube2T (blue). Top left panel: The pi stacking in the binding interface between Y311 of FANCL and R6 and R9 of Ube2T. Top right panel: The hydrophobic binding interface of the RING domain (magenta) and Ube2T (blue). Bottom panels: The electrostatic and hydrogen bonding network of the RING-Ube2T interface. Interactions are represented by dashed lines. Figure from Hodson et al., 2014.

In collaboration with Arno Alpi in the Unit, and Mark Howard at the University of Kent, we have also identified a crucial interaction between the N-terminal ELF domain of FANCL, and ubiquitin (Miles et al., 2015). This interaction is essential for FANCD2 ubiquitination in cells, via an as yet unknown mechanism.
The exquisite specificity of this ligase ensures the strict monoubiquitination of a single lysine on two related substrates thus making it an attractive model system to understand the underlying mechanisms of ubiquitination.

Parkin – a broad spectrum ligase mutated in Parkinson’s disease

At the opposite end of the specificity profile, Parkin is a broad spectrum, promiscuous ligase, with over 450 apparent substrates, and multiple E2 partners. It is also mutated in autosomal recessive juvenile Parkinsonism (ARJP), with pathogenic Parkin mutations being responsible for more than 50% of ARJP cases. The E3 ligase Parkin is a member of the RBR family of ubiquitin ligases. Each of the 12-14 RBR proteins contains the characteristic RBR module, which comprises a RING domain (R1), a catalytic domain (R2 – required for catalysis), and a linking benign, in-between domain (B). Outwith the RBR module, each family member has additional domains that are unique to itself. In the case of Parkin, these are an N-terminal ubiquitin-like (Ubl) domain and zinc-chelating RING0 domain.

Figure 5 Domain architecture of Parkin. Top, schematic representation of Parkin coloured by domain. Bottom, 1.8 Å resolution structure of human Parkin (Kumar et al., 2015).

In collaboration with Gary Shaw’s lab at the University of Western Ontario, we demonstrated that the biological effect of pathogenic mutations in the Ubl domain of Parkin lead to inappropriately-activated Parkin, the consequence of which is turnover by the proteasome (Spratt et al., 2013). Thus activating Parkin mutations lead to Parkin degradation, and therefore inactivation of Parkin ligase function. These findings provided a molecular explanation for the recessive inheritance of Parkin mutations. We determined the structure of the BR2 fragment of Parkin, and showed that the effect of Parkin mutations in the catalytic domains is to perturb the environment of the catalytic cysteine. In 2014,  an activator of Parkin was discovered by several groups including the Muqit and Alessi labs, in the form of phosphoubiquitin. PINK1 phosphorylates both Parkin Ubl and ubiquitin itself, at the equivalent Serine65 position. In 2015, we determined the 1.8 Å structure of human Parkin, UblR0RBR domains, the most complete and highest-resolution structure of Parkin, and the first to contain all domains (Kumar et al., 2015), containing all Parkin domains (Figure 5). The structure reveals the extensive inhibition by the Ubl domain, and mutations disrupt this. Furthermore we solved the structure of a S65D-phosphomimic mutant of Parkin, which revealed an allosteric switch in the Parkin domains whereby inclusion of a negative charge in the Ubl domain stabilises a histidine on the surface of the RING0, which is where phosphoubiquitin binds (Kazlauskaite et al., 2015). In collaboration with Gary Shaw, we found that the direct consequence of phosphoubiquitin binding is to displace the Ubl domain, and reveal a ubiquitin-binding patch on the surface of Parkin. These structures represent a major breakthrough in our understanding of Parkin mechanism, and allow us to put forward a model of how Parkin is inhibited in the absence of phospho signals, and how it is activated.  Our structures and supporting data also reveal how Parkin can function with multiple different E2s, thereby providing a molecular rationale for the apparent promiscuity at the E2/E3 level of ubiquitin conjugation. What we want to understand now is the mechanisms of the remaining steps of the Parkin activity cycle, including substrate selection, lysine selection, and modification.

We take a variety of approaches to address these questions, including biochemical, biophysical, cellular, and structural methodologies. Our primary tool is X-ray crystallography, but we also use both Nuclear Magnetic Resonance spectroscopy and Electron Microscopy to unlock the mechanisms of the fundamental systems of biology (Figure 6).

Figure 6 Techniques used to study protein ubiquitination. Top L-R In vitro assays, NMR spectroscopy, Isothermal Titration Calorimetry, cell-based assays, X-ray diffraction pattern. Bottom, gallery of crystals grown in the lab.



Chaugule, V.K. & Walden, H. Specificity and disease in the ubiquitin system.  Biochemical Society Transactions (2016) 44(1):212-27. doi: 10.1042/BST20150209.

Chaugule, V.K., Burchell, L., Barber, K.R.,  Sidhu, A., Leslie, S.J., Shaw, G.S. & Walden, H. Autoregulation of Parkin activity through its ubiquitin-like domain.  EMBO Journal (2011) 30, 2853-2867 

Cole, A.R.,  Lewis, L.P.C. & Walden, H.  The structure of FANCL, the catalytic subunit of the Fanconi Anemia core complex.  Nature Structural and Molecular Biology (2010) 17. p.294-298.

Hodson, C., Purkiss, A., Miles, J.A., & H. Walden.  Structure of the human FANCL RING-Ube2T complex reveals determinants of cognate E3-E2 selection. Structure (2014) 22, p337-344.

Kazlauskaite, A., Martínez-Torres, R.J., Wilkie, S., Kumar, A., Peltier, J., Gonzalez, A., Johnson, C., Zhang, J., Hope, A.G., Peggie, M., Trost, M., van Aalten, D.M., Alessi, D.R., Prescott, A.R., Knebel, A., H. Walden, & Muqit, M.M. Binding to serine 65-phosphorylated ubiquitin primes Parkin for optimal PINK1-dependent phosphorylation and activation. EMBO Reports (2015). 16, p.939-54. doi: 10.15252/embr.201540352

Kondapalli, C., Kazlauskaite, A., Campbell, D.G., Gourlay, R., Burchell, L., Walden, H., Macartney, T.J., Deak, M., Alessi, D.R  Muqit, M.M.  The Parkinson’s disease associated kinase PINK1 is activated by the mitochondrial uncoupler, CCCP, and phosphorylates Parkin at Ser65 Open Biology (2012) 2:120080.

Kumar, A.*, Aguirre, J.D.*, Condos, T.E.C.*,  Martinez-Torres, R.J.*, Chaugule, V.K.,  Toth, R.,  Sundaramoorthy, R., Mercier, P., Knebel, A.,  Spratt, D.E., Barber, K.R.,  Shaw, G.S.*, & Walden, H*. Disruption of the autoinhibited state primes the E3 ligase Parkin for activation and catalysis. EMBO Journal (2015) 34 p.2506-21 DOI:10.15252/embj.201592337..

Miles, J.A., Frost, M.G., Carroll, E., Rowe, M.L., Howard, M.J., Sidhu, A., Chaugule, V.K.,  Alpi, A.F.*, & Walden, H*. The Fanconi Anemia DNA repair pathway is regulated by an interaction between ubiquitin and the E2-like fold domain of FANCL. Journal of Biological Chemistry (2015) 10.1074/jbc.M115.675835

Spratt, D.E.*,  Martinez-Torres R.J.*, Noh, Y.J.*,  Mercier, P., Manczyk, N., Barber, K.R., Aguirre, J.D., Burchell, L., Purkiss, A.,  H.Walden*, & G.S.Shaw*.
A molecular explanation for the recessive nature of parkin-linked Parkinson’s disease Nature Communications (2013) 10.1038/ncomms2983 

* denotes equal contribution
Members of Walden lab highlighted in bold.


Top 10 Publications

Kumar, A.*, Aguirre, J.D.*, Condos, T.E.C.*,  Martinez-Torres, R.J.*, Chaugule, V.K.,  Toth, R.,  Sundaramoorthy, R., Mercier, P., Knebel, A.,  Spratt, D.E., Barber, K.R.,  Shaw, G.S.*, and Walden, H*. (2015)
Disruption of the autoinhibited state primes the E3 ligase Parkin for activation and catalysis.
EMBO J 34 p.2506-21 DOI:10.15252/embj.201592337.
(Cover Story and News & Views piece “pUBLically unzipping parkin” by Dove, Klevit & Rittinger)

Miles, J.A., Frost, M.G., Carroll, E., Rowe, M.L., Howard, M.J., Sidhu, A., Chaugule, V.K., Alpi, A.F.*, and Walden, H*. (2015)
The Fanconi Anemia DNA repair pathway is regulated by an interaction between ubiquitin and the E2-like fold domain of FANCL.
J Biol Chem 10.1074/jbc.M115.675835

Hodson, C., Purkiss, A., Miles, J.A. and H. Walden. (2014)
Structure of the human FANCL RING-Ube2T complex reveals determinants of cognate E3-E2 selection.
Structure  22, p337-344.

Spratt, D. E.*, Martinez-Torres, R. J.*, Noh, Y. J.*, Mercier, P., Manczyk, N., Barber, K. R., Aguirre, J. D., Burchell, L., Purkiss, A., Walden, H*. and Shaw, G. S*. (2013).
A molecular explanation for the recessive nature of parkin-linked Parkinson's disease.
Nat Commun 4, 1983
(with accompanying perspectives in EMBO J and Current Biology)

Burchell, L., Chaugule, V. K. and Walden, H. (2012).
Small, N-terminal tags activate Parkin E3 ubiquitin ligase activity by disrupting its autoinhibited conformation.
PloS one 7, e34748

Hodson, C., Cole, A. R., Lewis, L. P., Miles, J. A., Purkiss, A. and Walden, H. (2011).
Structural analysis of human FANCL, the E3 ligase in the Fanconi anemia pathway.
J Biol Chem 286, 32628-32637

Chaugule, V. K., Burchell, L., Barber, K. R., Sidhu, A., Leslie, S. J., Shaw, G. S. and Walden, H. (2011).
Autoregulation of Parkin activity through its ubiquitin-like domain.
Embo J 30, 2853-2867
(with an accompanying news and views article “Policing Parkin with a UblD” by Fen Liu and Kylie Walters, p. 2757-2758).

Cole, A. R., Lewis, L. P. and Walden, H. (2010).
The structure of the catalytic subunit FANCL of the Fanconi anemia core complex.
Nat Struct Mol Biol 17, 294-298

Walden, H., Podgorski, M. S. and Schulman, B. A. (2003).
Insights into the ubiquitin transfer cascade from the structure of the activating enzyme for NEDD8.
Nature 422, 330-334
(A news and views piece covering this paper is in Nature Structural Biology, (2003), 10 244-246: Two-stepping with E1.  A.P. VanDemark and C.P. Hill.)

Walden, H., Podgorski, M. S., Huang, D. T., Miller, D. W., Howard, R. J., Minor, D. L., Jr., Holton, J. M. and Schulman, B. A. (2003).
The structure of the APPBP1-UBA3-NEDD8-ATP complex reveals the basis for selective ubiquitin-like protein activation by an E1.
Mol Cell 12, 1427-1437



Commercial Impact:

In collaboration with the pharmaceutical industry via the Division of Signal Transduction Therapy collaboration with AstraZeneca, Boehringer Ingelheim, GlaxoSmithKline, Janssen Pharmaceutica, Merck Serono and Pfizer the research ouptuts from my group contribute to accelerating the development of company drug development programmes through access to research data and reagents. Reagents are also commercialised to provide access to the wider scientific community via license arrangements with companies such as Millipore, AbCam and Ubiquigent.