- Email: [email protected]
Emerging Features of Linear Motif-Binding Hub Proteins
TIBS 1644 No. of Pages 10
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
Opinion
Emerging Features of Linear Motif-Binding Hub Proteins Nathan Jespersen1 and Elisar Barbar1,* Hub proteins are important elements of interactomes within an organism; they bind diverse partners, display significant pleiotropy, and connect many cellular systems. Static hubs interact with their partners simultaneously, while dynamic hubs bind different partners at different locations and times. Although this distinguishes some features of hub protein/partner interactions, the increasing literature requires an expanded categorization of molecular and functional properties. Here, we focus on dynein light chain LC8 as a canonical example of dynamic hub proteins to develop a conceptual residue-level framework for hub–partner interactions and functions. We propose a new class of structural linear motif-binding hub proteins (LMB-hubs) with key common features. LMBhubs have structural plasticity yet conserved interfaces, can function as integral members of large multimolecular assemblies, and are self-regulating.
Highlights Hub proteins bind multiple diverse partner proteins and connect numerous components of a cellular system. LMB-hubs represent a newly defined class of hubs that target an 8–10-residue consensus motif on a partner protein. LMB-hubs have a highly conserved binding groove that is also remarkably flexible, suggesting that dynamic plasticity underlies its partner binding selectivity and diversity. Structural LMB-hubs stabilize partner– protein complexes that nucleate formation of large molecular assemblies, while enzymatic LMB-hubs transiently interact with consensus motifs.
LMB-Hubs An organism’s interactome describes the entire set of protein–protein interactions that occur within a cell. For example, in yeast, 6000 proteins participate in approximately 16 000–40 000 binary interactions [1,2]. Protein–protein interactions are not evenly distributed, and are best described by a power law, wherein most proteins interact with a few partners, while some sit at the center of complex interaction networks [3]. The centrality of a protein arises from both its essential function and its pleiotropy. Knockouts of central proteins, termed hubs (see Glossary), are either lethal, or lead to multiple unrelated deleterious effects [4]. Hubs are grouped as static (party) or dynamic (date). Static hubs bind many partners simultaneously at different sites, for example BRCA2 [5]. Dynamic hubs typically bind to short linear motifs (SLiMs) within intrinsically disordered regions (IDRs), and have multiple partners that compete for the same site [6]. While dynamic/date hubs are the most commonly used terms to define this class of proteins, here we suggest using the more descriptive and specific term linear motif-binding hubs (LMB-hubs) [7], due to their propensity to interact with short linear motifs at a single, highly conserved binding interface [8]. While excellent reviews describe the ubiquitous nature and functional importance of SLiMs for protein interaction networks [9,10], these reviews focus on properties of SLiMs, rather than the proteins that bind them. In this Opinion, we focus instead on these proteins and emphasize that LMB-proteins have some common traits that firmly place them within the category of hubs, and thereby we use the LMB-hubs nomenclature to connect the SLiM and hub fields. Well-known examples of LMB-hubs include calmodulin, 14-3-3, and SH3 domains [11,12], and many LMB-hubs and their consensus motifs are described on the Eukaryotic Linear Motif Database (ELM) [13]. A more recently discovered and characterized member of this class is the dynein light chain LC8 [14–17], which is a primary example in this Opinion. The more than 100 experimentally verified LC8-partner proteins [15] include many intrinsically disordered proteins (IDPs) regulating multifarious functions, such as intracellular transport Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
Cellular availability of the LMB-hub LC8 is exquisitely controlled by feedback regulation of transcription factor ASCIZ. Using recently developed motif prediction tools for LMB-hub 14-3-3, we propose autoregulation as a general mechanism for controlling cellular concentrations of LMB-hubs.
1
Department of Biochemistry and Biophysics, Oregon State University, Corvallis, OR 97331, USA
*Correspondence: [email protected] (E. Barbar).
https://doi.org/10.1016/j.tibs.2020.01.004 © 2020 Elsevier Ltd. All rights reserved.
1
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
and cell motility (dynein intermediate chain) [18,19], genome stability and DNA-end resection (MRE11) [20], transcription (ATMIN-substrate Chk-interacting Zn2+ finger; ASCIZ) [21–23], and apoptosis (Bim) [24]. Additionally, LC8 is highly expressed across a wide variety of cell types [25], and LC8 partners are broadly distributed within individual cells [15,26]. Clear residue-level descriptions of binding motifs and structural characteristics of LMB-hub– partner interactions are increasingly reported. The availability of tools to predict hub partners facilitates identification of new motif-bearing proteins and provides a commensurate increase in our understanding of LMB-hub functions in a given system. Despite this influx in data, the only described unifying feature of hubs is that they bind to many partners. Here, we suggest the usage of the term LMB-hubs to describe a central and essential class of hub proteins and propose three defining features of LMB-hubs inspired from our work on LC8. Specifically, we hypothesize that structural LMB-hubs have highly conserved sequences that are structurally plastic at the binding interfaces, are often stabilizers of large complex assemblies, and have tightly controlled intracellular concentrations, leading to various forms of self-regulation. Defining some common features of this essential class of proteins is an important step to facilitate our understanding of their function in cellular homeostasis.
LMB-Hubs Have Binding Interfaces That Are Structurally Plastic, but Conserved in Sequence A defining feature of LMB-hubs is that one conserved binding sequence accommodates many diverse partners. Prime examples are MHC I and II; a set of proteins integral to the adaptive immune system [27]. They function by extracellularly presenting antigens from potential pathogens for recognition by appropriate T cells [28]. Although MHCs are polygenic and polymorphic [29], it is not feasible for such a small number of proteins to bind and present the wide variety of peptides derived from foreign proteins without considerable promiscuity in their recognized linear motifs [30]. Studies on the cornucopia of MHC crystal structures note that B factors, which can act as a proxy for dynamics, are surprisingly high within the MHC binding groove [27]. This suggests that MHC binding of diverse peptides is facilitated by a plastic-binding interface. Indeed, analysis of the stability of the interface before and after binding indicates that free MHC molecules are significantly more dynamic, and have structures stabilized by bound peptides [31]. Although LMB-hub plasticity is perhaps best characterized in MHCs, this general feature is observed in many LMB-hubs [15,32,33]. LC8 interacts with an eight-amino-acid recognition motif within IDRs of its partners. While there is substantial variation in this recognition motif in various LC8 partners, it is most frequently characterized by a TQT sequence (termed the motif anchor) [15,34]. As in the case of MHCs, an abundance of structural data facilitates analysis of the structural heterogeneity in free and bound LC8 (Figure 1, top). By overlaying the protomers from five published crystal and NMR structures of free LC8, from multiple species, we observed that the β strand that directly binds to partners is highly variable, and has the highest root mean squared deviation (RMSD) value among structures. By contrast, an overlay of 17 structures of LC8 bound to peptides demonstrates that partners stabilize the binding interface. Despite the plasticity of the binding groove, a comparison of regions of conservation within LC8 reveals that the general features of the groove are strictly conserved across organisms as evolutionarily distant as Giardia lamblia and humans [15]. Residues near the dimeric interface/partner binding site are noticeably more conserved than peripheral regions [35]. Taken together, these observations lead to the hypothesis that structural plasticity of the binding groove is conserved in LMB-hub interactions, specifically for those involved in highly diverse interactions (Box 1). 2
Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
Glossary B factors: or temperature factors, used in X-ray scattering to describe uncertainty or spread of electron density for an atom. Higher B factors imply more flexibility. Dynamic/date hubs: protein hubs that bind multiple partners at a single interface. Dynein light chain LC8: a highly connected protein that was first discovered as an integral component of the dynein motor complex. Eukaryotic Linear Motif Database: database and tool that catalogs known eukaryotic linear motifs (or short linear motifs), and predicts motifs within proteins of interest. Hub proteins: highly connected proteins that interact with many (N50) different partners. Intrinsically disordered proteins: proteins that lack a regular 3D structure, or contain multiple intrinsically disordered regions. Intrinsically disordered regions: polypeptide segments within a protein sequence that do not fold into a regular structure, often rich in polar/charged residues, and depleted in hydrophobic residues. Linear motif-binding hub proteins: highly connected proteins that bind to partners containing a recognizable and definable consensus motif. Polygenic: multiple genes that are functionally interdependent. In some cases, such as for major histocompatibility complexes, protein products from polygenes form a complex. Polymorphic: genes that have a high amount of sequence variation within a given population. Root mean squared deviation: the average distance between atoms of superimposed proteins. Short linear motifs: functional motifs of typically 3–11 amino acids that are embedded within intrinsically disordered proteins. Static/party hubs: protein hubs that bind multiple partners simultaneously at multiple interfaces. Transducisome: a signaling complex involved in photoreception for Drosophila.
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
Trends in Biochemical Sciences
Figure 1. Sequence Conservation and Structural Variability in Linear Motif-Binding Hub Proteins (LMB-Hubs). A comparison of the conservation of sequence (left) and structure in free (middle) and bound (right) LMB-hubs. Some of the most conserved regions in terms of sequence are the most variable structurally for both free dynein light chain LC8 (top) and calcineurin (bottom). The binding region is significantly stabilized when bound to peptides, suggesting that the plasticity facilitates binding of diverse sequences. Conservation analyses were performed using ConSurf [92] on 500 sequences with sequence identities ranging from 40% to 95%. Structural conservation was measured using the Ensemblator’s global overlay tool [93], which aligns independently determined 3D structures to identify regions of structural conservation or plasticity. Analyses were performed on all available structures for free calcineurin (4 structures, including human, mouse, and rat orthologs), bound calcineurin (eight structures, including human and cow orthologs), free LC8 (five structures, including mouse, rat, Drosophila, and Plasmodium falciparum orthologs), and bound LC8 (17 structures, including human, rat, Drosophila, and yeast orthologs).
In support of this hypothesis, a similar conservation of structural plasticity is observed in the PxIxIT LMB subunit of the serine/threonine protein phosphatase calcineurin (PP2B or PPP3; Figure 1, bottom) [36,37]. Upon binding, disordered partners gain structural stability and form a β strand that fits within a small binding groove [38]. Like LC8, the available crystal structures indicate that the binding groove is the most plastic region of the free protein. In the bound form, although the LMB-hub on average becomes more rigid, its level of rigidity is partner dependent. For example, the binding interface of LC8 when bound to a peptide from Swallow is fully rigid, while it remains highly dynamic when bound to a peptide from dynein intermediate chain [39]. It is remarkable that some of the most sequence-conserved regions of calcineurin and LC8 are also the most structurally variable parts, supporting the hypothesis that structural plasticity allows accommodation of a diverse set of partners with a wide range of properties. Other examples of well-studied proteins that we categorize here as LMB-hubs, such as the PDZ domains Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
3
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
Box 1. Interface Plasticity of LMB-Hubs Correlates with Binding Diversity It is well recognized in the literature that the plasticity of interacting regions provides them with increased interaction capacity. However, in the retinoblastoma pocket domain (pRB) hub that interacts with the LXCXE linear motif found in viral proteins, a plastic interface in not readily apparent from comparison of temperature factors. The positions of the RB sidechains that create the binding pockets are nearly identical in both the unbound and bound structures and the crystallographic temperature factors of these sidechains are already low in the unliganded structure and remain low upon binding [81]. In human papillomavirus 16 E7, for example, the core motif residues bind tightly into hydrophobic pockets of the pRB binding groove [82]. A more ordered pocket in the unliganded structure would contribute to higher affinity (nM to sub-μM compared to the μM affinity of LC8 partners) and also to the reduced diversity of interactions (pRB is specific to the cell cycle compared to the multitude of systems with LC8). It is also important to note that while the crystal structures do not show plasticity at the interface, plasticity in the protein as a whole is inferred from unfolding and aggregation studies in solution. The Rb pocket is marginally stable and presents multiple conformations in solution but is significantly stabilized when bound to peptide ligands [83].
[33,40,41], WW domains [42,43], and SH2 domains [44,45], are also fully ordered, but have notable functional plasticity within their binding interfaces. Although the examples in Figure 1 show bound structures that are both significantly stabilized, and are the same regardless of the partner peptide, it is not uncommon for LMB-hubs to bind to partners in a more variable manner. For example, the bound form of the SH2 domain of the SLAM-associated protein (SAP) has a variety of different conformations [45,46]. Thus, although our examples suggest a rigidified bound form, there can be variable binding mechanisms for this class of proteins.
LMB-Hubs Can Have Either Enzymatic or Structural Roles LMB-hubs can be segregated into enzymatic proteins (such as phosphatases and kinases) and structural, chaperoning proteins (such as PDZ domains). We refer to these two groups as enzymatic and structural LMB-hubs, respectively. The key point of differentiation is the transient nature of the enzymatic interaction, which requires linear motifs to only bind during the catalytic process. One well-known example is the serine/threonine-specific protein kinase B (PKB, also known as AKT1). PKB is a key basophilic kinase that functions in cell survival, cell cycle progression, metabolism, and apoptosis [47] by targeting and phosphorylating proteins with a minimal consensus sequence of RxRxxS/T [48]. While we differentiate enzymatic from structural hubs based on whether the interaction with the partners is either transient at the catalytic site, or an obligate part of the complex, it is important to note that some enzymatic LMB-hubs have both catalytic and noncatalytic linear motif-binding sites. Mitogen-activated protein kinases, for example, belong to this group, as they mainly recognize their substrates not with the catalytic site but with auxiliary docking surfaces on their kinase domains [49–51]. In contrast to enzymatic LMB-hubs, structural LMB-hubs form stable interactions with partner linear motifs and promote formation and structural organization of higher order protein associations (Figure 2). For example, PDZ domains are 80–90-amino-acid domains that occur within a wide variety of proteins. Although there is variability in the binding motif based on PDZ class, they often bind to the four C-terminal residues of their partner protein, typically containing an xxΦ* motif (where Φ is a hydrophobic residue, and * is the C terminus); notable examples of internal PDZ motifs do exist [52,53]. PDZ domains often arise as repeating clusters within a protein, providing scaffolding for large, macromolecular assemblies [54]. For example, photoreception in the Drosophila eye is dependent on the INAD protein [55], which encodes five separate PDZ domains. These PDZ domains work together to recruit various key components of the photoreceptor transducisome, as well as to promote oligomerization of INAD [56], making them integral to transducisome structure. Unlike enzymatic LMB-hubs, stabilization 4
Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
Trends in Biochemical Sciences
Figure 2. Functions for Dynein Light Chain LC8 as a Structural Linear Motif-Binding Hub Protein (LMB-Hub). Models depicting the role for structural LMBhubs, using LC8 interactions as an example system. LC8 binds to disordered chains at the recognition motif, causing the formation of an adjacent coiled-coil (in Swallow, top) [94], a self-association domain [in the dynein intermediate chain (DIC), middle] [95], or an additional binding event (in dynein intermediate chain and Nup159, bottom) [58,96].
of large molecular assemblies often requires integration of structural LMB-hubs like PDZ domains within the complex, and partners of structural LMB-hubs are consequently competing directly for limited hub pools. In our opinion, LC8 can be firmly placed within the category of structural LMB-hubs that confer stability to their disordered binding partners (Figure 2). LC8 is perhaps best known for its role as a light chain in the dynein complex [57], where as a dimer with symmetrical binding grooves, it binds to a short linear TQT motif from each protomer of the dynein intermediate chain. LC8 binding brings together two chains of the partner protein, facilitating its self-association at a site distant from binding (Figure 2) and stabilizing the essential quaternary motor complex [58]. In addition to structural stabilization of the partner protein, the functional consequences of binding have been demonstrated on the yeast dynein complex where LC8 leads to a significant decrease in complex formation, decreased velocity, and processivity for the motor, as well as an increase in dynein heavy chain aggregation [59]. Parallel examples are the centriolar protein Ana2, and the syntaphilin protein involved in mitochondrial mobility and inhibition of syntaxin-1. In the case of Ana2, LC8 interactions lead to tetramerization of Ana2, facilitating interactions with downstream partners like Sas6 and Mud, and eventually contributing to correct spindle orientation [60,61]. For syntaphilin, interactions with LC8 stabilize a coiled coil structure proximal to the LC8 binding site, and allows it to recruit and anchor axonal mitochondria for synaptic regulation [26]. At the systems level, network analyses suggest that LMB-hubs help to organize and connect biological processes and pathways. Removal of hub protein nodes in interaction networks tends to make systems more sensitive to additional perturbations [62]. This suggests that the Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
5
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
structural plasticity of LMB-hub binding sites endows a functional adaptability that allows cells to efficiently respond to various requirements. We therefore propose that LMB-hubs are particularly critical to a cell’s ability to respond effectively to external stimuli such as increased salinity or temperature. Unfortunately, knockouts (or even knockdowns) of hubs are often lethal [63–65], which complicate experimental tests of the role of hubs in the response to external stimuli. Intriguingly, recent work by King et al. [66] reports the generation of a limb bud mesoderm-restricted conditional knockout of LC8, resulting in a potentially informative animal (and cell line) model. Although these cells are not yet fully characterized, preliminary analyses indicate that knockouts of LC8 lead to impairments in the Hippo signaling pathway [66], amongst other things.
LMB-Hubs Availability Is Tightly Controlled and Feedback-Regulation May Be Common Since LMB-hubs form 1:1 complexes with many partners, several questions arise. How can many different partners bind the same LMB-hub when they all compete for a common binding site? With a limited pool of each hub, how does a cell allocate sufficient protein for each function? For enzymatic LMB-hubs, the case can be made that transient binding facilitates the maintenance of sufficient pools of hub protein. However, structural LMB-hubs tend to be integrated into high-order complexes where they are inaccessible to other partners. For many hub proteins there is evidence for ‘just-in-time’ synthesis [65]; that is, protein expression is commensurate with the cell’s needs, suggesting that hub protein concentrations are delicately balanced and exquisitely controlled. For the majority of hub proteins, such regulatory systems are not well described; however, we hypothesize that feedback loops that enable LMB-hub self-regulation are common. A prime example of hub protein self-regulation is LC8 regulation by the transcription factor ASCIZ [22]. ASCIZ is an 88-kDa protein identified as a transcription factor for the LMB-hub protein LC8 [21]. Recent mouse experiments demonstrate that LC8-1 (the more highly expressed of two LC8 isoforms) germline knockouts lead to death early in embryogenesis, while a conditional knockout can proceed with severe developmental defects [66]. Additionally, ASCIZ knockouts display identical, but attenuated, phenotypes, suggesting that the most essential role for ASCIZ is regulation of LC8 concentrations within the cell [66]. Structural work on the LC8– ASCIZ complex reveals a remarkable number of LC8 binding sites (11 within human ASCIZ) and shows that this multivalency is important for regulation of LC8 concentrations within the cell [22]. Although the specifics of this regulation require further research, it appears that when LC8 concentrations are high, the full complement of ASCIZ binding sites are occupied, leading to inhibition of transcription of the LC8-1 gene (Figure 3). Conversely, when LC8 concentrations are low, the linear motifs on ASCIZ are largely unoccupied, leading to LC8 transcription and translation [22]. Seemingly, LC8 is part of a feedback loop tightly regulating its own concentrations within the cell. Due to their importance, LMB-hub motif binding preferences are often thoroughly studied, leading to the generation partner prediction tools [15,67]. It is therefore possible to combine information about gene regulation with partner predictions to identify potential feedback loops. For example, 14-3-3s are a family of structural LMB-hub proteins involved in cell-cycle control, signal transduction, protein trafficking, and apoptosis, among other things [68]. They are often dimeric, and usually bind to phosphorylated partners at a basic and proline-rich motif [69]. Binding facilitates partner protein folding, localization, and stimulation or inhibition of downstream interactions [70]. The 14-3-3-Pred [67] prediction tool has accurately identified 14-3-3 motifs in various partners/systems [71]. Using this tool, we identified a high-probability 14-3-3 binding motif within ATF-1, a cAMP response element binding transcription factor known to regulate 6
Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
Trends in Biochemical Sciences
Figure 3. Structural LMB-Hub Transcriptional Self-Regulation. A proposed model of LC8 regulation via the transcription factor ATMIN-substrate Chk-interacting Zn2+ finger (ASCIZ), where low concentrations of the structural LMB-hub (LC8, blue) would result in an active form of the transcription factor, facilitating increased hub protein production. When hub concentrations are high, binding to the transcription factor results in an inactive form of the transcription factor, turning off hub production. We hypothesize that this regulatory tactic is also shared by 14-3-3 whose transcription factor ATF-1 has a predicted site for binding 14-3-3. Red helices represent DNA-binding domains, green lines are IDRs, grey double-helices are DNA, Small ‘Ps’ are phosphorylated residues on ATF-1, and LMB-hubs are shown in dark blue.
intracellular 14-3-3ζ concentrations [72]. Additionally, the motif occurs at a known phosphorylation site (S63, UniProt code: P18846), previously associated with enhanced ATF-1 transcriptional activity [73,74]. It remains to be shown whether 14-3-3 indeed binds the predicted motif within ATF-1, but if this were the case, it suggests an intriguing self-regulation mechanism similar to what we observe with LC8 (Figure 3). This idea is speculative, but not out of the realms of possibility, considering the necessity of mechanisms for effective regulation of intracellular hub concentrations. Although we focus on direct transcriptional self-regulation, other mechanisms of LMB-hub feedback regulation are common (Box 2). Additionally, the presence of LMB-hubs within large macromolecular assemblies further supports the importance of tight regulation of protein concentrations. The gene balance hypothesis, developed by Papp et al. [75] and furthered in many other studies [76–78], states that changing the concentration of one subunit of a complex leads to unproductive subcomplexes that are missing key proteins. Therefore, a stoichiometric imbalance of members of the complex affects the amount of complete product. On the genomic scale, addition of a single chromosome is actually more detrimental than the addition of a complete genome to make a polyploid organism [79,80], purportedly due to the imbalance of macromolecular complex components [78]. As key components of many complexes, structural
Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
7
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
Box 2. Alternative Mechanisms for Self-Regulation of LMB-Hubs Indirect routes of self-regulation can be identified for many hubs. For example, the plant protein radical-induced cell death 1 (RCD1) interacts with a (DE)x(1,2)(YF)x(1,4)(DE)L motif in IDRs in many transcription factors [84]. Knockouts of RCD1 result in pleiotropic effects in stress and developmental pathways. One well studied system involves the generation of reactive oxygen species (ROS) from mitochondrial dysfunction. High concentrations of ROS cause inactivation, aggregation, and the eventual degradation of RCD1. RCD1, in turn, binds to and inhibits various transcription factors tasked with production of MDS (mitochondrial dysfunction stimulon) genes, whose role is to ameliorate the production, and control the impact of ROS [85]. Thus, when RCD1 concentrations are high, production of MDSs is decreased, leading to increased ROS, which in turn facilitate RCD1 degradation. We hypothesize that as the regulatory pathways for various LMB-hubs are more adequately described, similar feedback loops will be found. In addition to self-regulation at the transcriptional level, the activity of LMB-hubs can be modified post-translation. For enzymatic LMB-hubs like kinases, autophosphorylation is a common mechanism, and can even allow for tuning of enzymatic activity based on kinase concentrations [86]. Post-translational self-regulation of structural LMB-hubs is often less conspicuous. One interesting example is Disheveled (Dsh), an integral protein in the Wnt signalosome complex. The Wnt signaling pathway is important for animal development, and is involved in the regulation of cell polarity, movement, differentiation, survival, and proliferation [87]. Members of the pathway vary with function, but in all cases it contains a Wnt ligand, a membrane-bound receptor from the Frizzled family, and Dsh. Signal transduction relies upon a low-affinity interaction (about 10 μM) between the PDZ domain of Dsh with the linear motif from Frizzled [88]. Dsh self-associates via an N-terminal Dix domain, and forms large functional polymers with properties similar to membraneless organelles [89]. Polymerization leads to a concomitant increase in the local concentration of Dsh, allowing Dsh to overcome the low affinity interactions with Frizzled, and facilitates recruitment of signalosome subunits [89]. Dsh activity is partially regulated by an internal PDZ motif at its C terminus, which can associate with the Dsh PDZ domain and form a closed conformation [90,91]. The closed conformation displays decreased signaling activity and membrane localization, and is therefore an example of LMB-hub autoinhibition. We anticipate that self-regulation at this level is less common than self-regulation at the transcriptional level; however, autoinhibition and activation are likely to be important mechanisms for the mediation of active LMB-hub concentrations.
LMB-hubs would require fine tuning to keep concentrations from being too high or too low; both of which could lead to unproductive subcomplexes. We suggest that self-regulation is vital for balancing roles in their various macromolecular complexes.
Concluding Remarks In this work, we propose a newly defined category of hub proteins, the LMB-hubs. We are by no means the first to recognize that dynamic/date hubs typically bind linear motifs at a single interface, but we find this method of classification to be clearer and more concise when compared to the alternatives in common use. We also suggest further vernacular changes in order to differentiate between structural and enzymatic LMB-hubs, as these types of proteins are processed and used differently within cells. Current data indicate that structural LMB-hubs like calcineurin and LC8 often have dynamic binding interfaces in order to accommodate many motif variants, but are surprisingly conserved in sequence, leading us to hypothesize that the conserved feature is binding site plasticity. Although thorough analyses of LMB-hub regulatory systems are scarce, available data on LC8, RCD1, and Dsh, along with predictions for 14-3-3, lead us to speculate that self-regulation is an important and common feature for LMB-hubs specifically. Further work on LMB-hub transcriptional regulation can validate this hypothesis (see Outstanding Questions), and we propose using the existing motif predictors on known transcription factors to facilitate progress in this direction. References 1.
2.
3.
8
Ekman, D. et al. (2006) What properties characterize the hub proteins of the protein-protein interaction network of Saccharomyces cerevisiae? Genome Biol. 7, R45 Grigoriev, A. (2003) On the number of protein-protein interactions in the yeast proteome. Nucleic Acids Res. 31, 4157–4161 He, X. and Zhang, J. (2006) Why do hubs tend to be essential in protein networks? PLoS Genet. 2, e88
Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
4. 5.
6.
Yu, H. et al. (2008) High-quality binary protein interaction map of the yeast interactome network. Science 322, 104–110 Komurov, K. and White, M. (2007) Revealing static and dynamic modular architecture of the eukaryotic protein interaction network. Mol. Syst. Biol. 3, 110 Wu, X. et al. (2009) The properties of hub proteins in a yeastaggregated cell cycle network and its phase sub-networks. Proteomics 9, 4812–4824
Outstanding Questions Although hubs are often pleiotropic, are their interactions skewed? As in, are they active in one main functional pathway, with other minor roles? Autophosphorylation of enzymatic LMB-hubs is a common method for self-regulation. Do enzymatic LMBhubs also display feedback regulation at the transcriptional level? As hub protein concentrations are tightly controlled, what is the impact of a minor change in binding affinity for a specific partner? Would partner binding lead to a feedback change in hub protein expression? To verify self-regulation, one possibility is to transfect with a plasmid encoding an LMB-hub, and then monitor the expression of the genomic LMB-hub protein. If the hub is self-regulating, would transfection result in some inhibition of genomic expression of the LMB-hub? Can we describe the entropic benefit of a plastic-binding interface with respect to promiscuity? Is 14-3-3 indeed self-regulated as proposed here? Recently, a limb bud mesodermrestricted conditional knockout of LC8 was successfully generated, resulting in a potentially informative animal (and cell line) model. What type of experiments performed on this cell line could illuminate the impact of hub protein depletion on cellular homeostasis? Considering that LC8 dimerizes so many different client proteins in both the nucleus and the cytoplasm, how does the cell manage to keep all these client proteins separated? Can LC8 hetero-dimerize different client proteins?
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
7.
8.
9.
10. 11.
12.
13. 14. 15.
16. 17.
18.
19.
20.
21.
22. 23.
24.
25.
26.
27.
28. 29.
30. 31.
32.
Garamszegi, S. et al. (2013) Signatures of pleiotropy, economy and convergent evolution in a domain-resolved map of human–virus protein–protein interaction networks. PLoS Pathog. 9, e1003778 Kim, P.M. et al. (2006) Relating three-dimensional structures to protein networks provides evolutionary insights. Science 314, 1938–1941 Van Roey, K. et al. (2014) Short linear motifs: ubiquitous and functionally diverse protein interaction modules directing cell regulation. Chem. Rev. 114, 6733–6778 American Chemical Society Davey, N.E. et al. (2012) Attributes of short linear motifs. Mol. BioSyst. 8, 268–281 Uchikoga, N. et al. (2016) Specificity of broad protein interaction surfaces for proteins with multiple binding partners. Biophys. Physicobiol. 13, 105–115 Hu, G. et al. (2017) Functional analysis of human hub proteins and their interactors involved in the intrinsic disorder-enriched interactions. Int. J. Mol. Sci. 18, E2761 Gouw, M. et al. (2018) The eukaryotic linear motif resource – 2018 update. Nucleic Acids Res. 46, D428–D434 Barbar, E. (2008) Dynein Light Chain LC8 is a dimerization hub essential in diverse protein networks. Biochemistry 47, 503–508 Jespersen, N. et al. (2019) Systematic identification of recognition motifs for the hub protein LC8. Life Sci. Alliance 2, e201900366 Rapali, P. et al. (2011) DYNLL/LC8: a light chain subunit of the dynein motor complex and beyond. FEBS J. 278, 2980–2996 Erdős, G. et al. (2017) Novel linear motif filtering protocol reveals the role of the LC8 dynein light chain in the Hippo pathway. PLoS Comput. Biol. 13, e1005885 Nyarko, A. and Barbar, E. (2011) Light chain-dependent selfassociation of dynein intermediate chain. J. Biol. Chem. 286, 1556–1566 Toropova, K. et al. (2019) Structure of the dynein-2 complex and its assembly with intraflagellar transport trains. Nat. Struct. Mol. Biol. 26, 823–829 He, Y.J. et al. (2018) DYNLL1 binds to MRE11 to limit DNA end resection in BRCA1-deficient cells. Nature 563, 522–526 Jurado, S. et al. (2012) ATM substrate Chk2-interacting Zn 2+ finger (ASCIZ) is a bi-functional transcriptional activator and feedback sensor in the regulation of dynein light chain (DYNLL1) expression. J. Biol. Chem. 287, 3156–3164 Clark, S. et al. (2018) Multivalency regulates activity in an intrinsically disordered transcription factor. eLife 7, e36258 Becker, J.R. et al. (2018) The ASCIZ-DYNLL1 axis promotes 53BP1-dependent non-homologous end joining and PARP inhibitor sensitivity. Nat. Commun. 9, 5406 Singh, P.K. et al. (2019) Dynein light chain binding determines complex formation and posttranslational stability of the Bcl-2 family members Bmf and Bim. Cell Death Differ. 27, 434–450 Petryszak, R. et al. (2016) Expression Atlas update – an integrated database of gene and protein expression in humans, animals and plants. Nucleic Acids Res. 44, D746–D752 Chen, Y.-M. et al. (2009) dynein light chain LC8 regulates syntaphilin-mediated mitochondrial docking in axons. J. Neurosci. 29, 9429–9438 Wieczorek, M. et al. (2017) Major histocompatibility complex (MHC) class I and MHC class II proteins: conformational plasticity in antigen presentation. Front. Immunol. 8, 292 Dausset, J. (1981) The major histocompatibility complex in man. Science (80-. ). 213, 1469–1474 Charles A Janeway, J. et al. (2001) The major histocompatibility complex and its functions. In Immunobiology: The Immune System in Health and Disease (5th edition), Garland Science Rapin, N. et al. (2008) MHC motif viewer. Immunogenetics 60, 759–765 Reich, Z. et al. (1997) Stability of empty and peptide-loaded class II major histocompatibility complex molecules at neutral and endosomal pH: comparison to class I proteins. Proc. Natl. Acad. Sci. 94, 2495–2500 Agniswamy, J. et al. (2007) Plasticity of S2-S4 specificity pockets of executioner caspase-7 revealed by structural and kinetic analysis. FEBS J. 274, 4752–4765
33. Fujiwara, Y. et al. (2015) Crystal structure of afadin PDZ domainnectin-3 complex shows the structural plasticity of the ligandbinding site. Protein Sci. 24, 376–385 34. Clark, S. et al. (2016) The Anchored Flexibility Model in LC8 motif recognition: insights from the Chica complex. Biochemistry 55, 199–209 35. Holland, D.O. et al. (2017) Protein-protein binding selectivity and network topology constrain global and local properties of interface binding networks. Sci. Rep. 7, 5631 36. Roy, J. et al. (2007) A conserved docking site modulates substrate affinity for calcineurin, signaling output, and in vivo function. Mol. Cell 25, 889–901 37. Wigington, C.P. et al. (2019) Systematic discovery of short linear motifs decodes calcineurin phosphatase signaling. bioRxiv. Published online May 9, 2019. https://doi.org/10.1101/632547 38. Grigoriu, S. et al. (2013) The molecular mechanism of substrate engagement and immunosuppressant inhibition of calcineurin. PLoS Biol. 11, e1001492 39. Hall, J. et al. (2008) Differences in dynamic structure of LC8 monomer, dimer, and dimer–peptide complexes. Biochemistry 47, 11940–11952 40. Lee, J.H. et al. (2011) The structural flexibility of the shank1 PDZ domain is important for its binding to different ligands. Biochem. Biophys. Res. Commun. 407, 207–212 41. Kang, B.S. et al. (2003) Molecular roots of degenerate specificity in syntenin’s PDZ2 domain: Reassessment of the PDZ recognition paradigm. Structure 11, 845–853 42. Koepf, E.K. et al. (2008) WW: An isolated three-stranded antiparallel β-sheet domain that unfolds and refolds reversibly; evidence for a structured hydrophobic cluster in urea and GdnHCl and a disordered thermal unfolded state. Protein Sci. 8, 841–853 43. Iglesias-Bexiga, M. et al. (2015) WW domains of the Yes-kinaseassociated-protein (YAP) transcriptional regulator behave as independent units with different binding preferences for PPxY motif-containing ligands. PLoS One 10, e0113828 44. Kay, L.E. et al. (1996) Correlation between dynamics and high affinity binding in an SH2 domain interaction. Biochemistry 35, 361–368 45. Hwang, P.M. et al. (2002) A ‘three-pronged’ binding mechanism for the SAP/SH2D1A SH2 domain: structural basis and relevance to the XLP syndrome. EMBO J. 21, 314–323 46. Finerty, P.J. et al. (2002) Side-chain dynamics of the SAP SH2 domain correlate with a binding hot spot and a region with conformational plasticity. J. Mol. Biol. 322, 605–620 47. Hanada, M. et al. (2004) Structure, regulation and function of PKB/AKT – a major therapeutic target. Biochim. Biophys. Acta - Proteins Proteomics 1697, 3–16 48. Obata, T. et al. (2000) Peptide and protein library screening defines optimal substrate motifs for AKT/PKB. J. Biol. Chem. 275, 36108–36115 49. Tanoue, T. et al. (2000) A conserved docking motif in MAP kinases common to substrates, activators and regulators. Nat. Cell Biol. 2, 110–116 50. Zeke, A. et al. (2015) Systematic discovery of linear binding motifs targeting an ancient protein interaction surface on MAP kinases. Mol. Syst. Biol. 11, 837 51. Gógl, G. et al. Disordered protein kinase regions in regulation of kinase domain cores. Trends Biochem. Sci. 44, 300–311 52. Vaccaro, P. and Dente, L. (2002) PDZ domains: troubles in classification. FEBS Lett. 512, 345–349 53. Penkert, R.R. et al. (2004) Internal recognition through PDZ domain plasticity in the Par-6-Pals1 complex. Nat. Struct. Mol. Biol. 11, 1122–1127 54. Harris, B.Z. and Lim, W.A. (2001) Mechanism and role of PDZ domains in signaling complex assembly. J. Cell Sci. 114, 3219–3231 55. Shieh, B.H. and Zhu, M.Y. (1996) Regulation of the TRP Ca2+ channel by INAD in Drosophila photoreceptors. Neuron 16, 991–998 56. Ranganathan, R. and Ross, E.M. (1997) PDZ domain proteins: scaffolds for signaling complexes. Curr. Biol. 7, R770–R773 57. Pfister, K.K. et al. (1982) Purification and polypeptide composition of dynein ATPases from Chlamydomonas flagella. Cell Motil. 2, 525–547
Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
9
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
58. Hall, J. et al. (2009) Multivalency in the assembly of intrinsically disordered dynein intermediate chain. J. Biol. Chem. 284, 33115–33121 59. Rao, L. et al. (2013) The yeast dynein Dyn2-Pac11 complex is a dynein dimerization/processivity factor: structural and singlemolecule characterization. Mol. Biol. Cell 24, 2362–2377 60. Slevin, L.K. et al. (2014) The mechanism of dynein light chain LC8-mediated oligomerization of the Ana2 centriole duplication factor. J. Biol. Chem. 289, 20727–20739 61. Wang, C. et al. (2011) An Ana2/Ctp/Mud complex regulates spindle orientation in Drosophila neuroblasts. Dev. Cell 21, 520–533 62. Han, J.-D.J. et al. (2004) Evidence for dynamically organized modularity in the yeast protein–protein interaction network. Nature 430, 88–93 63. Jeong, H. et al. (2001) Lethality and centrality in protein networks. Nature 411, 41–42 64. Zotenko, E. et al. (2008) Why do hubs in the yeast protein interaction network tend to be essential: reexamining the connection between the network topology and essentiality. PLoS Comput. Biol. 4, e1000140 65. Batada, N.N. et al. (2006) Evolutionary and physiological importance of hub proteins. PLoS Comput. Biol. 2, e88 66. King, A. et al. (2019) Dynll1 is essential for development and promotes endochondral bone formation by regulating intraflagellar dynein function in primary cilia. Hum. Mol. Genet. 28, 2573–2588 67. Madeira, F. et al. (2015) 14-3-3-Pred: improved methods to predict 14-3-3-binding phosphopeptides. Bioinformatics 31, 2276–2283 68. Stevers, L.M. et al. (2018) Modulators of 14-3-3 protein-protein interactions. J. Med. Chem. 61, 3755–3778 69. Johnson, C. et al. (2010) Bioinformatic and experimental survey of 14-3-3-binding sites. Biochem. J. 427, 69–78 70. Aitken, A. (2006) 14-3-3 proteins: a historic overview. Semin. Cancer Biol. 16, 162–172 71. Yuan, L. et al. (2019) 14-3-3 signal adaptor and scaffold proteins mediate GPCR trafficking. Sci. Rep. 9, 11156 72. Kasinski, A. et al. (2014) Transcriptional regulation of YWHAZ, the gene encoding 14-3-3ζ. PLoS One 9, e93480 73. Gupta, P. and Prywes, R. (2002) ATF1 phosphorylation by the ERK MAPK pathway is required for epidermal growth factorinduced c- jun expression. J. Biol. Chem. 277, 50550–50556 74. Kobayashi, M. et al. (1997) Phosphorylation of ATF-1 enhances its DNA binding and transcription of the Na,K-ATPase alpha 1 subunit gene promoter. Nucleic Acids Res. 25, 877–882 75. Papp, B. et al. (2003) Dosage sensitivity and the evolution of gene families in yeast. Nature 424, 194–197 76. Veitia, R.A. (2004) Gene dosage balance in cellular pathways. Genetics 168, 569–574 77. Veitia, R.A. et al. (2008) Cellular reactions to gene dosage imbalance: genomic, transcriptomic and proteomic effects. Trends Genet. 24, 390–397 78. Birchler, J.A. and Veitia, R.A. (2012) Gene balance hypothesis: connecting issues of dosage sensitivity across biological disciplines. Proc. Natl. Acad. Sci. U. S. A. 109, 14746–14753
10
Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
79. Blakeslee, A.F. et al. (1920) Chromosomal duplication and Mendelian phenomena in Datura mutants. Science (80-.) 52, 388–390 80. Bridges, C.B. Sex in relation to chromosomes and genes. Am. Nat., 59, 127–137 81. Balog, E.R.M. et al. (2011) Crystal structure of the unliganded retinoblastoma protein pocket domain. Proteins 79, 2010–2014 82. Palopoli, N. et al. (2018) Short linear motif core and flanking regions modulate retinoblastoma protein binding affinity and specificity. Protein Eng. Des. Sel. 31, 69–77 83. Chemes, L.B. et al. (2013) Folding of a cyclin box. J. Biol. Chem. 288, 18923–18938 84. O’Shea, C. et al. (2017) Structures and short linear motif of disordered transcription factor regions provide clues to the interactome of the cellular hub protein radical-induced cell death. J. Biol. Chem. 292, 512–527 85. Shapiguzov, A. et al. (2019) Arabidopsis RCD1 coordinates chloroplast and mitochondrial functions through interaction with ANAC transcription factors. eLife 8, e43284 86. Dodson, C.A. et al. (2013) A kinetic test characterizes kinase intramolecular and intermolecular autophosphorylation mechanisms. Sci. Signal. 6, ra54–ra54 87. Cadigan, K.M. and Waterman, M.L. (2012) TCF/LEFs and Wnt signaling in the nucleus. Cold Spring Harb. Perspect. Biol. 4, a007906 88. Wong, H.-C. et al. (2003) Direct binding of the PDZ domain of Dishevelled to a conserved internal sequence in the C-terminal region of Frizzled. Mol. Cell 12, 1251–1260 89. Bienz, M. (2014) Signalosome assembly by domains undergoing dynamic head-to-tail polymerization. Trends Biochem. Sci. 39, 487–495 90. Lee, H.-J. et al. (2015) Conformational change of Dishevelled plays a key regulatory role in the Wnt signaling pathways. eLife 4, e08142 91. Qi, J. et al. (2017) Autoinhibition of Dishevelled protein regulated by its extreme C terminus plays a distinct role in Wnt/β-catenin and Wnt/planar cell polarity (PCP) signaling pathways. J. Biol. Chem. 292, 5898–5908 92. Ashkenazy, H. et al. (2016) ConSurf 2016: an improved methodology to estimate and visualize evolutionary conservation in macromolecules. Nucleic Acids Res. 44, W344–W350 93. Brereton, A.E. and Karplus, P.A. (2018) Ensemblator v3: robust atom-level comparative analyses and classification of protein structure ensembles. Protein Sci. 27, 41–50 94. Kidane, A.I. et al. (2013) Structural features of LC8-induced selfassociation of swallow. Biochemistry 52, 6011–6020 95. Benison, G. et al. (2006) Heteronuclear NMR identifies a nascent helix in intrinsically disordered dynein intermediate chain: implications for folding and dimerization. J. Mol. Biol. 362, 1082–1093 96. Stelter, P. et al. (2007) Molecular basis for the functional interaction of dynein light chain with the nuclear-pore complex. Nat. Cell Biol. 9, 788–796
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
Opinion
Emerging Features of Linear Motif-Binding Hub Proteins Nathan Jespersen1 and Elisar Barbar1,* Hub proteins are important elements of interactomes within an organism; they bind diverse partners, display significant pleiotropy, and connect many cellular systems. Static hubs interact with their partners simultaneously, while dynamic hubs bind different partners at different locations and times. Although this distinguishes some features of hub protein/partner interactions, the increasing literature requires an expanded categorization of molecular and functional properties. Here, we focus on dynein light chain LC8 as a canonical example of dynamic hub proteins to develop a conceptual residue-level framework for hub–partner interactions and functions. We propose a new class of structural linear motif-binding hub proteins (LMB-hubs) with key common features. LMBhubs have structural plasticity yet conserved interfaces, can function as integral members of large multimolecular assemblies, and are self-regulating.
Highlights Hub proteins bind multiple diverse partner proteins and connect numerous components of a cellular system. LMB-hubs represent a newly defined class of hubs that target an 8–10-residue consensus motif on a partner protein. LMB-hubs have a highly conserved binding groove that is also remarkably flexible, suggesting that dynamic plasticity underlies its partner binding selectivity and diversity. Structural LMB-hubs stabilize partner– protein complexes that nucleate formation of large molecular assemblies, while enzymatic LMB-hubs transiently interact with consensus motifs.
LMB-Hubs An organism’s interactome describes the entire set of protein–protein interactions that occur within a cell. For example, in yeast, 6000 proteins participate in approximately 16 000–40 000 binary interactions [1,2]. Protein–protein interactions are not evenly distributed, and are best described by a power law, wherein most proteins interact with a few partners, while some sit at the center of complex interaction networks [3]. The centrality of a protein arises from both its essential function and its pleiotropy. Knockouts of central proteins, termed hubs (see Glossary), are either lethal, or lead to multiple unrelated deleterious effects [4]. Hubs are grouped as static (party) or dynamic (date). Static hubs bind many partners simultaneously at different sites, for example BRCA2 [5]. Dynamic hubs typically bind to short linear motifs (SLiMs) within intrinsically disordered regions (IDRs), and have multiple partners that compete for the same site [6]. While dynamic/date hubs are the most commonly used terms to define this class of proteins, here we suggest using the more descriptive and specific term linear motif-binding hubs (LMB-hubs) [7], due to their propensity to interact with short linear motifs at a single, highly conserved binding interface [8]. While excellent reviews describe the ubiquitous nature and functional importance of SLiMs for protein interaction networks [9,10], these reviews focus on properties of SLiMs, rather than the proteins that bind them. In this Opinion, we focus instead on these proteins and emphasize that LMB-proteins have some common traits that firmly place them within the category of hubs, and thereby we use the LMB-hubs nomenclature to connect the SLiM and hub fields. Well-known examples of LMB-hubs include calmodulin, 14-3-3, and SH3 domains [11,12], and many LMB-hubs and their consensus motifs are described on the Eukaryotic Linear Motif Database (ELM) [13]. A more recently discovered and characterized member of this class is the dynein light chain LC8 [14–17], which is a primary example in this Opinion. The more than 100 experimentally verified LC8-partner proteins [15] include many intrinsically disordered proteins (IDPs) regulating multifarious functions, such as intracellular transport Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
Cellular availability of the LMB-hub LC8 is exquisitely controlled by feedback regulation of transcription factor ASCIZ. Using recently developed motif prediction tools for LMB-hub 14-3-3, we propose autoregulation as a general mechanism for controlling cellular concentrations of LMB-hubs.
1
Department of Biochemistry and Biophysics, Oregon State University, Corvallis, OR 97331, USA
*Correspondence: [email protected] (E. Barbar).
https://doi.org/10.1016/j.tibs.2020.01.004 © 2020 Elsevier Ltd. All rights reserved.
1
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
and cell motility (dynein intermediate chain) [18,19], genome stability and DNA-end resection (MRE11) [20], transcription (ATMIN-substrate Chk-interacting Zn2+ finger; ASCIZ) [21–23], and apoptosis (Bim) [24]. Additionally, LC8 is highly expressed across a wide variety of cell types [25], and LC8 partners are broadly distributed within individual cells [15,26]. Clear residue-level descriptions of binding motifs and structural characteristics of LMB-hub– partner interactions are increasingly reported. The availability of tools to predict hub partners facilitates identification of new motif-bearing proteins and provides a commensurate increase in our understanding of LMB-hub functions in a given system. Despite this influx in data, the only described unifying feature of hubs is that they bind to many partners. Here, we suggest the usage of the term LMB-hubs to describe a central and essential class of hub proteins and propose three defining features of LMB-hubs inspired from our work on LC8. Specifically, we hypothesize that structural LMB-hubs have highly conserved sequences that are structurally plastic at the binding interfaces, are often stabilizers of large complex assemblies, and have tightly controlled intracellular concentrations, leading to various forms of self-regulation. Defining some common features of this essential class of proteins is an important step to facilitate our understanding of their function in cellular homeostasis.
LMB-Hubs Have Binding Interfaces That Are Structurally Plastic, but Conserved in Sequence A defining feature of LMB-hubs is that one conserved binding sequence accommodates many diverse partners. Prime examples are MHC I and II; a set of proteins integral to the adaptive immune system [27]. They function by extracellularly presenting antigens from potential pathogens for recognition by appropriate T cells [28]. Although MHCs are polygenic and polymorphic [29], it is not feasible for such a small number of proteins to bind and present the wide variety of peptides derived from foreign proteins without considerable promiscuity in their recognized linear motifs [30]. Studies on the cornucopia of MHC crystal structures note that B factors, which can act as a proxy for dynamics, are surprisingly high within the MHC binding groove [27]. This suggests that MHC binding of diverse peptides is facilitated by a plastic-binding interface. Indeed, analysis of the stability of the interface before and after binding indicates that free MHC molecules are significantly more dynamic, and have structures stabilized by bound peptides [31]. Although LMB-hub plasticity is perhaps best characterized in MHCs, this general feature is observed in many LMB-hubs [15,32,33]. LC8 interacts with an eight-amino-acid recognition motif within IDRs of its partners. While there is substantial variation in this recognition motif in various LC8 partners, it is most frequently characterized by a TQT sequence (termed the motif anchor) [15,34]. As in the case of MHCs, an abundance of structural data facilitates analysis of the structural heterogeneity in free and bound LC8 (Figure 1, top). By overlaying the protomers from five published crystal and NMR structures of free LC8, from multiple species, we observed that the β strand that directly binds to partners is highly variable, and has the highest root mean squared deviation (RMSD) value among structures. By contrast, an overlay of 17 structures of LC8 bound to peptides demonstrates that partners stabilize the binding interface. Despite the plasticity of the binding groove, a comparison of regions of conservation within LC8 reveals that the general features of the groove are strictly conserved across organisms as evolutionarily distant as Giardia lamblia and humans [15]. Residues near the dimeric interface/partner binding site are noticeably more conserved than peripheral regions [35]. Taken together, these observations lead to the hypothesis that structural plasticity of the binding groove is conserved in LMB-hub interactions, specifically for those involved in highly diverse interactions (Box 1). 2
Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
Glossary B factors: or temperature factors, used in X-ray scattering to describe uncertainty or spread of electron density for an atom. Higher B factors imply more flexibility. Dynamic/date hubs: protein hubs that bind multiple partners at a single interface. Dynein light chain LC8: a highly connected protein that was first discovered as an integral component of the dynein motor complex. Eukaryotic Linear Motif Database: database and tool that catalogs known eukaryotic linear motifs (or short linear motifs), and predicts motifs within proteins of interest. Hub proteins: highly connected proteins that interact with many (N50) different partners. Intrinsically disordered proteins: proteins that lack a regular 3D structure, or contain multiple intrinsically disordered regions. Intrinsically disordered regions: polypeptide segments within a protein sequence that do not fold into a regular structure, often rich in polar/charged residues, and depleted in hydrophobic residues. Linear motif-binding hub proteins: highly connected proteins that bind to partners containing a recognizable and definable consensus motif. Polygenic: multiple genes that are functionally interdependent. In some cases, such as for major histocompatibility complexes, protein products from polygenes form a complex. Polymorphic: genes that have a high amount of sequence variation within a given population. Root mean squared deviation: the average distance between atoms of superimposed proteins. Short linear motifs: functional motifs of typically 3–11 amino acids that are embedded within intrinsically disordered proteins. Static/party hubs: protein hubs that bind multiple partners simultaneously at multiple interfaces. Transducisome: a signaling complex involved in photoreception for Drosophila.
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
Trends in Biochemical Sciences
Figure 1. Sequence Conservation and Structural Variability in Linear Motif-Binding Hub Proteins (LMB-Hubs). A comparison of the conservation of sequence (left) and structure in free (middle) and bound (right) LMB-hubs. Some of the most conserved regions in terms of sequence are the most variable structurally for both free dynein light chain LC8 (top) and calcineurin (bottom). The binding region is significantly stabilized when bound to peptides, suggesting that the plasticity facilitates binding of diverse sequences. Conservation analyses were performed using ConSurf [92] on 500 sequences with sequence identities ranging from 40% to 95%. Structural conservation was measured using the Ensemblator’s global overlay tool [93], which aligns independently determined 3D structures to identify regions of structural conservation or plasticity. Analyses were performed on all available structures for free calcineurin (4 structures, including human, mouse, and rat orthologs), bound calcineurin (eight structures, including human and cow orthologs), free LC8 (five structures, including mouse, rat, Drosophila, and Plasmodium falciparum orthologs), and bound LC8 (17 structures, including human, rat, Drosophila, and yeast orthologs).
In support of this hypothesis, a similar conservation of structural plasticity is observed in the PxIxIT LMB subunit of the serine/threonine protein phosphatase calcineurin (PP2B or PPP3; Figure 1, bottom) [36,37]. Upon binding, disordered partners gain structural stability and form a β strand that fits within a small binding groove [38]. Like LC8, the available crystal structures indicate that the binding groove is the most plastic region of the free protein. In the bound form, although the LMB-hub on average becomes more rigid, its level of rigidity is partner dependent. For example, the binding interface of LC8 when bound to a peptide from Swallow is fully rigid, while it remains highly dynamic when bound to a peptide from dynein intermediate chain [39]. It is remarkable that some of the most sequence-conserved regions of calcineurin and LC8 are also the most structurally variable parts, supporting the hypothesis that structural plasticity allows accommodation of a diverse set of partners with a wide range of properties. Other examples of well-studied proteins that we categorize here as LMB-hubs, such as the PDZ domains Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
3
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
Box 1. Interface Plasticity of LMB-Hubs Correlates with Binding Diversity It is well recognized in the literature that the plasticity of interacting regions provides them with increased interaction capacity. However, in the retinoblastoma pocket domain (pRB) hub that interacts with the LXCXE linear motif found in viral proteins, a plastic interface in not readily apparent from comparison of temperature factors. The positions of the RB sidechains that create the binding pockets are nearly identical in both the unbound and bound structures and the crystallographic temperature factors of these sidechains are already low in the unliganded structure and remain low upon binding [81]. In human papillomavirus 16 E7, for example, the core motif residues bind tightly into hydrophobic pockets of the pRB binding groove [82]. A more ordered pocket in the unliganded structure would contribute to higher affinity (nM to sub-μM compared to the μM affinity of LC8 partners) and also to the reduced diversity of interactions (pRB is specific to the cell cycle compared to the multitude of systems with LC8). It is also important to note that while the crystal structures do not show plasticity at the interface, plasticity in the protein as a whole is inferred from unfolding and aggregation studies in solution. The Rb pocket is marginally stable and presents multiple conformations in solution but is significantly stabilized when bound to peptide ligands [83].
[33,40,41], WW domains [42,43], and SH2 domains [44,45], are also fully ordered, but have notable functional plasticity within their binding interfaces. Although the examples in Figure 1 show bound structures that are both significantly stabilized, and are the same regardless of the partner peptide, it is not uncommon for LMB-hubs to bind to partners in a more variable manner. For example, the bound form of the SH2 domain of the SLAM-associated protein (SAP) has a variety of different conformations [45,46]. Thus, although our examples suggest a rigidified bound form, there can be variable binding mechanisms for this class of proteins.
LMB-Hubs Can Have Either Enzymatic or Structural Roles LMB-hubs can be segregated into enzymatic proteins (such as phosphatases and kinases) and structural, chaperoning proteins (such as PDZ domains). We refer to these two groups as enzymatic and structural LMB-hubs, respectively. The key point of differentiation is the transient nature of the enzymatic interaction, which requires linear motifs to only bind during the catalytic process. One well-known example is the serine/threonine-specific protein kinase B (PKB, also known as AKT1). PKB is a key basophilic kinase that functions in cell survival, cell cycle progression, metabolism, and apoptosis [47] by targeting and phosphorylating proteins with a minimal consensus sequence of RxRxxS/T [48]. While we differentiate enzymatic from structural hubs based on whether the interaction with the partners is either transient at the catalytic site, or an obligate part of the complex, it is important to note that some enzymatic LMB-hubs have both catalytic and noncatalytic linear motif-binding sites. Mitogen-activated protein kinases, for example, belong to this group, as they mainly recognize their substrates not with the catalytic site but with auxiliary docking surfaces on their kinase domains [49–51]. In contrast to enzymatic LMB-hubs, structural LMB-hubs form stable interactions with partner linear motifs and promote formation and structural organization of higher order protein associations (Figure 2). For example, PDZ domains are 80–90-amino-acid domains that occur within a wide variety of proteins. Although there is variability in the binding motif based on PDZ class, they often bind to the four C-terminal residues of their partner protein, typically containing an xxΦ* motif (where Φ is a hydrophobic residue, and * is the C terminus); notable examples of internal PDZ motifs do exist [52,53]. PDZ domains often arise as repeating clusters within a protein, providing scaffolding for large, macromolecular assemblies [54]. For example, photoreception in the Drosophila eye is dependent on the INAD protein [55], which encodes five separate PDZ domains. These PDZ domains work together to recruit various key components of the photoreceptor transducisome, as well as to promote oligomerization of INAD [56], making them integral to transducisome structure. Unlike enzymatic LMB-hubs, stabilization 4
Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
Trends in Biochemical Sciences
Figure 2. Functions for Dynein Light Chain LC8 as a Structural Linear Motif-Binding Hub Protein (LMB-Hub). Models depicting the role for structural LMBhubs, using LC8 interactions as an example system. LC8 binds to disordered chains at the recognition motif, causing the formation of an adjacent coiled-coil (in Swallow, top) [94], a self-association domain [in the dynein intermediate chain (DIC), middle] [95], or an additional binding event (in dynein intermediate chain and Nup159, bottom) [58,96].
of large molecular assemblies often requires integration of structural LMB-hubs like PDZ domains within the complex, and partners of structural LMB-hubs are consequently competing directly for limited hub pools. In our opinion, LC8 can be firmly placed within the category of structural LMB-hubs that confer stability to their disordered binding partners (Figure 2). LC8 is perhaps best known for its role as a light chain in the dynein complex [57], where as a dimer with symmetrical binding grooves, it binds to a short linear TQT motif from each protomer of the dynein intermediate chain. LC8 binding brings together two chains of the partner protein, facilitating its self-association at a site distant from binding (Figure 2) and stabilizing the essential quaternary motor complex [58]. In addition to structural stabilization of the partner protein, the functional consequences of binding have been demonstrated on the yeast dynein complex where LC8 leads to a significant decrease in complex formation, decreased velocity, and processivity for the motor, as well as an increase in dynein heavy chain aggregation [59]. Parallel examples are the centriolar protein Ana2, and the syntaphilin protein involved in mitochondrial mobility and inhibition of syntaxin-1. In the case of Ana2, LC8 interactions lead to tetramerization of Ana2, facilitating interactions with downstream partners like Sas6 and Mud, and eventually contributing to correct spindle orientation [60,61]. For syntaphilin, interactions with LC8 stabilize a coiled coil structure proximal to the LC8 binding site, and allows it to recruit and anchor axonal mitochondria for synaptic regulation [26]. At the systems level, network analyses suggest that LMB-hubs help to organize and connect biological processes and pathways. Removal of hub protein nodes in interaction networks tends to make systems more sensitive to additional perturbations [62]. This suggests that the Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
5
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
structural plasticity of LMB-hub binding sites endows a functional adaptability that allows cells to efficiently respond to various requirements. We therefore propose that LMB-hubs are particularly critical to a cell’s ability to respond effectively to external stimuli such as increased salinity or temperature. Unfortunately, knockouts (or even knockdowns) of hubs are often lethal [63–65], which complicate experimental tests of the role of hubs in the response to external stimuli. Intriguingly, recent work by King et al. [66] reports the generation of a limb bud mesoderm-restricted conditional knockout of LC8, resulting in a potentially informative animal (and cell line) model. Although these cells are not yet fully characterized, preliminary analyses indicate that knockouts of LC8 lead to impairments in the Hippo signaling pathway [66], amongst other things.
LMB-Hubs Availability Is Tightly Controlled and Feedback-Regulation May Be Common Since LMB-hubs form 1:1 complexes with many partners, several questions arise. How can many different partners bind the same LMB-hub when they all compete for a common binding site? With a limited pool of each hub, how does a cell allocate sufficient protein for each function? For enzymatic LMB-hubs, the case can be made that transient binding facilitates the maintenance of sufficient pools of hub protein. However, structural LMB-hubs tend to be integrated into high-order complexes where they are inaccessible to other partners. For many hub proteins there is evidence for ‘just-in-time’ synthesis [65]; that is, protein expression is commensurate with the cell’s needs, suggesting that hub protein concentrations are delicately balanced and exquisitely controlled. For the majority of hub proteins, such regulatory systems are not well described; however, we hypothesize that feedback loops that enable LMB-hub self-regulation are common. A prime example of hub protein self-regulation is LC8 regulation by the transcription factor ASCIZ [22]. ASCIZ is an 88-kDa protein identified as a transcription factor for the LMB-hub protein LC8 [21]. Recent mouse experiments demonstrate that LC8-1 (the more highly expressed of two LC8 isoforms) germline knockouts lead to death early in embryogenesis, while a conditional knockout can proceed with severe developmental defects [66]. Additionally, ASCIZ knockouts display identical, but attenuated, phenotypes, suggesting that the most essential role for ASCIZ is regulation of LC8 concentrations within the cell [66]. Structural work on the LC8– ASCIZ complex reveals a remarkable number of LC8 binding sites (11 within human ASCIZ) and shows that this multivalency is important for regulation of LC8 concentrations within the cell [22]. Although the specifics of this regulation require further research, it appears that when LC8 concentrations are high, the full complement of ASCIZ binding sites are occupied, leading to inhibition of transcription of the LC8-1 gene (Figure 3). Conversely, when LC8 concentrations are low, the linear motifs on ASCIZ are largely unoccupied, leading to LC8 transcription and translation [22]. Seemingly, LC8 is part of a feedback loop tightly regulating its own concentrations within the cell. Due to their importance, LMB-hub motif binding preferences are often thoroughly studied, leading to the generation partner prediction tools [15,67]. It is therefore possible to combine information about gene regulation with partner predictions to identify potential feedback loops. For example, 14-3-3s are a family of structural LMB-hub proteins involved in cell-cycle control, signal transduction, protein trafficking, and apoptosis, among other things [68]. They are often dimeric, and usually bind to phosphorylated partners at a basic and proline-rich motif [69]. Binding facilitates partner protein folding, localization, and stimulation or inhibition of downstream interactions [70]. The 14-3-3-Pred [67] prediction tool has accurately identified 14-3-3 motifs in various partners/systems [71]. Using this tool, we identified a high-probability 14-3-3 binding motif within ATF-1, a cAMP response element binding transcription factor known to regulate 6
Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
Trends in Biochemical Sciences
Figure 3. Structural LMB-Hub Transcriptional Self-Regulation. A proposed model of LC8 regulation via the transcription factor ATMIN-substrate Chk-interacting Zn2+ finger (ASCIZ), where low concentrations of the structural LMB-hub (LC8, blue) would result in an active form of the transcription factor, facilitating increased hub protein production. When hub concentrations are high, binding to the transcription factor results in an inactive form of the transcription factor, turning off hub production. We hypothesize that this regulatory tactic is also shared by 14-3-3 whose transcription factor ATF-1 has a predicted site for binding 14-3-3. Red helices represent DNA-binding domains, green lines are IDRs, grey double-helices are DNA, Small ‘Ps’ are phosphorylated residues on ATF-1, and LMB-hubs are shown in dark blue.
intracellular 14-3-3ζ concentrations [72]. Additionally, the motif occurs at a known phosphorylation site (S63, UniProt code: P18846), previously associated with enhanced ATF-1 transcriptional activity [73,74]. It remains to be shown whether 14-3-3 indeed binds the predicted motif within ATF-1, but if this were the case, it suggests an intriguing self-regulation mechanism similar to what we observe with LC8 (Figure 3). This idea is speculative, but not out of the realms of possibility, considering the necessity of mechanisms for effective regulation of intracellular hub concentrations. Although we focus on direct transcriptional self-regulation, other mechanisms of LMB-hub feedback regulation are common (Box 2). Additionally, the presence of LMB-hubs within large macromolecular assemblies further supports the importance of tight regulation of protein concentrations. The gene balance hypothesis, developed by Papp et al. [75] and furthered in many other studies [76–78], states that changing the concentration of one subunit of a complex leads to unproductive subcomplexes that are missing key proteins. Therefore, a stoichiometric imbalance of members of the complex affects the amount of complete product. On the genomic scale, addition of a single chromosome is actually more detrimental than the addition of a complete genome to make a polyploid organism [79,80], purportedly due to the imbalance of macromolecular complex components [78]. As key components of many complexes, structural
Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
7
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
Box 2. Alternative Mechanisms for Self-Regulation of LMB-Hubs Indirect routes of self-regulation can be identified for many hubs. For example, the plant protein radical-induced cell death 1 (RCD1) interacts with a (DE)x(1,2)(YF)x(1,4)(DE)L motif in IDRs in many transcription factors [84]. Knockouts of RCD1 result in pleiotropic effects in stress and developmental pathways. One well studied system involves the generation of reactive oxygen species (ROS) from mitochondrial dysfunction. High concentrations of ROS cause inactivation, aggregation, and the eventual degradation of RCD1. RCD1, in turn, binds to and inhibits various transcription factors tasked with production of MDS (mitochondrial dysfunction stimulon) genes, whose role is to ameliorate the production, and control the impact of ROS [85]. Thus, when RCD1 concentrations are high, production of MDSs is decreased, leading to increased ROS, which in turn facilitate RCD1 degradation. We hypothesize that as the regulatory pathways for various LMB-hubs are more adequately described, similar feedback loops will be found. In addition to self-regulation at the transcriptional level, the activity of LMB-hubs can be modified post-translation. For enzymatic LMB-hubs like kinases, autophosphorylation is a common mechanism, and can even allow for tuning of enzymatic activity based on kinase concentrations [86]. Post-translational self-regulation of structural LMB-hubs is often less conspicuous. One interesting example is Disheveled (Dsh), an integral protein in the Wnt signalosome complex. The Wnt signaling pathway is important for animal development, and is involved in the regulation of cell polarity, movement, differentiation, survival, and proliferation [87]. Members of the pathway vary with function, but in all cases it contains a Wnt ligand, a membrane-bound receptor from the Frizzled family, and Dsh. Signal transduction relies upon a low-affinity interaction (about 10 μM) between the PDZ domain of Dsh with the linear motif from Frizzled [88]. Dsh self-associates via an N-terminal Dix domain, and forms large functional polymers with properties similar to membraneless organelles [89]. Polymerization leads to a concomitant increase in the local concentration of Dsh, allowing Dsh to overcome the low affinity interactions with Frizzled, and facilitates recruitment of signalosome subunits [89]. Dsh activity is partially regulated by an internal PDZ motif at its C terminus, which can associate with the Dsh PDZ domain and form a closed conformation [90,91]. The closed conformation displays decreased signaling activity and membrane localization, and is therefore an example of LMB-hub autoinhibition. We anticipate that self-regulation at this level is less common than self-regulation at the transcriptional level; however, autoinhibition and activation are likely to be important mechanisms for the mediation of active LMB-hub concentrations.
LMB-hubs would require fine tuning to keep concentrations from being too high or too low; both of which could lead to unproductive subcomplexes. We suggest that self-regulation is vital for balancing roles in their various macromolecular complexes.
Concluding Remarks In this work, we propose a newly defined category of hub proteins, the LMB-hubs. We are by no means the first to recognize that dynamic/date hubs typically bind linear motifs at a single interface, but we find this method of classification to be clearer and more concise when compared to the alternatives in common use. We also suggest further vernacular changes in order to differentiate between structural and enzymatic LMB-hubs, as these types of proteins are processed and used differently within cells. Current data indicate that structural LMB-hubs like calcineurin and LC8 often have dynamic binding interfaces in order to accommodate many motif variants, but are surprisingly conserved in sequence, leading us to hypothesize that the conserved feature is binding site plasticity. Although thorough analyses of LMB-hub regulatory systems are scarce, available data on LC8, RCD1, and Dsh, along with predictions for 14-3-3, lead us to speculate that self-regulation is an important and common feature for LMB-hubs specifically. Further work on LMB-hub transcriptional regulation can validate this hypothesis (see Outstanding Questions), and we propose using the existing motif predictors on known transcription factors to facilitate progress in this direction. References 1.
2.
3.
8
Ekman, D. et al. (2006) What properties characterize the hub proteins of the protein-protein interaction network of Saccharomyces cerevisiae? Genome Biol. 7, R45 Grigoriev, A. (2003) On the number of protein-protein interactions in the yeast proteome. Nucleic Acids Res. 31, 4157–4161 He, X. and Zhang, J. (2006) Why do hubs tend to be essential in protein networks? PLoS Genet. 2, e88
Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
4. 5.
6.
Yu, H. et al. (2008) High-quality binary protein interaction map of the yeast interactome network. Science 322, 104–110 Komurov, K. and White, M. (2007) Revealing static and dynamic modular architecture of the eukaryotic protein interaction network. Mol. Syst. Biol. 3, 110 Wu, X. et al. (2009) The properties of hub proteins in a yeastaggregated cell cycle network and its phase sub-networks. Proteomics 9, 4812–4824
Outstanding Questions Although hubs are often pleiotropic, are their interactions skewed? As in, are they active in one main functional pathway, with other minor roles? Autophosphorylation of enzymatic LMB-hubs is a common method for self-regulation. Do enzymatic LMBhubs also display feedback regulation at the transcriptional level? As hub protein concentrations are tightly controlled, what is the impact of a minor change in binding affinity for a specific partner? Would partner binding lead to a feedback change in hub protein expression? To verify self-regulation, one possibility is to transfect with a plasmid encoding an LMB-hub, and then monitor the expression of the genomic LMB-hub protein. If the hub is self-regulating, would transfection result in some inhibition of genomic expression of the LMB-hub? Can we describe the entropic benefit of a plastic-binding interface with respect to promiscuity? Is 14-3-3 indeed self-regulated as proposed here? Recently, a limb bud mesodermrestricted conditional knockout of LC8 was successfully generated, resulting in a potentially informative animal (and cell line) model. What type of experiments performed on this cell line could illuminate the impact of hub protein depletion on cellular homeostasis? Considering that LC8 dimerizes so many different client proteins in both the nucleus and the cytoplasm, how does the cell manage to keep all these client proteins separated? Can LC8 hetero-dimerize different client proteins?
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
7.
8.
9.
10. 11.
12.
13. 14. 15.
16. 17.
18.
19.
20.
21.
22. 23.
24.
25.
26.
27.
28. 29.
30. 31.
32.
Garamszegi, S. et al. (2013) Signatures of pleiotropy, economy and convergent evolution in a domain-resolved map of human–virus protein–protein interaction networks. PLoS Pathog. 9, e1003778 Kim, P.M. et al. (2006) Relating three-dimensional structures to protein networks provides evolutionary insights. Science 314, 1938–1941 Van Roey, K. et al. (2014) Short linear motifs: ubiquitous and functionally diverse protein interaction modules directing cell regulation. Chem. Rev. 114, 6733–6778 American Chemical Society Davey, N.E. et al. (2012) Attributes of short linear motifs. Mol. BioSyst. 8, 268–281 Uchikoga, N. et al. (2016) Specificity of broad protein interaction surfaces for proteins with multiple binding partners. Biophys. Physicobiol. 13, 105–115 Hu, G. et al. (2017) Functional analysis of human hub proteins and their interactors involved in the intrinsic disorder-enriched interactions. Int. J. Mol. Sci. 18, E2761 Gouw, M. et al. (2018) The eukaryotic linear motif resource – 2018 update. Nucleic Acids Res. 46, D428–D434 Barbar, E. (2008) Dynein Light Chain LC8 is a dimerization hub essential in diverse protein networks. Biochemistry 47, 503–508 Jespersen, N. et al. (2019) Systematic identification of recognition motifs for the hub protein LC8. Life Sci. Alliance 2, e201900366 Rapali, P. et al. (2011) DYNLL/LC8: a light chain subunit of the dynein motor complex and beyond. FEBS J. 278, 2980–2996 Erdős, G. et al. (2017) Novel linear motif filtering protocol reveals the role of the LC8 dynein light chain in the Hippo pathway. PLoS Comput. Biol. 13, e1005885 Nyarko, A. and Barbar, E. (2011) Light chain-dependent selfassociation of dynein intermediate chain. J. Biol. Chem. 286, 1556–1566 Toropova, K. et al. (2019) Structure of the dynein-2 complex and its assembly with intraflagellar transport trains. Nat. Struct. Mol. Biol. 26, 823–829 He, Y.J. et al. (2018) DYNLL1 binds to MRE11 to limit DNA end resection in BRCA1-deficient cells. Nature 563, 522–526 Jurado, S. et al. (2012) ATM substrate Chk2-interacting Zn 2+ finger (ASCIZ) is a bi-functional transcriptional activator and feedback sensor in the regulation of dynein light chain (DYNLL1) expression. J. Biol. Chem. 287, 3156–3164 Clark, S. et al. (2018) Multivalency regulates activity in an intrinsically disordered transcription factor. eLife 7, e36258 Becker, J.R. et al. (2018) The ASCIZ-DYNLL1 axis promotes 53BP1-dependent non-homologous end joining and PARP inhibitor sensitivity. Nat. Commun. 9, 5406 Singh, P.K. et al. (2019) Dynein light chain binding determines complex formation and posttranslational stability of the Bcl-2 family members Bmf and Bim. Cell Death Differ. 27, 434–450 Petryszak, R. et al. (2016) Expression Atlas update – an integrated database of gene and protein expression in humans, animals and plants. Nucleic Acids Res. 44, D746–D752 Chen, Y.-M. et al. (2009) dynein light chain LC8 regulates syntaphilin-mediated mitochondrial docking in axons. J. Neurosci. 29, 9429–9438 Wieczorek, M. et al. (2017) Major histocompatibility complex (MHC) class I and MHC class II proteins: conformational plasticity in antigen presentation. Front. Immunol. 8, 292 Dausset, J. (1981) The major histocompatibility complex in man. Science (80-. ). 213, 1469–1474 Charles A Janeway, J. et al. (2001) The major histocompatibility complex and its functions. In Immunobiology: The Immune System in Health and Disease (5th edition), Garland Science Rapin, N. et al. (2008) MHC motif viewer. Immunogenetics 60, 759–765 Reich, Z. et al. (1997) Stability of empty and peptide-loaded class II major histocompatibility complex molecules at neutral and endosomal pH: comparison to class I proteins. Proc. Natl. Acad. Sci. 94, 2495–2500 Agniswamy, J. et al. (2007) Plasticity of S2-S4 specificity pockets of executioner caspase-7 revealed by structural and kinetic analysis. FEBS J. 274, 4752–4765
33. Fujiwara, Y. et al. (2015) Crystal structure of afadin PDZ domainnectin-3 complex shows the structural plasticity of the ligandbinding site. Protein Sci. 24, 376–385 34. Clark, S. et al. (2016) The Anchored Flexibility Model in LC8 motif recognition: insights from the Chica complex. Biochemistry 55, 199–209 35. Holland, D.O. et al. (2017) Protein-protein binding selectivity and network topology constrain global and local properties of interface binding networks. Sci. Rep. 7, 5631 36. Roy, J. et al. (2007) A conserved docking site modulates substrate affinity for calcineurin, signaling output, and in vivo function. Mol. Cell 25, 889–901 37. Wigington, C.P. et al. (2019) Systematic discovery of short linear motifs decodes calcineurin phosphatase signaling. bioRxiv. Published online May 9, 2019. https://doi.org/10.1101/632547 38. Grigoriu, S. et al. (2013) The molecular mechanism of substrate engagement and immunosuppressant inhibition of calcineurin. PLoS Biol. 11, e1001492 39. Hall, J. et al. (2008) Differences in dynamic structure of LC8 monomer, dimer, and dimer–peptide complexes. Biochemistry 47, 11940–11952 40. Lee, J.H. et al. (2011) The structural flexibility of the shank1 PDZ domain is important for its binding to different ligands. Biochem. Biophys. Res. Commun. 407, 207–212 41. Kang, B.S. et al. (2003) Molecular roots of degenerate specificity in syntenin’s PDZ2 domain: Reassessment of the PDZ recognition paradigm. Structure 11, 845–853 42. Koepf, E.K. et al. (2008) WW: An isolated three-stranded antiparallel β-sheet domain that unfolds and refolds reversibly; evidence for a structured hydrophobic cluster in urea and GdnHCl and a disordered thermal unfolded state. Protein Sci. 8, 841–853 43. Iglesias-Bexiga, M. et al. (2015) WW domains of the Yes-kinaseassociated-protein (YAP) transcriptional regulator behave as independent units with different binding preferences for PPxY motif-containing ligands. PLoS One 10, e0113828 44. Kay, L.E. et al. (1996) Correlation between dynamics and high affinity binding in an SH2 domain interaction. Biochemistry 35, 361–368 45. Hwang, P.M. et al. (2002) A ‘three-pronged’ binding mechanism for the SAP/SH2D1A SH2 domain: structural basis and relevance to the XLP syndrome. EMBO J. 21, 314–323 46. Finerty, P.J. et al. (2002) Side-chain dynamics of the SAP SH2 domain correlate with a binding hot spot and a region with conformational plasticity. J. Mol. Biol. 322, 605–620 47. Hanada, M. et al. (2004) Structure, regulation and function of PKB/AKT – a major therapeutic target. Biochim. Biophys. Acta - Proteins Proteomics 1697, 3–16 48. Obata, T. et al. (2000) Peptide and protein library screening defines optimal substrate motifs for AKT/PKB. J. Biol. Chem. 275, 36108–36115 49. Tanoue, T. et al. (2000) A conserved docking motif in MAP kinases common to substrates, activators and regulators. Nat. Cell Biol. 2, 110–116 50. Zeke, A. et al. (2015) Systematic discovery of linear binding motifs targeting an ancient protein interaction surface on MAP kinases. Mol. Syst. Biol. 11, 837 51. Gógl, G. et al. Disordered protein kinase regions in regulation of kinase domain cores. Trends Biochem. Sci. 44, 300–311 52. Vaccaro, P. and Dente, L. (2002) PDZ domains: troubles in classification. FEBS Lett. 512, 345–349 53. Penkert, R.R. et al. (2004) Internal recognition through PDZ domain plasticity in the Par-6-Pals1 complex. Nat. Struct. Mol. Biol. 11, 1122–1127 54. Harris, B.Z. and Lim, W.A. (2001) Mechanism and role of PDZ domains in signaling complex assembly. J. Cell Sci. 114, 3219–3231 55. Shieh, B.H. and Zhu, M.Y. (1996) Regulation of the TRP Ca2+ channel by INAD in Drosophila photoreceptors. Neuron 16, 991–998 56. Ranganathan, R. and Ross, E.M. (1997) PDZ domain proteins: scaffolds for signaling complexes. Curr. Biol. 7, R770–R773 57. Pfister, K.K. et al. (1982) Purification and polypeptide composition of dynein ATPases from Chlamydomonas flagella. Cell Motil. 2, 525–547
Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
9
Trends in Biochemical Sciences An official publication of the INTERNATIONAL UNION OF BIOCHEMISTRY AND MOLECUL AR BIOLOGY
58. Hall, J. et al. (2009) Multivalency in the assembly of intrinsically disordered dynein intermediate chain. J. Biol. Chem. 284, 33115–33121 59. Rao, L. et al. (2013) The yeast dynein Dyn2-Pac11 complex is a dynein dimerization/processivity factor: structural and singlemolecule characterization. Mol. Biol. Cell 24, 2362–2377 60. Slevin, L.K. et al. (2014) The mechanism of dynein light chain LC8-mediated oligomerization of the Ana2 centriole duplication factor. J. Biol. Chem. 289, 20727–20739 61. Wang, C. et al. (2011) An Ana2/Ctp/Mud complex regulates spindle orientation in Drosophila neuroblasts. Dev. Cell 21, 520–533 62. Han, J.-D.J. et al. (2004) Evidence for dynamically organized modularity in the yeast protein–protein interaction network. Nature 430, 88–93 63. Jeong, H. et al. (2001) Lethality and centrality in protein networks. Nature 411, 41–42 64. Zotenko, E. et al. (2008) Why do hubs in the yeast protein interaction network tend to be essential: reexamining the connection between the network topology and essentiality. PLoS Comput. Biol. 4, e1000140 65. Batada, N.N. et al. (2006) Evolutionary and physiological importance of hub proteins. PLoS Comput. Biol. 2, e88 66. King, A. et al. (2019) Dynll1 is essential for development and promotes endochondral bone formation by regulating intraflagellar dynein function in primary cilia. Hum. Mol. Genet. 28, 2573–2588 67. Madeira, F. et al. (2015) 14-3-3-Pred: improved methods to predict 14-3-3-binding phosphopeptides. Bioinformatics 31, 2276–2283 68. Stevers, L.M. et al. (2018) Modulators of 14-3-3 protein-protein interactions. J. Med. Chem. 61, 3755–3778 69. Johnson, C. et al. (2010) Bioinformatic and experimental survey of 14-3-3-binding sites. Biochem. J. 427, 69–78 70. Aitken, A. (2006) 14-3-3 proteins: a historic overview. Semin. Cancer Biol. 16, 162–172 71. Yuan, L. et al. (2019) 14-3-3 signal adaptor and scaffold proteins mediate GPCR trafficking. Sci. Rep. 9, 11156 72. Kasinski, A. et al. (2014) Transcriptional regulation of YWHAZ, the gene encoding 14-3-3ζ. PLoS One 9, e93480 73. Gupta, P. and Prywes, R. (2002) ATF1 phosphorylation by the ERK MAPK pathway is required for epidermal growth factorinduced c- jun expression. J. Biol. Chem. 277, 50550–50556 74. Kobayashi, M. et al. (1997) Phosphorylation of ATF-1 enhances its DNA binding and transcription of the Na,K-ATPase alpha 1 subunit gene promoter. Nucleic Acids Res. 25, 877–882 75. Papp, B. et al. (2003) Dosage sensitivity and the evolution of gene families in yeast. Nature 424, 194–197 76. Veitia, R.A. (2004) Gene dosage balance in cellular pathways. Genetics 168, 569–574 77. Veitia, R.A. et al. (2008) Cellular reactions to gene dosage imbalance: genomic, transcriptomic and proteomic effects. Trends Genet. 24, 390–397 78. Birchler, J.A. and Veitia, R.A. (2012) Gene balance hypothesis: connecting issues of dosage sensitivity across biological disciplines. Proc. Natl. Acad. Sci. U. S. A. 109, 14746–14753
10
Trends in Biochemical Sciences, Month 2020, Vol. xx, No. xx
79. Blakeslee, A.F. et al. (1920) Chromosomal duplication and Mendelian phenomena in Datura mutants. Science (80-.) 52, 388–390 80. Bridges, C.B. Sex in relation to chromosomes and genes. Am. Nat., 59, 127–137 81. Balog, E.R.M. et al. (2011) Crystal structure of the unliganded retinoblastoma protein pocket domain. Proteins 79, 2010–2014 82. Palopoli, N. et al. (2018) Short linear motif core and flanking regions modulate retinoblastoma protein binding affinity and specificity. Protein Eng. Des. Sel. 31, 69–77 83. Chemes, L.B. et al. (2013) Folding of a cyclin box. J. Biol. Chem. 288, 18923–18938 84. O’Shea, C. et al. (2017) Structures and short linear motif of disordered transcription factor regions provide clues to the interactome of the cellular hub protein radical-induced cell death. J. Biol. Chem. 292, 512–527 85. Shapiguzov, A. et al. (2019) Arabidopsis RCD1 coordinates chloroplast and mitochondrial functions through interaction with ANAC transcription factors. eLife 8, e43284 86. Dodson, C.A. et al. (2013) A kinetic test characterizes kinase intramolecular and intermolecular autophosphorylation mechanisms. Sci. Signal. 6, ra54–ra54 87. Cadigan, K.M. and Waterman, M.L. (2012) TCF/LEFs and Wnt signaling in the nucleus. Cold Spring Harb. Perspect. Biol. 4, a007906 88. Wong, H.-C. et al. (2003) Direct binding of the PDZ domain of Dishevelled to a conserved internal sequence in the C-terminal region of Frizzled. Mol. Cell 12, 1251–1260 89. Bienz, M. (2014) Signalosome assembly by domains undergoing dynamic head-to-tail polymerization. Trends Biochem. Sci. 39, 487–495 90. Lee, H.-J. et al. (2015) Conformational change of Dishevelled plays a key regulatory role in the Wnt signaling pathways. eLife 4, e08142 91. Qi, J. et al. (2017) Autoinhibition of Dishevelled protein regulated by its extreme C terminus plays a distinct role in Wnt/β-catenin and Wnt/planar cell polarity (PCP) signaling pathways. J. Biol. Chem. 292, 5898–5908 92. Ashkenazy, H. et al. (2016) ConSurf 2016: an improved methodology to estimate and visualize evolutionary conservation in macromolecules. Nucleic Acids Res. 44, W344–W350 93. Brereton, A.E. and Karplus, P.A. (2018) Ensemblator v3: robust atom-level comparative analyses and classification of protein structure ensembles. Protein Sci. 27, 41–50 94. Kidane, A.I. et al. (2013) Structural features of LC8-induced selfassociation of swallow. Biochemistry 52, 6011–6020 95. Benison, G. et al. (2006) Heteronuclear NMR identifies a nascent helix in intrinsically disordered dynein intermediate chain: implications for folding and dimerization. J. Mol. Biol. 362, 1082–1093 96. Stelter, P. et al. (2007) Molecular basis for the functional interaction of dynein light chain with the nuclear-pore complex. Nat. Cell Biol. 9, 788–796