Skip to main content
NCBI home page
Protein Science : A Publication of the Protein Society logoLink to Protein Science : A Publication of the Protein Society
. 2024 Apr 12;33(5):e4980. doi: 10.1002/pro.4980

Retromer‐mediated recruitment of the WASH complex involves discrete interactions between VPS35, VPS29, and FAM21

Miguel Romano‐Moreno 1,2, Elsa N Astorga‐Simón 1, Adriana L Rojas 1, Aitor Hierro 1,3,4,✉
PMCID: PMC11010949  PMID: 38607248

Abstract

Endosomal trafficking ensures the proper distribution of lipids and proteins to various cellular compartments, facilitating intracellular communication, nutrient transport, waste disposal, and the maintenance of cell structure. Retromer, a peripheral membrane protein complex, plays an important role in this process by recruiting the associated actin‐polymerizing WASH complex to establish distinct sorting domains. The WASH complex is recruited through the interaction of the VPS35 subunit of retromer with the WASH complex subunit FAM21. Here, we report the identification of two separate fragments of FAM21 that interact with VPS35, along with a third fragment that binds to the VPS29 subunit of retromer. The crystal structure of VPS29 bound to a peptide derived from FAM21 shows a distinctive sharp bend that inserts into a conserved hydrophobic pocket with a binding mode similar to that adopted by other VPS29 effectors. Interestingly, despite the network of interactions between FAM21 and retromer occurring near the Parkinson's disease‐linked mutation (D620N) in VPS35, this mutation does not significantly impair the direct association with FAM21 in vitro.

Keywords: endosome, membrane trafficking, protein complex, retromer, x‐ray crystallography

1. INTRODUCTION

Retromer is a protein complex that promotes the selective retrieval and recycling of hundreds of integral membrane proteins, referred to as cargo, from endosomes back to the trans‐Golgi network, the cell surface, or other specialized organelles (Burd & Cullen, 2014; Carosi et al., 2023; Cullen & Steinberg, 2018). The impairment of retromer function has indeed been associated with numerous pathologies, particularly neurodegenerative disorders, including Alzheimer's disease (AD) (Muhammad et al., 2008; Small et al., 2005), Parkinson's disease (PD) (Zimprich et al., 2011), Down's syndrome (DS) (Curtis et al., 2020), and amyotrophic lateral sclerosis (ALS) (Muzio et al., 2020; Perez‐Torres et al., 2022). Not surprisingly, there is a growing interest in exploring retromer as a target for therapeutic interventions (Chen et al., 2021; Mecozzi et al., 2014; Muzio et al., 2020). The mammalian retromer consists of an evolutionary conserved trimer of vacuolar protein sorting (VPS) 35, VPS29, and VPS26, which orchestrates cargo selection, membrane deformation, and cytoskeletal association through interactions with accessory proteins. The function of the retromer is linked with its recruitment to the endosomal membrane, mediated by a subset of sorting nexin (SNX) proteins (reviewed in Tilley et al., 2018; Yong et al., 2022), along with Rab5 and Rab7 GTPases (Langer et al., 1991). Once recruited to endosomes, retromer facilitates selective cargo sorting into distinct transport carriers by forming actin‐decorated microdomains (Puthenveedu et al., 2010; Steinberg et al., 2013).

Actin polymerization contributes to many membrane remodelling processes, including endocytosis, endosomal recycling, autophagy and vesicle trafficking. An important component in vesicle dynamics is the Arp2/3 complex, which controls actin nucleation and branching of filaments for the generation of membrane subdomains, as well as mechanical forces during vesicle formation and scission. Proper functioning of the Arp2/3 complex requires members of the Wiskott‐Aldrich Syndrome Protein (WASP) family, termed Nucleation‐Promoting Factors (NPFs), for the spatial and temporal organization of actin filaments. The WASP and SCAR Homolog WASH is the major NPF at the surface of endosomes and is an integral part of a pentameric complex which includes WASH (WASHC1), the family with sequence similarity 21 (FAM21 or WASHC2), the coiled coil domain containing protein 53 (CCDC53 or WASHC3), Strumpellin (WASHC5), and the Strumpellin and WASH interacting protein (SWIP or WASHC4) (Jia et al., 2010). WASH complex exhibits significant homology to another WASP family member, the WAVE regulatory complex (WRC) found at the plasma membrane (Jia et al., 2010). Nonetheless, the overall resemblance between WRC and the WASH complex is only applicable to four of their subunits (Jia et al., 2010). The major difference corresponds to FAM21 which is a much larger protein (∼1300 amino acids) compared to its counterpart subunit (ABI, ∼370 amino acids) in the WRC complex. FAM21 is formed by a small (∼200 amino acids) “head” domain required for the interaction with other WASH complex subunits and a long (∼1100 amino acids) “tail,” mostly unstructured, that mediates direct binding to several proteins including the acting‐capping proteins CAPZa and CAPZb (Hernandez‐Valladares et al., 2010), the DNAJC13 protein, known as receptor‐mediated endocytosis‐8 (RME‐8) (Freeman et al., 2014), the CCDC22 and CCDC93 subunits of the CCC (COMMD/CCDC22/CCDC93) complex (Harbour et al., 2012; Phillips‐Krawczak et al., 2015), the FK506‐binding protein 15 (FKBP15) (Harbour et al., 2012; Nooh & Bahouth, 2017), the cargo adaptor SNX27 (Lee et al., 2016; Steinberg et al., 2013), and the VPS35 subunit of the retromer complex (Harbour et al., 2012; Helfer et al., 2013; Jia et al., 2012) (Figure 1a). This network of interactions suggests different mechanisms for controlling actin polymerization on endosomes (Simonetti & Cullen, 2019). Yet, the localization of the WASH complex to endosomes largely relies on the interaction between FAM21 and retromer (Harbour et al., 2010; Harbour et al., 2012; Helfer et al., 2013). The large unstructured C‐terminal tail of FAM21 contains 21 repeats of the consensus L‐F‐[D/E/S]3–10‐L‐F sequence, leucine‐phenylalanine‐acidic (LFa) motif, that bind via multivalent interactions to VPS35 (Jia et al., 2012). Nonetheless, despite the general similarity among the 21 repeats, only those at the very C‐terminal end of the tail significantly contribute to WASH complex recruitment (Helfer et al., 2013; Jia et al., 2012). In addition, the interaction between FAM21 and VPS35 requires the presence of VPS29, suggesting a more complex pattern of interactions than initially anticipated (Helfer et al., 2013; Seaman et al., 2018). Interestingly, genetic studies of sporadic and familial forms of Parkinson's disease (PD) identified a point mutation in VPS35 (D620N) as the causative factor in the development of the disease (Vilarino‐Guell et al., 2011; Zimprich et al., 2011). The D620N mutation does not affect the stability or assembly of VPS35 with other retromer subunits, but impairs its association with FAM21 and the recruitment of the WASH complex to endosomes (Cui et al., 2021; McGough, Steinberg, Jia, et al., 2014; Zavodszky et al., 2014). The ankyrin‐repeat‐domain‐containing protein 50 (ANKRD50) is another retromer‐SNX27‐SHRC interacting protein that fails to associate with VPS35 harboring the D620N mutation (Kvainickas et al., 2017). These alterations lead to altered endosomal morphology (Cui et al., 2021) and impaired sorting of some retromer cargos such as the cation‐independent mannose 6‐phosphate receptor (CI‐MPR), the post‐synaptic AMPA receptor GluR1, and the Autophagy‐related protein 9A (ATG9a), but not retromer‐SNX27 cargo like the Glucose transporter 1 (GLUT1) (Cui et al., 2021; Follett et al., 2014; Follett et al., 2016; McGough, Steinberg, Jia, et al., 2014; Munsie et al., 2015; Zavodszky et al., 2014). Intriguingly, while VPS35 interacts robustly with endogenous FAM21 in cell culture models, the equivalent interaction has not been observed using similar co‐immunoprecipitation (co‐IP) assays from mouse brain extracts suggesting that it may form a distinct tissue‐specific network of interactions or a labile complex (Chen et al., 2019). While these findings suggest a delicate interplay between retromer and the WASH complex to control actin dynamics, most molecular details regarding how retromer interacts with FAM21 remain unknown.

FIGURE 1.

FIGURE 1

Open in a new tab

The long, disordered tail of FAM21 interacts with retromer through three regions. (a) Schematic cartoon depicting the association between various endosomal proteins and the FAM21 tail. Vertical black bars within the extended tail represent each of the 21 LFa repeats. (b) Representative ITC measurements illustrate the interaction of distinct segments of the FAM21 tail titrated into retromer. The top three panels correspond to the titration curves for the indicated mutants within the R20 or R21 LFa motifs. The lower panel illustrates the binding isotherms for each fragment of FAM21. Raw titration curves for the binding isotherms and the thermodynamic parameters are summarized in Figure S1 and Table S1, respectively. (c) Schematic representation of the FAM21 constructs used, with green areas indicating the binding regions toward retromer.

In the present study, we provide a detailed biochemical and structural characterization of the interactions between retromer and FAM21, required for the endosomal association of the WASH complex. We demonstrate that FAM21 harbors two distinct binding sites toward VPS35 and a third one toward VPS29 within a conserved pocket shared by other cellular ligands. The network of FAM21 interactions occurs in proximity to the apex of the arch formed by retromer dimers. Interestingly, the Parkinson's VPS35(D620N) mutation, situated near the VPS35:VPS35 interface, does not significantly reduce the association with FAM21 in vitro.

2. RESULTS

2.1. Retromer complex interacts with distinct segments of the FAM21 tail

The 21 repeats (R1–R21) of the LFa motif are distributed along the long unstructured tail of FAM21 (Jia et al., 2012). However, the analysis of distinct segments of the tail have shown strong variations with respect to their contributions to retromer binding (Helfer et al., 2013; Jia et al., 2012). These findings suggest a more complex recognition mechanism than initially anticipated, possibly directed by multiple interactions and binding modes. To define more precisely the interactions between FAM21 and retromer we first confirmed several of the previously described interactions by isothermal titration calorimetry (ITC), and then expanded the analysis to new engineered fragments encompassing distinct LFa motifs. We generated five main fragments harboring multiple LFa motifs. The largest fragment corresponded to the entire tail region with the 21 repeats of the LFa motif (denoted as R1–21),whereas shorter fragments included R1–19, R20–21, R1–4 and R5‐19 (Figure 1b,c). The ITC analysis confirmed direct interactions between two distant regions in the amino acid sequence of FAM21, R20–21 and R1–4, with retromer (Figure 1b,c, and Figure S1A). In contrast, the large middle region R5‐19 did not yield any measurable interaction in solution. Out of the two regions that interact with retromer, the R20–21 exhibited the strongest contribution to the binding in agreement with previous studies (Helfer et al., 2013; Jia et al., 2012). To further delineate these interactions, we narrowed the search using synthetic and recombinantly expressed peptides covering R1–4 and R20–21 regions (Figure 1b,c). Interestingly, while R20 and R21 fragments exhibited significant binding to retromer, neither of the peptides encompassing the R1–4 region, nor the fragments R1–2, R2–4 or R3–4 showed any interaction (Figures 1b,c and S1). Indeed, the interactions of R20 and R21 with retromer were abolished by single or double alanine substitutions within their respective LFa motifs, which evidenced for a direct physical interaction by short sequence patches (Figure 1b,c). These observations hinted at distinct binding modes between the two stretches. In this sense, whilst R20 and R21 functioned as independent linear motifs, the R1–4 region seemed to rely on conformational constraints for binding.

2.2. FAM21 interacts with retromer through VPS35 and VPS29 subunits

We next aimed at identifying the retromer sites to which R1–4, R20 and R21 bind. First, considering that retromer is an elongated complex where the VPS26 and VPS29 subunits associate with the N‐terminal and C‐terminal ends of the VPS35 subunit, respectively, we divided VPS35 into roughly two halves: VPS35N (aa 14–470) and VPS35C (aa 476–780). This division allowed us to produce two subcomplexes, namely the VPS26–VPS35N subcomplex and the VPS29–VPS35C subcomplex. We then evaluated their affinities toward R1–4, R20, and R21. We found that R1–4 and R21 bound to VPS29–VPS35C subcomplex with similar affinity as to the full retromer, but unexpectedly, R20 did not interact with any of the subcomplexes (Figure 2a‐d and S2). We reasoned that R20 could bind the middle section of VPS35 that was not included in the constructs. To test this possibility, we designed an extended VPS35C construct (aa 204–780) and, as expected, the R20 bound to the VPS29–VPS35204–780 subcomplex with equivalent affinity as to the whole retromer. Next, we analyzed the interaction of R1–4 and R21 with each component of the VPS29–VPS35C subcomplex. Remarkably, R1–4 exhibited binding to VPS35C, whereas the R21 associated with VPS29, indicating distinct binding subunits for each region.

FIGURE 2.

FIGURE 2

Open in a new tab

Mapping FAM21‐interacting fragments on retromer. (a) Binding isotherms for the R1‐4 fragment, (b) R20 fragment, and (c) R21 fragment to individual retromer subunits or truncated subcomplexes. Raw titration curves and thermodynamic parameters are summarized in Figure S2  and Table S1, respectively. (d) Schematic representation of the retromer subunits and truncated subcomplexes used for mapping the interactions.

2.3. Identification of R1–4 , R20 , and R21 partner‐binding surfaces on VPS35 and VPS29

Our previous experiments narrowed the interaction of the FAM21 tail down to two regions in VPS35 and one site on VPS29. Next, to identify amino acid residues important for the interaction, we selected several double and triple mutants on the basis of their solvent accessibility, conservation and spatial proximity in the three‐dimensional structure. For mapping the R1–4‐binding site in VPS35 we selected the mutants KL659/661AE, EP643/648AA, and KKK544/548/552AAA (Figure 3a and video S1). The only mutant that significantly altered the affinity for R1–4 was KL659/661AE (Figure 3a,b and S3). Similarly, for mapping the R20‐binding site in VPS35 we evaluated the substitutions YI447/451AA, LK508/515AA, and PPV472/475/476AAA. In this case, only the substitution YI447/451AA precluded the interaction with R20 (Figure 3a,c). Finally, to map the region within VPS29 responsible for binding to R21, we focused on an evolutionary conserved pocket (Figure S4) located on the opposite side from the contact with VPS35. This pocket has been shown to be important for binding to the Rab7 GTPase activating protein TBC1 (Tre‐2/USP6, BUB2, Cdc16) Domain Family, Member 5 (TBC1D5) (Jia et al., 2016), the VPS9‐domain ankyrin repeated protein (VARP) (Crawley‐Snowdon et al., 2020), the Vacuolar Protein Sorting 35‐Like (VPS35L) (Healy et al., 2023), and the Legionella pneumophila effector protein RidL (Barlocher et al., 2017; Romano‐Moreno et al., 2017; Yao et al., 2018). Thus, we introduced mutations to various residues within the conserved pocket of VPS29, including L152E, Y163A, and Y165A, known to play a crucial role in interactions with TBC1D5, VARP, VPS35L, and RidL. Notably, all three mutations proved indispensable for the interaction with R21 (Figure 3a,d), highlighting a shared binding pocket in VPS29 among multiple retrieval complexes.

FIGURE 3.

FIGURE 3

Open in a new tab

Mapping of FAM21's binding sites on retromer using a structurally guided site‐directed mutagenesis approach. (a) Semi‐transparent surface overlaying a cartoon representation of full‐length retromer (PDB 6VAC; Kendall et al., 2020). Mutated residues are shown in a ball‐and‐stick representation, color‐coded in red for mutations interfering with binding and green for mutations not affecting binding toward the respective FAM21 fragments. (b–d) ITC analysis showing the binding of (b) R1–4 fragment, (c) R20 fragment, and (d) R21 fragment to both wildtype retromer subunits and selected mutants. Raw thermograms for negative interactions are represented on top of the integrated heats and data fit, while raw thermograms for positive interactions and thermodynamic parameters are shown in Figure S3 and Table S1, respectively.

Previous studies have suggested that the VPS35 (D620N) mutation linked to Parkinson's disease impairs the recruitment of the pentameric WASH complex to endosomes (McGough, Steinberg, Jia, et al., 2014; Zavodszky et al., 2014). This impairment was attributed to a reduced interaction between VPS29–VPS35C and the R20–21 motifs of FAM21 (McGough, Steinberg, Jia, et al., 2014). Given the close proximity between D620 and the binding area for R1–4 (Figure 3a), we investigated whether D620N could influence the association with R1–4 using ITC. The results indicated that the retromer complex bearing the VPS35 (D620N) mutation exhibited no significant differences compared to the wild type in binding to either R1–4 or R20 (Figure 3b,c). We did not assess the binding toward R21, as it binds to VPS29, which is far from the D620 site (Figure 3a).

2.4. Structure of VPS29 bound to the R21 motif of FAM21

Previous crystallographic studies have solved the structure of retromer from two subcomplexes, the C‐terminal region of VPS35 (VPS35C, amino acids 476–780) bound to VPS29 (Hierro et al., 2007), and the N‐terminal region of VPS35 (VPS35N, amino acids 14–470) bound to VPS26 and SNX3 (Lucas et al., 2016). Taking advantage of the crystallization ability of these constructs we attempted to co‐crystallize R1–4, R20 and R21 with their corresponding retromer subcomplex. Unfortunately, despite extensive efforts we were unable to obtain crystals of R1–4 and R20 in complex with retromer. On the contrary, we succeeded on the crystallization of R21 with VPS35C‐VPS29. We determined the crystal structure by molecular replacement using the coordinates of VPS35C‐VPS29 (PDB 2R17) (Hierro et al., 2007) as the search model, and refined to 3.01 Å resolution with Rfactor and Rfree of 25.08% and 28.77%, respectively (Table 1). The difference Fourier map clearly showed a portion of R21 comprising amino acids 1331FDDPLNA1337 (Figure 4a). The remaining residues were not modeled due to missing electron density which most probably reflected intrinsic disorder. Overall, the central region of R21 forms a sharp turn inserted into the conserved pocket of VPS29 (Figure 4b and S4). In agreement with the biochemical data described before, R21 anchors to VPS29 by the burial of P1334 and L1335 into a hydrophobic cavity formed by L2, L25, L26, F150, L152 and Y163. More peripherally F1331 of R21 stacks against the R176 aliphatic side‐chain of VPS29 (Figure 4b,c). In addition, the interface includes a network of sidechain and backbone hydrogen bonds that further stabilize the association between R21 and VPS29 (Figure 4c). In this regard, the crystal structure provides key structural determinants for the interaction of R21 with VPS29.

TABLE 1.

Summary of x‐ray crystal data collection, phasing, and refinement statistics for the retromer complex bound to the R21 motif of FAM21.

FAM21
Data collection
Wavelength [Å] 0.97623
Space group P21
Resolution [Å] 146.15–3.1 (3.315–3.1) a
Cell dimensions
a, b, c [Å] 59.27139.10146.15
α, β, γ [°] 90.0 90.0 90.0
CC1/2 (%) 99.7 (49.8)
Completeness (spherical) (%) 65.4 (11.7)
Completeness (ellipsoidal) (%) 91.1 (66.9)
I/σ 7.4 (1.3)
Number of unique reflexions 31,911 (1584)
Redundancy 7.0 (7.0)
Refinement
R‐factor (%) 25.08
R‐free (%) 28.77
R.m.s deviations
Bond lengths (Å) 0.0022
Bond angles (°) 0.523
PDB CODE 8RKS
Open in a new tab
a

Highest resolution shell is shown in parenthesis.

FIGURE 4.

FIGURE 4

Open in a new tab

Crystal structure of R21 bound to retromer. (a) The overall structure of the complex formed by the R21 motif and VPS29‐VPS35C in cartoon representation with transparent surfaces. Left insets show the omit difference electron density map (Fo‐Fc) in green, contoured at 2.0σ, and the final refined 2Fo‐Fc electron density map in blue, contoured at 1.5σ. (b) Detailed view of the R21‐VPS29 interaction with relevant residues shown as sticks. (c) Interaction diagram generated with LigPlot+ (Laskowski & Swindells, 2011), illustrating hydrophobic contacts, hydrogen bonds, and solvent accessibility of the R21 (aa 1331‐FDDPLNA‐1337) binding motif.

2.5. Structural comparison of R21 with other VPS29‐effector complexes

Previous structural studies have uncovered the binding mode of other VPS29 effectors such as TBC1D5 (Jia et al., 2016), VARP (Crawley‐Snowdon et al., 2020), VPS35L (Healy et al., 2023), and RidL (Barlocher et al., 2017; Romano‐Moreno et al., 2017; Yao et al., 2018). Superposition of the crystal structure of R21 with these effectors revealed a similar pattern for binding where a highly conserved proline, denoted as P0 position, represents the linchpin of the interaction (Figure 5a,b). The constrained geometry of this proline favors a sharper bend of the loop and introduces rigidity to the neighboring residues. In particular at position P1, there is a conserved leucine or isoleucine that, together with the proline at P0, form a conformationally restricted hydrophobic tip, optimizing the fit between R21 and the VPS29 pocket. Interestingly, the P‐L amino acid tandem in R21 is unique among the 21 LFa repeats in FAM21 and agrees with the above results on its exclusive interaction with VPS29.

FIGURE 5.

FIGURE 5

Open in a new tab

Structural comparison of the R21‐VPS29 complex and other ligand complexes. (a) Superimposition of available structures from VPS29 ligands onto the VPS29 pocket. (b) WebLogo analysis of the superimposed structures. Note that despite RidL aligns in reverse direction, the tight P‐L/I turn is conserved. (c) Comparative description of per‐residue contributions from electrostatics (blue), van der Waals (vdW) (gray), and desolvation (orange) for residues inserted into the VPS29 pocket, extending from P‐2 to P2.

Given that VPS29, either alone or in complex with VPS35, has shown distinct affinities toward TBC1D5 (Kd ~ 0.22–0.45 μM) (Jia et al., 2016; Romano‐Moreno et al., 2017), VARP (Kd ~ 2.7–5 μM) (Crawley‐Snowdon et al., 2020), VPS35L (Kd ~ 1.87 μM) (Healy et al., 2023), RidL (Kd ~ 0.15–0.5 μM) (Barlocher et al., 2017; Romano‐Moreno et al., 2017; Yao et al., 2018), and the R21 motif of FAM21 (Kd ~ 20 μM, this work), we examined the energy contribution at the residue level for each of the loops that contact the conserved hydrophobic pocket in VPS29 extending from amino acids P−2 to P2 (Figure 5c). With the exception of VARP, the remaining VPS29 effectors exhibited a similar pattern of energy contribution at each position, with P−1 to P1 identified as the hot‐spot residues. Discrepancies with VARP may have arisen due to the methodological source of structures used for the analysis. While the energetic values for VARP originated from averaged NMR conformers that genuinely reflect the solution state, the remaining structures were derived from single crystallographic conformers, typically considered to adopt minimal energy landscapes within the lattice. In summary, this analysis indicates that the anchoring strength at the tip of each of the loops inserted into the hydrophobic pocket of VPS29 is very similar. The differences in affinity between distinct VPS29 effectors may, therefore, arise from the expansion of the interface area around the hydrophobic pocket or from the inclusion of additional intermolecular binding sites. As such, TBC1D5 contacts both VPS29 and the VPS35 subunits of retromer, whereas the VPS35L subunit of retriever clamps VPS29 through the N‐ and C‐terminal regions (Healy et al., 2023). On the other hand, RidL does not have additional intermolecular interactions with retromer but exhibits an extended buried surface next to the hairpin loop that further stabilizes the association with VPS29 (Barlocher et al., 2017; Romano‐Moreno et al., 2017; Yao et al., 2018). VARP, on the contrary, holds two Zn‐fingernails with conformationally restrained scaffolds for binding VPS29 individually (Crawley‐Snowdon et al., 2020; Hesketh et al., 2014; McGough, Steinberg, Gallon, et al., 2014). In the case of FAM21, it employs the R21 motif exclusively for binding to VPS29 with low affinity through a relatively small buried interaction, while R20 and R1–4, on the other hand, associate with distinct sites on VPS35.

3. DISCUSSION

Since the identification of the 21 LFa repeats within the C‐terminal unstructured tail of FAM21 (Jia et al., 2012), their multivalency and binding attributes have remained enigmatic. In the current study, we expanded the analysis of FAM21 and retromer interactions, identifying three distinct regions with diverse binding modes. While R20 and R21 bind to VPS35 and VPS29 subunits of retromer, respectively, as short linear motifs, the R1‐4 segment, which contains four LFa repeats, requires the entire stretch of amino acids for binding to VPS35, indicating the presence of certain spatial constraints within this segment. In line with earlier findings (Jia et al., 2012), we observed that R20 and R21 contribute most significantly to the binding with retromer. Indeed, our data suggests that the R20–R21 segment contributes roughly half of the total affinity, while the R1–4 segment contributes only about one‐fifth of the total affinity toward retromer, assuming independent binding events. On the other hand, despite previous indications of R19 playing a role in retromer binding (Jia et al., 2012), we couldn't detect a measurable association when using the R5–19 fragment. In this regard, it is possible that R19 functions as a relatively weak binder in R5–19, while in the context of R19–21, it assumes a more prominent interaction with retromer, possibly due to inter‐motif distances at which favorable binding occurs.

Previous studies indicated that FAM21 binds to the VPS35 subunit of the retromer complex (Helfer et al., 2013; Jia et al., 2012), and that this binding relies on the VPS35‐VPS29 association (Helfer et al., 2013). Here, we identified a direct interaction between FAM21 and VPS29 mediated by the R21 motif. This motif includes a distinctive P‐L amino acid tandem, which is unique among the 21 LFa repeats. The crystal structure showed that the P‐L motif forms the tip of a tight turn that inserts into a conserved hydrophobic pocket on the surface of VPS29. This pocket is located opposite to the VPS29‐VPS35C interface and is an important site for interaction with other endosomal effectors such as TBC1D5 (Jia et al., 2016), VARP(Crawley‐Snowdon et al., 2020), VPS35L (Healy et al., 2023), and RidL (Barlocher et al., 2017; Romano‐Moreno et al., 2017; Yao et al., 2018). Indeed, siRNA knockdown of TBC1D5 in HeLa cells leads to an increased interaction between retromer and the WASH complex (Seaman et al., 2018). Interestingly, most of these effectors share a similar binding mode and per‐residue energy contribution within the hydrophobic pocket, suggesting that the variations in affinity may arise from either the enlargement of the interface area surrounding the hydrophobic pocket and/or the incorporation of additional intermolecular binding sites. Retromer forms dimeric arch‐like structures (Figure S5a,b) where the VPS26 subunit, in combination with distinct SNXs and cargo proteins, localizes at the base of the arch near the membrane surface (Kovtun et al., 2018; Leneva et al., 2021; Lucas et al., 2016). Meanwhile, VPS29 localizes away from the crowded membrane surface at the apex of the arch, reinforcing the idea of a central scaffold for the assembly of accessory proteins to establish distinct sorting stations on endosomes (Banos‐Mateos et al., 2019). In this context, the binding of R21 to VPS29, along with the interaction of both R1‐4 and R20 with the C‐terminal half of VPS35, occurs away from the membrane surface (Figure S5c), potentially facilitating the recruitment of the WASH complex to endosomes. Interestingly, the D620N mutation in VPS35, linked to PD, is localized at the apex where VPS35 homodimerizes. This mutation did not significantly reduce the direct binding toward R1–4 or R20 in vitro, suggesting that the perturbed association with FAM21 and the WASH complex (Cui et al., 2021; McGough, Steinberg, Jia, et al., 2014; Zavodszky et al., 2014) might arise from indirect effects or changes in the dynamic behavior of the retromer complex, such as the transition from a pseudoplanar to a tubular organization.

While this manuscript was in preparation, a preprint with similar observations was shared with the community (Guo et al., https://doi.org/10.1101/2023.08.15.553351). Both studies confirm the presence of discrete binding regions within the FAM21 tail toward VPS35 and VPS29 subunits of retromer and provide the crystal structure of VPS29 bound to the R21 motif. While Guo et al. provide additional data showing that R19 and R20 motifs can bind at distinct sites on VPS35 and to the FERM domain of SNX27, our study identifies the R1–R4 segment as an additional engagement to VPS35. In this regard, both studies present a comprehensive and complementary overview of FAM21 binding modes toward retromer and SNX27.

Our observations define discrete interactions between the unstructured tail of FAM21 and retromer. Importantly, the results align with the notion that the LFa motifs in FAM21, rather than promoting local clustering of retromer through tandem repetitions, might instead enable specific interactions with a number of factors using minor sequence modifications (Deatherage et al., 2020). We speculate that the consensus L‐F‐[D/E/S]3‐10‐L‐F motif could display specificity through defined sequences and spatial distribution. As such, the spacing between the two L‐F tandem repeats within each LFa motif, using variable stretches of negatively charged residues, could fine‐tune their binding specificities. Meanwhile, the global positioning of the 21 LFa repeats could serve as an “adaptor” system where inter‐motif distances might favor networks of interactions for the establishment of sorting subdomains. Future work should provide valuable information about the individual LFa binders and how such interactions evolve over time.

4. MATERIALS AND METHODS

4.1. Recombinant DNA procedures

DNA sequences encoding different fragments of FAM21 (R1–19, R1–21, R1–4, R20–21, R20, R21, R5–19, R1–2, R2–4, R3–4) and retromer subunit VPS35C204–780 were cloned by Gibson assembly into pGST‐Parallel2 vector (Sheffield et al., 1999) with a cleavable N‐terminal glutathione S‐transferase (GST) tag (Hierro et al., 2007). Site‐directed mutations in FAM21 R20, FAM21 R21 and VPS35 coding sequences were introduced using the Phusion Site‐Directed Mutagenesis Kit (Thermo Fisher) according to the manufacturer's directions. All constructs were verified by DNA sequencing. For the expression of retromer complex (VPS29‐VPS35‐VPS26), the following plasmids were used: pMR101A‐VPS29 (Hierro et al., 2007), pGST‐Parallel2‐VPS35 (Hierro et al., 2007) and pET28‐Sumo3‐VPS26 (Lucas et al., 2016). In order to express the amino or carboxy terminal fragments of VPS35, pGST‐Parallel2‐VP35N (Lucas et al., 2016) and pGST‐Parallel2‐VPS35C (Hierro et al., 2007) vectors were used. For the expression of VPS29 mutants on Y165, Y163 and L152, the following vectors were used: pGST‐Parallel2‐VPS29, pGST‐Parallel2‐VPS29‐Y165A, pGST‐Parallel2‐VPS29‐Y163A and pGST‐Parallel2‐VPS29‐L152E (Romano‐Moreno et al., 2017).

4.2. Protein expression and purification

Proteins were expressed in E. coli BL21 (DE3) cells and grown in Luria–Bertani (LB) broth supplemented with the appropriate antibiotic at 37°C. Protein expression was induced at an OD600 of 0.8 by adding 1 mM isopropyl‐β‐D‐thiogalactopyranoside (IPTG). Cells were harvested after 16 h of growth at 18°C. Purification steps were carried out at 4°C, and the final protein concentration was determined using the theoretical extinction coefficient.

For the purification of VPS35 and FAM21 constructs, the cellular pellet was lysed using high‐pressure homogenization (25 kpsi) in 50 mM Tris–HCl pH 7.5, 300 mM NaCl, 1 mM dithiothreitol (DTT) (buffer A), supplemented with 0.1 mM phenylmethylsulfonyl fluoride (PMSF) and 1 mM benzamidine. After clearing the bacterial lysates by centrifugation for 45 min at 50,000g, the soluble fraction was incubated in batch for 2 h with glutathione Sepharose 4B (GE Healthcare), previously equilibrated in 50 mM Tris–HCl pH 7.5, 150 mM NaCl, 1 mM DTT (buffer B). Protein was released from the resin by overnight cleavage of the N‐terminal GST tag with TEV protease in buffer B. Subsequently, ion‐exchange chromatography (HiTrap Q HP) was performed using a linear gradient of 20–1000 mM NaCl, followed by size‐exclusion chromatography on a Superdex 75 16/60 or Superdex 200 16/60 column based on the molecular weight of the protein, in buffer B. For the purification of VPS29 constructs, we followed the procedure described in Romano‐Moreno et al. (2017).

To purify the different protein complexes in all cases, cell pellets were mixed and purified together following the procedures described in Lucas et al. (2016), for the retromer complex (VPS35‐VPS29‐VPS26) and the amino‐terminal complex part (VPS35N‐VPS26). For the purification of the carboxyl‐terminal part (VPS35C‐VPS29) the procedure described in Hierro et al. (2007), was used. In the case of the extended carboxyl complex purification (VPS35204–780‐VPS29), the procedure described in Hierro et al. (2007), was also employed.

The peptide FAM21‐R21 (1328SNIFDDPLNAFGGQ1341) was synthesized by GenScript with a purity percentage greater than 98%.

4.3. Protein crystallization, data collection and structure determination

The VPS35476‐780‐VPS29‐FAM21 R21 complex was crystallized using the hanging drop vapor diffusion method. Crystals were obtained by mixing the VPS35476‐780‐VPS29 complex at 143 μM with an excess of the FAM21 R21 peptide (1328SNIFDDPLNAFGGQ1341) at 1.5 mM. Native crystals appeared after 3 days at 18°C by mixing 1ul of protein sample with 1ul of the precipitant solution containing 20% (w/v) PEG 3350, 0.1 M NaCl and 0.1 M Tris–HCl pH 8.5. Crystals were cryoprotected by quick‐soaking into mother liquor supplemented with 25% (v/v) ethylene glycol before being flash‐frozen in liquid nitrogen.

Crystallographic native datasets were collected with the software MxCuBE at XALOC beamline in the ALBA synchrotron facility (Cerdanyola del Valles, Spain) using a Pilatus 6 M detector, and at Diamond Light Source (Oxfordshire, UK) with the software GDA. Diffraction images were indexed, integrated and scaled using XDS software (Kabsch, 2010). The VPS35476–780‐VPS29‐FAM21 R21 structure was solved by molecular replacement using as a template the complex VPS29‐VPS35C (PDB 2R17) in PHASER (McCoy et al., 2007).

The asymmetric unit contained four copies of VPS35476–780‐VPS29, with one FAM21 R21 molecule bound to each VPS29. The Peptide sequence was manually built in COOT (Emsley et al., 2010) through iterative refinement with REFMAC5 (Murshudov et al., 2011). Data collection statistics for each dataset are shown in Table 1.

4.4. Isothermal titration calorimetry

ITC measurements were carried out at 25°C on a MicroCal PEAQ‐ITC titration microcalorimeter (Malvern Panalytical). All proteins and peptides used in this work were dialyzed overnight at 4°C against 300 mM NaCl, 0.5 mM TCEP, 25 mM HEPES pH 7.5 buffer. Before titration, samples were tempered at 25°C and degassed for 5 min in a Thermo Vac. The titration sequence consisted of an initial 2 μL injection to prevent artifacts (not used in data fitting), followed by 18 injections of 2 s and 2 μL, with a spacing of 150 s between them. Heat of dilution used to correct the experimental data was performed under the same conditions. Results were fitted and integrated to a one‐site model using the MicroCal PEAQ‐ITC software (Malvern Panalytical). Final graphs were prepared using Origin ITC software (MicroCal). Values for the binding constant (Ka, Kd = 1/Ka), the molar binding stoichiometry, binding enthalpy, free energy and entropy of binding were obtained after data analysis. For ITC analysis of the interaction between different FAM21 motifs and retromer, 250 μM of FAM21 R1–19, R1–21, R1–4, R20–21, R20, R21, R5–19, R1–2, R2–4, R3–4 and the corresponding mutants, were titrated into 10 μM of retromer complex. For studying the region of retromer involved in the interaction with FAM21 regions R1–4, R20 and R21, 250 μM of each region was titrated into 10 μM of retromer, VPS35, VPS26‐VPS35N, VPS29‐VPS35C, VPS29‐VPS35204‐780, or VPS35C, VPS29 and its different point mutants. The data are representative of a minimum of two replicate titrations for each assay.

4.5. Computation of per‐residue docking energy

We estimated the residue contribution to the binding energy using the pyDockEneRes Server (https://life.bsc.es/pid/pydockeneres) with default parameters. The input PDBs for VPS29‐TBC1D5, VPS29‐VPS35L, and VPS29‐RidL crystal structures were 5GTU, 8ESE, and 5OSH, respectively. For the VPS29‐VARP complex (6TL0), the energy contribution was calculated based on the average of 10 NMR ensembles.

AUTHOR CONTRIBUTIONS

Miguel Romano‐Moreno: Formal analysis; investigation; writing – original draft; validation. Elsa N. Astorga‐Simón: Investigation; formal analysis; writing – original draft; validation. Adriana L. Rojas: Investigation; validation; formal analysis. Aitor Hierro: Conceptualization; writing – original draft; project administration; funding acquisition; supervision.

FUNDING INFORMATION

This work was supported by The Ministry of Science and Innovatión Grant PID2020‐119132GB‐I00 (to A.H.).

CONFLICT OF INTEREST STATEMENT

The authors declare no competing interests.

Supporting information

Data S1. Supporting Information.

PRO-33-e4980-s002.docx (1.5MB, docx)

Video S1. Cartoon representation of the full‐length retromer (PDB 6VAC (57)) is depicted, highlighting mutated residues in stick format. Red‐colored residues indicate mutations that interfere with binding, while green‐colored residues denote mutations that do not affect binding toward the respective FAM21 fragments as illustrated in Figure 3.

Download video file (29.1MB, mp4)

ACKNOWLEDGMENTS

This study made use of the Diamond Light Source (Oxfordshire, UK) proposal MX20113, and ALBA synchrotron beamline BL13‐XALOC. We express our gratitude to all beamline staff for their valuable assistance.

Romano‐Moreno M, Astorga‐Simón EN, Rojas AL, Hierro A. Retromer‐mediated recruitment of the WASH complex involves discrete interactions between VPS35, VPS29, and FAM21 . Protein Science. 2024;33(5):e4980. 10.1002/pro.4980

Reviewing Editor: John Kuriyan

DATA AVAILABILITY STATEMENT

Atomic coordinates and structure factors of the crystallographic complexes are available in the Protein Data Bank (PDB) with accession code 8RKS listed in Table 1.

REFERENCES

  1. Banos‐Mateos S, Rojas AL, Hierro A. VPS29, a tweak tool of endosomal recycling. Curr Opin Cell Biol. 2019;59:81–87. [DOI] [PubMed] [Google Scholar]
  2. Barlocher K, Hutter CAJ, Swart AL, Steiner B, Welin A, Hohl M, et al. Structural insights into legionella RidL‐Vps29 retromer subunit interaction reveal displacement of the regulator TBC1D5. Nat Commun. 2017;8:1543. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Burd C, Cullen PJ. Retromer: a master conductor of endosome sorting. 2014. Cold Spring Harb Perspect Biol 6. [DOI] [PMC free article] [PubMed]
  4. Carosi JM, Denton D, Kumar S, Sargeant TJ. Receptor recycling by retromer. Mol Cell Biol. 2023;43:317–334. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chen KE, Guo Q, Hill TA, Cui Y, Kendall AK, Yang Z, et al. De novo macrocyclic peptides for inhibiting, stabilizing, and probing the function of the retromer endosomal trafficking complex. Sci Adv. 2021;7:eabg4007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Chen X, Kordich JK, Williams ET, Levine N, Cole‐Strauss A, Marshall L, et al. Parkinson's disease‐linked D620N VPS35 knockin mice manifest tau neuropathology and dopaminergic neurodegeneration. Proc Natl Acad Sci U S A. 2019;116:5765–5774. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Crawley‐Snowdon H, Yang JC, Zaccai NR, Davis LJ, Wartosch L, Herman EK, et al. Mechanism and evolution of the Zn‐fingernail required for interaction of VARP with VPS29. Nat Commun. 2020;11:5031. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cui Y, Yang Z, Flores‐Rodriguez N, Follett J, Ariotti N, Wall AA, et al. Formation of retromer transport carriers is disrupted by the Parkinson disease‐linked Vps35 D620N variant. Traffic. 2021;22:123–136. [DOI] [PubMed] [Google Scholar]
  9. Cullen PJ, Steinberg F. To degrade or not to degrade: mechanisms and significance of endocytic recycling. Nat Rev Mol Cell Biol. 2018;19:679–696. [DOI] [PubMed] [Google Scholar]
  10. Curtis ME, Yu D, Pratico D. Dysregulation of the retromer complex system in down syndrome. Ann Neurol. 2020;88:137–147. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Deatherage CL, Nikolaus J, Karatekin E, Burd CG. Retromer forms low order oligomers on supported lipid bilayers. J Biol Chem. 2020;295:12305–12316. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Emsley P, Lohkamp B, Scott WG, Cowtan K. Features and development of coot. Acta Crystallogr D Biol Crystallogr. 2010;66:486–501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Follett J, Bugarcic A, Yang Z, Ariotti N, Norwood SJ, Collins BM, et al. Parkinson disease‐linked Vps35 R524W mutation impairs the endosomal association of retromer and induces alpha‐synuclein aggregation. J Biol Chem. 2016;291:18283–18298. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Follett J, Norwood SJ, Hamilton NA, Mohan M, Kovtun O, Tay S, et al. The Vps35 D620N mutation linked to Parkinson's disease disrupts the cargo sorting function of retromer. Traffic. 2014;15:230–244. [DOI] [PubMed] [Google Scholar]
  15. Freeman CL, Hesketh G, Seaman MN. RME‐8 coordinates the activity of the WASH complex with the function of the retromer SNX dimer to control endosomal tubulation. J Cell Sci. 2014;127:2053–2070. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Harbour ME, Breusegem SY, Antrobus R, Freeman C, Reid E, Seaman MN. The cargo‐selective retromer complex is a recruiting hub for protein complexes that regulate endosomal tubule dynamics. J Cell Sci. 2010;123:3703–3717. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Harbour ME, Breusegem SY, Seaman MN. Recruitment of the endosomal WASH complex is mediated by the extended 'tail' of Fam21 binding to the retromer protein Vps35. Biochem J. 2012;442:209–220. [DOI] [PubMed] [Google Scholar]
  18. Healy MD, McNally KE, Butkovic R, Chilton M, Kato K, Sacharz J, et al. Structure of the endosomal commander complex linked to Ritscher‐Schinzel syndrome. Cell. 2023;186:2219–2237.e29. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Helfer E, Harbour ME, Henriot V, Lakisic G, Sousa‐Blin C, Volceanov L, et al. Endosomal recruitment of the WASH complex: active sequences and mutations impairing interaction with the retromer. Biol Cell. 2013;105:191–207. [DOI] [PubMed] [Google Scholar]
  20. Hernandez‐Valladares M, Kim T, Kannan B, Tung A, Aguda AH, Larsson M, et al. Structural characterization of a capping protein interaction motif defines a family of actin filament regulators. Nat Struct Mol Biol. 2010;17:497–503. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hesketh GG, Perez‐Dorado I, Jackson LP, Wartosch L, Schafer IB, Gray SR, et al. VARP is recruited on to endosomes by direct interaction with retromer, where together they function in export to the cell surface. Dev Cell. 2014;29:591–606. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hierro A, Rojas AL, Rojas R, Murthy N, Effantin G, Kajava AV, et al. Functional architecture of the retromer cargo‐recognition complex. Nature. 2007;449:1063–1067. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Jia D, Gomez TS, Billadeau DD, Rosen MK. Multiple repeat elements within the FAM21 tail link the WASH actin regulatory complex to the retromer. Mol Biol Cell. 2012;23:2352–2361. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Jia D, Gomez TS, Metlagel Z, Umetani J, Otwinowski Z, Rosen MK, et al. Wash and WAVE actin regulators of the Wiskott‐Aldrich syndrome protein (WASP) family are controlled by analogous structurally related complexes. Proc Natl Acad Sci U S A. 2010;107:10442–10447. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Jia D, Zhang JS, Li F, Wang J, Deng Z, White MA, et al. Structural and mechanistic insights into regulation of the retromer coat by TBC1d5. Nat Commun. 2016;7:13305. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kabsch W. XDS. Acta Crystallogr D Biol Crystallogr. 2010;66:125–132. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kendall AK, Xie B, Xu P, Wang J, Burcham R, Frazier MN, et al. Mammalian retromer is an adaptable scaffold for cargo sorting from endosomes. Structure. 2020;28:393–405.e394. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kovtun O, Leneva N, Bykov YS, Ariotti N, Teasdale RD, Schaffer M, et al. Structure of the membrane‐assembled retromer coat determined by cryo‐electron tomography. Nature. 2018;561:561–564. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kvainickas A, Orgaz AJ, Nagele H, Diedrich B, Heesom KJ, Dengjel J, et al. Retromer‐ and WASH‐dependent sorting of nutrient transporters requires a multivalent interaction network with ANKRD50. J Cell Sci. 2017;130:382–395. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Langer M, Fiegl J, Riegel V, Prohaska R, Kubista E, Ringler M. Psychosomatic aspects of galactorrhea. Arch Gynecol Obstet. 1991;248:167–173. [DOI] [PubMed] [Google Scholar]
  31. Laskowski RA, Swindells MB. LigPlot+: multiple ligand‐protein interaction diagrams for drug discovery. J Chem Inf Model. 2011;51:2778–2786. [DOI] [PubMed] [Google Scholar]
  32. Lee S, Chang J, Blackstone C. FAM21 directs SNX27‐retromer cargoes to the plasma membrane by preventing transport to the Golgi apparatus. Nat Commun. 2016;7:10939. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Leneva N, Kovtun O, Morado DR, Briggs JAG, Owen DJ. Architecture and mechanism of metazoan retromer:SNX3 tubular coat assembly. Sci Adv. 2021;7: eabf8598. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Lucas M, Gershlick DC, Vidaurrazaga A, Rojas AL, Bonifacino JS, Hierro A. Structural mechanism for cargo recognition by the retromer complex. Cell. 2016;167:1623–1635.e14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. McCoy AJ, Grosse‐Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. Phaser crystallographic software. J Appl Cryst. 2007;40:658–674. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. McGough IJ, Steinberg F, Gallon M, Yatsu A, Ohbayashi N, Heesom KJ, et al. Identification of molecular heterogeneity in SNX27‐retromer‐mediated endosome‐to‐plasma‐membrane recycling. J Cell Sci. 2014;127:4940–4953. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. McGough IJ, Steinberg F, Jia D, Barbuti PA, McMillan KJ, Heesom KJ, et al. Retromer binding to FAM21 and the WASH complex is perturbed by the parkinson disease‐linked VPS35(D620N) mutation. Curr Biol. 2014;24:1678. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Mecozzi VJ, Berman DE, Simoes S, Vetanovetz C, Awal MR, Patel VM, et al. Pharmacological chaperones stabilize retromer to limit APP processing. Nat Chem Biol. 2014;10:443–449. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Muhammad A, Flores I, Zhang H, Yu R, Staniszewski A, Planel E, et al. Retromer deficiency observed in Alzheimer's disease causes hippocampal dysfunction, neurodegeneration, and Abeta accumulation. Proc Natl Acad Sci U S A. 2008;105:7327–7332. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Munsie LN, Milnerwood AJ, Seibler P, Beccano‐Kelly DA, Tatarnikov I, Khinda J, et al. Retromer‐dependent neurotransmitter receptor trafficking to synapses is altered by the Parkinson's disease VPS35 mutation p.D620N. Hum Mol Genet. 2015;24:1691–1703. [DOI] [PubMed] [Google Scholar]
  41. Murshudov GN, Skubak P, Lebedev AA, Pannu NS, Steiner RA, Nicholls RA, et al. REFMAC5 for the refinement of macromolecular crystal structures. Acta Crystallogr D Biol Crystallogr. 2011;67:355–367. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Muzio L, Sirtori R, Gornati D, Eleuteri S, Fossaghi A, Brancaccio D, et al. Retromer stabilization results in neuroprotection in a model of amyotrophic lateral sclerosis. Nat Commun. 2020;11:3848. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Nooh MM, Bahouth SW. Two barcodes encoded by the type‐1 PDZ and by phospho‐Ser(312) regulate retromer/WASH‐mediated sorting of the ss(1)‐adrenergic receptor from endosomes to the plasma membrane. Cell Signal. 2017;29:192–208. [DOI] [PubMed] [Google Scholar]
  44. Perez‐Torres EJ, Utkina‐Sosunova I, Mishra V, Barbuti P, De Planell‐Saguer M, Dermentzaki G, et al. Retromer dysfunction in amyotrophic lateral sclerosis. Proc Natl Acad Sci U S A. 2022;119:e2118755119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Phillips‐Krawczak CA, Singla A, Starokadomskyy P, Deng Z, Osborne DG, Li H, et al. COMMD1 is linked to the WASH complex and regulates endosomal trafficking of the copper transporter ATP7A. Mol Biol Cell. 2015;26:91–103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Puthenveedu MA, Lauffer B, Temkin P, Vistein R, Carlton P, Thorn K, et al. Sequence‐dependent sorting of recycling proteins by actin‐stabilized endosomal microdomains. Cell. 2010;143:761–773. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Romano‐Moreno M, Rojas AL, Williamson CD, Gershlick DC, Lucas M, Isupov MN, et al. Molecular mechanism for the subversion of the retromer coat by the legionella effector RidL. Proc Natl Acad Sci U S A. 2017;114:E11151–E11160. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Seaman MNJ, Mukadam AS, Breusegem SY. Inhibition of TBC1D5 activates Rab7a and can enhance the function of the retromer cargo‐selective complex. J Cell Sci. 2018;131: jcs217398. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Sheffield P, Garrard S, Derewenda Z. Overcoming expression and purification problems of RhoGDI using a family of "parallel" expression vectors. Protein Expr Purif. 1999;15:34–39. [DOI] [PubMed] [Google Scholar]
  50. Simonetti B, Cullen PJ. Actin‐dependent endosomal receptor recycling. Curr Opin Cell Biol. 2019;56:22–33. [DOI] [PubMed] [Google Scholar]
  51. Small SA, Kent K, Pierce A, Leung C, Kang MS, Okada H, et al. Model‐guided microarray implicates the retromer complex in Alzheimer's disease. Ann Neurol. 2005;58:909–919. [DOI] [PubMed] [Google Scholar]
  52. Steinberg F, Gallon M, Winfield M, Thomas EC, Bell AJ, Heesom KJ, et al. A global analysis of SNX27‐retromer assembly and cargo specificity reveals a function in glucose and metal ion transport. Nat Cell Biol. 2013;15:461–471. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Tilley FC, Gallon M, Luo C, Danson CM, Zhou J, Cullen PJ. Retromer associates with the cytoplasmic amino‐terminus of polycystin‐2. J Cell Sci. 2018;131: jcs211342. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Vilarino‐Guell C, Wider C, Ross OA, Dachsel JC, Kachergus JM, Lincoln SJ, et al. VPS35 mutations in Parkinson disease. Am J Hum Genet. 2011;89:162–167. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Yao J, Yang F, Sun X, Wang S, Gan N, Liu Q, et al. Mechanism of inhibition of retromer transport by the bacterial effector RidL. Proc Natl Acad Sci U S A. 2018;115:E1446–E1454. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Yong X, Mao L, Seaman MNJ, Jia D. An evolving understanding of sorting signals for endosomal retrieval. iScience. 2022;25:104254. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Zavodszky E, Seaman MN, Moreau K, Jimenez‐Sanchez M, Breusegem SY, Harbour ME, et al. Mutation in VPS35 associated with Parkinson's disease impairs WASH complex association and inhibits autophagy. Nat Commun. 2014;5:3828. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Zimprich A, Benet‐Pages A, Struhal W, Graf E, Eck SH, Offman MN, et al. A mutation in VPS35, encoding a subunit of the retromer complex, causes late‐onset Parkinson disease. Am J Hum Genet. 2011;89:168–175. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Data S1. Supporting Information.

PRO-33-e4980-s002.docx (1.5MB, docx)

Video S1. Cartoon representation of the full‐length retromer (PDB 6VAC (57)) is depicted, highlighting mutated residues in stick format. Red‐colored residues indicate mutations that interfere with binding, while green‐colored residues denote mutations that do not affect binding toward the respective FAM21 fragments as illustrated in Figure 3.

Download video file (29.1MB, mp4)

Data Availability Statement

Atomic coordinates and structure factors of the crystallographic complexes are available in the Protein Data Bank (PDB) with accession code 8RKS listed in Table 1.

Articles from Protein Science : A Publication of the Protein Society are provided here courtesy of The Protein Society

RESOURCES