A LINC between the nucleus and the cytoskeleton takes form(in)

Abstract

Dynamic nuclear positioning requires the formation of robust connections between the cytoskeleton and components of the LINC complex, a protein assembly that spans the nuclear envelope. A new study by Lim and colleagues reveals the mechanism of association between the LINC complex proteins Nesprin-1/2 Giant and the cytoplasmic formin FHOD1.

Active positioning of the nucleus within the cytoskeletal framework is a crucial component of cellular division, migration and differentiation (Gundersen and Worman, 2013). During nuclear migration, force is exerted across the nuclear envelope through direct connections with the cytoskeleton. These connections are primarily mediated by a complex known as the Linker of Nucleoskeleton and Cytoskeleton (LINC) that spans the inner and outer nuclear membranes (Starr and Fridolfsson, 2010). Within the perinuclear space, the core of the LINC complex is established through binding of Sad1-UNC84 (SUN) proteins, which pass through the inner nuclear membrane, to Klarsicht/ANC-1/Syne homology (KASH) proteins, which cross the outer nuclear membrane (Figure 1A) (Starr and Fridolfsson, 2010).

Figure 1. Interactions between the LINC complex protein Nesprin-2 Giant and the formin FHOD1 establish robust connections between the nuclear envelope and the actin cytoskeleton.

(A) A simplified model depicting components of the LINC complex and their interactions with TAN lines. The C-terminal KASH domain of the KASH family protein Nesprin-2G (Nes2G, blue) interacts with the SUN domain of a SUN family protein (green), forming an oligomeric complex within the perinuclear space located between the inner and outer nuclear membranes (INM and ONM). Nes2G contains 56 spectrin repeats (dashed lines indicate repeats omitted for clarity) that extend outward from the ONM and into the cytoplasm to establish connections with actin filaments (gray) that are localized in TAN lines. Spectrin repeats 11 and 12 (SR11–12) of Nes2G bind to the formin FHOD1 (yellow) through direct interactions with its N-terminal spectrin repeat binding module (SRBM) and formin homology 3 (FH3) domain. The N-terminal domain of Nes2G also contains two calponin homology domains (CHD) that bind actin filaments. (B) The crystal structure (PDB ID 6XF1) of the N-terminal domain of human FHOD1 (SRBM, orange; FH3, yellow) complexed with spectrin repeats 11–12 of human Nes2G (SR11, SR12; blue) determined by Lim and colleagues. (C) A magnified view of the structure in the same orientation as in (B) highlighting major sites of interaction between the FH3 domain of FHOD1 and SR12 of Nes2G (left) and the SRBM of FHOD1 and SR11–12 of Nes2G (right). The DxWLD[IVLA]xE motif is shown in gray. Residues identified as participating in interactions between the domains are shown in stick representation with CPK coloring.

In migrating fibroblasts, the KASH family protein Nesprin-2 Giant (Nes2G) plays an integral role in nuclear positioning through its interactions with SUN proteins at its C-terminus and cytoskeletal elements called transmembrane actin-associated (TAN) lines at its N-terminus (Figure 1A) (Luxton et al., 2010). Nes2G’s cytoplasmic domain contains two actin-binding calponin homology domains followed by 56 spectrin repeats (Kutscheidt et al., 2014). A recent study identified a specific interaction between Nes2G’s spectrin repeats 11 and 12 (SR11–12) and the formin FHOD1, which is a key component of TAN lines (Kutscheidt et al., 2014). Formins are multidomain proteins that promote actin filament nucleation and elongation (Pruyne, 2017). In addition to its actin polymerization activity (Patel et al., 2018), FHOD1 also mediates actin filament bundling (Schönichen et al., 2013). Consistent with these functions, FHOD1 localizes along the lengths of TAN lines and other bundled actin structures (Kutscheidt et al., 2014; Schönichen et al., 2013).

In this issue of Structure, a study by Lim and colleagues uncovers the molecular details of the interaction between Nes2G and FHOD1 (Lim et al., 2021). This work elucidates the structural mechanism underlying the association between SR11–12 of human Nes2G and an N-terminal fragment of FHOD1 (FHOD1-N) that contains the formin’s conserved formin homology 3 (FH3) domain. A crystal structure of the Nes2G:FHOD1 interacting complex reveals canonical folds for Nes2G’s two spectrin repeats and for the five armadillo repeats that comprise FHOD1’s FH3 domain (Figure 1B). The proteins associate with a 1:1 stoichiometry via a continuous binding surface that extends longitudinally along the spectrin repeats. The majority of the interaction occurs between Nes2G’s SR12 and FHOD1’s FH3 domain. However, additional contacts made by the long helix that connects SR11 and SR12 reveal a key mechanistic function for a small domain located directly N-terminal to the FH3 domain (Figure 1C). This domain adopts a ubiquitin super-fold, which is structurally divergent from the GTPase Binding Domains that are commonly located at this position in formin sequences (Pruyne, 2017). In light of its role in Nes2G:FHOD1 complex assembly, the authors name this domain the “SR-binding module” (SRBM).

Lim and colleagues perform a careful mutational analysis to confirm that the charged network located at the interface between SR12 and the FH3 domain is necessary for binding. In contrast, mutations of interacting residues at the SR11/12:SRBM interface still exhibit residual binding, suggesting that FHOD1’s SRBM functions as a modulator of binding rather than as an autonomous domain with an independent function. Based on their mutational studies, the authors identify the sequence DxWLD[IVLA]xE as the signature motif within a spectrin repeat that enables FHOD1 binding (Figure 1C). This motif is also 100% conserved in spectrin repeat 18 in the related KASH family protein Nesprin-1 Giant (Nes1G), and the authors confirm its interaction with FHOD1-N using a combination of rigorous biochemical and structural approaches.

Identification of the essential interfacial residues enables Lim and colleagues to assay the significance of the SR11–12:FH3 interaction in vivo. A previous study showed that expression of Nes2G SR11–13 in fibroblasts inhibits rearward nuclear movement and the establishment of polarity as measured by centrosome orientation (Antoku et al., 2019). This effect likely arises from a competitive interaction between Nes2G SR11–13 and FHOD1 that inhibits binding of wild-type Nes2G to the formin. In this study, the authors report that expression of Nes2G SR11–12 is sufficient to reproduce this effect in fibroblasts. Further, expression of a variant of Nes2G SR11–12 containing point-mutations that disrupt SR12:FH3 binding does not inhibit nuclear movement or centrosome orientation. Similarly, the authors demonstrate that expression of FHOD1 mutants that do not bind Nes2G does not rescue nuclear movement or centrosome orientation defects in FHOD1-depleted cells despite proper localization of these mutant formins to actin bundles.

Like many metazoan formins, FHOD1 is autoinhibited through the association of its FH3 domain (which is also commonly known as the formin’s diaphanous inhibitory domain, or DID) with a helical, diaphanous autoregulatory domain (DAD) sequence located at its C-terminus (Pruyne, 2017). In FHOD1, autoinhibition can be released through the phosphorylation of a number of serine and threonine residues within the DAD helix by Rho kinase (Takeya et al., 2008). Since both Nes2G SR11–12 and the DAD helix bind the FH3 domain, the authors explore the possibility that they might do so competitively. They generate a structural model that incorporates a putatively bound DAD helix onto the Nes2G:FHOD1 complex. This structure suggests that there are two separate and distinct binding sites for SR11–12 and the DAD helix on FHOD1’s FH3 domain. The authors demonstrate that the SR11–12:FH3 complex and FHOD1’s DAD helix co-elute following size exclusion chromatography, further supporting their model for two independent binding sites and suggesting that FHOD1 DID:DAD autoinhibition does not interfere with the interaction with Nes1G or Nes2G.

Formins that possess an FH3/DID domain are known to also encode either a GTPase Binding Domain (GBD) or a G2 domain, which Lim and colleagues now reveal to be an SRBM (Pruyne, 2017). Binding of a small GTPase to the GBD and FH3/DID can disrupt the DID:DAD association and partially relieve autoinhibition (Pruyne, 2017). A phylogenetic analysis carried out by the authors reveals distinct conservation of key residues within the binding sites of FH3 domains that are tethered to either a GBD or an SRBM. Their analysis suggests that the FH3 domain has evolved to either bind a small GTPase or spectrin repeats, depending on which domain (GBD or SRBM) it is tethered to. Further supporting their identification of the SRBM as a modulator of FH3 binding specificity, the authors detect no interaction between FHOD1 and the small GTPase Rac1.

Lim and colleagues integrate their findings into a model that accounts for the multiple interactions that occur among Nes1G/Nes2G, FHOD1 and actin. In addition to their interactions with FHOD1, Nes1G and Nes2G bind actin filaments directly via their two N-terminal calponin homology domains (Figure 1A). Dimerization of FHOD1 further doubles the number of actin-binding sites. This clustering of multiple actin-binding domains via Nes1G and Nes2G’s association with FHOD1 likely enhances the overall stability of their interactions with actin filaments in TAN lines. Taken together, the elegant structural, biochemical and phylogenetic analyses presented by the authors provide direct insight into the establishment of connections between the nucleus and the cytoskeleton that are stable enough to withstand the large mechanical forces involved in nuclear positioning.

This model also opens interesting avenues for future exploration and reveals important questions that remain to be addressed. For example, the spectrin repeat domains of Nes1G and Nes2G are very large, yet the FHOD1-recognition sequence is confined to a single spectrin repeat. Does a change in the position of this binding site still support stable connections between the LINC complex and TAN lines? It would also be interesting to establish whether the number of FHOD1- and actin-binding sites located in the sequences of Nes1G and Nes2G influence the dynamics of TAN line assembly. It further remains to be determined whether FHOD1’s interactions with Nes1G and Nes2G alter its actin polymerization or bundling activities. Finally, the authors identify the SRBM as a key determinant of formin binding specificity. Can this domain serve as a modulator of interactions mediated by formins in other contexts? Further investigation into the implications of these interactions on cytoskeletal dynamics will undoubtedly yield novel insights. While many mechanistic aspects remain to be discovered, the elegant work presented by Lim and colleagues brings the field one step closer towards understanding the molecular basis for the intricate process of nuclear migration.