Structure of the Sec14 domain of Kalirin reveals a distinct class of lipid-binding module in RhoGEFs

Li, Yunfeng; Pustovalova, Yulia; Doukov, Tzanko I.; Hoch, Jeffrey C.; Mains, Richard E.; Eipper, Betty A.; Hao, Bing

doi:10.1038/s41467-022-35678-4

Download PDF

Article
Open access
Published: 06 January 2023

Structure of the Sec14 domain of Kalirin reveals a distinct class of lipid-binding module in RhoGEFs

Nature Communications volume 14, Article number: 96 (2023) Cite this article

2261 Accesses
1 Altmetric
Metrics details

Subjects

Abstract

Gated entry of lipophilic ligands into the enclosed hydrophobic pocket in stand-alone Sec14 domain proteins often links lipid metabolism to membrane trafficking. Similar domains occur in multidomain mammalian proteins that activate small GTPases and regulate actin dynamics. The neuronal RhoGEF Kalirin, a central regulator of cytoskeletal dynamics, contains a Sec14 domain (Kal^bSec14) followed by multiple spectrin-like repeats and catalytic domains. Previous studies demonstrated that Kalirin lacking its Sec14 domain fails to maintain cell morphology or dendritic spine length, yet whether and how Kal^bSec14 interacts with lipids remain unknown. Here, we report the structural and biochemical characterization of Kal^bSec14. Kal^bSec14 adopts a closed conformation, sealing off the canonical ligand entry site, and instead employs a surface groove to bind a limited set of lysophospholipids. The low-affinity interactions of Kal^bSec14 with lysolipids are expected to serve as a general model for the regulation of Rho signaling by other Sec14-containing Rho activators.

The structural basis of PTEN regulation by multi-site phosphorylation

Article 08 October 2021

Palmitoylation targets the calcineurin phosphatase to the phosphatidylinositol 4-kinase complex at the plasma membrane

Article Open access 18 October 2021

Cryo-EM structures of the TTYH family reveal a novel architecture for lipid interactions

Article Open access 12 August 2021

Introduction

The Sec14 superfamily is defined by the presence of a CRAL_TRIO domain (Pfam, PF00650), best known for its ability to bind small lipophilic molecules^1,2,3,4,5. The functions of family members that contain only a Sec14 domain range from coordinating lipid metabolism and vesicular trafficking in yeast to providing 11-cis-retinal to retinal pigment epithelial cells and Mueller glia for delivery to photoreceptor cells^3,6,7. A third of the ~30 human Sec14 superfamily members also contain enzymatic domains that control activities of the Rho family small GTP-binding proteins^8,9.

Based on their sequences and structural features, Sec14 superfamily members fall into two sub-families, CRAL_TRIO and BCH¹. Structural, biochemical and functional studies of yeast Sec14p, the prototypical CRAL_TRIO sub-family member, have guided subsequent studies^10,11,12. Despite relatively limited sequence identity, members of the CRAL_TRIO sub-family for which structural information is available adopt a characteristic two-lobed globular structure. The CRAL_TRIO motif harbors a large hydrophobic pocket in the center of its α/β fold with a mobile helical gate controlling access of lipophilic ligands to the internal pocket^3,11. A small N-terminal lobe, referred to as a CRAL_TRIO_N domain (Pfam, PF03765), closes off one end of the lipid-binding pocket with an all-helical tripod-like fold that can interact with bound lipid head groups. The CRAL_TRIO_N domain also participates in subcellular targeting of Sec14 domain proteins^13,14. The only BCH sub-family member for which structural data are available is Schizosaccharomyces pombe p50RhoGAP, in which intertwined asymmetric monomers contribute to a uniquely different lipid/RhoA-binding pocket¹⁵.

The Sec14-like domains of five mammalian Rho GDP/GTP exchange factors (RhoGEFs) share limited sequence identity with both BCH and CRAL_TRIO sub-family members^1,3,8. SESTD1, a synaptic regulator of Rho GTPases, falls into this sub-group, but lacks a RhoGEF domain¹⁶. Each of these six proteins lacks a CRAL_TRIO_N domain but contains a putative CRAL_TRIO domain followed by multiple spectrin-like repeats (SRs), a common structural element consisting of a three-helix bundle.

The N-terminal CRAL_TRIO domain of Kalirin is followed by nine SRs and a variable number of RhoGEF (consisting of DH and PH domains), SH3 and protein kinase domains¹⁷ (Fig. 1a). The ability of Kalirin to participate in downstream signaling pathways initiated by multiple receptor tyrosine kinases, G-protein coupled receptors, cell/cell adhesion complexes and ligand-gated ion channels identify Kalirin as a signaling hub in the nervous system^17,18,19, heart¹⁷, bone²⁰ and skeletal muscle²⁰. Mutations in KALRN and changes in its expression have been associated with schizophrenia, ischemic stroke, Alzheimer’s disease and vascular disease^17,18,19. Kal7, the major isoform expressed in the adult brain, supports the formation and function of excitatory synapses¹⁹. Kal7 lacking its Sec14 domain still stimulates spine formation, but the spines produced are short; consistent with this, expression of the Sec14 domain alone increases spine length but not spine number²¹. The mechanisms through which the Kalirin Sec14 domain interacts with specific membranes, lipids or proteins to affect spine length and synaptic transmission remain elusive.

**Fig. 1: Crystal structure of Kal^bSec14.**

The tissue-specific and developmentally regulated use of alternate initiation exons (Ex1) places different peptides immediately upstream of the Kalirin CRAL_TRIO domain. Ex1B encodes a hydrophilic negatively charged front peptide, producing bKal7, while Ex1C generates a positively charged front peptide that forms an amphipathic helix, producing cKal7²⁰. Attachment of the Ex1C front peptide to the CRAL_TRIO domain of Kalirin enables its direct interaction with liposomes in a phosphoinositide-dependent manner; attaching it to EGFP localizes EGFP to the Golgi region²⁰. When preceded by the Ex1B front peptide, the ability of the CRAL_TRIO domain to interact with phosphoinositide-containing liposomes is eliminated²⁰. The ability of these front peptides to alter Sec14 domain-mediated membrane interactions is essential to their function.

In this report, we present the crystal structure of the CRAL_TRIO domain of Kalirin preceded by the Ex1B front peptide (Kal^bSec14), the only Sec14 structure for a RhoGEF. Kal^bSec14 alone adopts a Sec14p homologous CRAL_TRIO domain fold, but with a closed conformation impeding access to the canonical ligand entry site. Our biochemical, biophysical and mutagenesis studies demonstrate that the amino acid residues lining a surface groove on Kal^bSec14 are directly involved in the binding of specific lysophosphatidylcholine (LPC) lipids and that the Ex1B peptide enhances lipid binding affinity. Both secondary-structure predictions and structure-based alignments identify Kal^bSec14 as the prototype of a Sec14 fold sub-class, with a unique ligand-binding site and ligand specificity (hereafter called SecSR sub-class). Coupled with interactions mediated by their other domains, linear peptide motifs and numerous phosphorylation sites, our findings suggest that the low-affinity interactions of SecSR domains with membrane lipids play an important role in the ability of Kalirin and other SecSR-containing RhoGEFs to participate in rapid regulation of their subcellular localization and signaling functions.

Results

Production and structure determination

The CRAL-TRIO domain of Kalirin is sandwiched between alternative start sites in Ex1A, B, C, and D of the KALRN gene and the nine sequential SRs that follow its C-terminus (Fig. 1a). To optimize crystallization, we designed multiple CRAL-TRIO-containing constructs of Kalirin and examined their protein expression level as well as crystallization propensity. We obtained diffraction-quality crystals of Kal^bSec14, a construct consisting of the Ex1B front peptide (residues 1–25), the putative CRAL-TRIO domain (residues 26–182) and a part of the predicted A-helix (residues 183–192) of the first SR (Fig. 1a). Kal^bSec14 was produced in Escherichia coli and the recombinant protein is monomeric as determined by size exclusion chromatography. The X-ray crystal structure of Kal^bSec14 was determined at 1.89-Å resolution by single-wavelength anomalous dispersion using a selenomethionine derivative and refined to an R factor of 19.5% and an R_free value of 21.8% (Supplementary Table 1). The asymmetric unit contains eight nearly identical protein monomers, with an overall root mean square deviation (rmsd) of 0.8 Å for 175 C_α atoms (Supplementary Fig. 1). The refined model for monomer B in the PDB file, which is complete except for the intrinsically disordered first 17 residues of the Ex1B front peptide, was chosen for detailed structural analysis.

Kal^bSec14 structure

Kal^bSec14 adopts an αβα sandwich fold similar to the prototypical CRAL-TRIO domain in yeast Sec14p¹⁰, but with unique features (Fig. 1b, c, and Supplementary Fig. 2a). The central β sheet is composed of five strands (S1–S5), with S1 the only antiparallel strand, and is flanked by six α helices (H1–H6) and a 3₁₀ helix (H7) on one side and by two α helices, H8 and H9, on the other. An N-terminal 10-amino-acid extension (residues 18–27) upstream of the H1 helix projects ~18 Å away from the αβα core, confirming the notion that the Sec14 domain of Kalirin lacks the CRAL_TRIO_N domain found in many stand-alone Sec14 proteins. Interestingly, Ile186 and Leu190 of helix H9 engage in van der Waals interactions with Ala34, Val37, Leu38, Ile40 and Leu41 of helices H1 and H2. In addition, the bulky side chains of Tyr180, His182 and Trp185 on the H9 helix stack with the benzene ring of Phe47 from strand S1. As such, the interaction of the C-terminal H9 helix, which is part of the first SR, with the central αβα fold buries 441 Å² of solvent-accessible surface area. In the absence of the H9 helix, diffraction-quality crystals of the Kalirin Sec14 domain were not obtained, suggesting that this interaction stabilizes the core fold of the Kalirin CRAL_TRIO domain.

Despite their low degrees of sequence homology, the CRAL-TRIO domains of Kal^bSec14 and several CRAL_TRIO sub-family proteins can be superimposed with a C_α rmsd of <5.0 Å (Supplementary Fig. 2b). For example, Kal^bSec14 and yeast Sec14p¹⁰ share only 16.1% sequence identity but display a 4.0-Å rmsd for 149 equivalent C_α atoms (Supplementary Fig. 2b, c). A comparison of Sec14p paralogs from yeast and orthologs from various fungal species demonstrated conservation of surface charge in a dipole pattern, with conspicuous electropositive regions localized near the ligand binding site and gating helix, thereby facilitating potential interactions of each protein with an electronegative membrane surface¹². Consistent with this notion, Kal^bSec14 exhibits an asymmetric electrostatic potential distribution on its solvent-exposed surface, with one side largely electropositive and the other electronegative (Fig. 1d). Therefore, the CRAL-TRIO domain of Kal^bSec14 is a structurally conserved element that may govern membrane binding specificity.

A closed conformation and unique surface groove

Previous crystallographic studies revealed the presence of a phospholipid/detergent-binding pocket with a volume of ~2,000 Å³ in an open yeast Sec14p conformer¹⁰ (Supplementary Fig. 3a). This hydrophobic pocket, a hallmark of Sec14 proteins examined to date, is required for Sec14p to coordinate regulation of lipid metabolism, lipid signaling and vesicular trafficking¹¹. Access to this binding pocket is controlled by a helical gate that is flipped open in Sec14 domain apo-structures and closed in ligand-bound structures¹¹. Unexpectedly, Kal^bSec14 has only several small, discontinuous hydrophobic pockets (~100 Å³ each) in the center of its αβα fold, none of which is large enough to accommodate even a single phospholipid molecule (Fig. 2a, b). A string of hydrophobic residues from strands S1–S4, helices H3 and H6, as well as the H6–S4 and H7–S5 loop regions line these pockets and render the entire region extremely hydrophobic. Moreover, the 3₁₀ helix H7, along with an extended loop, is positioned across any potential openings into these internal pockets, mimicking the closed conformation of the so-called mobile gate as often observed for holo-forms of other Sec14 proteins (Supplementary Fig. 3b). Helix H7 makes extensive van der Waals contacts with helix H5 and strands S3 and S4, further stabilizing this conformation (Fig. 2a). We therefore conclude that Kal^bSec14 exists in a closed conformation even in the absence of ligand.

**Fig. 2: Kal^bSec14 adopts a closed conformation containing a surface groove.**

A number of computer programs^22,23,24 were used to analyze potential ligand-binding sites in Kal^bSec14; none detected any interior pocket with a volume comparable to the pockets found in other Sec14 family proteins (Fig. 2b). Nevertheless, an elongated groove is present on the surface of Kal^bSec14, directly above the S2 and S3 strands. The groove has a volume of ~1200 Å³ and is ~20 Å long, 8 Å wide and 8 Å deep (Fig. 2c and Supplementary Fig. 4a). Segments of two strands (Phe62–Ala64 of S2 and Ile100–Met102 of S3) constitute the floor of the groove, while the S2–H3 loop (Arg65–Arg72) together with a segment of the H3 helix (Asn73–Lys78) frames one side of the groove, and the other side is flanked by the S3–H5 loop (Arg103–Ser105) and segments of the H5 and H6 helices (Lys106–Lys111 and Lys113–Leu117, respectively). Interestingly, this groove displays a dipole-like pattern of surface charge distribution, with the S2–H3 and S3–H5 loop regions at one end being largely electronegative and the H3 and H6 helices at the other being electropositive (Fig. 2d).

A comparison of the distribution of internal and exposed pockets of Kal^bSec14 and other CRAL_TRIO sub-family proteins reveals that canonical CRAL_TRIO domains lack a surface groove-like feature. Each of the six structures determined for CRAL_TRIO sub-family proteins encloses a large internal hydrophobic pocket extending from the gating helix to cover most part of the central β sheet, to which most ligands bind (Supplementary Fig. 4b). None of these CRAL_TRIO domains possess an exposed cavity in a location similar to the Kal^bSec14 surface groove. In α-TTP complexed with both α-tocopherol (α-Toc) and phosphatidylinositol 4,5-bisphosphate (PIP2), α-Toc is bound deep in its hydrophobic core, while PIP2 is found in close proximity to the gating helix²⁵ (Supplementary Fig. 4b). Although the binding site for α-Toc is right below the site equivalent to the surface groove in Kal^bSec14, the α-Toc site is entirely sequestered from bulk solvent. Among the five Sec14 sub-family proteins with a CRAL_TRIO_N domain preceding their CRAL_TRIO domain (Sec14p, CARALBP, Sfh1, Sfh5 and α-TTP), each CRAL_TRIO_N domain is distant from the location of the presumed surface groove but adjacent to the bottom of the gating helix, allowing it to interact with the head group of the bound ligand (Supplementary Fig. 4b).

Expanding on the finding of the Kal^bSec14 surface groove, we explored its ability to interact with water-soluble phosphatidycholine (PC) mimetics (Supplementary Table 2). These PC mimetics, also called phosphocholine detergents, were selected for the study because PC is the most abundant phospholipid in mammalian membranes²⁶. We were unable to obtain cocrystals of Kal^bSec14–phosphocholine complexes. Their interactions were therefore characterized using solution nuclear magnetic resonance (NMR) spectroscopy, which allows detection and characterization of weak, transient interactions²⁷.

Binding of PC mimetics to Kal^bSec14 via its surface groove

We used NMR chemical shift perturbation (CSP) measurements of Kal^bSec14 in the absence and presence of PC mimetics to determine ligand-interacting residues and ligand binding affinities. As expected for a well-folded structural unit, well-dispersed signals with adequate intensity were detected in the acquired two-dimensional (2D) ¹H-¹⁵N heteronuclear single quantum coherence (HSQC) spectrum of Kal^bSec14 in native-like conditions (Supplementary Fig. 5a). 178 out of 187 expected backbone amide resonances were observed. Because NMR chemical shift mapping experiments are only capable of defining protein-ligand interactions when highly complete specific backbone assignments are available, we next sought to determine sequence-specific backbone assignments for Kal^bSec14. However, a high degree of spectral crowding in the N-terminal disordered region of Kal^bSec14 (between 8.0 and 8.5 ppm on the ¹H axis) made it difficult to unambiguously identify individual cross-peaks. To simplify NMR data collection and subsequent data analysis, a shorter construct referred to as Kal^Sec14 was generated. Kal^Sec14 (residues 25–192) consists of the CRAL-TRIO domain and the A-helix of the first SR but lacks the Ex1B front peptide. The 2D ¹H-¹⁵N-HSQC spectrum of Kal^Sec14 features an extremely well-dispersed set of resonances (Supplementary Fig. 5b). Importantly, the vast majority of the amide resonance peaks of Kal^bSec14 and Kal^Sec14 are well matched, facilitating the confident transfer of backbone assignments (Supplementary Fig. 5c).

A total of 92.7% of the backbone ¹H and ¹⁵N resonances of the 160 non-proline residues, and 92% of the ¹³C_α, 91% of the C_O and 88% of the expected ¹³C_β resonances were unambiguously assigned for Kal^Sec14 based on a standard set of triple resonance spectra (Fig. 3a and Supplementary Fig. 5d). The absence of several peaks and the presence of weak peaks prevented a higher percentage of resonance assignments. The backbone amide residues that could not be assigned are localized in two clusters, which include residues in the S2 strand and the S2–H3 loop and in the H7–S5 loop region. Importantly, the secondary-structure elements of Kal^Sec14 predicted based on the backbone and C^β chemical shifts are in excellent agreement with the secondary structures observed in Kal^bSec14, as determined by X-ray crystallography, indicating that Kal^Sec14 adopts a fold similar to that of Kal^bSec14 (Fig. 3b).

**Fig. 3: Sequence-specific NMR backbone assignments for Kal^Sec14.**

We used CSP mapping to probe interactions between ¹⁵N-labeled Kal^bSec14 and PC mimetics of varying lengths (Supplementary Table 2). Overlays of the ¹H-¹⁵N HSQC spectra of Kal^bSec14 display gradual shifts in the amide resonance positions for a number of peaks upon titration of increasing amounts of tetradecylphosphocholine (FC14) or dodecylphosphocholine (FC12), indicative of specific interactions between Kal^bSec14 and FC14/FC12 and fast exchange between the free and bound states on the NMR timescale (Fig. 4a, b, and Supplementary Fig. 6). In striking contrast, no peak shifts are observed when decylphosphocholine (FC10) or hexadecylphosphocholine (FC16) is added. Per-residue CSPs are calculated as the distance between the free and FC14-bound peaks (Fig. 4c). Of 160 assigned peaks, 19 experience CSPs greater than two standard deviations (2σ) of the average chemical shift distribution. Binding of FC12 to Kal^bSec14 exhibits large CSPs to a similar set of residues as those with FC14. The fitting of the NMR titration data obtained with a two-state binding model results in dissociation constants (K_Ds) of 17.1 ± 4.9 μM and 245 ± 36 μM for FC14 and FC12, respectively, revealing a 14-fold higher binding affinity for FC14 (Fig. 4b and Supplementary Fig. 6d). In the absence of the Ex1B front peptide, Kal^Sec14 also exhibits significant CSPs for a similar set of residues (Supplementary Fig. 7). The approximately two-fold attenuation in the affinities of FC14 and FC12 for Kal^Sec14 (K_D = 35.6 ± 7.2 μM and 514 ± 66 μM, respectively) suggests a role for the front peptide in ligand binding (Supplementary Fig. 7c, f).

**Fig. 4: Kal^bSec14 binds FC14 via the surface groove.**

NMR chemical shift mapping of the FC14-induced backbone CSPs onto the Kal^bSec14 structure reveals that the majority of the residues most sensitive to FC14 binding (CSP > 2σ) are those forming the afore-described surface groove (Fig. 4d). These residues form patches that line both sides of the groove: Ala64, Arg65, Asp69, Arg70, Ile71, Arg72 and Leu79 from strand S2, helix H3, and the loop between them are on one side, and Ile100, Met102, Ser105, Lys106, Asp108, Leu109, Lys111, Leu113, Leu114, Lys115 and Leu117 along the entire face of helices H5 and H6 are on the other side (Fig. 4e, f). Similar CSPs are induced in Kal^Sec14 upon complex formation with FC14 (Supplementary Fig. 8). Taken together, these results reveal a mode of ligand recognition by the Sec14 domain of Kalirin via its surface groove.

Kal^bSec14–phospholipid interactions

To determine whether a similar binding mode was employed under more physiologically relevant conditions, we used an in vitro liposome flotation binding assay to evaluate the capability of Kal^bSec14/Kal^Sec14 to interact with lipid bilayers of varying composition. A similar assay was used to study the binding of phosphoinositides to the CRAL_TRIO domain of Kalirin preceded by the Ex1C front peptide (Kal^cSec14)²⁰. Unilamellar liposomes formed by mixing synthetic dioleoyl phospholipids and cholesterol in proportions mimicking the membranes of the trans-Golgi network (Golgi-mixture) were incubated with Kal^bSec14. An Accudenz gradient was used to separate liposome-bound proteins, which were recovered from the top of the gradient (fraction 1), from unbound proteins, which remained at the bottom (fraction 4). Protein levels in each fraction were analyzed by Western blotting with an antibody to the Sec14 domain of Kalirin. As observed previously, <2% of the total Kal^bSec14 bound to the control liposome (Fig. 5a; Lipo).

**Fig. 5: Interaction of Kal^bSec14 with phospholipids.**

Lipids to be tested for binding to Kal^bSec14 were added to the Golgi mixture, accounting for 8% of the total lipid content of the liposomes; although detergents like FC14 cannot be incorporated into liposomes, lysolipids with a single fatty acyl chain attached to the glycerol moiety can be. Taking into account the dimensions and properties of the lipid binding surface groove detected using FC12 and FC14, we first tested two lysophosphatidylcholine (LPC) lipids, 1-lauroyl-2-hydroxy-sn-glycero-3-phosphocholine (LPC12:0) and 1-myristoyl-2-hydroxy-sn-glycero-3- phosphocholine (LPC14:0) (see Supplementary Table 2 for their structures). Both LPC12:0 and LPC14:0 exhibit significant binding to Kal^bSec14 (Fig. 5). In contrast, LPCs with longer acyl chains (1-palmitoyl-2-hydroxy-sn-glycero-3-phosphocholine, LPC16:0; 1-stearoyl-2-hydroxy-sn-glycero-3-phosphocholine, LPC18:0; 1-oleoyl-2-hydroxy-sn-glycero-3-phosphocholine, LPC18:1) fail to interact with Kal^bSec14 (Fig. 5). Strikingly, the presence of a second 14-carbon acyl chain (1,2-dimyristoyl-sn-glycero-3-phosphocholine, DMPC) also precludes binding (Fig. 5). Evidently, the single aliphatic hydrocarbon moiety plays an important role in the Kal^bSec14–lipid interaction.

We next evaluated the ability of liposomes containing lysophospholipids with different head groups to interact with Kal^bSec14. Like LPC, lysophosphoethanolamine (LPE) possesses a positively charged head group (Supplementary Table 2). Liposomes containing 1-myristoyl-2-hydroxy-sn-glycero-3-phosphoethanolamine (LPE14:0) bind Kal^bSec14 to a level similar to that of LPC14:0 (Fig. 5). On the other hand, liposomes containing 1-myristoyl-2-hydroxy-sn-glycero-3-phospho-(1'-rac-glycerol) (LPG14:0), 1-tridecanoyl-sn-glycero-3-phospho-L-serine (LPS13:0), or 1-tridecanoyl-2-hydroxy-sn-glycero-3-phospho-(1'-myo-inositol) (LPI13:0) fail to interact with Kal^bSec14 (Fig. 5), indicating that the positively charged head groups of LPC and LPE are essential determinants for Kal^bSec14 recognition. Our previous studies showed that Kal^cSec14 specifically interacts with liposomes containing L-α-phosphatidylinositol-4-phosphate (PI4P), a phospholipid with two acyl chains and a PI head group²⁰. Strikingly, Kal^bSec14 does not interact with liposomes that contain PI4P (Fig. 5). Similar results were obtained when liposomes containing this same set of test lipids were incubated with Kal^Sec14 (Supplementary Fig. 9). Together, these data suggest that phospholipids with a single, relatively short acyl chain and a positively charged head group are recognized by Kal^bSec14 and Kal^Sec14, and that its front peptide affects its lipid binding specificity.

LPC lipid binding to the surface groove

We again turned to NMR spectroscopy to determine whether the interaction of Kal^bSec14 with LPC lipids involves its surface groove; LPE14:0 was precluded from this experiment due to its limited solubility in aqueous solution. As observed using phosphocholine (FC14 and FC12), ¹H and ¹⁵N chemical shifts in Kal^bSec14 change continuously upon titration with either LPC14:0 or LPC12:0, but not with LPC16:0 (Fig. 6a and Supplementary Fig. 10). The K_D values calculated for the titrations with LPC14:0 and LPC12:0 are 18.9 ± 2.9 μM and 316 ± 26 μM, respectively (Fig. 6b and Supplementary Fig. 10d); the ~15-fold effect of acyl chain length on binding affinity is similar to that of hydrocarbon chain length with FC14 and FC12. Chemical shift mapping again revealed that the Kal^bSec14 residues most sensitive to the addition of LPC14:0 cluster around the surface groove (Fig. 6c, d). Twelve residues have significant CSP values (>2σ) while 11 show noteworthy changes, with CSP values > 1σ. Interestingly, the side chains of several “2σ” residues (Arg65, Asp69, Arg70, Arg72, Lys106, Asp108, and Lys111) create an ionic surface around the groove opening, while seven “1σ” residues (Val45, Ile71, Leu79, Ile100, Gly104, Leu109 and Leu117) together with three “2σ” residues (Met102, Leu113 and Leu114) form the hydrophobic floor of the groove (Fig. 6e, f).

**Fig. 6: LPC14:0 binds to the surface groove of Kal^bSec14.**

We next examined the interaction of Kal^Sec14 with LPC lipids; its affinities for LPC14:0 and LPC12:0 are weaker than those observed for Kal^bSec14 (K_D = 57 ± 8 μM and 562 ± 55 μM, respectively; Supplementary Fig. 11). However, the set of residues in Kal^Sec14 that exhibit the largest CSPs upon complex formation with LPC14:0 are similar to those identified in Kal^bSec14 (Supplementary Fig. 12). These results are consistent with the suggested binding mode of FC14/FC12 to Kal^bSec14/ Kal^Sec14, confirming that the surface groove observed in Kal^bSec14 serves as a unique ligand binding site not seen in stand-alone Sec14 proteins.

To further probe the importance of its surface groove in lipid binding, we generated two paired Kal^bSec14 mutants (D69A/R70A and S105A/K106A) that carry alanine substitutions lining each side of the groove. These four amino acid residues were chosen not only because of their large CSPs during NMR titrations, but also for their locations in loops, thus minimizing potential disturbance of protein folding (Fig. 6). Size exclusion chromatography and NMR spectra confirmed that both mutant proteins remained well folded in solution. NMR titrations of the mutants with LPC14:0 show that Kal^bSec14D69A/R70A and Kal^bSec14S105A/K106A exhibit substantially reduced binding affinities to LPC14:0, with K_Ds of 129 ± 7 μM and 244 ± 21 μM, respectively (Fig. 7 and Supplementary Fig. 13), as compared to the K_D of 18.9 ± 2.9 μM for the wild-type protein. These results strongly support for our conclusion that the surface groove of Kal^bSec14 serves as the ligand binding site.

**Fig. 7: Effects of the surface groove mutations of Kal^bSec14 on LPC14:0 binding.**

As multiple attempts to crystallize Kal^bSec14 complexed with a LPC lipid failed, we used molecular docking to examine the probability of accommodating LPC14:0 in the surface cavity. LPC14:0 is docked as a flexible entity using AutoDock Vina²⁸. Most of its top-ranked docked conformations are projected to bind to Kal^bSec14 at the surface cavity, although the exact locations and conformations of the docked LPC14:0 molecules differ (Supplementary Fig. 14). Taken together, our structural observations, lipid binding data and mutagenesis studies indicate that Kal^bSec14 defines a distinct class of Sec14-containing proteins that employ a surface groove to interact with ligands that have a single acyl chain.

Role of the Ex1B front peptide in LPC binding

Using the structural and NMR titration data for Kal^bSec14 and Kal^Sec14, we sought to understand how the Ex1B front peptide might affect lipid binding. The 24 amino-acid Ex1B front peptide, including two proline residues, is negatively charged (Supplementary Fig. 15a). We identified 22 unique peaks in the HSQC spectrum of Kal^bSec14 but not in that of Kal^Sec14 (Supplementary Fig. 5c). None of these 22 unassigned peaks were shifted upon titration of increasing amounts of LPC14:0 (Supplementary Fig. 10a), indicating that the Ex1B front peptide does not interact directly with the lipid.

Chemical shift mapping of the superimposed HSQC spectra of Kal^bSec14 and Kal^Sec14 revealed that the residues displaying significant CSPs (>2σ) are located in four regions surrounding the rear portion of the surface groove (Supplementary Fig. 15b, c). A network of basic residues located adjacent to these CSP sensitive regions, including Arg52, Lys54, Arg55, Lys93, Arg94 and His126, forms a large positively charged surface patch that could be readily accessed by the Ex1B front peptide (Supplementary Fig. 15d, e). It is thus possible that the Ex1B front peptide affects lipid binding by transiently interacting with the core domain of the protein to change protein dynamics to favor a conformation optimal for ligand binding. In contrast, the positively charged, hydrophobic Ex1C front peptide (Supplementary Fig. 15a) forms an amphipathic helix that interacts directly with phosphoinositides, altering the function of cKal7²⁰. Different front peptides may employ distinctive strategies to affect ligand binding. Unfortunately, the limited solubility of the Ex1C front peptide precluded our many attempts to crystallize Kal^cSec14 or to measure its binding affinity for lipids.

Structural and sequence comparison to other mammalian SecSR proteins

In addition to Kalirin, the SecSR sub-class of the Sec14 superfamily contains four other mammalian RhoGEFs (TRIO, MCF2/DBL, MCF2L/DBS and MCF2L2) and SESTD1. With the Kal^bSec14 structure in hand, we searched for sequence and structural features conserved amongst these proteins. First, Kal^bSec14 shares a much higher degree of sequence identity with SecSR members (~35–77%) than it does with other Sec14 family proteins (Supplementary Fig. 16a). The residues engaged in specific phosphatidylinositol head group binding in Sec14p and its close homologs¹² are not conserved in SecSR family members. Second, as for Kalirin, each SecSR protein lacks a predicted CRAL_TRIO_N domain. Third, like Kalirin, isoforms of TRIO, MCF2L and MCF2L2 include an N-terminal front peptide preceding their CRAL_TRIO domains (Supplementary Fig. 16b). As observed for KALRN, alternative splicing generates isoforms of TRIO and MCF2L with front peptides that have very different biophysical properties. For example, in isoforms 4 and 9 of MCF2L, the CRAL_TRIO domain is preceded by a peptide predicted to form an amphipathic helix²⁹. Fourth, secondary-structure and structure predictions suggest that the overall fold of the CRAL_TRIO domain and the first helix of the neighboring SR of these proteins resembles the fold observed for Kal^bSec14 (Supplementary Fig. 16c). The structures predicted using AlphaFold³⁰ share a 1.3–1.7 Å rmsd with Kal^bSec14 for equivalent C_α positions (Supplementary Fig. 16a). Last, residues exhibiting large ligand-induced CSPs in Kal^bSec14 are generally conserved, maintaining their hydrophobicity, charge or size in the other SecSR RhoGEFs and in SESTD1, suggesting formation of a similar ligand-binding site (Supplementary Fig. 16b). Collectively, our data identify Kal^bSec14 as an example for a distinct class of Sec14-like domains with ligand binding sites compatible with their roles in dynamic signaling pathways.

Discussion

Structural studies of multiple stand-alone Sec14-domain proteins have been integral in understanding how ligand binding induces specific biological outcomes within this large and diverse protein family. Sec14 domains also play essential roles in a number of highly-conserved multidomain mammalian RhoGEFs known to participate in the protein complexes that control signaling, membrane trafficking and cytoskeletal organization. The results presented here provide structural and biochemical evidence that Kalirin possesses a CRAL_TRIO domain with a unique ligand binding site. Preceded by alternate front peptides and followed by multiple SRs, the Sec14 domain of Kalirin contains structural and functional features evolving from both CRAL_TRIO and BCH sub-group proteins and serves as a prototype for other SecSR sub-class family members.

Crystallographic analysis reveals that Kal^bSec14 exhibits clear structural homology to the αβα fold of canonical CRAL_TRIO domains; a similar fold is observed in the BCH domain of p50RhoGAP even though one of its five strands is contributed by the other monomer present in its intertwined dimer¹⁵. However, Kal^bSec14 adopts a closed conformation in which H7, the presumed gate helix, blocks access to its fragmented hydrophobic interior, leading to the discovery of its alternate ligand binding site. By assigning NMR resonances of Kal^Sec14, we developed a powerful tool to study the interactions of Kal^bSec14/Kal^Sec14 with ligands in solution. Our NMR CSP data and mutagenesis studies demonstrate that the residues lining a distinct surface groove of the Kalirin Sec14 domain are responsible for its interaction with selected detergents and LPC lipids. Pocket distribution analyses using the structures of six CRAL_TRIO sub-family proteins confirm the notion that each contains an internal hydrophobic pocket and lacks an exposed groove similar to that observed in Kal^bSec14. Two exposed pockets were identified in the BCH domain of p50RhoGAP, one in a location similar to that of Kal^bSec14 and the other close to the canonical gating helix (PDB ID: 7E0W; Supplementary Fig. 17) A tetraethylene glycol (TEG) molecule co-purified with the protein occupies the first pocket, in a straight up pose with one end diving into the hydrophobic core and the other projecting out of the surface. Our comparison of the CRAL_TRIO domains of Kal^bSec14, BCH and other Sec14 proteins suggests that the ligand binding pocket in the BCH domain of p50RhoGAP is an intermediate form between the exposed surface cavity in Kal^bSec14 and the buried pocket in canonical CRAL_TRIO domains.

Unlike many stand-alone Sec14 domain proteins, the Kalirin CRAL_TRIO domain lacks a helical CRAL_TRIO_N lobe. In both phosphatidylinositol- and phosphatidylcholine-bound structures of yeast Sfh1, the lipid head groups are oriented towards the N-terminal lobe while the acyl chains remain sequestered in the hydrophobic pocket¹¹. The residues engaged in lipid head group binding (i.e. binding barcodes) are located in both the CRAL_TRIO_N and CRAL_TRIO domains and are conserved in Sec14p homologs¹² and in α-tocopherol transfer protein²⁵, indicating that these two domains function as a unit. As both the CRAL_TRIO_N domain and the phospholipid binding barcodes are absent from Kalirin, our structural and biochemical studies are consistent with the inability of Kal^bSec14 to recognize phosphatidylinositides²⁰. Kal^bSec14 binds LPC12:0 and LPC14:0 about twice as tightly as Kal^Sec14. Comparisons of the NMR spectra of Kal^bSec14 and Kal^Sec14 indicate that the Ex1B front peptide might mimic at least some of the functions attributed to CRAL_TRIO_N domains by populating the Kal^bSec14 core fold in a conformation optimal for ligand binding.

Our search for lipids that bind to Kal^bSec14 and Kal^Sec14 led to the identification of lysophospholipids with a single, relatively short acyl chain and a positively charged head group, with K_Ds in the micromolar ranges. The ~20 μM K_D of Kal^bSec14 for LPC14:0 is distinctly different from the nanomolar K_D values determined for the interactions of stand-alone Sec14 proteins such as α-TTP/Sec14p/SPF/CRALBP and their ligands³¹. Identification of an exposed groove suggests a major role for protein-lipid interactions on the Kal^bSec14 surface, rather than lipid transfer activity via the canonical buried hydrophobic pocket observed in stand-alone Sec14 proteins¹¹.

It is widely recognized that rapid responses require low affinity (1–100 μM), readily reversible interactions amongst multiple proteins; modularity of recognition domains and the simultaneous use of multiple low-affinity interactions ensures more readily modified, specific interaction^{32,33,34,35,36}. Functional studies using knockout mice and knockdown technologies indicate that Kalirin can play a role in downstream signaling from receptor tyrosine kinases, GPCRs and ion channels^17,18,19,20. Consistent with these observations, tissue-specific, regulated phosphorylation of multiple Ser/Thr and Tyr residues in Kalirin has been documented^17,18,19,20. In addition, multiple protein-protein interactions involving the Sec14, SR, PH and SH3 domains of Kalirin have been identified, along with interactions of its short, linear peptide motifs with PDZ-, SH3- and SH2-domains. Coupled with these additional interactions, the low-affinity interaction of Kal^bSec14 with lysolipids may contribute to the essential role the Kalirin Sec14 domain plays in controlling spine length and endocytic trafficking^18,19,20. While the SRs and PDZ-binding motif of Kal7 may ensure its stable enrichment in the post-synaptic density, changes in the lipids accessible to the Kalirin Sec14 domain at the post-synaptic membrane or in tension on its SRs may contribute to the inhibitory effect these regions have on its GEF activity (see below).

For the vast majority of Sec14 superfamily members, the identities of their physiologically relevant ligands remain unknown³. Lysophospholipids, integral components of the plasma membrane, are usually present only in small amounts³⁷. They can be generated rapidly by phospholipases (A1 and A2), serving as intermediates in phospholipid metabolism and as second messengers in GPCR-mediated signaling pathways^37,38,39. With a relatively large hydrophilic head group and a cone shape, incorporation of lysolipids into membrane bilayers introduces packing stresses and drives lipid assemblies into structures with more positive curvature^40,41. Interestingly, it should be noted that Kal^bSec14 and Kal^Sec14 bind selectively to a very narrow range of the LPC species (with hydrocarbon tails from 12:0 to 14:0). While our data suggest a role for phospholipase activation in generating a lysophospholipid that might bind to Kal^bSec14, acyl chains with 14 carbons are generally not prevalent. Whether membrane curvature or membrane disruption associated with vesicular trafficking might expose a limited region of longer acyl chains has not yet been assessed. New methods are needed to allow the assessment of lipid/protein interactions in a cellular environment with subcellular specificity^32,42.

Our structure-based sequence alignment analyses demonstrate that the Kalirin CRAL_TRIO domain can serve as the prototype for a distinct functional subclass of the Sec14 superfamily. The six members of the SecSR sub-group are multidomain proteins with an N-terminal CRAL_TRIO domain followed by a variable number of SRs. Kalirin, TRIO, MCF2, MCF2L and MCF2L2 are Dbl family RhoGEFs, while SESTD1 partners with multiple GAPs and GEFs to modulate RhoA GTPase activities⁴³. Unlike conventional CRAL_TRIO domain containing proteins, SecSR family members share high sequence homology and are implicated in control of actin cytoskeletal organization and dynamics by modulating the activity of small GTPases. None of the other Sec14 sub-groups contains a RhoGEF, and RhoGAPs are found exclusively in BCH domain sub-families¹. Besides a role in subcellular localization, the function of the SecSR domain in five RhoGEF proteins seems to be inhibition of GEF activity. The catalytic activity of the isolated GEF domain of Kal7 is much higher than that of intact Kal7 or Kal7 lacking its Sec14 domain and first four SRs¹⁷. In agreement with this observation, Sec14 domain and SR removal reveals the full transforming activity of MCF2 and MCF2L, identifying an autoinhibitory intramolecular interaction of the Sec14 domain with the GEF PH domain^44,45. A subset of BCH-domain containing RhoGAPs, much like their GEF counterparts, use an autoinhibitory interaction of their BCH and GAP domains to control activity^15,46,47. As the postulated non-protein ligands for the SecSR-containing proteins are currently unknown, an exciting possibility is that the SecSR domain may directly interact with the small GTPase or its post-translationally modified protein terminus to regulate signaling cascades. Further structural and functional studies are needed to understand the potential implications for small GTPase signaling mediated by the previously uncharacterized SecSR domain-containing proteins.

Methods

Protein expression and purification

The Kalirin constructs were designed based on the rat KALRN sequence [NCBI Accession #: U88156.1 with two single nucleotide sequence corrections (C197G and A244T) that lead to two amino-acid changes (T66S and N82Y)]; constructs were verified by DNA sequence analysis. Amino acid numbers are based on the initiator Met encoded by Ex1B as residue 1. Inserts encoding Kal^bSec14 (residues 2–192; 21.7 kDa) and Kal^Sec14 (residues 25–192; 19.4 kDa) were cloned into a modified pET15b vector containing a removable tobacco etch virus (TEV) protease recognition site. Primer sequences are: 5'-ATATGCGGCCGCAACCCCCCGGAGGGGGCA (Kal^bSec14-Fwd), 5'-ATATGGTACCTCACAGGGAGAGGCGCAACTCAA (Kal^bSec14-Rev), and 5'-ATATGCGGCCGCGGATCTTTTCGGAATGATGGTTT (Kal^Sec14-Fwd). The surface groove mutations of Kal^bSec14 were introduced using an overlap PCR method and confirmed by sequencing. Each of the constructs was expressed in E. coli BL21(DE3) and purified by Ni²⁺-nitriloacetic acid (GE Healthcare) affinity chromatography. After removal of the N-terminal His₆ tag by overnight cleavage with TEV protease, the proteins were further purified by anion-exchange (Source Q; GE Healthcare) and gel-filtration (SD200; GE Healthcare) chromatography. For crystallization, Kal^bSec14 was concentrated to 1.1 mM by ultrafiltration in a buffer containing 10 mM Tris-HCl (pH 7.6), 150 mM NaCl, and 5 mM dithiothreitol (DTT). SeMet-substituted Kal^bSec14 was produced following established procedures⁴⁸ and purified as described above. For NMR chemical shift perturbation experiments, each protein was expressed in M9 medium supplemented with ¹⁵NH₄Cl (1 g/L; Cambridge Isotope Laboratories) as the sole nitrogen source. For NMR backbone resonance assignment experiments, Kal^Sec14 was expressed in M9 medium prepared with ¹⁵NH₄Cl and [¹³C]-D-glucose (2 g/L; Cambridge Isotope Laboratories) as the sole nitrogen and carbon sources, respectively. The ¹⁵N- and ¹⁵N/¹³C-labeled proteins were purified as described above and concentrated to 0.4 mM in NMR buffer containing 20 mM Na/K phosphate (pH 7.2), 120 mM NaCl, 2 mM β-mercaptoethanol, 0.2 mM EDTA, 8% D₂O and 0.02% NaN₃, immediately prior to data acquisition.

Crystallization and structure determination

Kal^bSec14 was crystallized in a solution consisting of 100 mM sodium cacodylate (pH 6.5), 0.2 M ammonium sulfate, 20%-25% PEG 3350, 0.2 M NaCl at 4 °C using the hanging-drop vapor diffusion method. Crystals were cryoprotected in reservoir solution supplemented with glycerol to a final concentration of 10% and flash-cooled in liquid nitrogen. Diffraction data were collected using the Life Science Data Collection (LSDC) software at the Stanford Synchrotron Radiation Lightsource (SSRL) beamlines 14-1 and 12-2 and 17-ID-1 and 17-ID-2 of National Synchrotron Light Source II (NSLS-II). The diffraction data were processed with autoPROC⁴⁹ and Fast DP^50,51. The crystals contain eight molecules in the asymmetric unit. The structure of Kal^bSec14 was determined by single-wavelength anomalous dispersion using the data collected at the selenium peak wavelength. About 16 selenium sites were located by HySS as implemented in AutoSol/PHENIX⁵². Initial phases calculated from these sites were improved by density modification using Phaser/PHENIX⁵². The resulting electron density map was readily interpretable and used to build >90% of the molecules with the program COOT⁵³. Iterative cycles of refinement using REFMAC5/CCP4⁵⁴ and BUSTER (Global Phasing Limited) followed by manual rebuilding in COOT were carried out until no further improvement of the R_free factor was observed. X-ray data collection and phasing and refinement statistics are summarized in Supplementary Table 1. Ramachandran statistics were calculated using MolProbity⁵⁵. Molecular graphics were rendered using PyMOL (Schrödinger LLC).

NMR resonance assignment

NMR spectra for backbone assignment were recorded using VnmrJ (https://openvnmrj.org/) at 25 °C on an Agilent VNMRS 800 MHz (¹H) spectrometer equipped with a cryogenic probe. A set of standard double and triple resonance experiments were performed using 0.4 mM uniformly ¹³C/¹⁵N-labeled Kal^Sec14 in NMR buffer. The spectra used for sequence specific assignment (¹H, ¹⁵N, ¹³C_α, ¹³C_β, and ¹³C_O) included 2D ¹H-¹⁵N HSQC, and 3D HNCO, HNCACB, HN(CO)CACB, HN(CA)CO and ¹⁵N-filtered NOESY-HSQC^56,57,58. All spectra were processed with NMRPipe⁵⁹ and analyzed using CARA (http://cara.nmr.ch) and CcpNmr Analysis⁶⁰, which were made available through NMRbox⁶¹. The near-complete resonance assignment was obtained with 92.7% of all expected chemical shifts assigned (93% HN, 88% N, 91% C_O, 92% C_α, 88% C_β). The assignments for Kal^Sec14 peaks in the ¹H-¹⁵N HSQC spectrum were readily transferred to the cross peaks for Kal^bSec14. Secondary structure predictions for Kal^Sec14 were performed using chemical shift assignments of five atoms (HN, N, C_O, C_α and C_β) for a given residue in the sequence with TALOS-N⁶². The predicted ρ(α) and ρ(β) values for each residue are provided in the Source Data file.

NMR titration experiments

¹H-¹⁵N HSQC spectra of the uniformly ¹⁵N-labeled Kal^bSec14 and Kal^Sec14 were recorded at 25 °C on an Agilent VNMRS 800 MHz (¹H) spectrometer equipped with a cryogenic probe. ¹H-¹⁵N HSQC spectra of the uniformly ¹⁵N-labeled Kal^bSec14 mutants were recorded using the TopSpin software (https://www.bruker.com/en/products-and-solutions/mr/nmr-software/topspin.html) at 25 °C on a Bruker Avance NEO 600 MHz spectrometer equipped with a cryogenic probe. All detergents and lipids used in this study were purchased from Cube Biotech Inc. (Wayne, Pennsylvania) and Avanti Polar Lipids, Inc. (Alabaster, Alabama), respectively; details are listed in Supplementary Table 2. After collecting an initial protein-only spectrum with a starting protein concentration of 0.4 mM, each detergent or lipid was dissolved in the same NMR buffer and gradually added to the protein until no further CSP was observed. Each series usually comprised 8-12 titration points, with ligand concentrations ranging from 0 to 8 mM and final protein:ligand ratios from 1:7 to 1:25 (the protein concentration in each mixture during titration was confirmed by UV-vis absorbance at 280 nm). The HSQC spectrum was recorded for each titration point using the same instrument setup as that of the apo protein. All spectra were processed with NMRPipe⁵⁹, and peak heights and locations were analyzed in CcpNmr Analysis⁶⁰.

For each titration series, observed CSPs for a single amino acid residue were calculated using the combined shift change (Δδ) of the backbone amide nitrogen and proton based on the equation \(\triangle \delta \,({ppm})=\sqrt{[\triangle {\delta }_{H}^{2}+{(\alpha \cdot {\triangle \delta }_{N})}^{2}]}\), where Δδ_H and Δδ_N are the ¹H and ¹⁵N chemical shift differences between the free and bound states, respectively, and α denotes the chemical shift scaling factor for ¹⁵N⁶³. The scaling factor α, which was set to 0.154⁶⁴ for this study, was determined from the ratio of the average variances of the ¹H and ¹⁵N chemical shifts observed for the 20 common amino acid residues in proteins as deposited with the Biological Magnetic Resonance Bank (BMRB)⁶⁵.

In each titration series, the peaks for all residues were ranked based on the CSP observed upon titration at the highest ligand concentration. For those peaks that showed a significant CSP, defined as those larger than two standard deviations (2σ) of the average shift for all residues, nearest neighbor chemical shift difference analysis using custom python scripts was performed to confirm shifted peak locations. The well-resolved and non-overlapped CSPs, usually including most of the 2σ CSPs and some of the 1σ CSPs, were then plotted as a function of the ligand concentration and the dissociation constant (K_D) was extracted by nonlinear least-square fitting of the following equation: \({\triangle \delta }_{{obs}}={\triangle \delta }_{\max }\frac{({K}_{D}+{[L]}_{t}{+[P]}_{t})-\sqrt{{({K}_{D}+{[L]}_{t}{+[P]}_{t})}^{2}-4{[L]}_{t}{[P]}_{t}}}{2{[P]}_{t}}\), where \({[P]}_{t}\) and \({[L]}_{t}\) are the total protein and ligand concentrations, respectively, and \({\triangle \delta }_{\max }\) is the chemical shift difference at saturation⁶⁶. The bound fraction plotted as the y-axis in the Figures was defined as \({\triangle \delta }_{{obs}}/{\triangle \delta }_{\max }\) based on the average of the normalized CSPs from all the selected peaks. Titration data for each residue and each peak used for K_D calculations are provided in the Source Data file.

Molecular docking

Three-dimensional coordinates of the lipids were generated by Open Babel⁶⁷ using the ChemDraw files downloaded from the website of Avanti Polar Lipids (Alabaster, Alabama). Docking of the lipids into the Kal^bSec14 structure was performed using the iterated Local Search Global Optimization algorithm provided by AutoDock Vina⁶⁸. The PDBQT format files (required as input) of both the lipids and Kal^bSec14 were generated using the AutoDock Tools package provided by AutoDock 4. The flexible ligand docking protocol was adopted in this work to allow rotation of any rotatable bonds in the lipid when it is docked to the binding site in the protein. Once the rotatable bonds are identified, the ligand is allowed to be flexible by moving different rigid parts through the iterated local search global optimizer. During the global energy optimization, the ligand binding score is evaluated by an empirical scoring function, which is a weighted sum of terms representing van der Waals, hydrogen bonds, electrostatic and desolvation contributions. The entire surface of the Kal^bSec14 structure was searched for possible binding sites without bias. All other parameters were set at the default values defined by AutoDock Vina. In particular, a cutoff distance of 5 Å for van der Waals interactions and 3.7 Å for hydrogen bonds between ligand atoms and protein acceptors were used. The resulting poses were scored based on the free energy of binding in which the least energetic poses would be ranked as the top hits. Total ten hits were outputted and top four hits were shown in Supplementary Fig. 14.

Liposome preparation and flotation assays

The synthetic and natural lipids used in flotation assays were purchased from Avanti Polar Lipids, Inc. (Supplementary Table 2). All powdered lipids were dissolved in chloroform per the manufacturer’s instructions. Liposomes were prepared as previously described²⁰. In brief, stocks of synthetic 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC), -phosphoethanolamine (DOPE), and -phospho-L-serine (DOPS) were mixed with cholesterol in molar proportions mimicking those observed in secretory pathway membranes (control lipid mix). Individual test lipids were dried together with the control lipid mix under a nitrogen stream and lyophilized overnight. Each dried lipid film was reconstituted to a final concentration of 1 mm in HN buffer containing 20 mm HEPES (pH 7.4) and 150 mm NaCl. Each final lipid mixtures consisted of 52% DOPC, 20% DOPE, 5% DOPS, 15% cholesterol and 8% tested lipid; inclusion of a trace amount of fluorescent lipid (Liss Rhod PE) allowed for localization of liposomes following gradient fractionation (see below). Each lipid mixture underwent five freeze-thaw cycles separated by vigorous vortexing each time it was thawed. Using a mini-extruder, unilamellar liposomes were prepared by passing each lipid mixture 21 times through a polycarbonate membrane of 100-nm pore size. Hydrated lipid mixtures were stored at −20 °C for up to 1 month; liposomes were freshly prepared for each experiment.

Flotation assays were conducted following the protocol previously described²⁰. Accudenz (Accurate Chemical and Scientific Corp., Westbury, NY) solutions (80 and 20% weight/vol) were prepared in HN buffer. To remove any particulate material, the protein to be tested was first centrifuged at 436,000 g at 4 °C for 15 min in a TL-100 ultracentrifuge using a Beckman TLA100.2 fixed angle rotor. Freshly prepared unilamellar liposomes (50 μl) were combined with 0.2 μg of the centrifuged protein and incubated at room temperature for 30 min. The entire liposome/protein mixture was then combined with 50 μl of 80% Accudenz at the bottom of an ultracentrifuge tube and carefully mixed to completion by pipetting. The resulting mixture was overlayered with 250 μl of 20% Accudenz and then with 50 μl of HN buffer. Samples were centrifuged at 259,000 g at 4 °C for 30 min in the same ultracentrifuge using a Beckman TLS55 swinging bucket rotor. Gradients were then divided into four 100-μl fractions, which were carefully collected from the bottom of the tube using gel loading micropipette tips. Each fraction was analyzed for liposome and protein content using spectrophotometry (546-nm excitation/571-nm emission) and Western blot, respectively. Only those assays with at least 90% of the Liss Rhod PE marker recovered in the top fraction (fraction 1) were used for Western bloting analysis.

For Western bloting analysis, equal aliquots of each gradient fraction were denatured in 1× Laemmli sample buffer, fractionated on 15% acrylamide gels, and transferred to PVDF membranes. The protein content was detected using a custom-made rabbit antibody to the Sec14 domain of Kalirin (CT302⁶⁹, 1:3000) and an HRP-tagged goat anti-rabbit IgG (H + L) secondary antibody (catalog # 31460, 1:6000; ThermoFisher Scientific). SuperSignal West chemiluminescent substrate (Thermo Scientific) and a Syngene PXi digital imaging system (Synoptics Group) were used to visualize blots, with exposures adjusted to be in the linear range. After background subtraction, bands were quantified using the GeneSys software provided with the imaging system, and data were represented as a proportion of total protein recovered. Uncropped blots, quantification of each data point and exact P-values are provided in the Source Data file.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The data that support this study are available from the corresponding authors upon reasonable request. The atomic coordinates and structure factors of Kal^bSec14 have been deposited in the Protein Data Bank (PDB) under accession code 7UR2 (Sec14 domain of Kalirin). The backbone ¹H, ¹³C and ¹⁵N resonance assignments of Kal^Sec14 have been deposited to BMRB under the accession number 51382 (NMR backbone assignments of the Kalirin Sec14 domain). Atomic coordinates for the previously determined Sec14 domain containing proteins were accessed from Protein Data Bank under accession codes 1AUA, 1OIZ, 1O6U, 2D4Q, 3B7N, 3HY5, 3W67, 6W32, and 7E0W. All other supporting data are available within the article and its Supplementary Figures and Tables. The source data underlying Figs. 3–7, Supplementary Figs. 6–11 and Supplementary Fig. 15 are provided as a Source Data file. Specific data P-values are also included within the Source Data file. Source data are provided with this paper.

References

Gupta, A. B., Wee, L. E., Zhou, Y. T., Hortsch, M. & Low, B. C. Cross-species analyses identify the BNIP-2 and Cdc42GAP homology (BCH) domain as a distinct functional subclass of the CRAL_TRIO/Sec14 superfamily. PLoS One 7, e33863 (2012).
Article ADS CAS Google Scholar
Grabon, A., Bankaitis, V. A. & McDermott, M. I. The interface between phosphatidylinositol transfer protein function and phosphoinositide signaling in higher eukaryotes. J. Lipid Res. 60, 242–268 (2019).
Article CAS Google Scholar
Saito, K., Tautz, L. & Mustelin, T. The lipid-binding SEC14 domain. Biochim. Biophys. Acta 1771, 719–726 (2007).
Article CAS Google Scholar
Lipp, N. F., Ikhlef, S., Milanini, J. & Drin, G. Lipid exchangers: cellular functions and mechanistic links with phosphoinositide metabolism. Front. Cell Dev. Biol. 8, 663 (2020).
Article Google Scholar
Kf de Campos, M. & Schaaf, G. The regulation of cell polarity by lipid transfer proteins of the SEC14 family. Curr. Opin. Plant Biol. 40, 158–168 (2017).
Article CAS Google Scholar
He, X., Lobsiger, J. & Stocker, A. Bothnia dystrophy is caused by domino-like rearrangements in cellular retinaldehyde-binding protein mutant R234W. Proc. Natl Acad. Sci. USA 106, 18545–18550 (2009).
Article ADS CAS Google Scholar
Bolze, C. S. et al. Human cellular retinaldehyde-binding protein has secondary thermal 9-cis-retinal isomerase activity. J. Am. Chem. Soc. 136, 137–146 (2014).
Article CAS Google Scholar
Curwin, A. & McMaster, C. Structure and function of the enigmatic Sec14 domain-containing proteins and the etiology of human disease. Future Lipido. 3, 399–410 (2008).
Article CAS Google Scholar
Wong, L. H., Copic, A. & Levine, T. P. Advances on the transfer of lipids by lipid transfer proteins. Trends Biochem. Sci. 42, 516–530 (2017).
Article CAS Google Scholar
Sha, B., Phillips, S. E., Bankaitis, V. A. & Luo, M. Crystal structure of the Saccharomyces cerevisiae phosphatidylinositol-transfer protein. Nature 391, 506–510 (1998).
Article ADS CAS Google Scholar
Bankaitis, V. A., Mousley, C. J. & Schaaf, G. The Sec14 superfamily and mechanisms for crosstalk between lipid metabolism and lipid signaling. Trends Biochem. Sci. 35, 150–160 (2010).
Article CAS Google Scholar
Tripathi, A. et al. Functional diversification of the chemical landscapes of yeast Sec14-like phosphatidylinositol transfer protein lipid-binding cavities. J. Biol. Chem. 294, 19081–19098 (2019).
Article CAS Google Scholar
Skinner, H. B. et al. Phospholipid transfer activity is relevant to but not sufficient for the essential function of the yeast SEC14 gene product. EMBO J. 12, 4775–4784 (1993).
Article CAS Google Scholar
Cleves, A. E. et al. Mutations in the CDP-choline pathway for phospholipid biosynthesis bypass the requirement for an essential phospholipid transfer protein. Cell 64, 789–800 (1991).
Article CAS Google Scholar
Chichili, V. P. R. et al. Structural basis for p50RhoGAP BCH domain-mediated regulation of Rho inactivation. Proc. Natl Acad. Sci. USA 118, e2014242118 (2021).
Article Google Scholar
Lee, C. C., Huang, C. C. & Hsu, K. S. The phospholipid-binding protein SESTD1 negatively regulates dendritic spine density by interfering with Rac1-Trio8 signaling pathway. Sci. Rep. 5, 13250 (2015).
Article ADS CAS Google Scholar
Bircher, J. E. & Koleske, A. J. Trio family proteins as regulators of cell migration and morphogenesis in development and disease—mechanisms and cellular contexts. J. Cell. Sci. 134, jcs248393 (2021).
Article CAS Google Scholar
Paskus, J. D., Herring, B. E. & Roche, K. W. Kalirin and Trio: RhoGEFs in synaptic transmission, plasticity, and complex brain disorders. Trends Neurosci. 43, 505–518 (2020).
Article CAS Google Scholar
Parnell, E. et al. KALRN: A central regulator of synaptic function and synaptopathies. Gene 768, 145306 (2021).
Article CAS Google Scholar
Miller, M. B. et al. Alternate promoter usage generates two subpopulations of the neuronal RhoGEF Kalirin-7. J. Neurochem. 140, 889–902 (2017).
Article CAS Google Scholar
Ma, X. M. et al. Nonenzymatic domains of Kalirin7 contribute to spine morphogenesis through interactions with phosphoinositides and Abl. Mol. Biol. Cell. 25, 1458–1471 (2014).
Article Google Scholar
Gaudreault, F., Morency, L. P. & Najmanovich, R. J. NRGsuite: a PyMOL plugin to perform docking simulations in real time using FlexAID. Bioinformatics 31, 3856–3858 (2015).
CAS Google Scholar
Ho, B. K. & Gruswitz, F. HOLLOW: generating accurate representations of channel and interior surfaces in molecular structures. BMC Struct. Biol. 8, 49 (2008).
Article Google Scholar
Dundas, J. et al. CASTp: computed atlas of surface topography of proteins with structural and topographical mapping of functionally annotated residues. Nucleic Acids Res 34, W116–W118 (2006).
Article CAS Google Scholar
Kono, N. et al. Impaired alpha-TTP-PIPs interaction underlies familial vitamin E deficiency. Science 340, 1106–1110 (2013).
Article ADS CAS Google Scholar
Vance, J. E. Phospholipid synthesis and transport in mammalian cells. Traffic 16, 1–18 (2015).
Article CAS Google Scholar
O’Connell, M. R., Gamsjaeger, R. & Mackay, J. P. The structural analysis of protein-protein interactions by NMR spectroscopy. Proteomics 9, 5224–5232 (2009).
Article Google Scholar
Yanamala, N., Tirupula, K. C. & Klein-Seetharaman, J. Preferential binding of allosteric modulators to active and inactive conformational states of metabotropic glutamate receptors. BMC Bioinforma. 9, S16 (2008).
Article Google Scholar
Gautier, R., Douguet, D., Antonny, B. & Drin, G. HELIQUEST: a web server to screen sequences with specific alpha-helical properties. Bioinformatics 24, 2101–2102 (2008).
Article CAS Google Scholar
Jumper, J. et al. Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589 (2021).
Article ADS CAS Google Scholar
Panagabko, C. et al. Ligand specificity in the CRAL-TRIO protein family. Biochemistry 42, 6467–6474 (2003).
Article CAS Google Scholar
Bagheri, Y., Ali, A. A. & You, M. Current methods for detecting cell membrane transient interactions. Front. Chem. 8, 603259 (2020).
Article CAS Google Scholar
Gibson, T. J. Cell regulation: determined to signal discrete cooperation. Trends Biochem. Sci. 34, 471–482 (2009).
Article CAS Google Scholar
Stein, A., Pache, R. A., Bernado, P., Pons, M. & Aloy, P. Dynamic interactions of proteins in complex networks: a more structured view. FEBS J. 276, 5390–5405 (2009).
Article CAS Google Scholar
Mayer, B. J. Perspective: Dynamics of receptor tyrosine kinase signaling complexes. FEBS Lett. 586, 2575–2579 (2012).
Article CAS Google Scholar
Perkins, J. R., Diboun, I., Dessailly, B. H., Lees, J. G. & Orengo, C. Transient protein-protein interactions: structural, functional, and network properties. Structure 18, 1233–1243 (2010).
Article CAS Google Scholar
Pineiro, R. & Falasca, M. Lysophosphatidylinositol signalling: new wine from an old bottle. Biochim. Biophys. Acta 1821, 694–705 (2012).
Article CAS Google Scholar
Law, S. H. et al. An updated review of lysophosphatidylcholine metabolism in human diseases. Int. J. Mol. Sci. 20, 1149 (2019).
Article CAS Google Scholar
Corda, D., Iurisci, C. & Berrie, C. P. Biological activities and metabolism of the lysophosphoinositides and glycerophosphoinositols. Biochim. Biophys. Acta 1582, 52–69 (2002).
Article CAS Google Scholar
Fuller, N. & Rand, R. P. The influence of lysolipids on the spontaneous curvature and bending elasticity of phospholipid membranes. Biophys. J. 81, 243–254 (2001).
Article CAS Google Scholar
Arouri, A. & Mouritsen, O. G. Membrane-perturbing effect of fatty acids and lysolipids. Prog. Lipid Res. 52, 130–140 (2013).
Article CAS Google Scholar
De Keersmaecker, H. et al. Mapping transient protein interactions at the nanoscale in living mammalian cells. ACS Nano 12, 9842–9854 (2018).
Article Google Scholar
Yang, X. Y. et al. Sestd1 encodes a developmentally dynamic synapse protein that complexes with BCR Rac1-GAP to regulate forebrain dendrite, spine and synapse formation. Cereb. Cortex 29, 505–516 (2019).
Article Google Scholar
Ognibene, M. et al. Identification of a novel mouse Dbl proto-oncogene splice variant: evidence that SEC14 domain is involved in GEF activity regulation. Gene 537, 220–229 (2014).
Article CAS Google Scholar
Kostenko, E. V., Mahon, G. M., Cheng, L. & Whitehead, I. P. The Sec14 homology domain regulates the cellular distribution and transforming activity of the Rho-specific guanine nucleotide exchange factor Dbs. J. Biol. Chem. 280, 2807–2817 (2005).
Article CAS Google Scholar
Jiang, T., Pan, C. Q. & Low, B. C. BPGAP1 spatially integrates JNK/ERK signaling crosstalk in oncogenesis. Oncogene 36, 3178–3192 (2017).
Article CAS Google Scholar
Eberth, A. et al. A BAR domain-mediated autoinhibitory mechanism for RhoGAPs of the GRAF family. Biochem. J. 417, 371–377 (2009).
Article CAS Google Scholar
Van Duyne, G. D., Standaert, R. F., Karplus, P. A., Schreiber, S. L. & Clardy, J. Atomic structures of the human immunophilin FKBP-12 complexes with FK506 and rapamycin. J. Mol. Biol. 229, 105–124 (1993).
Article Google Scholar
Vonrhein, C. et al. Data processing and analysis with the autoPROC toolbox. Acta Crystallogr. D. Biol. Crystallogr. 67, 293–302 (2011).
Article CAS Google Scholar
Winter, G. & McAuley, K. E. Automated data collection for macromolecular crystallography. Methods 55, 81–93 (2011).
Article CAS Google Scholar
Kabsch, W. Xds. Acta Crystallogr. D. Biol. Crystallog. r. 66, 125–132 (2010).
Article CAS Google Scholar
Terwilliger, T. C. et al. Decision-making in structure solution using Bayesian estimates of map quality: the PHENIX AutoSol wizard. Acta Crystallogr. D. Biol. Crystallogr. 65, 582–601 (2009).
Article CAS Google Scholar
Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. Features and development of Coot. Acta Crystallogr. D. Biol. Crystallogr. 66, 486–501 (2010).
Article CAS Google Scholar
Winn, M. D. et al. Overview of the CCP4 suite and current developments. Acta Crystallogr. D. Biol. Crystallogr. 67, 235–242 (2011).
Article CAS Google Scholar
Williams, C. J. et al. MolProbity: More and better reference data for improved all-atom structure validation. Protein Sci. 27, 293–315 (2018).
Article CAS Google Scholar
Pervushin, K., Riek, R., Wider, G. & Wuthrich, K. Attenuated T2 relaxation by mutual cancellation of dipole-dipole coupling and chemical shift anisotropy indicates an avenue to NMR structures of very large biological macromolecules in solution. Proc. Natl Acad. Sci. USA 94, 12366–12371 (1997).
Article ADS CAS Google Scholar
Sattler, M., Schleucher, J. & Griesinger, C. Heteronuclear multidimensional NMR experiments for the structure determination of proteins in solution employing pulsed field gradients. Prog. Nucl. Magn. Reson. Spectrosc. 34, 93–158 (1999).
Article CAS Google Scholar
Kay, L. E., Keifer, P. & Sarrinen, T. Pure absorption gradient enhanced heteronuclear single quantum correlation spectroscopy with improved sensitivity. J. Am. Chem. Soc. 114, 10663–10665 (1992).
Article CAS Google Scholar
Delaglio, F. et al. NMRPipe: a multidimensional spectral processing system based on UNIX pipes. J. Biomol. NMR 6, 277–293 (1995).
Article CAS Google Scholar
Vranken, W. F. et al. The CCPN data model for NMR spectroscopy: development of a software pipeline. Proteins 59, 687–696 (2005).
Article CAS Google Scholar
Maciejewski, M. W. et al. NMRbox: a resource for biomolecular NMR computation. Biophys. J. 112, 1529–1534 (2017).
Article ADS CAS Google Scholar
Shen, Y. & Bax, A. Protein backbone and sidechain torsion angles predicted from NMR chemical shifts using artificial neural networks. J. Biomol. NMR 56, 227–241 (2013).
Article CAS Google Scholar
Schumann, F. H. et al. Combined chemical shift changes and amino acid specific chemical shift mapping of protein-protein interactions. J. Biomol. NMR 39, 275–289 (2007).
Article CAS Google Scholar
Mulder, F. A., Schipper, D., Bott, R. & Boelens, R. Altered flexibility in the substrate-binding site of related native and engineered high-alkaline Bacillus subtilisins. J. Mol. Biol. 292, 111–123 (1999).
Article CAS Google Scholar
Ulrich, E. L. et al. BioMagResBank. Nucleic Acids Res 36, D402–D408 (2008).
Article CAS Google Scholar
Baxter, N. J., Lilley, T. H., Haslam, E. & Williamson, M. P. Multiple interactions between polyphenols and a salivary proline-rich protein repeat result in complexation and precipitation. Biochemistry 36, 5566–5577 (1997).
Article CAS Google Scholar
O’Boyle, N. M. et al. Open Babel: An open chemical toolbox. J. Cheminform. 3, 33 (2011).
Article Google Scholar
Trott, O. & Olson, A. J. AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J. Comput. Chem. 31, 455–461 (2010).
CAS Google Scholar
Yan, Y., Eipper, B. A. & Mains, R. E. Kalirin-9 and Kalirin-12 play essential roles in dendritic outgrowth and branching. Cereb. Cortex 25, 3487–3501 (2015).
Article Google Scholar
Baker, N. A., Sept, D., Joseph, S., Holst, M. J. & McCammon, J. A. Electrostatics of nanosystems: application to microtubules and the ribosome. Proc. Natl Acad. Sci. USA 98, 10037–10041 (2001).
Article ADS CAS Google Scholar

Download references

Acknowledgements

We thank W. Shi, J. Jakoncic and A. Soares at the beamlines 17-ID-1 and 17-ID-2 of NSLS-II for assistance with the X-ray data collection, I. Bezsonova, D. Korzhnev and M. Maciejewski for NMR data collection and processing, and A. K. Pozhidaeva and P. Setlow for critical reading of the manuscript. This work was supported by National Institutes of Health (NIH) Grants DK032948 (to R.E.M. and B.A.E.), GM123249 (to J.C.H.), GM099948 (to B.H.) and GM135592 (to B.H.). This study made use of NMRbox: National Center for Biomolecular NMR Data Processing and Analysis, a Biomedical Technology Research Resource, which is supported by NIH Grant P41GM111135 (to J.C.H.). Support was also provided by the Dr. Daniel Schwartzberg Fund (to R.E.M. and B.A.E.). This research used resources of NSLS-II, a U.S. Department of Energy (DOE) Office of Science User Facility operated by Brookhaven National Laboratory under Contract No. DE-SC0012704. The Center for BioMolecular Structure (CBMS) is primarily supported by the NIH, National Institute of General Medical Sciences (NIGMS) through a Center Core P30 Grant (P30GM133893), and by the DOE Office of Biological and Environmental Research (KP1605010). Use of the SSRL, SLAC National Accelerator Laboratory, is supported by the DOE Office of Basic Energy Sciences under Contract No. DE-AC02-76SF00515. The SSRL Structural Molecular Biology Program is supported by the DOE Office of Biological and Environmental Research, and by the NIGMS (P30GM133894). The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of NIGMS or NIH.

Author information

Authors and Affiliations

Department of Molecular Biology and Biophysics, University of Connecticut Health Center, Farmington, CT, 06030, USA
Yunfeng Li, Yulia Pustovalova, Jeffrey C. Hoch, Betty A. Eipper & Bing Hao
Macromolecular Crystallography Group, Stanford Synchrotron Radiation Light Source, SLAC National Accelerator Laboratory, Stanford University, Stanford, CA, 94309, USA
Tzanko I. Doukov
Department of Neuroscience, University of Connecticut Health Center, Farmington, CT, 06030, USA
Richard E. Mains & Betty A. Eipper

Authors

Yunfeng Li
View author publications
You can also search for this author in PubMed Google Scholar
Yulia Pustovalova
View author publications
You can also search for this author in PubMed Google Scholar
Tzanko I. Doukov
View author publications
You can also search for this author in PubMed Google Scholar
Jeffrey C. Hoch
View author publications
You can also search for this author in PubMed Google Scholar
Richard E. Mains
View author publications
You can also search for this author in PubMed Google Scholar
Betty A. Eipper
View author publications
You can also search for this author in PubMed Google Scholar
Bing Hao
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

R.E.M., B.A.E. and B.H. conceived the study. Y.L., T.I.D. and B.H. performed the crystallographic studies., Y.P., Y.L. and B.H. performed the NMR analyses. Y.L. and B.H. performed the biochemical experiments. Y.P. and J.C.H. provided the NMR data analysis resources. All authors analyzed the results. B.A.E. and B.H. wrote the manuscript. All authors edited and approved the final manuscript.

Corresponding author

Correspondence to Bing Hao.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. Peer reviewer reports are available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Peer Review File

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Li, Y., Pustovalova, Y., Doukov, T.I. et al. Structure of the Sec14 domain of Kalirin reveals a distinct class of lipid-binding module in RhoGEFs. Nat Commun 14, 96 (2023). https://doi.org/10.1038/s41467-022-35678-4

Download citation

Received: 28 May 2022
Accepted: 16 December 2022
Published: 06 January 2023
DOI: https://doi.org/10.1038/s41467-022-35678-4

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.