Article https://doi.org/10.

1038/s41467-024-54583-6

Structural basis for Retriever-SNX17

assembly and endosomal sorting

Received: 21 March 2024 Amika Singla 1,11, Daniel J. Boesch2,11, Ho Yee Joyce Fung 3,11, Chigozie Ngoka2,
Avery S. Enriquez2, Ran Song 4, Daniel A. Kramer 2, Yan Han 3,
Accepted: 15 November 2024
Esther Banarer1, Andrew Lemoff 5, Puneet Juneja6,10, Daniel D. Billadeau7,
Xiaochen Bai 3, Zhe Chen 3, Emre E. Turer 4 , Ezra Burstein 1,8 &
Baoyu Chen 2,9
Check for updates

During endosomal recycling, Sorting Nexin 17 (SNX17) facilitates the transport

1234567890():,;
1234567890():,;

of numerous membrane cargo proteins by tethering them to the Retriever

complex. Despite its importance, the mechanisms underlying this interaction
have remained elusive. Here, we provide biochemical, structural, cellular, and
proteomic analyses of the SNX17-Retriever interaction. Our data reveal that
SNX17 adopts an autoinhibited conformation in the basal state, with its FERM
domain sequestering its C-terminal tail. The binding of cargo proteins to the
FERM domain displaces the C-terminal tail through direct competition. The
released tail engages with Retriever by binding to a highly conserved interface
between its VPS35L and VPS26C subunits, as revealed by cryogenic electron
microscopy (cryo-EM). Disrupting this interface impairs the Retriever-SNX17
interaction, subsequently affecting the recycling of SNX17-dependent cargoes
and altering the composition of the plasma membrane proteome. Intriguingly,
the SNX17-binding pocket on Retriever can be utilized by other ligands con-
taining a consensus acidic C-terminal tail motif. Together, our findings
uncover a mechanism underlying endosomal trafficking of critical cargo pro-
teins and reveal how Retriever can potentially engage with other regulatory
factors or be exploited by pathogens.

Plasma membrane (PM) proteins undergo frequent internalization into involves intricate regulatory systems. Among these, the trimeric pro-
the endosomal compartment, where they are either routed back to the tein complex Retriever plays a crucial role in identifying PM proteins,
cell surface for reuse or to lysosomes for degradation. The main- also called cargoes, for recycling from endosomes. Composed of
tenance of this trafficking process is vital for cellular homeostasis and VPS35L, VPS26C, and VPS29 (Fig. 1a), Retriever is remotely related to

1
Department of Internal Medicine, University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, TX 75390, USA. 2Roy J. Carver
Department of Biochemistry, Biophysics & Molecular Biology, Iowa State University, 2437 Pammel Drive, Ames, IA 50011, USA. 3Department of Biophysics,
University of Texas Southwestern Medical Center, 6001 Forest Park Road, Dallas, TX 75390, USA. 4Center for the Genetics of Host Defense, University of Texas
Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, TX 75390, USA. 5Department of Biochemistry, University of Texas Southwestern Medical
Center, 5323 Harry Hines Boulevard, Dallas, TX 75230, USA. 6Cryo-EM facility, Office of Biotechnology, Iowa State University, 2437 Pammel Drive, Ames, IA
50011, USA. 7Division of Oncology Research, College of Medicine, Mayo Clinic, Rochester, MN 55905, USA. 8Department of Molecular Biology, University of
Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, TX 75390, USA. 9On sabbatical leave at Department of Biophysics, University of
Texas Southwestern Medical Center, 6001 Forest Park Road, Dallas, TX 75390, USA. 10Present address: Thermo Fisher Scientific, 5350 NE Dawson Creek Drive,
Hillsboro, OR 97124, USA. 11These authors contributed equally: Amika Singla, Daniel J. Boesch, Ho Yee Joyce Fung.
e-mail: [email protected]; [email protected]; [email protected]

Nature Communications | (2024)15:10193 1

Article https://doi.org/10.1038/s41467-024-54583-6

a b GST-SNX17
VPS26C

0
Retriever

7
-4
GST FL CT

∆4 0
VPS35L

47

67
VPS29

FL

∆L
1 109 432 470 GST-SNX17 Retriever + + + -
GST PX FERM SNX17 FL kDa kDa
100 VPS35L 100 VPS35L
“Belt” linker GST-SNX17 FL
? 75 GST-SNX17 75
GST PX FERM ∆L470
SNX17 PX FERM CT 50 50

GST PX FERM ∆467-470 37 37

VPS26C
mbrane NxxY/F VPS26C GST-SNX17 CT
endosome me 25 25
VPS29 VPS29
GST CT (451-470) 20 20
1 2 3 4 5 6 7
Cargo

c d e loading
peptide (μM): 0 0 0.2 1.4 8.0 1 control GST GST-SNX17
Retriever + + + + +

Fractional occupancy
GST: + GST-SNX17 CT
GST-SNX17: + + + + 0.8 KD = 0.11 ± 0.01 μM CCDC22-CCDC93 VBD + + +
kDa kDa
100 VPS35L 0.6 100 VPS35L
75 MBP-CCDC22 VBD
75 GST-SNX17
GST-SNX17
0.4 MBP-CCDC93 VBD
50 50
0.2
37 37
GST
VPS26C 0 VPS26C
25 25
20 VPS29 VPS29
−0.2 20
1 2 3 4 5 0.001 0.01 0.1 1.0 10 1 2 3 4 5 6
GST bait (μM)

f loading
control GST-SNX17 GST
Retriever + + +
VPS35L-VPS29 + + +
VPS26C + + +
kDa
100 VPS35L
75 GST-SNX17

50
37

VPS26C
25
20 VPS29
1 2 3 4 5 6 7 8

Fig. 1 | SNX17 uses its C-terminal tail to bind Retriever. a Cartoon depiction of independent experiments and globally fitted to a one-binding site model to obtain
Retriever and the domain architecture of SNX17. b Cartoon representation of GST- the KD and fitting error48. GST pull-down as a negative control was from one
SNX17 constructs used (left panel) and Coomassie blue-stained SDS PAGE gels experiment. Representative Coomassie blue-stained SDS-PAGE gels from the EPD
showing in vitro GST pull-down between indicated GST-SNX17 constructs and experiments are shown in Supplementary Fig. 1a. Coomassie blue-stained SDS
Retriever (right panel). c Coomassie blue-stained SDS PAGE gels showing in vitro PAGE gels showing in vitro GST pull-down between GST-SNX17, and Retriever
GST pull-down between GST-SNX17 and Retriever in the presence of increasing complexed with CCDC22-CCDC93 VBD dimer (e) or isolated subunits of Retriever
concentrations of a competing peptide consisting of the last 20 amino acids of (f). Representative results from at least two independent experiments are shown
SNX17. d Binding isotherms obtained from EPD assays measuring the binding for each pull-down. Source data for b–f are provided as a Source Data file.
affinity between GST-SNX17 CT and Retriever. Data were pooled from three

the well-studied endosomal recycling complex Retromer1–3, which Sorting nexin proteins represent crucial regulatory factors
handles a separate subset of cargoes. Recent studies have revealed that responsible for tethering Retromer or Retriever to endosomal mem-
while Retriever shares a similar overall architecture with Retromer, it branes and their specific cargoes23. Categorized into seven groups
possesses distinct structural features and regulatory mechanisms4–7. based on their domain organization, sorting nexins participate in many
Retriever manages the recycling of a broad spectrum of cargoes, facets of protein trafficking within cells23. Retromer-associated sorting
including integrins, tyrosine receptor kinases, G-protein coupled nexins, such as the PX domain-only SNX324 and the BAR domain-
receptors (GPCRs), and lipoprotein receptors4,8,9. In contrast, Retro- containing SNX1, SNX2, SNX5, and SNX625, mediate membrane defor-
mer handles a distinct subset of cargoes, such as various transporters mation. Among these, some sorting nexins like SNX3, SNX5, and
(DMT1, ATP7A/B, GLUT1), GPCRs (β2AR), and SorL1 (a sorting factor SNX626–28 also contribute to cargo recognition. Another important
implicated in Alzheimer’s disease)10–15. Both Retriever and Retromer sorting nexin, SNX27, which contains both PDZ and FERM domains,
cooperate with additional factors to ensure efficient cargo sorting. serves to tether Retromer to over 100 specific cargoes10,15. SNX27
Integral to the function of both complexes is the WASH regulatory accomplishes this by binding to the VPS26 subunit of Retromer and,
complex, which promotes Arp2/3-mediated actin polymerization at through its PDZ domain, connecting Retromer to PDZ binding motifs
endosomal membranes16–19. In addition, Retriever associates with the in the cytoplasmic tails of these cargoes29. In contrast, SNX17, a distant
COMMD/CCDC22/CCDC93 complex (CCC)8 to form a larger structure homolog of SNX27 that also contains a FERM domain, is specifically
called the Commander assembly20–22. In this assembly, the ten COMMD associated with Retriever. Unlike SNX27, SNX17 uses its FERM domain
proteins form a ring-like structure5–7, while the CCDC22-CCDC93 to recognize the NxxY/F motif in the cytoplasmic tails of over 100
dimer connects this ring to Retriever using different domains. The distinct cargo proteins30,31. The interaction between Retriever and
dimer also interacts with DENND10, a putative Rab guanine nucleotide SNX17 is dependent on SNX17’s C-terminal tail4, but the precise
exchange factor (GEF) whose functions are not yet clearly defined. mechanism underlying their binding has yet to be deciphered (Fig. 1a).

Nature Communications | (2024)15:10193 2

Article https://doi.org/10.1038/s41467-024-54583-6

It also remains unknown if other regulatory factors connect Retriever containing an NxxY/F motif31 (Fig. 2a–c, lanes 3, 5), indicating that the
to additional cargoes or recycling processes. CT tail and cargo peptide bind to the same surface on the FERM
In this study, we present the cryo-EM structure of a Retriever- domain. We also tested whether the N-terminal PX domain binding to
SNX17 complex, along with comprehensive validation of the binding the headgroup of phosphoinositide 3-phosphate [PI(3)P] influences
mechanism through biochemical, cellular, and proteomic analyses. the FERM-CT interaction in SNX17 and observed no changes (Fig. 2c,
Furthermore, we report the discovery of additional ligands for lanes 4, 6), aligning with the predicted structural model showing that
Retriever, which similarly interact with the complex through the con- the PX domain and the PI(3)P binding pocket are not involved in
served SNX17-binding pocket. This finding expands the repertoire of sequestering the CT tail (Fig. 2a). We further validated the predicted
regulatory factors of Retriever and suggests versatile connections of structure by mutating the NxxY/F motif in the CT tail, finding that
Retriever with other potential targets. mutations in N459 and F462, but not F460, abolished the binding to
the FERM domain, confirming that the sequestering relies on the NxxY/
Results F motif in SNX17 CT (Fig. 2d).
SNX17 uses its C-terminal tail to bind Retriever Given the proximity of the NxxY/F motif in SNX17 to its extreme
Previous cellular and co-immunoprecipitation studies showed that the CT end, we next tested if this intramolecular interaction prevents
C-terminal (CT) unstructured tail of SNX17 is important for interacting SNX17 CT from binding to Retriever. We found that Retriever com-
with Retriever4 (Fig. 1a). We first used recombinantly purified proteins peted off the FERM domain binding to GST-CT in a dose-dependent
to determine whether the interaction is direct. Our GST pull-down manner (Fig. 2e, left panel). This competition is due to the CT tail
assays showed that GST-SNX17 directly interacted with Retriever and, binding to Retriever, as a mutant GST-CT (L470G), which is incapable
consistent with previous cell-based results4, the in vitro interaction of binding to Retriever, showed no impact of Retriever on its binding
relied on the C-terminal tail of SNX17 (Fig. 1b). Deleting the last four to the FERM domain (Fig. 2e, right panel). Conversely, the addition of
residues (Δ467-470) or the last residue (Δ470) of the tail abolished the the FERM domain also competed off Retriever from binding to GST-CT
interaction (Fig. 1b, lanes 2-3). We found that the tail was both neces- in a dose dependent manner (Fig. 2f, lanes 1–3). These results
sary and sufficient for the interaction, as a GST-tagged tail peptide demonstrate that the SNX17 CT tail can only bind to either the FERM
comprising the last 20 residues similarly pulled down Retriever domain or Retriever, but not both simultaneously. Thus, SNX17 CT
(Fig. 1b, lane 6), and a chemically synthesized peptide of the same 20 binding to FERM domain can prevent its binding to Retriever.
residues of the tail could compete off the binding of GST-SNX17 in a We then tested if cargo peptide binding to the FERM domain
dose-dependent manner (Fig. 1c). Using an equilibrium pull-down promotes the binding of the CT tail to Retriever. Using a pull-down
assay, we determined that the binding has a dissociation constant (KD) condition where GST-CT shows similar levels of pull-down signals for
of ~0.11 µM in our buffer condition (Fig. 1d; Supplementary Fig. 1a). In Retriever and the FERM domain (Fig. 2f, lanes 3, 4), we added
addition, in the same in vitro conditions, we found that SNX17 could increasing concentrations of the cargo peptide and observed that the
similarly bind to Retriever complexed with the VPS35L binding domain interaction between the CT tail and Retriever increased, while the
(VBD) of the CCDC22-CCDC93 dimer (Fig. 1e, lane 6), the key scaffold interaction between the CT tail and the FERM domain decreased
required for CCC complex assembly6, suggesting that SNX17 interacts simultaneously (Fig. 2f, lanes 4-7). This indicates that cargo peptide
similarly with Retriever alone or with the Retriever-CCC complex. binding releases the CT tail, enabling it to interact with Retriever. To
Intriguingly, SNX17 could not bind to individual subunits of further test this model with FL SNX17, we immobilized a His6-tagged
Retriever, including a VPS35L-VPS29 binary subcomplex or the VPS26C Retriever to pull down untagged FL SNX17 and observed that the
subunit in isolation, and only bound to fully assembled Retriever addition of the cargo peptide enhanced the Retriever-SNX17 interac-
(Fig. 1f, lanes 4-6). This is not due to misfolding or mis-assembly of the tion (Fig. 2g, lanes 3, 4), while the addition of the short-chain PI(3)P did
isolated components, as the interaction was readily recovered when not have this effect (Fig. 2g, lanes 3, 5). Together, the above data
the individually purified VPS35L-VPS29 subcomplex and VPS26C were suggest that SNX17 is basally autoinhibited from binding to Retriever,
freshly mixed in the reaction (Fig. 1f, lane 7; also see Supplementary and cargo binding dislodges the CT tail to enhance the efficiency of
Fig. 1b for size exclusion chromatography profiles of individual com- SNX17 in tethering Retriever to cargo proteins (Fig. 2a, right panel).
ponents indicating monodispersed, well-behaving materials). The
above results confirm the requirement of VPS26C for binding4 and Cryo-EM structure of Retriever bound to SNX17 CT
suggest that SNX17 only directly interacts with fully assembled To understand how the CT tail of SNX17 interacts with Retriever, we
Retriever in vivo. next determined the structure of the Retriever-SNX17 complex using
cryo-EM. After exhaustively surveying protein constructs and grid
FERM domain and cargo binding regulate the CT tail of SNX17 conditions, we were able to obtain a cryo-EM map with a resolution of
AlphaFold structures32 suggest that SNX17 may also engage in an ~3.4 Å by using Retriever mixed with saturating concentrations of the
intramolecular interaction between an 459NFAF462 sequence, located SNX17 tail peptide (Fig. 3a, Table 1, Supplementary Fig. 2). We used
−11 to −8 residues from its CT end, and the FERM domain (Fig. 2a). This local refinement and local resolution-based map sharpening33 to
interaction is similar to how NxxY/F motifs in SNX17-specific cargoes improve map quality and built the structural model starting with one
bind to the same surface of the FERM domain, as shown in previous generated by AlphaFold prediction (Fig. 3a and Supplementary Figs. 2
crystal structures30,31 (Fig. 2a, red sequences). This model suggests an and 3a). The overall crescent-shaped structure of Retriever is slightly
intriguing regulatory mechanism where the CT tail is sequestered by extended compared to its apo form, with an average root-mean-square
binding to the FERM domain and can be displaced by cargo binding to deviation of ~1.9 Å (Supplementary Fig. 3b). Due to potential con-
facilitate its interaction with Retriever (Fig. 2a, right panel). formational dynamics, we could not obtain a well-resolved map for the
To test this model, we used individually purified CT tail and the VPS29-bound end, where the N-terminal “belt” sequence of VPS35L was
FERM domain of SNX17 to assess their interaction and the effect of found to stabilize the bound VPS29 and the CT region of VPS35L in our
cargo peptide binding (Fig. 2b). We found that GST-CT interacted with previous work6 (Fig. 3a, represented by dashed line, and Supplemen-
the FERM domain as well as a longer construct containing PX-FERM, tary Figs. 2 and 3b).
but not with full-length (FL) SNX17 (Fig. 2c, lanes 1, 2, 4). This suggests Nevertheless, the map unambiguously located the density of
that in FL SNX17, its CT tail binds to the FERM domain, blocking it from SNX17’s CT tail over a conserved surface nestled between the VPS35L
accessing the free CT tail. The binding of GST-CT to FERM or PX-FERM and VPS26C subunits of Retriever (Fig. 3a, b, with map quality shown in
was blocked by the addition of a synthetic cargo peptide from KRIT1 Supplementary Fig. 3c, d). This density readily accommodated 12

Nature Communications | (2024)15:10193 3

Article https://doi.org/10.1038/s41467-024-54583-6

a SNX17 CT (NFAF) b1
KRIT1 (PDB 4TKN) (NPLF) 109 432 470
P-selectin (PDB 4GXB) (NAAY) PX FERM SNX17 FL
Retriever NFAF
PI(3)P Autoinhibited PX FERM PX-FERM

SNX17 CT
Opened FERM FERM
PX PX FERM
NFAF PX FERM
PI(3)P GST GST-CT (451-470)
NxxY/F
mbr ane
endosome me
FERM Cargo Cargo peptide (KRIT1, 225-237)
FERM (PDB 4TKN) NPLF
FERM (PDB 4GXB)

c GST-CT pull-down Loading control d 451 459 462

ASASDVHGNFAFEGIGDEDL
470 e GST-CT pull-down FERM domain

SNX17: FL FERM PX-FERM GST-CT: WT L470G

GST-CT pull-down FERM domain
M
PI(3)P diC4: + E R
PX M
-F
R

+ + Retriever(nM): 0 25 50 100 200 500 0 25 50 100 200 500

F4 A
Cargo peptide:

F4 A
A
FE
FL

59
60
62
KDa: 100

T
N4
GST-CT: VPS35L

W
kDa: 75
FL 75
PX-FERM kDa: 50
50
50
FERM 37 FERM
37 37 FERM
GST-CT 25 GST-CT GST-CT /
25 25 VPS26C
20 20
20 VPS29
1 2 3 4 5 6 7 8 9 1 2 3 4 1 2 3 4 5 6 7 8 9 10 11 12

f GST-CT pull-down Retriever g His6-Retriever pull-down SNX17 FL

PI(3)P diC4: + +
Cargo peptide (μM): 0 0 0 0 2.5 5 10 Cargo peptide: + +
SNX17 FL: + + + + + +
FERM (μM): 0 1.5 5 5 5 5 5 His6-Retriever: + + + +
KDa: 100 VPS35L
KDa: 100 VPS35L
75
75
50
50 SNX17 FL
37 FERM
GST-CT / 37
25 VPS26C
25 VPS26C
20 VPS29
20 VPS29
1 2 3 4 5 6 7 1 2 3 4 5 6 7

Fig. 2 | SNX17 CT tail is sequestered by FERM domain and displaced by cargo d shows GST-SNX17 CT tail mutants pulling down the SNX17 FERM domain. e shows
binding. a Left: Overlay of an AlphaFold model of FL SNX17 with crystal structures GST-SNX17 CT tails pulling down the FERM domain in the presence of increasing
of the SNX17 FERM domain bound to cargo peptides. The NxxY/F motifs are concentrations of Retriever. f shows GST-SNX17 CT tail pulling down Retriever in
highlighted in red. Right: Cartoon representation show how cargo binding dis- the presence of FERM domain and/or increasing concentrations of cargo peptide.
places the sequestered CT tail to promote Retriever binding. b Cartoon repre- g shows His6-tagged Retriever on TALON beads pulling down untagged FL SNX17 in
sentation of SNX17 and cargo constructs used. c–g Coomassie blue-stained SDS the presence of 6 µM cargo peptide or 200 µM PI(3)P diC4. Representative results
PAGE gels showing in vitro pull-down under the indicated conditions. c shows GST- from at least two independent experiments are shown for each pull-down. Source
SNX17 CT tail pulling down indicated SNX17 constructs in the presence of 0.6 µM data for c–g are provided as a Source Data file.
KRIT1 cargo peptide or 200 µM PI(3)P diC4, a short chain, soluble PI(3)P analog.

residues at the C-terminus of the peptide (Fig. 3a, b), which also carboxyl group engages with residues K204, R248, and T276 in
encompasses the NxxF motif. The peptide adopts a uniquely twisted VPS35L, while its side chain fits into a deep hydrophobic pocket
conformation containing two short, distorted 1-turn helices separated formed by V205, W280, and the alkyl chains of K204 and K283 of
by a short loop (Fig. 3c, d). The majority of the interaction is mediated VPS35L. These interactions explain why mutating L470 to G or deleting
by a conserved and positively charged surface on VPS35L, contributed this residue abolished the Retriever-SNX17 interaction4 (Fig. 1b, lane 2).
largely by residues from helices α2, α3, and α4 (Fig. 3b, c; Supple- In addition to L470, the structure also explains how other residues
mentary Fig. 4a, b). In addition, a conserved and slightly positively in SNX17’s tail contribute to the binding (Fig. 3c, d). At the C-terminal
charged surface on VPS26C, mainly contributed by residues on Loop 1, portion of the peptide, D469 in SNX17 is oriented towards K14 from
Loop 13, and strand β12, interacts with the SNX17 peptide from the VPS26C, while D467 in SNX17 and K204 in VPS35L engage with each
opposite side (Fig. 3c; Supplementary Fig. 4a, b). This interaction is other’s backbone. At the N-terminal portion of the peptide, F462 in
unique to Retriever, as Retromer uses distinct surfaces to interact with SNX17, which is part of the NxxF motif crucial for binding to the FERM
adaptors, such as SNX3 and SNX27, or directly with cargoes, such as domain (Fig. 2a–d), interacts with residues L208, I212, I287, and K283
DMT126,29,34 (Supplementary Fig. 4c). This experimentally derived cryo- in VPS35L, with the backbone of F462 and I465 in SNX17 further
EM model of the Retriever-SNX17 complex is consistent with Alpha- interacting with K283 in VPS35L. In addition, residues N459, F460, and
Fold predictions, with some differences in the residues leading to the A461 of SNX17 may form van der Waals interactions with the VPS26C
C-terminal tip of SNX17 (Supplementary Fig. 3a). surface.
The structural model elucidates the significance of the last few It is remarkable that the sequences of both VPS35L and VPS26C
amino acids of SNX17 previously shown to be critical to the binding to that contribute to the SNX17 binding pocket, especially the residues
Retriever4. Remarkably, the last residue of SNX17, L470, uses both its directly involved in the interaction, are conserved across a diverse
side chain and the carboxyl group to establish a network of interac- range of organisms from human to amoeba and Arabidopsis (Fig. 3c–e;
tions with VPS35L critical for binding (Fig. 3c, d). Specifically, L470’s Supplementary Fig. 4a). This suggests that the SNX17-Retriever

Nature Communications | (2024)15:10193 4

Article https://doi.org/10.1038/s41467-024-54583-6

Fig. 3 | Structure of SNX17 C-terminal tail binding to Retriever. a Schematic and lines indicate polar interactions. Residues mutated in this study are indicated by a
overall colored cryo-EM map of the SNX17 CT peptide (golden) complexed with black box. e Sequence alignment of human VPS35L and VPS26C with orthologs
Retriever (VPS35L in green, VPS29 in magenta, and VPS26C in cyan). Dotted lines from indicated representative species. Residues contacting SNX17 are indicated
indicate structural elements not resolved in cryo-EM. N-terminal domains of SNX17 with yellow boxes and arrowheads. Deleterious somatic mutations found in the
are shown as a reference. b–d Close-up views showing key interactions between the COSMIC database and mutations tested for binding to SNX17 are indicated. Coo-
SNX17 CT peptide (carbon in green, oxygen in red, and nitrogen in blue) and its massie blue-stained SDS PAGE gels showing GST-SNX17 (f) or MBP-CCDC22-
binding surface on VPS35L and VPS26C: b shows cryo-EM density of the SNX17 CT CCDC93 VBD dimer (g) pulling down purified Retriever bearing the indicated point
peptide; c, d shows surface conservation calculated with ConSurf63, with color-to- mutations in VPS35L or VPS26C. Representative results from at least two inde-
white gradients representing the most (ConSurf score = 9) to the least conserved pendent experiments are shown. Source data for f and g are provided as a Source
residues (ConSurf score = 1). Contacting residues are shown as sticks. Dotted yellow Data file.

Nature Communications | (2024)15:10193 5

Article https://doi.org/10.1038/s41467-024-54583-6

Table 1 | Cryo-EM data collection and model statistics

VPS35L-VPS29-VPS26C- VPS35L (partial)-VPS26C- VPS35L (partial)-VPS29 Composite map (EMD-43872)
SNX17 (EMD-43873) SNX17 (EMD-43870) (EMD-43871) (PDB 9AU7)
Data collection and processing
Magnification 105,000
Voltage (kV) 300
Electron exposure (e–/Å2) 64
Defocus range (μm) −0.9 to −2.4
Pixel size (Å) 0.83
Symmetry imposed C1
Initial particle images (no.) 1,009,886
Final particle images (no.) 227,973
Map resolution (Å) 3.4 3.35 3.75
FSC threshold 0.143
Refinement
Initial model used AlphaFold Multimer model
(ma-swt4h)
Model composition
Nonhydrogen atoms 8601
Protein residues 1079
Ligand 0
R.m.s. deviations
Bond lengths (Å) 0.005
Bond angles (°) 1.026
Validation
MolProbity score 1.93
Clashscore 13.06
Poor rotamers (%) 0.21
Ramachandran plot
Favored (%) 95.65
Allowed (%) 4.35
Disallowed (%) 0.00
Protein residues included in VPS35L:115-131, 180-253, 264-
the model 349, 352-738, 742-755, 768-786
VPS29: 3-186
VPS26C: 1-53, 63-146, 150-297
SNX17: 458-470

interaction mechanism is conserved through evolution. Moreover, at Disrupting the SBP alters SNX17 binding in cells
least three of these SNX17-interacting residues have been noted to be Having defined the SBP as required for in vitro binding between
mutated in cancer (Fig. 3e, indicated by pink dots)6, with the resulting Retriever and SNX17, we examined whether this interaction mechan-
missense change predicted to be deleterious. ism held true in cells using co-immunoprecipitation experiments. First,
To validate our structural model, we purified a series of Retriever we observed that in transfected HEK293T cells, different mutations in
complexes in which we mutated individual residues that make critical the SBP impacted the binding between SNX17 and Retriever to varied
contacts with the SNX17 tail, either from the VPS35L or VPS26C side. extents. The mutations V205D and R248M in VPS35L substantially
We then used GST pull-down experiments to examine how these weakened the interaction, while other mutations (N279L and W280Y)
mutations affect the SNX17-Retriever interaction. Consistent with our had minimal effects (Fig. 4a). Combining the V205D and R248M
structure, all mutations abolished the binding to GST-SNX17 (Fig. 3f). mutations (denoted as DM for double mutant hereafter) had a more
Importantly, the disruption of the binding was not due to mis- profound impact on SNX17-Retriever binding (Fig. 4b). Next, using
assembly of Retriever, as the mutant complexes behaved similarly to immunoprecipitation in the reciprocal direction, we further confirmed
their wild-type (WT) counterparts during protein expression, pur- the significant contribution of VPS35L SBP residues (N279, W280,
ification, and size-exclusion chromatography (Supplementary V205, and R248) to the interaction between Retriever and SNX17
Fig. 1b, c). Furthermore, the mutations did not affect the binding of (Fig. 4c), with VPS35L DM displaying the most robust impairment.
Retriever to the CCDC22/CCDC93 VPS35L binding domain (VBD) Importantly, all the mutants tested retained normal interactions in
dimer (Fig. 3g; Supplementary Fig. 1d), which is mediated by differ- cells with the Retriever subunits VPS29 and VPS26C, as well as normal
ent conserved surfaces on the VPS29-bound end of the complex, interactions with the CCC complex (CCDC22 and DENND10) (Fig. 4c),
away from the SNX17 binding pocket (Supplementary Fig. 4a), fur- confirming our in vitro results that the mutations specifically disrupted
ther supporting that the mutations were specific in disrupting the Retriever binding to SNX17 without affecting other regions of
binding to SNX17. Thus, we postulate that the identified SNX17- Retriever.
binding pocket (hereafter named SBP) is an evolutionarily conserved Finally, we complemented a previously established VPS35L
surface required for binding to SNX17. knockout (KO) HeLa cell line8 and generated polyclonal sublines stably

Nature Communications | (2024)15:10193 6

Article https://doi.org/10.1038/s41467-024-54583-6

a IP: FLAG (SNX17)

b c
IP: FLAG (SNX17) IP: HA (VPS35L)

0Y
N 8M
R 5D

W 9L
28
24
27
0
T
EV EV

V2
VPS35L-2HA: VPS35L-2HA: WT WT DM

D 0D

9 W
28

27
T

T
HA-VPS26C: + + + + + HA-VPS26C: + + +

EV

EV
VPS35L-2HA:

W
W

W
N
FLAG-GFP-SNX17: — + + + + + FLAG-GFP-SNX17: — + + FLAG-GFP-SNX17: + + + + + + +
kDa kDa
kDa
FLAG (SNX17) 80
115 HA (VPS35L) 115
HA (VPS35L)
80
80
70 CCDC22
70 65

50
50 DENND10
30

HA (VPS26C)
HA (VPS26C) 30
VPS29 25
30
25
25
VPS26C
30

FLAG (SNX17) 80 FLAG (SNX17) 80 HA (VPS35L)

70 80
70 1 2 3 4 5 6 7
1 2 3 4 5 6 1 2 3

d e HeLa VPS35L KO f HeLa VPS35L KO + VPS35L-2HA

HeLa + VPS35L-2HA
WT N279W W280D DM

W 9W
D 0D
VPS35L KO
28
27
T

M
EV
W
N

+ VPS35L-2HA 1 2 3 4 5 6 7 8
FLAG-SNX17: + + + + +
IP: FLAG
(SNX17)

kDa
W 9W
0D

HA (VPS35L)
28
27
T
M

80
EV

5 μm
W
D
N

kDa 80
HA FLAG (SNX17)
80

VPS35L
80
50 FLAG-SNX31: + + + + +
Actin
IP: FLAG
(SNX31)

HA (VPS35L)
1 2 3 4 5 6 7 80

FLAG (SNX31) 50 2 μm

1 2 3 4 5 VPS35L (HA) / FAM21

g n = 49 45 35 48
h HeLa VPS35L KO + VPS35L-2HA
Spearman's Rank Correlation

1.0 p = 0.18 p = 0.65 p = 0.08

WT EV N279W W280D DM
(VPS35L and FAM21)

0.8

0.6

2 μm
0.4
GFP-SNX17 / FAM21
0.2

0.0
T

9W

0D

M
W

D
28
27

W
N

Fig. 4 | Disrupting the SBP impairs SNX17 and SNX31 binding in cells. VPS35L (HA). Representative confocal images (f) and quantification of colocaliza-
a, b Immunoprecipitation of SNX17 (FLAG) followed by immunoblotting for VPS35L tion (g) derived from concurrent immunofluorescence staining for VPS35L (HA,
and VPS26C (HA) in HEK293T cells transfected with the indicated expression vec- green) and the endosomal marker FAM21 (red) in HeLa cells shown in e. In g, each
tors. EV, empty vector; DM, double mutant. c Immunoprecipitation of VPS35L (HA) dot represents an individual cell, with number of cells in each group indicated
followed by immunoblotting for SNX17 (FLAG) and indicated protein components above the graph. Mean and standard deviation are shown. One-way ANOVA with
of the CCC and Retriever complexes in HEK293T cells transfected with indicated Dunnett’s correction was used. NS not significant. h Representative confocal ima-
SNX17 and VPS35L variants. d Immunoblotting analysis for endogenous and stably ges showing concurrent immunofluorescence staining for GFP-SNX17 (green) and
expressed VPS35L in the indicated HeLa cell lines derived from a VPS35L knockout the endosomal marker FAM21 (red) in HeLa cells shown in e and transfected with
(KO) rescued with the indicated variants of VPS35L or an empty vector (EV) control. GFP-SNX17. All western blot experiments and imaging experiments were per-
The parental HeLa cell line used to derive the VPS35L knockout line is included for formed at least twice. Representative results are shown. Source data for a–e, g are
comparison. e Immunoprecipitation of SNX17 (top) or SNX31 (bottom) after provided as a Source Data file.
transfection in the indicated HeLa cell lines, followed by immunoblotting for

re-expressing different HA-tagged VPS35L variants, including WT, and SNX31 bound to VPS35L WT but not to the SBP mutants (Fig. 4e),
N279W, W280D, and DM, or a control line transfected with an empty indicating that the SBP is required for Retriever interactions with both
vector (EV) (Fig. 4d). Using these lines, we examined whether the SBP is proteins.
required for SNX17 as well as SNX31 binding. SNX31 is a homolog of
SNX17 (40% identity between human proteins) expressed only in very Retriever-SNX17 binding is not required for their endosomal
few cell types. SNX31 was previously found to bind to Retriever in a localization
manner that also required its CT leucine residue4. Our co- Next, we examined the potential impact of decoupling Retriever from
immunoprecipitation experiments demonstrated that both SNX17 SNX17 on the localization of these proteins in cells. Using

Nature Communications | (2024)15:10193 7

Article https://doi.org/10.1038/s41467-024-54583-6

a ALPP VPS35L b
5.0 ITGA3 5.0 ITGA2 ITGB1
ITGB1 ITGA5

ITGA1 ITGA3
4.0 4.0
LRP1
ITGA1
- log(p-value)

- log(p-value)
LRP1
ALPI MYORG
3.0 3.0
PODXL2 ALPP
KARS1
2.0 SLC7A1 2.0 FN1

CTSZ
1.0 1.0

0.0 0.0
-3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5
Molecule abundance log2(WT / EV) Molecule abundance log2(WT / DM)

c HeLa VPS35L KO

N O1 IP1
KA 5 8 2

C R2 6

IT RT 1

E P O1 A

TMGT 2
IT P1 2

D D

L 15
L 1

O A
L G
Y L

G 1

O B

N C

Y 3
M L
G 4

T 1

T
+ VPS35L-2HA

L 6

Y 2

M M2A
ENPEP

G 1
A 7A

LNRS1

C PP1
C ST

S I P2
M S35

R 39
A OR

FN DM

IT ZO

N PA
LR A3

SL B1

PO A1

TA A2

SL A5

C A1

M OD
M DX

LP C1

SOAR

M KN
A RC

P C O6

SB D1

R 5A

9
IT PP

LR P2

M PL

C SZ

M L6
IT P1

PL 8 1

O F1
IT PI

YH
C

A 9
C

P
PI 1
G

R
E

R
A

D
N

D
SC
Gene Symbol:
VP

C
Fold change WT vs EV
WT vs DM
FDR corrected WT vs EV
p value WT vs DM
Value (max-min)
p value < 0.05 (min-max)
p value > 0.05

Fig. 5 | Disrupting the SNX17-Retriever interaction impairs PM protein home- statistical significance). c Heat map of PM protein abundance quantified by TMT-
ostasis. Volcano plots summarizing TMT-based proteomic quantification of based proteomics after surface biotinylation and streptavidin purification, using
plasma membrane proteins, comparing WT to EV cells (a), and WT to DM cells (b). indicated HeLa stable cell lines (shown in Fig. 4d). Only proteins whose abundance
This experiment was performed once, with each group consisting of six biological was significantly different between WT and EV lines (p value < 0.05) are displayed.
replicates. Two-sided Student’s t test and Benjamini–Hochberg false discovery rate Membrane proteins are denoted in brown font. Detailed proteomic data are pro-
(FDR) correction were applied in data analyses. Colored dots (blue or green) denote vided in Supplementary Data 1 and deposited in MassIVE repository, as detailed in
proteins with FDR-corrected p value < 0.05; green dots depict hits that also had >2- the data availability section.
fold change. Dot size is proportional to the position on the y axis (depicting

immunofluorescence staining in the aforementioned stable lines expressing VPS35L with SBP mutations (Fig. 6c, d). Direct assessment
shown in Fig. 4d, we found that the re-expressed VPS35L was localized of PM levels of ITGB1 by flow cytometry analysis confirmed that VPS35L
to FAM21-positive endosomes regardless of mutations in the SBP EV and DM cells exhibit reduced surface levels of this integrin
(Fig. 4f, g). Reciprocally, SNX17 localization in these cells was assessed (Fig. 6e–g). Associated with the endosomal trapping phenotypes was
after transfection of GFP-SNX17, showing that endosomal localization an altered morphology of FAM21-positive endosomes, displaying
of SNX17 appeared normal in EV cells and was not impacted by dis- enlarged endosomal domains and coalescence in the perinuclear
ruption of the SBP (Fig. 4h). Thus, endosomal recruitment of VPS35L region (Fig. 6h). The coalescence phenotype, quantified as area of
and SNX17 are both independent of Retriever-SNX17 complex FAM21-positive endosomes per cell, showed significant alterations in
formation. VPS35L deficiency (EV) as well as in all SBP mutants tested (Fig. 6i).
Thus, decoupling Retriever from SNX17 had a profound effect on the
Disruption of the SBP alters PM homeostasis endosomal recycling of various PM proteins and the associated mor-
Next, we examined the functional consequence of disrupting the phological alterations within endosomal compartments.
Retriever-SNX17 interaction on endosomal protein sorting and PM
protein homeostasis. To assess this, we first utilized surface biotiny- SBP mutations reveal other acidic tail partners of Retriever
lation, protein isolation, and mass spectrometry, coupled with tandem Next, we assessed the range of protein-protein interactions for VPS35L
mass tag (TMT) quantification, to compare the PM proteomes in HeLa WT and compared its interacting partners with the DM variant. To
cells re-expressing VPS35L WT or the DM mutant, or the knockout line accomplish this, we immunoprecipitated VPS35L from the corre-
(complemented with EV). This method detected 40 proteins that were sponding HeLa cell lines and identified interacting partners using mass
significantly reduced in VPS35L-deficient cells (EV) after surface bioti- spectrometry in an unbiased manner. Using a 10-fold enrichment over
nylation (Fig. 5a, c; Supplementary Data 1). These proteins are primarily the EV knockout as a threshold, coupled with statistical analysis cor-
PM proteins and include six integrins and LRP1, which are known rected for multiple testing, we identified 14 potential VPS35L inter-
cargoes of SNX174,31,35,36. Remarkably, many cargoes reduced in EV cells acting proteins (Fig. 7a; Supplementary Data 2). This analysis readily
were similarly reduced in VPS35L DM mutant cells (Fig. 5a–c), sug- identified known components of the Retriever and CCC complexes
gesting that the Retriever-SNX17 interaction plays a major role in (Fig. 7a), although we could not detect SNX17, possibly due to its low
Retriever-mediated cargo sorting and recycling to the PM. cellular abundance, low stoichiometry or affinity of binding, or poor
To validate the PM proteomics result, we used immuno- peptide ionization. Our method also identified several proteins not
fluorescence staining to directly evaluate the cellular localization of previously reported to be partners of Retriever, such as LRMDA and
Integrin α5 (ITGA5). Consistent with the proteomics data, ITGA5 ADGRE3 (Fig. 7a).
exhibited reduced staining at the PM and accumulation in FAM21- Intriguingly, we found that LRMDA (leucine rich melanocyte dif-
positive endosomes of EV cells (Fig. 6a). Importantly, SBP mutations in ferentiation associated) preferentially bound to VPS35L WT but not the
VPS35L led to a comparable phenotype, with EV and SBP mutants DM mutant (Fig. 7b). LRMDA contains an NT leucine-rich repeat (LRR)
showing significant endosomal trapping of ITGA5 (Fig. 6b). The same domain and a CT unstructured tail (Fig. 7c). Immunoprecipitation and
analysis of another SNX17 cargo, Integrin β1 (ITGB1), revealed a similar Western blot confirmed that LRMDA only interacted with VPS35L WT
pattern of endosomal trapping in cells lacking VPS35L (EV) or but not the SBP mutants (Fig. 7d). Immunoprecipitation in Lenti-X

Nature Communications | (2024)15:10193 8

Article https://doi.org/10.1038/s41467-024-54583-6

a HeLa VPS35L KO + VPS35L-2HA b n = 33 49 42 80 56

WT EV N279W W280D DM

Spearman's Rank Correlation

1.0 p < 0.0001 p < 0.0001 p < 0.0001 p < 0.0001

(ITGA5 and FAM21)

0.8

0.6
5 μm
0.4

0.2

0.0

EV

9W

0D

M
W

D
28
27
2 μm

W
N
ITGA5 / FAM21

c HeLa VPS35L KO + VPS35L-2HA d n = 36 47 52 50 46

Spearman's Rank Correlation

WT EV N279W W280D DM 1.0
p < 0.0001 p < 0.0001 p < 0.0001 p < 0.0001

(ITGB1 and FAM21)

0.8

0.6
5 μm
0.4

0.2

0.0

EV

9W

0D

M
W

D
28
27
2 μm

W
N
ITGB1 / FAM21

e f g n=8
4.0M 4.0M 100 Unstained
Isotype control

intensity of ITGB1 (%)

100
Normalized to mode

Mean fluorescence

01

01
80 ITGB1 Ab-WT

00

00
0.

0.
3.0M 3.0M
ITGB1 Ab-EV 80

<

<
p

p
SSC-H

FSC-H

60 ITGB1 Ab-DM
2.0M 2.0M 60
Single cells 40
HeLa cells 80.8 40
1.0M 62.8 1.0M
20 20

0 0 0 0
0 1.0M 2.0M 3.0M 4.0M 0 1.0M 2.0M 3.0M 4.0M 104 105 106 WT EV DM
FSC-H FSC-A ITGB1 (FITC)

h HeLa VPS35L KO + VPS35L-2HA i n = 40 59 48 80 61

WT EV N279W W280D DM
Area of FAM21-positive vesicles

280
p < 0.0001 p < 0.0001 p < 0.0001 p < 0.0001
240
per cell (square pixels)

200
160
5 μm
120
80
40
0
T

EV

9W

0D

M
W

D
28
27

2 μm
W
N

FAM21

Fig. 6 | Disrupting the SNX17-Retriever interaction impairs PM protein traf- the indicated HeLa cell lines (g). In g, aggregate data from 2 independent experi-
ficking and endosomal morphology. Representative confocal images (a) and ments are shown, with each dot representing a biological replicate. Representative
quantification of colocalization (b) derived from concurrent immunofluorescence confocal images (h) and quantification of the area of FAM21-positive endosomes (i)
staining for the cargo protein ITGA5 (green) and the endosomal marker FAM21 derived from immunofluorescence staining of FAM21 in the indicated HeLa stable
(red) in the indicated cell lines. c, d Similar to a, b but focusing on the cargo protein cell lines. In all quantifications, each dot represents an individual cell, with number
ITGB1 (green). Flow cytometry analysis of surface ITGB1 in indicated HeLa cell lines, of cells in each group indicated above the graph. Mean and standard deviation are
showing scatter plots depicting gating (e), a representative histogram depicting shown. One-way ANOVA with Dunnett’s correction was used. All imaging and FACS
ITGB1 surface staining (f), and mean fluorescence intensity compared to WT cells in experiments were performed at least twice. Representative results are shown.

293T cells transfected with the full length (FL), NT LRR, and CT tail of species revealed significant homology among their extreme
LRMDA revealed that the CT tail is both necessary and sufficient for C-terminus (Fig. 8a), suggesting a potentially shared mechanism of
binding to Retriever (Fig. 7e), analogous to SNX17. binding. This homology can be summarized as an evolutionarily con-
Comparing the CT tail sequences of validated SBP-dependent served consensus motif comprising the last six residues, which we
binders (SNX17, SNX31, and LRMDA) across various representative initially denoted as [ILF]-x-a-a-a-L, with “x” for any amino acid and “a”

Nature Communications | (2024)15:10193 9

Article https://doi.org/10.1038/s41467-024-54583-6

a DENND10
2.4 CCDC22

VPS26C CCDC93
2.0
COMMD7 COMMD5
VPS35L

- log(p-value)
VPS29
1.6 ADGRE3 COMMD3
COMMD8
LRMDA COMMD2
1.2 COMMD4

0.8

0.4

0.0
-6 -4 -2 0 2 4 6 8 10
Molecule abundance log2(WT / EV)

b
C S29 10
VP D C 5

C MM 3
C MM 7
C MM 8
L R MM 4
AD D 2

E3
C S35 2

D S26 3
C D

O D
O D
O D
O D
M D
EN C
O L

G A
2

9
C MM

C MM
VP D C

R
VP N
O
C

Gene Symbol:
C

Value
WT/EV
WT/DM min 10 100

c 1 166 229 e IP: FLAG (LRMDA)

LRMDA LRR
FLAG-LRMDA: — FL NT CT
NT zsGreen: + — — —
CT VPS35L-2HA: + + + +
kDa
HA (VPS35L) 100
d IP: HA (VPS35L)
75
CCDC93
W 9W
0D
28
27
T
M
EV

VPS35L-2HA:
W
D
N

25
kDa COMMD1
LRMDA 25

25
CCDC93 65
FLAG (LRMDA) 20

HA (VPS35L) 15
80 10
1 2 3 4 5 1 2 3 4

Fig. 7 | Interactome analysis reveals that LRMDA similarly binds to Retriever. domain organization and constructs used in e. d Immunoprecipitation of VPS35L
Volcano plot (a) and heat map (b) of VPS35L-interacting proteins quantified by (HA) followed by immunoblotting for LRMDA and CCDC93 in indicated stable HeLa
mass spectrometry after coimmunoprecipitation of HA-tagged VPS35L, using cell lines. e Immunoprecipitation of LRMDA full-length (FL), NT, and CT in Lenti-X
indicated HeLa stable cell lines shown in Fig. 4d. Two-sided Student’s t test and 293 T cells, followed by immunoblotting for indicated Retriever and CCC subunits.
Benjamini–Hochberg false discovery rate (FDR) correction were applied in data The experiment in a and b was performed once, with three biological replicates
analyses. Colored dots (blue or green) denote proteins with FDR-corrected p included in each group. Detailed proteomic data are provided in Supplementary
value < 0.05; green dots depict hits that also had >10-fold change between WT and Data 2 and deposited in MassIVE repository as detailed in the data availability
EV control samples. Dot size is proportional to the position on the y axis (depicting section. Western blot experiments in d and e were performed three times. Repre-
statistical significance). Protein spectral counts, fold change between indicated sentative results are shown.
cells, and corresponding p-values are depicted. c Cartoon presentation of LRMDA

for acidic residues (Fig. 8a, black box). The sequences preceding this species. The three central acidic residues could be a combination of D,
motif lack a discernable pattern, suggesting the decisive role of this six- E, N, and Q, while the “x” position could be various amino acids. To
residue motif in binding to Retriever. Note that this six-residue motif is define this motif’s composition more precisely, we performed exten-
located just distal to the conserved NxxY/F motif in SNX17, which is sive mutagenesis across the CT tail and tested how mutations affected
absent in SNX31 and LRMDA (Fig. 8a, pink). the binding to Retriever (Fig. 8b). We found that mutations of the
In the CT six-residue motif, the last L seemed the most invariant, terminal L abrogated the binding, consistent with the sequence ana-
followed by the first I residue, which could also be L or F in several lysis in Fig. 8a. However, the −2 and −3 positions tolerated various

Nature Communications | (2024)15:10193 10

Article https://doi.org/10.1038/s41467-024-54583-6

a -20 NxxF -6 -1
c Name CT sequence Primary location / note
-20 -12 -6 -1
Human ASASDVHGNFAFEGIGDEDL SNX17 ASASDVHGNFAFEGIGDEDL endosome
Mouse ASASAVHGNFAFEGIGDEDL SNX31 SKIKIAKDDCVFGNIKEEDL endosome
Chicken VGANDFHGNYAFEGIGDEDL LRMDA RYVYYGKNSEGNRFIRDDQL melanocyte differentiation
Frog MSADDFHGNYAFEGIGDDDL TIMM23 LYALYNNWEHMKGSLLQQSL mitochondria
Zebrafish ISGNDFHGNYAFEGIGDDDL PATE1 SVYLVNFRCCRSHDLCNEDL secreted
Lamprey GSGDGYLENDAFEGIGDDDL ARHGEF25 PTPKTPPCQARLAKLDEDEL cell membrane; cytoskeleton
SNX17
Fruit fly VDNGARVANGAFEGIGDDDL HYOU1 EPEQKEQSTGQKRPLKNDEL endoplasmic reticulum
Nematode ISDGIPQRNQAFTDITNDDL CCDC192 PEAPVFSTHDIPPVVSDENL unknown
Sponge ERDDAAGNKLFQMNIGDDDL BCAR3 QILTALSRKLEPPPVKQAEL cytoplasm; cell junction
Trichoplax EEVVQNEIFEDVNVIGDDDL CD34 QATSRNGHSARQHVVADTEL cell membrane
Capsaspora GKSETYDNAMYDGTIGDDDL USH1G AVRRRRQAMERPPALEDTEL cell membrane; nucleus
Choanoflag ISQGHHFGLTQQDVFGDNDL CCND2 KSEDELDQASTPTDVRDIDL cytoplasm; nucleus
Human SKIKIAKDDCVFGNIKEEDL MTO1 ESSKTDQYLCDADRLQEREL mitochondria
SNX31 Mouse GKMKRSEGDYVWDTLMEEGL SCART1 SRPVSQGYDEAAFPLEEMTL cell membrane
Frog TLLKDKAEYCLIDDISDLNL Human
SCARB1 YSESLMTSAPKGSVLQEAKL cell membrane
Human RYVYYGKNSEGNRFIRDDQL ABCC4 MVTNTSNGQPSTLTIFETAL cell membrane
Mouse RYFYYGRNSEGNRFIRDDQL SLC5A6 LDGTAYQGSSSTCILQETSL cell membrane
Chicken RYIYYGKHSEGNRFIRDDQL EREP1 RRVEAERPHSLIGVIRETVL endosome/cell membrane
Frog HYIYYGKHSEGNRFIRDDQL PIP5K1B AEPNTLEVQDDNASVLDVYL endosome/cell membrane
LRMDA Zebrafish RYVYYGKHSEGNRFIRNDQL ICOS YMFMRAVNTAKKSRLTDVTL cell membrane
Lamprey RYIYQGRQSEGNRFIYDGDL SLC15A5 KFYGSIQEFSSSIDLWETAL cell membrane
Sponge KYVYYGKHSEGNRFIRNNQL GRP2 PEIREEEVQTVEDGVFDIHL cell membrane
Trichoplax RYVYFGRHSEGNRFIRNADL PDGFRA IDMMDDIGIDSSDLVEDSFL golgi/cell membrane
Choanoflag RYVYYGRHSEGNRFIRDRDL KCNMA1 NRPKSRESRDKQKYVQEERL cell membrane
Capsaspora KYVYYGRHSEGNRFIRNNDL CDK5 YNRTNRSRMPNLNDLKETAL endosome
Amoeba TYVYYGRQSEGNRFIMNDDL LPAR2 GASTRIMLPENGHPLMDSTL endosome; cell membrane
PCDH19 LKEGRNKESPGVKRLKDIVL cell membrane
CFTR KPQIAALKEETEEEVQDTRL cell membrane
FLVCR2 EEEEESNTSKVPTAVSEDHL mitochondrial membrane; ER; cell membrane
TYR EKQPLLMEKEDYHSLYQSHL melanosome
nrdB YLVGQIDSEVSADDLSDFEL Yersinia pseudotuberculosis
gspD VIPSIRKDINNFYTLLDSEL Yersinia pseudotuberculosis
b -6 -5 -4 -3 -2 -1
Pathogen SR1 KSGSYLRSFAFVTKLSQQEL Toxoplasma gondii
ABC IEKRVKKENTRFTSIFDIEL Listeria monocytogenes
[ILV]-x-[DEQN]-x-x-L>
N( 3)A
F( 2)A

F( 0)D
A( 1)A

)P
E( )N

I(-6) G(-5) D(-4) E(-3) D(-2) L(-1)

G( )A
G( )A
F( )A

-7
-9
-1

-7
-1

-1

-8
-1

-9
T

GST-SNX17 CT: A M F S H D L V A R L E T R A L N Q A R L G T Q T R A L N Q S E G M F V I G
G(
W

kDa VPS35L
100
75

50
37
VPS26C
25 GST-CT
20 VPS29
1
Binding
relative 0.75
to WT: 0.25
0

d 25 po
rte
r
7 1 A 23 1 EF 1 19
2
3 G 2 rans
NX
1
NX
3 MD IMM ATE RHG YOU 17
D C CAR 34 SH1 CND O1 pD Ct Loading
GST-CT: S S LR T P A H SNX CC B CD U C MT nrdB gs S R1 AB control
Retriever: WT DMWT DM WT DM WT DM WT DM WT DM WT DM WTDM WT DM WT DM WT DM WT DM WT DM WT DM WT DM WT DM WT DM WT DM WT DM
kDa kDa
100 100 VPS35L
VPS35L
75 75

50 50
37 37
VPS26C VPS26C
25 25 GST-CT GST-CT
20 VPS29
VPS29
1 2 3 4 5 6 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 33 34 35 36 37 38

e S
GST-CT: S S-H H- L-S
-20 -12 -6 -1 kDa
SNX17 ASASDVHGNFAFEGIGDEDL 100 VPS35L
HYOU1 EPEQKEQSTGQKRPLKNDEL 75
SNX17-HYOU1 ASASDVHGNFAFEGLKNDEL 50
HYOU1-SNX17 EPEQKEQSTGQKRPIGDEDL 37
VPS26C
linker-SNX17 GGSGGSGGSGGSGGIGDEDL 25 GST-CT
20 VPS29
1 2 3 4

Fig. 8 | Sequence analysis reveals a consensus C-terminal Retriever- represents a repeat (n = 2 or 3). c Sequence alignment of the CT tail of indicated
binding motif. a Sequence alignment of the CT tail of indicated proteins across proteins in humans or pathogens containing a RICT motif at the C-termini of
representative species. Gold, light gold, and white shading denote sequences unstructured tails. Gold, light gold, and white shading denote sequences identical
identical to, similar to, and not conserved with the human SNX17 CT sequence, to, similar to, and not conserved with the human SNX17 CT sequence, respectively.
respectively. The identified 6-residue acidic sequences are highlighted by the black Black and gray arrowheads indicate strong and weak binding in the pull-down assay
box, with each position denoted by a colored dot corresponding to the residues shown in d, respectively. White arrowheads indicate no detectable interaction.
mutated in b. The conserved NxxF motifs are marked in pink, with the assignment Proteins without an arrowhead were not tested. d Coomassie blue-stained SDS
of QDVF in the Choanoflagellate sequence based on AlphaFold predictions. PAGE gels showing in vitro pull-down of Retriever WT vs. the DM mutant by GST-
b Definition of the consensus RICT motif and Coomassie blue-stained SDS PAGE tagged CT of the indicated proteins. e Sequences and Coomassie blue-stained SDS
gels showing in vitro pull-down of Retriever by GST-SNX17 CT tails containing PAGE gel showing in vitro pull-down of Retriever by indicated chimera CT tails.
indicated point mutations. Binding signals of VPS35L bands are quantified and Representative results from two independent experiments are shown for RICT
normalized to the WT CT and shown beneath corresponding mutations. Each dot motif screening and chimera CT tail pull-down.

Nature Communications | (2024)15:10193 11

Article https://doi.org/10.1038/s41467-024-54583-6

Retriever VPS26C
VPS35L
VPS29

SNX17-mediated recycling SBP Other trafficking pathways

SNX17 PX FERM Other

RICT motif factors RICT motif
NxxY/F

Cargo Cargo

Fig. 9 | Schematic of Retriever-mediated endosomal recycling. Our study sug- through NxxY/F motifs in their cytoplasmic tails (left), while the interaction with
gests that Retriever could use the same SNX17-binding pocket (SBP) to interact with other factors can potentially link Retriever to distinct cargoes, recycling pathways,
additional factors that contain the Retriever interacting C-terminal tail (RICT) motif. or cellular destinations (right).
The interaction with SNX17 tethers Retriever to many cargoes recognized by SNX17

other amino acids, with D(−2) tolerating most mutations except for T, that although the six-residue RICT motif is both necessary and suffi-
R, and N, and E(−3) tolerating most mutations except for R. In contrast, cient for binding to Retriever, the overall interaction is determined by
the −4 position could only accommodate D, N, Q (and likely E). The −5 two factors: 1) the exact amino acid combination within the RICT motif,
position tolerated all tested mutations, while the −6 position could and 2) the sequence preceding the RICT motif. This emphasizes the
only tolerate small hydrophobic residues, including I, L, and V, con- importance of experimentally validating the predicted RICT-
sistent with the sequence analysis. Outside of the last six residues, NT containing proteins in future studies.
mutations at most positions did not impact binding, except at the −9
position and, to a less extent, the −10 position (Fig. 8b), suggesting that Discussion
the sequences preceding the CT motif also play a role in binding, in Our study provides a pivotal advance in our understanding of
agreement with the observation of 12 residues of SNX17 CT in the SBP. Retriever-mediated endosomal cargo recycling. Unlike the well-
Combining these results, we redefined the consensus motif as [ILV]-x- studied Retromer25,26,34, the precise mechanisms of cargo selection
[DEQN]-x-x-L and named it Retriever Interacting C-terminal Tail (RICT). by Retriever have remained elusive37. Our findings elucidate how the
Based on the RICT motif, we searched transmembrane or cargo-recognition factor SNX17 uses its C-terminal tail to anchor into a
membrane-associated proteins from humans and several pathogens conserved surface formed by the VPS35L and VPS26C subunits of
and identified additional proteins containing a sequence in an Retriever (Fig. 9, left). This is in line with a recent study derived from
unstructured C-terminal tail that conforms with the consensus for AlphaFold structural prediction and mutagenesis validation38. Fur-
RICT (Fig. 8c). These proteins have diverse membrane localizations thermore, we have identified other regulatory factors that interact with
and were previously not shown to bind to Retriever. To validate our Retriever through the SBP and via similar CT tails, suggesting that the
prediction, we selected several proteins and purified their last 20 SBP is a critical surface that connects Retriever to other cellular func-
amino acid sequences fused to GST. Similar to SNX17, the CT tails of tions beyond SNX17-dependent cargo recycling8 (Fig. 9, right).
SNX31, LRMDA, and PATE1 showed robust binding to Retriever in vitro, We emphasize that despite sharing remote homology with Ret-
while TIMM23, ARHGEF25, USH1G, and pathogenic proteins nrdB (Y. romer and SNX27, Retriever and SNX17 operate through very distinct
pseudotuberculosis), gspD (Y. pseudotuberculosis), and amino acid ABC mechanisms. Notably, the residues surrounding the binding pocket on
transporter permease (L. monocytogenes) showed weaker interactions both the VPS35L and VPS26C sides are highly conserved across spe-
(Fig. 8c, d). Notably, all interactions were abrogated by the DM cies, representing one of the most conserved surfaces on Retriever
mutation, confirming that they use the same mechanism to interact (Supplementary Fig. 4a). This remarkable conservation underscores
with the SBP of Retriever (Fig. 8d). the evolutionary and functional significance of the interaction between
To understand why some CT tails containing a well-defined RICT Retriever and SNX17, as well as other RICT-containing factors.
motif did not show binding, we speculated that the overall interaction Another striking observation is that disrupting the Retriever-
is influenced not only by the RICT motif but also by the entire sequence SNX17 interaction has profound consequences on PM protein home-
context. To test this hypothesis, we designed chimeric CT tails by ostasis, affecting cellular signaling and potentially having clinical
swapping the NT sequence and the CT RICT motif from a binding implications. This is evidenced by our proteomic and cellular analyses,
protein, SNX17, with corresponding sequences from a non-binder, as well as the association between somatic mutations at the SNX17
HYOU1, or with a featureless flexible (GGS) linker sequence (Fig. 8e, left binding pocket and human cancers. Mutations at these residues dis-
panel). We found that replacing either the NT sequence or the CT RICT rupt the Retriever-SNX17 interaction in our experimental system,
motif of the SNX17 CT tail with corresponding HYOU1 sequences suggesting that the cancer-associated mutations may act by perturb-
abolished Retriever binding (Fig. 8e, lanes 2, 3). This demonstrates that ing the homeostasis of crucial cargoes involved in cell adhesion, pro-
the RICT motif of HYOU1 cannot bind to Retriever and that the NT liferation, or metabolism. This binding pocket, therefore, offers a
sequence preceding the RICT motif in the HYOU1 CT tail prevents promising target for the development of therapeutic interventions or
SNX17’s RICT motif from binding to Retriever. Interestingly, replacing small molecule drugs to modulate cellular signaling dynamics.
the NT sequence of the SNX17 tail with a (GGS) linker only weakened its Our proteomic and cellular studies indicate that SNX17 does not
binding to Retriever but did not abolish it, indicating that the native NT constitutively bind to Retriever. It is plausible that the binding could be
sequence of SNX17 tail promotes the binding, while the NT sequence of modulated by various parameters, such as SNX17’s expression level,
HYOU1 disrupts the binding (Fig. 8e, lane 4). Together, we conclude post-translational modifications, cargo binding, cargo membrane

Nature Communications | (2024)15:10193 12

Article https://doi.org/10.1038/s41467-024-54583-6

density, and cellular localization. In particular, our data indicate that MTO1, nrdB2, gspD, SR1, and amino acid ABC transporter were codon-
SNX17’s ability to bind to Retriever is autoinhibited through intramo- optimized and cloned into a pGexTev vector using PCR.
lecular interactions, and this autoinhibition is relieved by cargo bind-
ing, indicating that cargo recognition and Retriever recruitment to E. coli strains for protein expression
cargoes are synergistic processes. This notion is also supported by a Standard, commercial E. coli strains used in this study include Mach1T1R
recent study similarly showing autoinhibition of SNX1738. Interestingly, (Thermo Fisher) and BL21 (DE3)T1R (Sigma), and are grown in Luria-
various sorting nexin proteins, such as SNX2729, SNX326,39, SNX1- Bertani medium using standard molecular biology conditions.
SNX628, and SNX1-SNX540, operate similarly, being autoinhibited in the
basal state, with cargo binding enhancing their association with Ret- Insect cell lines for protein expression
romer or membranes. Although SNX31 and LRMDA, two other con- Sf9 cells (Expression System) were maintained in Sf-900™ II serum-free
firmed SBP-binding proteins, lack the NxxY/F motif and are not medium (Thermo Fisher) and used for baculovirus preparation and
predicted by AlphaFold to be autoinhibited, they may use distinct large-scale expression.
mechanisms to regulate the accessibility of their CT tails. It’s worth
noting that although our in vitro results suggest that the PX domain Cell culture
binding to PI(3)P does not directly impact the autoinhibitory state, it HEK293T (Cat # CRL-3216) and HeLa (Cat # CCL-2) cell lines were
remains possible that membrane recruitment through PI(3)P binding obtained from the American Type Culture Collection (Manassas, VA).
can indirectly promote Retriever-cargo interaction by facilitating Lenti-X 293T cells (Cat #632180) were obtained from Takara. All cell
SNX17 recruitment to microdomains enriched with both Retriever and lines were cultured in high-glucose Dulbecco’s modified Eagle’s med-
cargo proteins. Our structure also suggests that the SNX17-Retriever ium (DMEM) containing 10% fetal bovine serum (FBS) and 1% penicillin/
interaction can be regulated by post-translational modifications. For streptomycin at 37 °C with 5% CO2. Periodic PCR-based testing for
example, based on PhosphoSitePlus41, two key residues at the SBP Mycoplasma spp. was conducted to ensure culture purity. A HeLa line
could be potentially modified, including acetylation of K14 in VPS26C with VPS35L deficiency was previously described8 and these cells were
and ubiquitylation of K204 in VPS35L, which could disrupt SNX17 complemented using a lentiviral vector to express HA tagged VPS35L
binding. In addition, several residues in SNX17 CT near the NxxY/F protein versions as indicated.
motif could be phosphorylated, including S434, S43742, and S440,
which may modulate the autoinhibition mechanism. Finally, a recent Transfection and lentiviral methods
study showed that SNX17 CT could interact with the PDZ domain of a HEK293T or Lenti-X 293T cells were transfected using Lipofectamine
group of PDLIM proteins, which could provide a regulatory mechan- 2000 (Life Technologies) or PolyJet (SignaGen), respectively, and
ism to decouple SNX17 and its cargoes from Retriever43. cultured for either 24 or 48 h before analysis. VPS35L HeLa knockout
The identification of other factors containing the SNX17 homo- cells were reconstituted with HA empty vector or various HA-tagged
logous acidic tail sequences (i.e., the RICT motif) suggests a versatile VPS35L using a lentivirus system. Lentivirus experiments followed a
role for the binding pocket. These additional factors may act as com- standard protocol as previously described for viral vector production
petitors of SNX17 and connect Retriever to a broader range of recy- and selection46,47.
cling pathways, cargoes, or other cellular locations and functions.
These include not only proteins in the host cells, but also effectors Immunofluorescence staining
from pathogens, which might exploit the host trafficking system by We followed protocols previously described8,21. Briefly, cells were fixed
hijacking the Retriever-SNX17 interaction and compromising host with cold fixative (4% paraformaldehyde in PBS) for 18 min at room
cellular functions to create a niche or augment pathogen fitness. In temperature in the dark, followed by 3-min permeabilization using
summary, our research not only elucidates a key mechanism in 0.15% Surfact-Amps X-100 (28314, Thermo Fisher) in PBS. Samples
endosomal trafficking but also opens the door for further exploration were then incubated overnight at 4 °C in a humidified chamber with
into the biological significance of other Retriever-ligand interactions. primary antibodies in immunofluorescence (IF) buffer (Tris-buffered
saline plus human serum cocktail). After three washes in PBS, samples
Methods were incubated with secondary antibodies (1:500 dilution in IF buffer)
Plasmids for 1 h at room temperature or overnight at 4 °C in a humidified
All constructs were created using standard molecular biology proce- chamber. After four washes in PBS, coverslips were mounted on slides
dures and verified by Sanger sequencing. Detailed information about with SlowFade Anti-fade reagent (Life Technologies). Primary and
constructs for recombinant protein production and mammalian secondary antibodies used are provided in Supplementary Table 4.
expression, recombinant protein sequences, and DNA oligonucleo- Images were obtained using an A1R confocal microscope (Nikon, ×60
tides for construct generation can be found in Supplementary Tables 1, /1.4 oil immersion objective) operated by the NIS-Elements A1R (Nikon)
2, and 3, respectively. The ORFs of human VPS35L and VPS26C were software v5.42.03. Fluorescence signal values were quantified using Fiji
previously described6,21,44. The mammalian expression vector for v1.54 f (ImageJ, NIH). Data were processed with Excel (Microsoft) and
SNX17 was previously described45. SNX31 mammalian expression vec- plotted with Prism v9.5.1 (GraphPad) or a Python web application
tor was designed by GeneArt at Thermo Fisher Scientific. For insect cell https://biochempy.bb.iastate.edu. Each dot in the graphs represents
expression of Retriever, human full-length VPS35L (untagged, syn- the value from a single cell, with the horizontal bar indicating the mean
thesized as a codon-optimized GeneString from Thermo Fisher to and the error bars representing the standard error of the mean (SEM).
improve expression), VPS26C (untagged), and VPS29 (isoform 2) Spearman’s Rank correlation coefficient was measured using EzColo-
containing a C-terminal (GGS)2-His6 tag were cloned in a modified calization Fiji Plugin within manually outlined regions of inter-
pFastBacTM vector for insect cell expression as previously described6. est (ROIs).
For bacterial expression of isolated VPS26C, codon-optimized Gene-
String (Thermo Fisher) was cloned in a pMalC2Tev vector6. Bacterial Mammalian protein extraction, immunoblotting, and
expression vector of GST-SNX17 was previously described31. Bacterial immunoprecipitation
expression vectors of CCDC22 VBD and CCDC93 VBD were previously For most experiments, whole cell lysates were prepared using Tri-
described6. The CT 20 amino acids of SNX17, SNX31, LRMDA, TIMM23, ton X-100 lysis buffer (25 mM HEPES, 100 mM NaCl, 10 mM DTT,
PATE1, ARHGEF25, HYOU1, CCDC192, BCAR3, CD34, USH1G, CCND2, 1 mM EDTA, 10% Glycerol, 1% Triton X-100) supplemented with

Nature Communications | (2024)15:10193 13

Article https://doi.org/10.1038/s41467-024-54583-6

protease inhibitors (Roche). Immunoprecipitation, SDS-PAGE, and For the plasma membrane and interaction proteomics samples,
immunoblotting experiments were performed largely as previously raw MS data were analyzed using Proteome Discoverer v3.0 (Thermo),
described8. Specifically, for LRMDA immunoprecipitation, 48 h with peptide identification performed using Sequest HT searching
after transfection, cells were harvested in NP-40 lysis buffer and against the reviewed human protein database from UniProt. We set
mixed with anti-FLAG M2 magnetic beads (Sigma-Aldrich, Cat# fragment and precursor tolerances at 10 ppm and 0.6 Da, respectively,
F4799) for 2 hours at 4 °C. The beads were washed 4 times in NP-40 and allowed three missed tryptic cleavages. Cysteine carbamido-
lysis buffer, and the bound proteins were eluted with 150 μg/mL methylation was set as a fixed peptide modification and methionine
3×Flag (Sigma-Aldrich, Cat# F4799) at 4 °C for 1.5 h. Western blot oxidation as a variable modification. We applied a false-discovery rate
images were collected using ChemiDoc and Image Lab v6.1.0 (Bio- (FDR) cutoff of 1% for all peptides.
Rad). Antibodies used are detailed in Supplementary Table 4. To analyze protein complex composition in native gel samples,
raw MS data were analyzed using MaxQuant v.2.0.3.0, with peptide
Cell surface biotinylation identification performed against the human protein database from
Cell surface biotinylation was performed as previously reported21. UniProt. We set fragment and precursor tolerances at 20 ppm and
Briefly, cells were incubated at 4 °C with Sulfo-NHS-SS-biotin (Pierce) in 0.5 Da, respectively, and allowed three missed cleavages. We set
biotinylation buffer (10 mM triethanolamine, 150 mM NaCl, 2 mM cysteine carbamidomethylation as a fixed peptide modification, and
CaCl2, pH 8.0). After 30 min, cells were lysed in Tris-lysis buffer (50 mM methionine oxidation and N-terminal acetylation as a variable mod-
Tris-HCl, pH 7.4, 150 mM NaCl, 1% NP-40, 0.5% Na deoxycholate, 5 mM ification. We used iBAQ quantitation for protein quantitation within
EDTA, 5 mM EGTA) supplemented with Halt Protease/Phosphatase each sample.
inhibitor (Thermo Fisher). Biotinylated proteins were captured using
nanolink Streptavidin magnetic beads (Solulink) and washed three TMT proteomics
times with the same lysis buffer, once with high salt buffer (50 mM Tris- For TMT-based analysis, samples were thoroughly mixed with 25 μL of
HCl, pH 7.4, 500 mM NaCl), and once with low salt buffer (10 mM Tris- 10% SDS and 100 mM triethylammonium bicarbonate (TEAB) by vor-
HCl, pH 7.4, 5 μM Biotin). Proteins on the beads were eluted using at texing, followed by reduction with 2 μL of 0.5 M tris(2-carboxyethyl)
the elution buffer (PBS, 6 M urea, 0.2% SDS (v/w) containing 100 mM phosphine (TCEP) and incubation at 56 °C for 30 min. Free cysteine
DTT at 65 °C for 30 min. For TMT proteomics, the eluted proteins were thiol groups were then alkylated by adding 2 μL of 500 mM iodoace-
directly submitted in solution to the UT Southwestern Proteomics core tamide and incubating in the dark at room temperature for 30 min.
facility. Afterwards, 5.4 μL of 12% phosphoric acid and 300 μL of S-Trap (Pro-
tifi) binding buffer were added before being loaded onto an S-Trap
Protein affinity purification column. Samples were digested by 1 μg of trypsin and incubation
Knockout cells expressing HA-tagged VPS35L were grown on culture overnight at 37 °C. Digested peptides were eluted, dried and recon-
dishes and lysed in Triton-X lysis buffer. Clarified cell lysates contain- stituted in 26 μL of 50 mM TEAB buffer. Equal amounts of peptides
ing equal amounts of protein were added to HA-resin to capture HA- from each sample were labeled with TMTpro 18plex reagents (Thermo)
tagged proteins. HA beads were washed using lysis buffer and eluted based on absorbance at 205 nm using a NanoDrop (Thermo). The six
using 3 x LDS/DTT gel loading buffer at 95 °C. Eluted proteins were WT samples were labeled with TMTpro reagents 126–129, the six EV
analyzed by SDS-PAGE and LC-MS/MS mass spectrometry at the UT samples were labeled with TMTpro reagents 130–132, and the six DM
Southwestern Proteomics core. samples were labeled with TMTpro reagents 133–136. After labeling,
the reaction was quenched with 5% hydroxylamine, and samples were
Proteomic interactome and cell surface analysis combined and dried in a SpeedVac. Samples were then cleaned using
Potential interacting proteins were identified by comparing protein an Oasis HLB microelution plate (Waters), dried, and reconstituted in
abundance ratios to reveal which proteins were enriched relative to the 50 μL of 2% acetonitrile, 0.1% TFA.
empty vector (EV) samples, with an arbitrary cutoff of at least 10-fold Peptide samples (1.5 ug) were injected onto an Orbitrap Eclipse
enrichment. The experiment was performed once, with 3 biological mass spectrometer coupled to an Ultimate 3000 RSLC-Nano liquid
replicates per group. Plasma membrane protein quantification was chromatography system with a 75 μm i.d., 75-cm long EasySpray col-
performed through TMT-based quantification (see below), comparing umn (Thermo). Peptides were eluted with a gradient from 1-28% buffer
protein abundance between WT cells and either EV cells or DM B over 180 min, followed by 28–45% buffer B over 25 min. Buffer A
expressing cells, focusing on at least 2-fold differences between contained 2% (v/v) ACN and 0.1% formic acid in water, and buffer B
groups. The experiment was performed once, with 6 biological repli- contained 80% (v/v) ACN, 10% (v/v) trifluoroethanol, and 0.1% formic
cates per group. After reduction with DTT and alkylation with iodoa- acid in water. The mass spectrometer operated in positive ion mode
cetamide (Sigma–Aldrich), samples were digested overnight with with a source voltage of 2.5 kV and an ion transfer tube temperature of
trypsin (Pierce). After solid-phase extraction cleanup with an Oasis HLB 300 °C. MS scans were acquired at 120,000 resolution in the Orbitrap
plate (Waters), digested samples were injected into an Orbitrap Fusion over a mass range of m/z = 400–1600, and top speed mode was used
Lumos mass spectrometer coupled to an Ultimate 3000 RSLC-Nano for SPS-MS3 analysis with a cycle time of 2.5 s. MS2 was performed
liquid chromatography system with a 75 μm i.d., 75-cm long EasySpray using collisionally-induced dissociation (CID) with a collision energy of
column (Thermo). Peptides were eluted with a gradient from 1 to 28% 35% for ions with charges 2–6. Dynamic exclusion was set for 25 s after
buffer B over 90 min. Buffer A contained 2% (v/v) ACN and 0.1% formic an ion was selected for fragmentation. Real-time search was performed
acid in water, and buffer B contained 80% (v/v) ACN, 10% (v/v) tri- using the reviewed human protein database from UniProt. We set
fluoroethanol, and 0.1% formic acid in water. The mass spectrometer cysteine carbamidomethylation and TMT pro 18plex modification of
operated in positive ion mode with a source voltage of 1.8–2.4 kV and lysine and peptide N-termini as fixed modifications, and methionine
an ion transfer tube temperature of 275 °C. MS scans were acquired at oxidation as a variable modification. We allowed two missed cleavages
120,000 resolution in the Orbitrap. Up to 10 MS/MS spectra were and up to 3 modifications per peptide. The top 10 fragments for MS/
obtained in the ion trap for each full spectrum acquired using higher- MS spectra corresponding to peptides identified by real-time search
energy collisional dissociation (HCD) for ions with charges 2–7. were selected for MS3 fragmentation using high-energy collisional
Dynamic exclusion was set for 25 s after an ion was selected for dissociation (HCD), with a collision energy of 65%. Raw MS data files
fragmentation. were analyzed using both the Sequest HT and Comet nodes within

Nature Communications | (2024)15:10193 14

Article https://doi.org/10.1038/s41467-024-54583-6

Proteome Discoverer v3.0 (Thermo), searching against the reviewed 0–400 mM NaCl developed over 40 mL) and cleaved using TEV pro-
human protein database from UniProt. Fragment and precursor tol- tease to remove the MBP tag. Cleaved VPS26C was polished by size
erances of 10 ppm and 0.6 Da were specified, and two missed clea- exclusion chromatography using a 24-mL Superdex Increase 75 col-
vages were allowed. The same modifications were used in the search as umn [10 mM HEPES pH 7.0, 100 mM NaCl, 5% (w/v) glycerol, and 1 mM
for the real-time search. The false-discovery rate (FDR) cutoff was 1% DTT]. MBP-CCDC22 VBD and MBP-CCDC93 VBD were purified as
for all peptides. Impurity corrections were used for the TMTpro described6. All chromatography steps were performed using Cytiva
reagents based on the documentation included with the TMTpro columns on an ÄKTATM Pure protein purification system. SNX17
reagent kit. P-values for protein fold-change across different sample C-terminal peptide, corresponding to amino acid sequence 451–470,
types were adjusted for multiple comparisons using a ASASDVHGNFAFEGIGDEDL, was synthesized from GenScript at ≥98%
Benjamini–Hochberg correction. purity. The lyophilized peptide was dissolved in 100 mM HEPES pH 7.0
buffer at a stock concentration of 40 mg/mL (19.5 mM), aliquoted in
Flow cytometry small volumes, and stored at −80 °C.
Cells were detached from plates using a cell scraper in 1× PBS and
centrifuged at 300 g for 5 min. The cells were resuspended in fresh PBS In vitro pull-down assays
and rinsed once with a repeat centrifugation step. For ITGB1 staining, GST pull-down experiments followed previous procedures48. Briefly,
cells were immediately processed and resuspended in FACS buffer bait (100-200 pmol of GST-tagged proteins) and prey (50-200 pmol
(PBS, 1% BSA) containing an anti-ITGB1 (CD29) antibody conjugated to for Retriever) were mixed with 20 µL of Glutathione Sepharose beads
FITC for 30 min on ice protected from light. After this, cells were rinsed (Cytiva) in 1 mL of binding buffer [10 mM HEPES pH 7, 50 mM NaCl, 5%
three times by centrifugation and resuspension in FACS buffer. Sam- (w/v) glycerol, 0.05% (w/v) Triton-X100, and 5 mM BME] at 4 °C for
ples were processed by the Flow Cytometry core at UT Southwestern 30 min. After three 1-mL washes with the binding buffer, bound pro-
using a Cytek Aurora instrument. Data analysis was performed using teins were eluted with 100 mM Tris pH 8.5, 50 mM NaCl, and 30 mM
FlowJo software. reduced glutathione and examined by SDS-PAGE. Where it is indicated,
200 pmol of MBP-CCDC22-CCDC93 VBD dimer, various amounts of
Recombinant protein purification SNX17 CT peptide, various amounts of KRIT1 cargo peptide, or 200
The Retriever complex or the VPS35L-VPS29 subcomplex was expres- nmol of PI(3)P diC4 (Echelon Biosciences) were added in pull-down
sed from Sf9 cells using the Bac-to-Bac system and purified through Ni- assays. For mutagenesis pull-downs, VPS35L intensity was quantified
NTA affinity, cation exchange, and anion exchange, and size exclusion and normalized to GST-CT intensity using ImageJ v2.3.0/1.53q. His6
chromatography essentially as previously described6. To improve pull-down assays were performed by mixing 50 pmol of His6-tagged
expression, VPS35L ORF was changed to a codon-optimized sequence Retriever and 2000 pmol of SNX17 with 20 µL of TALON beads
from Thermo Fisher. Typical yield was ~1 mg of purified Retriever from (Clontech Labs) in 1 mL of binding buffer [10 mM HEPES pH 7, 50 mM
3 liters of Sf9 culture. SNX17, isolated VPS26C, and SNX17-homologous NaCl, 5% (w/v) glycerol, 0.05% (w/v) Triton-X100, and 5 mM BME] at
CT tails were expressed and purified following procedures essentially 4 °C for 30 min. After three 1-mL washes with 20 mM Imidazole pH 7,
as previously described6. Briefly, proteins were expressed in BL21 10 mM HEPES pH 7, 50 mM NaCl, 5% (w/v) glycerol, 0.05% (w/v) Triton-
(DE3)T1R cells (Sigma) at 18 °C overnight after induction with 1 mM X100, and 5 mM BME, bound proteins were eluted with 200 mM Imi-
IPTG. GST-tagged SNX17 proteins were purified using Glutathione dazole pH7, 10 mM HEPES pH 7, 50 mM NaCl, 5% (w/v) glycerol, and
Sepharose beads (Cytiva) and eluted using 100 mM Tris pH 8.5, 50 mM 5 mM BME and examined by SDS-PAGE. Where it is indicated, 6000
NaCl, and 30 mM reduced glutathione. The resulting GST-SNX17 pro- pmol of SNX17 CT peptide, or 200 nmol of PI(3)P diC4 were added in
teins were further purified by anion exchange chromatography using a pull-down assays.
2-mL Source 15Q column (10 mM Tris pH 8.0 and 5 mM BME in a gra-
dient of 0–400 mM NaCl developed over 40 mL) and size exclusion In vitro equilibrium pull-down (EPD) assays
chromatography using a 24-mL Superdex Increase 200 column Equilibrium pull-down assays were performed as previously
[10 mM HEPES pH 7.0, 100 mM NaCl, 5% (w/v) glycerol, and 1 mM DTT]. described48. Briefly, 60 µL of Glutathione Sepharose beads (50% slurry
Untagged SNX17 full-length, PX-FERM, and FERM were expressed with equilibrated in a pull-down buffer [10 mM HEPES pH 7, 50 mM NaCl, 5%
an N-terminal GST tag, purified using Glutathione Sepharose beads as (w/v) glycerol, 0.05% (w/v) Triton-X100, and 5 mM BME] were mixed
described above. The GST tags were removed using TEV cleavage with 0.1 µM Retriever and various amounts of GST-tagged protein (up
overnight, further purified by a 2-mL Source 15S column (10 mM MES to 45 µM, stored in the same pull-down buffer) and brought to 100 µL
pH 6.0 and 5 mM BME in a gradient of 0–400 mM NaCl developed over final reaction volume using the pull-down buffer. The reactions were
60 mL), and dialyzed into 10 mM HEPES pH 7.0, 50 mM NaCl, 5% (w/v) allowed to mix for 30 min at 4 °C, and four reactions at a time were
glycerol, and 1 mM DTT for pull-down assays. SNX17 point mutants spun at 15 krpm for 15 seconds. The supernatant was immediately
were purified using Glutathione Sepharose beads as described above removed and examined by SDS-PAGE and Coomassie blue staining.
and then dialyzed into 10 mM HEPES pH 7.0, 100 mM NaCl, 5% (w/v) The VPS35L intensity was quantified using ImageJ v2.3.0/1.53q to cal-
glycerol, and 1 mM DTT for pull-down assays. SNX17-homologous C- culate the fractional occupancy. The data from all repeats were pooled
terminal tails were purified using Glutathione Sepharose beads as and globally fitted in DynaFit v4.08.187 using a single binding site
described above, further purified by a 2-mL Source 15Q column model49,50.
(10 mM Tris pH 8.0 and 5 mM BME in a gradient of 0–400 mM NaCl
developed over 40 mL), and finally dialyzed into 10 mM HEPES pH 7.0, Sample preparation for electron microscopy
50 mM NaCl, 5% (w/v) glycerol, and 1 mM DTT for pull-down assays. Purified Retriever and synthesized SNX17 CT peptide were mixed
Various SNX17-homologous C-terminal tails were purified using Glu- freshly at a final concentration of 1.4 µM Retriever and 6.5 mM peptide
tathione Sepharose beads and directly dialyzed into 10 mM HEPES pH in a final buffer containing 10 mM HEPES pH 7.0, 150 mM NaCl,
7.0, 50 mM NaCl, 5% (w/v) glycerol, and 1 mM DTT for pull-down 5% (w/v) glycerol, and 1 mM DTT. The mixture was centrifuged for at
assays. Isolated MBP-tagged VPS26C was purified using Amylose beads least 10 min at 4 °C before 3 µL was applied to a glow-discharged
(New England Biolabs) and eluted using 20 mM Tris pH 8.0, 200 mM Quantifoil 300-mesh R1.2/1.3 Copper grid (Micro Tools GmbH). After a
NaCl, 2% (w/v) maltose, and 5 mM BME. The protein was further pur- 10-second preincubation under 100% humidity at 4 °C, the grid was
ified by anion exchange chromatography using a 2-mL Source 15S blotted for 3.5 sec and plunge-frozen in liquid ethane using Vitrobot
column (10 mM HEPES pH 7.0 and 5 mM BME in a gradient of Mark IV (Thermo Fisher).

Nature Communications | (2024)15:10193 15

Article https://doi.org/10.1038/s41467-024-54583-6

Electron microscopy data acquisition (p < 0.05; * in Figure panels) was considered to imply statistical sig-
Sample grids were screened on a 200 kV Talos Artica or Glacios nificance. All imaging and co-precipitation experiments were per-
microscope (Thermo Fisher) at the Cryo Electron Microscopy Facility formed in two to four independent iterations. All in vitro pull-down
by Structural Biology Laboratory at University of Texas Southwestern assays were performed at least twice, unless otherwise indicated. Large
Medical Center (UTSW) or at the cryo-EM Facility at Iowa State Uni- scale proteomics were performed once, with key results confirmed
versity. The final cryo-EM data were acquired on a Titan Krios micro- using other methods.
scope (Thermo Fisher) at PNCC operated at 300 kV, with a post-
column energy filter (Gatan) and a K3 direct detection camera (Gatan) Reporting summary
in non-CDS mode. Movies were acquired using SerialEM v4.051 at a Further information on research design is available in the Nature
pixel size of 0.4133 Å in super-resolution counting mode, with an Portfolio Reporting Summary linked to this article.
accumulated total dose of 60 e−/Å2 over 60 frames. The defocus range
of the images was set between −0.9 to −2.5 μm. In total, 10,009 movies Data availability
were collected for data processing. Cryo-EM maps and models have been deposited in the EMDB, with
accession number EMD-43870, EMD-43871, EMD-43872, and EMD-
Electron microscopy data processing 43873, and PDB, with accession number 9AU7. AlphaFold Multimer-
Cryo-EM data were processed using cryoSPARC52 v4.4.1. Pre- derived models are available in ModelArchive, with the accession code
processing was performed in cryoSPARC Live, including motion cor- ma-swt4h [https://doi.org/10.5452/ma-swt4h]. Mass spectrometry
rection with a binning factor of 2, resulting in a pixel size of 0.8266 Å/ data have been deposited at the MassIVE repository, with accession
pixel and Contrast Transfer Function (CTF) estimation. Blob picking number MSV000095676 [https://doi.org/10.25345/C5M902F39] and
was used in cryoSPARC Live and 1,009,886 particles were selected MSV000094101 [https://doi.org/10.25345/C5BV7B635]. Source data
after initial 2D classification. After extensive 2D classification, 559,719 are provided for all uncropped western blots, Coomassie-blue gels,
particles were selected for ab initio 3D reconstruction and hetero- and all quantitative datasets presented in this paper. All information
geneous refinement (Supplementary Fig. 2). The best resolved 3D required to reanalyze the data reported here is publicly available. Any
class, containing 227,973 particles, underwent global and local CTF- additional data will be shared upon request. Source data are provided
refinement and a final non-uniform refinement, producing a full map with this paper.
with an overall resolution of 3.4 Å with a binned pixel size of 1.03 Å/
pixel (deposited in EMD-43873 with its associated half maps). To References
improve the map quality on both ends of the map, two masks were 1. Seaman, M. N., McCaffery, J. M. & Emr, S. D. A membrane coat
generated—one for VPS29-bound half of VPS35L, and another for complex essential for endosome-to-Golgi retrograde transport in
VPS26C-bound half. Signals outside the masks were subtracted and yeast. J. Cell Biol. 142, 665–681 (1998).
local refinement of the two regions were performed independently. 2. Haft, C. R. et al. Human orthologs of yeast vacuolar protein sorting
The resulted maps were deposited with their half maps at EMD-43871 proteins Vps26, 29, and 35: assembly into multimeric complexes.
and EMD-43870, respectively (Supplementary Fig. 2d–f). Dee- Mol. Biol. Cell 11, 4105–4116 (2000).
pEMhancer v20220530_cu1133 was then used with the unfiltered half 3. Edgar, A. J. & Polak, J. M. Human homologues of yeast vacuolar
maps to generate sharpened maps of the two locally refined maps, protein sorting 29 and 35. Biochem. Biophys. Res. Commun. 277,
respectively, and a composite of the two was generated in UCSF Chi- 622–630 (2000).
meraX v1.7.1 by the vop maximum command53. This composite map 4. McNally, K. E. et al. Retriever is a multiprotein complex for retromer-
(EMD-43872/PDB-9AU7) was used for modeling and shown in Fig. 3 independent endosomal cargo recycling. Nat. Cell Biol. 19, 1214–1225
and Supplementary Fig 3c, d. All reported resolutions followed the (2017).
gold-standard Fourier shell correlation (FSC) using the 0.143 5. Healy, M. D. et al. Structure of the endosomal Commander complex
criterion54. linked to Ritscher-Schinzel syndrome. Cell 186, 2219–2237.e29 (2023).
6. Boesch, D. J. et al. Structural organization of the retriever-CCC
Atomic model building endosomal recycling complex. Nat. Struct. Mol. Biol. 31,
A model of the Retreiver-SNX17 complex predicted by AlphaFold 910–924 (2023).
Multimer v2.3.1 was used as the initial model55 for model building. 7. Laulumaa, S., Kumpula, E. P., Huiskonen, J. T. & Varjosalo, M.
Model was first docked and fitted into the sharpened composite map Structure and interactions of the endogenous human Commander
using ISOLDE56 in ChimeraX, followed by iterations of real-space complex. Nat. Struct. Mol. Biol. 31, 925–938 (2024).
refinement in Phenix v1.2157 with reference model and secondary 8. Singla, A. et al. Endosomal PI(3)P regulation by the COMMD/
structure restraints and manual building in COOT v0.9.8.858,59. Model CCDC22/CCDC93 (CCC) complex controls membrane protein
geometries were assessed by MolProbity in Phenix60 (http:// recycling. Nat. Commun. 10, 4271 (2019).
molprobity.biochem.duke.edu/), and the PDB Validation server61 9. Bartuzi, P. et al. CCC- and WASH-mediated endosomal sorting of
(www.wwpdb.org). Figures were generated using PyMOL v2.5.4 or LDLR is required for normal clearance of circulating LDL. Nat.
ChimeraX v1.7.162. Commun. 7, 10961 (2016).
10. Steinberg, F. et al. A global analysis of SNX27-retromer assembly
AlphaFold prediction and analysis and cargo specificity reveals a function in glucose and metal ion
AlphaFold version 2.3.1 (https://github.com/deepmind/alphafold) was transport. Nat. Cell Biol. 15, 461–471 (2013).
installed on local NVidia A100 80GB GPU computers at Iowa State 11. Harrison, M. S. et al. A mechanism for retromer endosomal coat
University ResearchIT or High-Performance Computing for AlphaFold complex assembly with cargo. Proc. Natl Acad. Sci. USA 111,
Multimer prediction using standard AlphaFold procedures55,58 as pre- 267–272 (2014).
viously described6. 12. Yong, X. et al. SNX27-FERM-SNX1 complex structure rationalizes
divergent trafficking pathways by SNX17 and SNX27. Proc. Natl
Reproducibility and statistical analysis Acad. Sci. USA 118, e2105510118 (2021).
To assess statistical significance, one-way ANOVA with Dunnett’s post- 13. Simonetti, B. et al. SNX27-Retromer directly binds ESCPE-1 to
hoc test was applied to compare multiple groups with one control transfer cargo proteins during endosomal recycling. PLoS Biol. 20,
group, using Prism v9.5.1 (GraphPad). An error probability below 5% e3001601 (2022).

Nature Communications | (2024)15:10193 16

Article https://doi.org/10.1038/s41467-024-54583-6

14. Fjorback, A. W. et al. Retromer binds the FANSHY sorting motif in 37. Yong, X., Mao, L., Seaman, M. N. J. & Jia, D. An evolving under-
SorLA to regulate amyloid precursor protein sorting and proces- standing of sorting signals for endosomal retrieval. iScience 25,
sing. J. Neurosci. 32, 1467–1480 (2012). 104254 (2022).
15. Temkin, P. et al. SNX27 mediates retromer tubule entry and 38. Butkovic, R. et al. Mechanism and regulation of cargo entry into the
endosome-to-plasma membrane trafficking of signalling receptors. Commander endosomal recycling pathway. Nat. Commun. 15,
Nat. Cell Biol. 13, 715–721 (2011). 7180 (2024).
16. Gomez, T. S. & Billadeau, D. D. A FAM21-containing WASH complex 39. Purushothaman, L. K. & Ungermann, C. Cargo induces retromer-
regulates retromer-dependent sorting. Dev. Cell 17, 699–711 (2009). mediated membrane remodeling on membranes. Mol. Biol. Cell 29,
17. Gomez, T. S., Gorman, J. A., de Narvajas, A. A., Koenig, A. O. & 2709–2719 (2018).
Billadeau, D. D. Trafficking defects in WASH-knockout fibroblasts 40. Lopez-Robles, C. et al. Architecture of the ESCPE-1 membrane coat.
originate from collapsed endosomal and lysosomal networks. Mol. Nat. Struct. Mol. Biol. 30, 958–969 (2023).
Biol. Cell 23, 3215–3228 (2012). 41. Hornbeck, P. V. et al. PhosphoSitePlus, 2014: mutations, PTMs and
18. Jia, D., Gomez, T. S., Billadeau, D. D. & Rosen, M. K. Multiple repeat recalibrations. Nucleic Acids Res. 43, D512-20 (2015).
elements within the FAM21 tail link the WASH actin regulatory 42. Schaffer, B. E. et al. Identification of AMPK Phosphorylation Sites
complex to the retromer. Mol. Biol. Cell 23, 2352–2361 (2012). Reveals a Network of Proteins Involved in Cell Invasion and
19. Derivery, E. et al. The Arp2/3 activator WASH controls the fission of Facilitates Large-Scale Substrate Prediction. Cell Metab. 22,
endosomes through a large multiprotein complex. Dev. Cell 17, 907–921 (2015).
712–723 (2009). 43. Healy, M. D. et al. Proteomic identification and structural basis for
20. Wan, C. et al. Panorama of ancient metazoan macromolecular the interaction between sorting nexin SNX17 and PDLIM family
complexes. Nature 525, 339–344 (2015). proteins. Structure 30, 1590–1602.e6 (2022).
21. Phillips-Krawczak, C. A. et al. COMMD1 is linked to the WASH 44. Starokadomskyy, P. et al. CCDC22 deficiency in humans blunts
complex and regulates endosomal trafficking of the copper trans- activation of proinflammatory NF-kB signaling. J. Clin. Invest 123,
porter ATP7A. Mol. Biol. Cell 26, 91–103 (2015). 2244–2256 (2013).
22. Mallam, A. L. & Marcotte, E. M. Systems-wide Studies Uncover 45. Osborne, D. G., Piotrowski, J. T., Dick, C. J., Zhang, J. S. & Billadeau,
Commander, a Multiprotein Complex Essential to Human Devel- D. D. SNX17 affects T cell activation by regulating TCR and integrin
opment. Cell Syst. 4, 483–494 (2017). recycling. J. Immunol. 194, 4555–4566 (2015).
23. Hanley, S. E. & Cooper, K. F. Sorting Nexins in Protein Homeostasis. 46. Li, H. et al. Endosomal sorting of Notch receptors through COMMD9
Cells 10, 17 (2020). dependent pathways modulates Notch signaling. J. Cell Biol. 211,
24. Leneva, N., Kovtun, O., Morado, D. R., Briggs, J. A. G. & Owen, D. J. 605–617 (2015).
Architecture and mechanism of metazoan retromer:SNX3 tubular 47. Mao, X. et al. GCN5 is a required cofactor for a ubiquitin ligase that
coat assembly. Sci Adv 7, eabf8598 (2021). targets NF-kB/RelA. Genes Dev. 23, 849–861 (2009).
25. Kovtun, O. et al. Structure of the membrane-assembled retromer 48. Chen, B. et al. Rac1 GTPase activates the WAVE regulatory complex
coat determined by cryo-electron tomography. Nature 561, through two distinct binding sites. Elife 6 e29795 (2017).
561–564 (2018). 49. Kuzmic, P. Program DYNAFIT for the analysis of enzyme kinetic data:
26. Lucas, M. et al. Structural Mechanism for Cargo Recognition by the application to HIV proteinase. Anal. Biochem. 237, 260–273 (1996).
Retromer Complex. Cell 167, 1623–1635.e14 (2016). 50. Kuzmic, P. DynaFit–a software package for enzymology. Methods
27. Simonetti, B. et al. Molecular identification of a BAR domain- Enzymol. 467, 247–280 (2009).
containing coat complex for endosomal recycling of transmem- 51. Mastronarde, D. N. Automated electron microscope tomography
brane proteins. Nat. Cell Biol. 21, 1219–1233 (2019). using robust prediction of specimen movements. J. Struct. Biol. 152,
28. Yong, X. et al. Mechanism of cargo recognition by retromer-linked 36–51 (2005).
SNX-BAR proteins. PLoS Biol. 18, e3000631 (2020). 52. Punjani, A., Rubinstein, J. L., Fleet, D. J. & Brubaker, M. A. cryoS-
29. Gallon, M. et al. A unique PDZ domain and arrestin-like fold PARC: algorithms for rapid unsupervised cryo-EM structure deter-
interaction reveals mechanistic details of endocytic recycling mination. Nat. Methods 14, 290–296 (2017).
by SNX27-retromer. Proc. Natl Acad. Sci. USA 111, 53. Pettersen, E. F. et al. UCSF Chimera–a visualization system for
E3604–E3613 (2014). exploratory research and analysis. J. Comput Chem. 25, 1605–1612
30. Ghai, R. et al. Structural basis for endosomal trafficking of diverse (2004).
transmembrane cargos by PX-FERM proteins. Proc. Natl Acad. Sci. 54. Henderson, R. et al. Outcome of the first electron microscopy
USA 110, E643–E652 (2013). validation task force meeting. Structure 20, 205–214 (2012).
31. Stiegler, A. L., Zhang, R., Liu, W. & Boggon, T. J. Structural deter- 55. Evans, R. et al. Protein complex prediction with AlphaFold-
minants for binding of sorting nexin 17 (SNX17) to the cytoplasmic Multimer. BioRxiv https://doi.org/10.1101/2021.10.04.
adaptor protein Krev interaction trapped 1 (KRIT1). J. Biol. Chem. 463034 (2022).
289, 25362–25373 (2014). 56. Croll, T. I. ISOLDE: a physically realistic environment for model
32. Abramson, J. et al. Accurate structure prediction of biomolecular building into low-resolution electron-density maps. Acta Crystal-
interactions with AlphaFold 3. Nature 630, 493–500 (2024). logr. D. Struct. Biol. 74, 519–530 (2018).
33. Sanchez-Garcia, R. et al. DeepEMhancer: a deep learning solution 57. Liebschner, D. et al. Macromolecular structure determination using
for cryo-EM volume post-processing. Commun. Biol. 4, 874 (2021). X-rays, neutrons and electrons: recent developments in Phenix.
34. Kendall, A. K., Chandra, M., Xie, B., Wan, W. & Jackson, L. P. Acta Crystallogr. D. Struct. Biol. 75, 861–877 (2019).
Improved mammalian retromer cryo-EM structures reveal a new 58. Jumper, J. et al. Highly accurate protein structure prediction with
assembly interface. J. Biol. Chem. 298, 102523 (2022). AlphaFold. Nature 596, 583–589 (2021).
35. Steinberg, F., Heesom, K. J., Bass, M. D. & Cullen, P. J. SNX17 pro- 59. Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. Features and
tects integrins from degradation by sorting between lysosomal and development of Coot. Acta Crystallogr. D. Biol. Crystallogr 66,
recycling pathways. J. Cell Biol. 197, 219–230 (2012). 486–501 (2010).
36. Stockinger, W. et al. The PX-domain protein SNX17 interacts with 60. Chen, V. B. et al. MolProbity: all-atom structure validation for
members of the LDL receptor family and modulates endocytosis of macromolecular crystallography. Acta Crystallogr. D. Biol. Crystal-
the LDL receptor. EMBO J. 21, 4259–4267 (2002). logr. 66, 12–21 (2010).

Nature Communications | (2024)15:10193 17

Article https://doi.org/10.1038/s41467-024-54583-6

61. Burley, S. K. et al. Protein Data Bank (PDB): The Single Global Mac- experiments related to LRMDA. B.C., H.Y.J.F., Z.C., and D.J.B. analyzed
romolecular Structure Archive. Methods Mol. Biol. 1607, 627–641 structures. E.B. and B.C. drafted the manuscript and prepared the fig-
(2017). ures with assistance from all other authors.
62. Pettersen, E. F. et al. UCSF ChimeraX: Structure visualization for
researchers, educators, and developers. Protein Sci. 30, 70–82 Competing interests
(2021). The authors declare no competing interests.
63. Ashkenazy, H. et al. ConSurf 2016: an improved methodology to
estimate and visualize evolutionary conservation in macro- Additional information
molecules. Nucleic Acids Res. 44, W344–W350 (2016). Supplementary information The online version contains
supplementary material available at
Acknowledgements https://doi.org/10.1038/s41467-024-54583-6.
GST-SNX17 FL construct was a gift from Titus Boggon at Yale University
School of Medicine. We thank the Research IT at Iowa State University for Correspondence and requests for materials should be addressed to
hardware resources, installation of AlphaFold, and ongoing computa- Emre E. Turer, Ezra Burstein or Baoyu Chen.
tional & diagnostic support. We also thank the Proteomics core and the
Flow Cytometry core at UT Southwestern. Electron Microscopy data Peer review information Nature Communications thanks Oleksiy Kov-
were collected in collaboration with the Structural Biology Laboratory tun, Momo Sae-Lee and the other, anonymous, reviewer(s) for their
using the Cryo Electron Microscopy Facility at UT Southwestern Medical contribution to the peer review of this work. A peer review file is
Center (partially supported by grant RP220582 from the Cancer Pre- available.
vention & Research Institute of Texas [CPRIT] for cryo-EM studies) and
the Iowa State University Cryo-EM Facility (supported by the Roy J. Reprints and permissions information is available at
Carver Structural Initiative). We also thank Omar Davulcu at the Pacific http://www.nature.com/reprints
Northwest Cryo-EM Center (PNCC, supported by NIH grant
U24GM129547) for his help with data collection performed at the PNCC Publisher’s note Springer Nature remains neutral with regard to
at OHSU and accessed through EMSL (grid.436923.9), a DOE Office of jurisdictional claims in published maps and institutional affiliations.
Science User Facility sponsored by the Office of Biological and Envir-
onmental Research. Research was supported by funding from the Open Access This article is licensed under a Creative Commons
National Institutes of Health (R35 GM128786 to B.C., R01 DK107733 to Attribution-NonCommercial-NoDerivatives 4.0 International License,
E.B. and D.D.B, and R01 DK133229, R01 DK119360 to E.E.T.), the Welch which permits any non-commercial use, sharing, distribution and
Foundation (I-1944 to X.B.), and the National Science Foundation (CDF reproduction in any medium or format, as long as you give appropriate
2047640 to B.C. and D.D.B.). credit to the original author(s) and the source, provide a link to the
Creative Commons licence, and indicate if you modified the licensed
Author contributions material. You do not have permission under this licence to share adapted
B.C. and E.B. conceived the project. E.B. oversaw cell biological and material derived from this article or parts of it. The images or other third
proteomic experiments performed by A.S. B.C. oversaw protein pur- party material in this article are included in the article’s Creative
ification, biochemical experiments, and AlphaFold predictions per- Commons licence, unless indicated otherwise in a credit line to the
formed by D.J.B. with the help from A.S.E., C.N., and D.A.K. Z.C. oversaw material. If material is not included in the article’s Creative Commons
cryo-EM grid preparation, data collection, single particle reconstruction, licence and your intended use is not permitted by statutory regulation or
and atomic-model building by B.C., Y.H., and H.Y.J.F. X.B. helped with exceeds the permitted use, you will need to obtain permission directly
cryo-EM data processing and supported cryo-EM grid preparation and from the copyright holder. To view a copy of this licence, visit http://
screening performed by B.C. at UTSW. Es.B. and A.L. performed pro- creativecommons.org/licenses/by-nc-nd/4.0/.
teomic data analyses. P.J. supervised cryo-EM grid preparation and data
collection performed by D.J.B. at Iowa State. D.D.B. helped with cellular © The Author(s) 2024
experiments and data interpretation. R.S. and E.E.T. performed cellular

Nature Communications | (2024)15:10193 18