Lipid turnover and GEF recruitment collectively determine Rab5 recruitment and activation during the first step of early endosome formation

Yongtao Du; Xilin Miao; Yu Gao; Yu Liang; Xue Bai; Song Dang; Zhongsheng Wu; Zhengyang An; Xinyi Zhang; Yaran Yang; Zhe Zhang; Huaqing Cai; Kangmin He

doi:10.1038/s41467-026-70543-8

. 2026 Mar 30;17:6896. doi: 10.1038/s41467-026-70543-8

Lipid turnover and GEF recruitment collectively determine Rab5 recruitment and activation during the first step of early endosome formation

Yongtao Du ^1,^2,^#, Xilin Miao ^2,^3,^#, Yu Gao ^1,², Yu Liang ^1,², Xue Bai ⁴, Song Dang ¹, Zhongsheng Wu ^1,², Zhengyang An ^1,², Xinyi Zhang ^1,², Yaran Yang ¹, Zhe Zhang ^4,⁵, Huaqing Cai ^2,^3,^✉, Kangmin He ^1,^2,^✉

PMCID: PMC13388671 PMID: 41912541

Abstract

Early endosomes are the pivotal sorting station in eukaryotic cells. A longstanding critical question is how the small GTPase Rab5 is precisely targeted to the correct membrane to initiate early endosome formation. Here, we identify Rabex5 and hRME6 as the two guanine-nucleotide exchange factors (GEFs) that together regulate Rab5 recruitment during early endosome formation. Single-molecule imaging of genome-edited cells reveals that Rabex5 and hRME6 are recruited continuously or transiently to nascent uncoated endocytic carriers, respectively. However, in contrast to uncoated endocytic carriers and other intracellular organelles, directing Rabex5 or its GEF domain to clathrin-coated pits or the plasma membrane fails to trigger Rab5 recruitment. Both in vivo and in vitro experiments show that the plasma membrane-enriched phospholipid PI(4,5)P₂ prevents Rab5 association with the plasma membrane. Importantly, we found that impaired hydrolysis of PI(4,5)P₂ led to reduced early endosome formation in Lowe syndrome cells. Therefore, the spatiotemporal recruitment and activation of Rab5 during early endosome formation are collectively determined by Rabex5/hRME6 recruitment and PI(4,5)P₂ depletion during uncoated endocytic carrier formation.

Subject terms: Endosomes, Lipid signalling, Membrane trafficking

Du, Miao et al identify that the spatiotemporal recruitment and activation of Rab5 during early endosome formation are collectively determined by Rabex5/hRME6 recruitment and PI(4,5)P2 depletion during uncoated endocytic carrier formation.

Introduction

Endocytosis is essential for cells to take up extracellular nutrients, internalize plasma membrane proteins and lipids, and control bacterial and viral infection¹. The internalized cargo is soon delivered to early endosomes, where it can be sorted to recycling endosomes or Golgi for retrieval, or lysosomes for degradation^2,3. Dysregulation of the biogenesis, distribution, and function of early endosomes is associated with various human diseases including Alzheimer’s disease⁴. Despite significant advances in unveiling the structural complexity, signaling functions, and identifying cargo sorting machinery of early endosomes^3,5, our understanding of the dynamics and mechanisms underlying early endosome biogenesis remains limited.

Rab5, a Rab protein specific to early endosome, plays a particularly critical role in membrane trafficking from the plasma membrane to early endosomes⁵. Like other small GTPases, Rab5 functions as a molecular switch that cycles between the GTP-bound active and GDP-bound inactive form. The conversion from the GDP-bound to the GTP-bound form is catalyzed by guanine-nucleotide exchange factors (GEFs), whereas the conversion from the GTP-bound to the GDP-bound form is driven by GTPase activating proteins (GAPs)^6,7. Rab5 is post-translationally modified at its C-terminus by covalent attachment of two lipophilic geranylgeranyl chains, which allows it to associate with membranes⁸. The GDP-bound Rab5 is extracted from membranes and kept soluble in the cytosol in the inactive state by association with the guanine dissociation inhibitor (GDI)^6,7. The release from GDI and the subsequent GEF-catalyzed GDP-to-GTP exchange enable Rab5 to interact with membranes and recruit effector proteins^6,7.

Our understanding of GDP/GTP cycling catalyzed by GEFs/GAPs mainly comes from in vitro experiments. Detecting and tracking the dynamic recruitment and activation of Rab5 during early endosome formation in live mammalian cells remains technically challenging. First, Rab5 overexpression leads to early endosome enlargement^9,10. Second, at least seven putative GEF molecules for Rab5 have been identified, but the precise subcellular localization and relative contribution of each GEF during Rab5 recruitment/activation remain poorly understood. Third, mammalian cells express three Rab5 isoforms, namely Rab5a, Rab5b, and Rab5c. Previous studies suggest that the three isoforms may have distinct subcellular distributions and roles in different tissues or cellular processes^11,12. The simultaneous knockout of all three isoforms was unsuccessful, and their conditional knockout caused cell death after 48 h^13,14. Lastly, early endosome formation begins with the generation of endocytic vesicles from the plasma membrane. The high mobility and small number of molecules recruited to these small, budded vesicles make it challenging to track the recruitment of Rab5 and other regulatory molecules during early endosome biogenesis¹⁵. Together, these technical challenges prevent the mechanistic understanding of when and how Rab5 is recruited to initiate early endosome formation in live cells.

By imaging the endogenous Rab5 in genome-edited cells with single-molecule sensitivity, our previous study showed that Rab5a and Rab5c appeared in newly budded endocytic vesicles, indicating that nascent endocytic carriers recruit Rab5 before their fusion with other carriers or endosomes¹⁵. However, the molecular mechanism that determines the precise spatial and temporal targeting of Rab5 to endocytic vesicles remains unknown. In this study, we developed a live cell system in which endogenous Rab5 is fluorescently tagged, and normally expressed, but cannot get activated. We discovered that the recruitment of Rabex5/hRME6 and the depletion of PI(4,5)P₂ from endocytic vesicles together determine the recruitment and activation of Rab5 during the first step of early endosome formation.

Results

Rab5 is recruited specifically to uncoated endocytic carriers in genome-edited cells

Clathrin assembles at the plasma membrane to form clathrin-coated pits, which subsequently pinch off to form clathrin-coated vesicles¹⁶. The disassembly of clathrin coats, initiated by a burst of auxilin, occurs a few seconds after vesicle release and generates uncoated endocytic carriers^17,18. Our previous work showed that Rab5a and Rab5c are recruited specifically to endocytic carriers soon after their release from the plasma membrane¹⁵. To investigate whether all three Rab5 isoforms exhibit similar recruitment dynamics, we generated an additional genome-edited cell line with endogenous Rab5b tagged with EGFP in SUM159 cells (Supplementary Fig. 1a–c). We then tagged endogenous clathrin light chain A with a near-infrared fluorescent protein miRFP670nano¹⁹ (clathrin-670nano^+/+) to create cell lines dual-labeled for clathrin and Rab5a/b/c. To determine the precise timing of Rab5a/b/c recruitment during endocytic carrier generation, we transiently expressed mScarlet-I-auxilin2/GAK in the EGFP-Rab5a/b/c^+/+ clathrin-670nano^+/+ cells and performed triple-color live-cell imaging by total internal reflection fluorescence (TIRF) microscopy (Supplementary Fig. 1d–f and Supplementary Movie 1). Consistent with previous studies^17,18, we observed the transient burst-like recruitment of GAK specifically to coated vesicles at the time of clathrin uncoating. Endogenous Rab5a/b/c were recruited after the uncoating but not the assembly stage of clathrin-coated pit formation (Supplementary Fig. 1d–f). Despite the wide range of lifetime distributions of endocytic events, quantitative analysis of the recruitment timing of Rab5a/b/c and GAK to individual endocytic vesicles further confirmed that Rab5a/b/c is recruited after the burst-like signal of GAK (Fig. 1a–c). In our previous work, we have shown that these newly uncoated endocytic carriers contain PI(3,4)P₂ and recruit Rab5, yet lack EEA1 and PI3P prior to fusing with endosomes¹⁵. The specific recruitment of endogenous Rab5c after uncoating was also observed in gene-edited HeLa and hTERT RPE-1 (hRPE-1) cells (Supplementary Fig. 1g).

Fig. 1 — a–c SUM159 cells gene-edited for EGFP-Rab5a^+/+ and clathrin-670nano^+/+ (a), EGFP-Rab5b^+/+ and clathrin-670nano^+/+ (b), or EGFP-Rab5c^+/+ and clathrin-670nano^+/+ (c) were transfected with mScarlet-I-GAK and then imaged at the bottom surface using triple-color TIRF microscopy every 1 s for 300 s. Left: Clathrin-mediated endocytic events with clathrin lifetimes between 20 and 100 s were identified in 31 (a), 26 (b) or 28 (c) cells. To account for the broad distribution of lifetimes, events were grouped into cohorts based on specific clathrin-670nano lifetime ranges. For each cohort, the fluorescence intensity traces for clathrin-670nano (blue), EGFP-Rab5 (green), and mScarlet-I-GAK (red) were averaged and are shown as mean ± SEM (intensity-lifetime cohorts). The number of traces analyzed is shown above each cohort. Right: Comparison of the relative timing of GAK and Rab5a/b/c recruitment during clathrin uncoating. SUM159 cells were used in this study unless otherwise specified. d EGFP-Rab5c^+/+ clathrin-TagRFP^+/+ cells were subjected to knockout of Rabex5 alone (Rabex5-KO), hRME6 alone (hRME6-KO), or both Rabex5 and hRME6 (DKO). Expression of the indicated proteins in the wild-type (WT) and KO cells was analyzed by western blot. Clone B1 of DKO was used in this study. e EGFP-Rab5c^+/+ clathrin-TagRFP^+/+ cells lacking Rabex5 and/or hRME6 expression were imaged using spinning-disk confocal microscopy. Left: distribution of EGFP-Rab5c at the bottom surface and the middle plane of the cells. Right: numbers of identified EGFP-Rab5c-positive spots in 36 WT, 37 Rabex5-KO, 37 hRME6-KO, and 39 DKO cells (mean ± SEM). f EGFP-Rab5c^+/+ clathrin-TagRFP^+/+ cells lacking Rabex5 and/or hRME6 expression were imaged at the bottom surface by TIRF microscopy every 1 s for 300 s. Shown are a single frame from a representative time series and the intensity-lifetime cohorts for 13 WT, 10 Rabex5-KO, 10 hRME6-KO, and 10 DKO cells (mean ± SEM). g, h EGFP-Rab5c^+/+ clathrin-TagRFP^+/+ cells lacking Rabex5 and/or hRME6 expression were transfected with Halo-EEA1 (g) or Halo-2xFYVE(Hrs) (h) stained with the JFX₆₅₀-HaloTag ligand, and then imaged using spinning-disk confocal microscopy. Representative images show the distribution of EGFP-Rab5c with Halo-tagged proteins at the middle plane of the cells. Experiments were repeated three (a–f) or four times (g, h) with similar results. Statistical analysis was performed using the ordinary one-way ANOVA with Tukey’s multiple comparisons test in (e); ****P < 0.0001; ns, not significant. Scale bars, 10 μm in (e), (g), and (h), 5 μm in (f). Source numerical data are available in source data.

Rab5 switches between the GTP-bound active form and GDP-bound inactive form⁶. To investigate how the nucleotide-bound status affects Rab5 recruitment, we transiently expressed mEGFP-tagged wild-type, GTP-bound constitutively active, or GDP-bound constitutively inactive forms of Rab5a/b/c in the clathrin-670nano^+/+ cells and performed live-cell TIRF imaging. Like the endogenous Rab5, the wild-type and constitutively GTP-bound forms of Rab5a/b/c were actively recruited to uncoated endocytic carriers (Supplementary Fig. 2a–c). However, compared to wild-type Rab5, the constitutively GTP-bound mutants caused enlargement of intracellular endosomes but showed relatively weaker recruitment to uncoated endocytic carriers. As expected, the constitutively GDP-bound mutants exhibited diffuse cytosolic distribution and were not recruited to uncoated endocytic carriers (Supplementary Fig. 2a-c). These results indicate that GEF-mediated nucleotide exchange facilitates the specific recruitment of Rab5 to uncoated endocytic carriers.

Simultaneous depletion of hRME6 and Rabex5 eliminates Rab5 recruitment to uncoated endocytic carriers and abolishes Rab5-positive early endosome formation

GEFs have been suggested to determine the subcellular targeting of Rab proteins²⁰. To identify proteins involved in recruiting Rab5 to uncoated endocytic carriers, we transiently expressed low levels of the putative Rab5 GEFs—Rabex5, hRME6, ALS2, ALS2CL, RIN1, RIN2, and RIN3—in clathrin-670nano^+/+ cells and analyzed their recruitment dynamics during clathrin-mediated endocytosis. Encouragingly, Rabex5, hRME6, and ALS2CL showed apparent recruitment to uncoated endocytic carriers (Supplementary Fig. 2d). Rabex5 and hRME6 were expressed in SUM159, HeLa, and hRPE-1 cells (Supplementary Fig. 2e). ALS2CL showed high expression in heart and liver tissues²¹ but was not detectable in the three human cell lines (Supplementary Fig. 2f). We then focused on the potential role of the two GEFs, Rabex5 and hRME6, in mediating Rab5 recruitment during early endosome formation.

To investigate the functions of Rabex5 and hRME6, we knocked out the expression of either Rabex5, hRME6, or both simultaneously (Rabex5/hRME6-DKO) in gene-edited cells expressing EGFP-Rab5c^+/+ clathrin-TagRFP^+/+ (Fig. 1d). Using immunoblotting, we confirmed that the knockout of hRME6, Rabex5, or both has a minor effect on the expression of Rab5c and other endosomal proteins, including Rabaptin5 and EEA1 (Fig. 1d). However, Rabex5 knockout resulted in a significant decrease in the number of mEGFP-Rab5c puncta around the bottom surface and perinuclear region (Fig. 1e). Knockout of hRME6 did not affect the number and distribution of mEGFP-Rab5c. Remarkably, Rabex5/hRME6-DKO led to the near complete elimination of EGFP-Rab5c puncta from the cells (Fig. 1e). Instead, we observed some fuzzy structures concentrated around the perinuclear region. Knockdown of hRME6, Rabex5, or both using siRNA had a similar inhibitory effect on the numbers and distribution of EGFP-Rab5c in gene-edited SUM159 and HeLa cells (Supplementary Fig. 3a-d).

By imaging the cells at the basal surfaces using TIRF microscopy, we further showed that simultaneous knockout of hRME6 and Rabex5 was necessary to fully abolish EGFP-Rab5c recruitment to uncoated endocytic carriers (Fig. 1f). To further investigate whether the depletion of hRME6 and/or Rabex5 affected the formation or distribution of early endosomes, we transiently expressed various endolysosomal markers in the cells. We found that hRME6/Rabex5-DKO resulted in the disappearance of EEA1-positive early endosomes (Fig. 1g). Additionally, we observed the disappearance of PI3P- and Rab5a-positive early endosomes in the DKO cells (Fig. 1h and Supplementary Fig. 3e). Rabex5/hRME6-DKO also lead to the significant reduction in the number of the Rab5 subfamily of GTPases Rab21- or Rab22a-positive endosomes (Supplementary Fig. 3f, g). Interestingly, Rabex5/hRME6-DKO did not lead to the disappearance of LAMP1-positive lysosomes (Supplementary Fig. 3h).

As GEF molecules, Rabex5 and hRME6 catalyze the conversion of Rab5 from the GDP-bound state to the GTP-bound form. Consequently, we wondered whether ectopically expressed GTP-bound Rab5 could be targeted to uncoated endocytic carriers and rescue early endosome formation in Rabex5/hRME6-DKO cells. Surprisingly, the constitutively GTP-bound Rab5a/b/c failed to be recruited to uncoated endocytic carriers, and their expression could not rescue the recruitment of endogenous EGFP-Rab5c to uncoated endocytic carriers (Supplementary Fig. 4a-c). The expression of constitutively GTP-bound Rab5a/b/c also failed to restore the formation of EGFP-Rab5c-positive early endosomes (Supplementary Fig. 4d-f). As expected, the wild-type or the constitutively GDP-bound forms of Rab5 were not recruited to uncoated endocytic carriers and were incapable of rescuing early endosome formation in Rabex5/hRME6-DKO cells (Supplementary Fig. 4a-f). These findings suggest that Rabex5- and hRME6-catalyzed GDP-to-GTP conversion is crucial for the recruitment of Rab5 to uncoated endocytic carriers and the formation of Rab5-positive early endosomes.

Distinct recruitment dynamics of endogenous Rabex5 and hRME6 to uncoated endocytic carriers

The finding that Rabex5 and hRME6 collectively control Rab5 recruitment prompted us to investigate how the spatiotemporal recruitment of these two GEFs is coordinated. Earlier studies have shown that ectopically overexpressed Rabex5 induced the enlargement of early endosomes^22–24. The overexpressed Rabex5 colocalized with Rab5 or EEA1 on these enlarged early endosomes around the perinuclear region^22–24. To overcome the limitations arising from overexpression, we fluorescently tagged endogenous Rabex5 with mEGFP by genome editing (Fig. 2a and Supplementary Fig. 5a). The functionality of the Rabex5-mEGFP^+/+ cells was confirmed by the uptake and intracellular sorting of fluorescently labeled transferrin and EGF (Supplementary Fig. 5b).

By imaging the Rabex5-mEGFP^+/+ cells using TIRF microscopy, we found that endogenous Rabex5 was recruited after clathrin uncoating (Fig. 2b and Supplementary Movie 2). Specifically, Rabex5 was recruited after the burst-like recruitment of GAK and remained associated with uncoated endocytic carriers (Fig. 2c, d). To further characterize the spatiotemporal correlation of endogenous Rabex5 with Rab5, we generated a triple gene-edited cell line expressing Rabex5-mEGFP^+/+ mScarlet-I-Rab5c^+/+ clathrin-670nano^+/+. Triple-color TIRF imaging revealed that endogenous Rabex5 and Rab5c were recruited almost simultaneously and remained associated with uncoated endocytic carriers (Fig. 2e, f). Additionally, endogenous Rabex5 was observed on Rab5c-positive endosomes within the cells (Supplementary Fig. 5c).

RME6 was originally identified as a GEF for Rab5 in Caenorhabditis elegans (C. elegans)²⁵. Interestingly, instead of colocalizing with Rab5 or EEA1, GFP-tagged RME6 concentrated primarily in clathrin-coated pits at the plasma membrane in C. elegans²⁵. More recently, by carefully re-examining the subcellular localization of RME6 in C. elegans, it was found that GFP-tagged RME6 localized primarily in the vicinity of the plasma membrane and colocalized with Rab5²⁶. While the exact subcellular localization of human RME6 (hRME6; also called RAP6 or Gapex-5) has been unclear, it was suggested that hRME6 is involved in the regulation of AP2 uncoating at the plasma membrane of mammalian cells²⁷. To reveal the genuine subcellular locations and dynamics of hRME6 in mammalian cells, we tagged endogenous hRME6 with mEGFP through genome editing (Fig. 2g and Supplementary Fig. 5d) and verified the functionality of the cells (Supplementary Fig. 5e).

By imaging mEGFP-hRME6^+/+ cells using TIRF microscopy, we unexpectedly observed the transient binding of many individual mEGFP-hRME6 spots at the plasma membrane (Fig. 2h and Supplementary Movie 3). Interestingly, mEGFP-hRME6 exhibited a transient burst of recruitment to the nascent endocytic carriers (Fig. 2h and Supplementary Movie 3), occurring immediately after the burst recruitment of GAK (Fig. 2i, j). To elucidate the spatiotemporal relationship between hRME6 and Rab5, we generated a triple gene-edited cell line expressing mEGFP-hRME6^+/+ mScarlet-I-Rab5c^+/+ clathrin-670nano^+/+. TIRF imaging revealed the simultaneous recruitment of hRME6 and Rab5c to the same uncoated endocytic carriers (Fig. 2k and Supplementary Movie 4). However, unlike Rabex5, hRME6 dissociated from uncoated endocytic carriers after its burst-like recruitment (Fig. 2h–l) and was rarely observed on Rab5c-positive endosomes within the cells (Supplementary Fig. 5f). Therefore, despite the distinct recruitment dynamics of Rabex5 and hRME6 to uncoated endocytic carriers, both proteins were recruited alongside Rab5 during the initial formation of uncoated endocytic carriers. Moreover, depletion of either Rabex5 or hRME6 did not affect the recruitment of the other protein (Supplementary Fig. 5g–i), suggesting that these two GEFs are independently recruited to uncoated endocytic carriers.

The Vps9 domain of Rabex5 is necessary for Rabex5 recruitment to uncoated endocytic carriers and Rab5-positive early endosome formation

To reveal the intrinsic connection between the recruitment dynamics and catalytic function of Rabex5, we generated a series of Rabex5 constructs with domain deletions or point mutations as shown in Fig. 3a. Rabex5 is a multidomain protein that consists of N-terminal ubiquitin-binding domains (UBDs), a long linker, a 4-helix bundle domain (4-HB), a Vps9 domain, a Rabaptin5-binding domain (RpBD), and a C-terminal proline-rich domain (PR)^28,29. We tagged the mutants with mEGFP and transiently expressed them in cells with Rabex5-KO, and then conducted TIRF imaging to analyze their recruitment dynamics during endocytic vesicle formation. Meanwhile, to evaluate the catalytic function of each mutant, we tagged the mutants with HaloTag and transiently expressed them in EGFP-Rab5c^+/+ clathrin-TagRFP^+/+ cells with Rabex5/hRME6-DKO. EGFP-Rab5c recruitment to uncoated endocytic carriers and early endosome formation was subsequently analyzed using TIRF microscopy and spinning-disk confocal microscopy, respectively.

Fig. 3 — a Schematic representation of the domain organization of various Rabex5 constructs used in (b–e). UBDs, ubiquitin-binding domains; 4-HB, 4-helix bundle domain; RpBD, Rabaptin5-binding domain; PR, proline-rich domain. b SUM159 cells with Rabex5-KO were transfected with clathrin-TagRFP and the specified mEGFP-tagged constructs in (a) and then imaged by TIRF microscopy every 1 s for 300 s. Left to right: a single frame from a representative time series; kymographs from a representative time series; and intensity-lifetime cohorts from 8 to 11 cells per construct. c EGFP-Rab5c^+/+ clathrin-TagRFP^+/+ cells with Rabex5/hMRE6-DKO were transfected with the specified Halo-tagged constructs in (a) and then imaged by TIRF microscopy every 1 s for 300 s. Shown are a single frame from a representative time series and the intensity-lifetime cohorts from 9 to 11 cells per construct. d EGFP-Rab5c^+/+ clathrin-TagRFP^+/+ cells with Rabex5/hMRE6-DKO were transfected with the specified Halo-tagged constructs in (a) and then imaged in 3D by spinning-disk confocal microscopy. Shown are maximum-intensity Z projections compiled from four sequential optical sections (starting from the bottom surface) of EGFP-Rab5c and Halo-tagged construct (corner). e Numbers of EGFP-Rab5c spots in cells expressing the specified Halo-tagged Rabex5 constructs and then imaged by spinning-disk confocal microscopy as described in (d) (n = 42, 35, 34, 33, 34, 29, 32, 33, 31, 34, 41, 34, 33, 32, 37, 34, and 35 cells). f Schematic representation of the domain organization of hRME6 and hRME6(D1429A). PXXP, proline-rich motif containing region. g SUM159 cells with hRME6-KO were transfected with clathrin-TagRFP and mEGFP-hRME6 or mEGFP-hRME6(D1429A), and then imaged by TIRF microscopy every 1 s for 300 s. Shown are a single frame and the kymographs from a representative time series, and the intensity-lifetime cohorts from 12 and 12 cells. h, i EGFP-Rab5c^+/+ clathrin-TagRFP^+/+ cells with Rabex5/hMRE6-DKO were transfected with Halo-hRME6 or Halo-hRME6(D1429A). The cells were imaged by TIRF microscopy (h) and spinning-disk confocal microscopy (i) as described in (c) and (d). Intensity-lifetime cohorts in (h) are from 12 and 10 cells. j Numbers of EGFP-Rab5c spots in cells expressing the specified Halo-tagged hRME6 constructs and then imaged by spinning-disk confocal microscopy (n = 50, 39, 41, 33, 34, 34, 33, and 17 cells). k EGFP-Rab5c^+/+ clathrin-TagRFP^+/+ cells with Rabex5/hMRE6-DKO were transfected with Halo-hRME6 and then imaged by TIRF microscopy every 1 s for 300 s. Kymographs from a representative time series are shown on the left. The montage and corresponding fluorescence intensity trace of a representative endocytic event (arrow) are shown. Experiments were repeated twice (d, e) or three times (b, c, g–k) with similar results. Halo-tagged constructs in (c–e) and (h–k) were stained with the JFX₆₅₀-HaloTag ligand. Data are shown as mean ± SEM in (b), (c), (e), (g), (h), and (j). Statistical analysis was performed using the ordinary one-way ANOVA, followed by Dunnett’s multiple comparisons test with the vector-expressed group; ***P < 0.001, ****P < 0.0001; ns, not significant. Scale bars, 5 μm in (b), (c), (g), and (h), or 10 μm in (d) and (i). Source numerical data are available in source data.

Consistent with previous studies²², we observed that removing the N-terminal UBDs or the C-terminal RpBD did not affect Rabex5 recruitment to uncoated endocytic carriers (Supplementary Fig. 5j). Expression of these truncated Rabex5 mutants restored recruitment of Rab5 to uncoated endocytic carriers and formation of Rab5-positive early endosomes in the DKO cells (Fig. 3e and Supplementary Fig. 5k,l). Intriguingly, the removal of the Vps9 domain reduced Rabex5 recruitment to uncoated endocytic carriers, and this mutant failed to restore Rab5 recruitment and early endosome formation (Fig. 3b–e). Rabex5 carrying the D313A mutation in the Vps9 domain, which is known to reduce the GEF activity³⁰, was still recruited to uncoated endocytic carriers but failed to restore Rab5 recruitment to uncoated endocytic carriers and early endosome formation (Fig. 3b–e).

Previous studies suggest that Rabex5 is recruited to early endosomes in a complex with Rabaptin5^29,31. Indeed, like Rabex5, Rabaptin5 was recruited to uncoated endocytic carriers in gene-edited cells expressing mEGFP-Rabaptin5^+/+ (Supplementary Fig. 6a–d). Knockdown of Rabex5 impaired the recruitment of Rabaptin5 to uncoated endocytic carriers (Supplementary Fig. 6e,f). However, the depletion of Rabaptin5 did not affect the recruitment of Rabex5 to uncoated endocytic carriers (Supplementary Fig. 6g, h). Nevertheless, Rabaptin5 depletion reduced the number of Rab5c-positive early endosomes (Supplementary Fig. 6i). These results suggest that while Rabaptin5 enhances the GEF activity of Rabex5 to facilitate early endosome formation^24,28,29,31, it is not required for the initial association of Rabex5 with uncoated endocytic carriers.

A burst of hRME6 recruitment triggers Rab5 recruitment to uncoated endocytic carriers and promotes Rab5-positive early endosome formation

hRME6 contains an N-terminal RasGAP domain, a proline-rich region, and a C-terminal Vps9 domain³² (Fig. 3f). By transiently expressing mEGFP-tagged hRME6 constructs with deletions of individual domains in cells with hRME6-KO, we found that only the full-length hRME6 was recruited in a transient burst to uncoated endocytic carriers (Fig. 3g). Consistent with this observation, only the full-length hRME6 was capable of restoring the recruitment of Rab5 to uncoated endocytic carriers (Fig. 3h) and the formation of Rab5-positive early endosomes (Fig. 3i, j). Mutation of hRME6 at D1429, which corresponds to the conserved D313 residue critical for Rabex5’s GEF activity³⁰, did not affect hRME6 recruitment to uncoated endocytic carriers (Fig. 3g). However, this mutant failed to restore Rab5 recruitment to uncoated endocytic carriers (Fig. 3h) and early endosome formation (Fig. 3i, j).

The observations that hRME6 is recruited in a transient burst and the re-expression of hRME6 can restore Rab5 recruitment to uncoated endocytic carriers in the Rabex5/hRME6-DKO cells indicate that a burst of hRME6 recruitment is sufficient to induce continuous Rab5 recruitment and activation. Indeed, by transiently expressing hRME6 in the DKO cells, we observed concurrent recruitment of both hRME6 and Rab5 to the same endocytic carrier. However, even after hRME6 dissociation, Rab5c remains associated with the endocytic carrier (Fig. 3k).

Targeting the Vps9 domain of Rabex5 to clathrin-coated pits fails to induce Rab5 recruitment and Rab5-positive early endosome formation

Although Rabex5 and hRME6 exhibit distinct dynamics, both are recruited after uncoating. The next question is whether this specific recruitment timing correlates with the absence of Rab5 recruitment to clathrin-coated pits/vesicles. It has been reported that artificially directing Rabex5 or its Vps9 domain to mitochondria induced the translocation of Rab5a to the mitochondrial membrane²⁰. This observation led to the hypothesis that the subcellular localization of GEFs determines the specific membrane targeting of Rab5²⁰, suggesting that Rab5 could potentially be recruited to clathrin-coated pits or vesicles by misdirecting the Vps9 domain to these structures. To explore this possibility, we generated a series of chimeras designed to target the Vps9 domain of Rabex5 to clathrin-coated pits, vesicles, or uncoated endocytic carriers (Fig. 4a).

Fig. 4 — a Schematic representation of the relative timing of recruitment of HIP1R and Epsin1 to clathrin-coated pits, GAK to clathrin-coated vesicles, and OCRL and Sac2 to uncoated endocytic carriers. The PTEN and clathrin-binding domains of GAK were used to recruit the Vps9 of Rabex5 to clathrin-coated vesicles. b, c EGFP-Rab5c^+/+ clathrin-TagRFP^+/+ cells with Rabex5/hMRE6-DKO were transfected with the specified Halo-tagged constructs, and then imaged by TIRF microscopy every 1 s for 300 s. Shown are a single frame from a representative time series and the intensity-lifetime cohorts from 9 to 12 cells. d EGFP-Rab5c^+/+ clathrin-TagRFP^+/+ cells with Rabex5/hMRE6-DKO were transfected with the specified Halo-tagged constructs, and then imaged in 3D by spinning-disk confocal microscopy. The maximum-intensity Z projections from four sequential optical sections (starting from the bottom surface) are shown for each construct. The signals from the Halo-tagged constructs are shown in the corners. e Numbers of EGFP-Rab5c spots in cells expressing the specified Halo-tagged constructs and then imaged by spinning-disk confocal microscopy as described in (d) (n = 36, 39, 20, 32, 35, 30, 37, 33, 37, 21, 22, 19, 20, 23, and 20 cells). f WT and Rabex5/hMRE6-DKO EGFP-Rab5c^+/+ clathrin-TagRFP^+/+ cells were transfected with either HIP1R-Halo-Rabex5 or HIP1R-Halo-Rabex5(D313A), and then imaged at the bottom surface using TIRF microscopy. g mEGFP-Rabaptin5^+/+ clathrin-670nano^+/+ cells were transfected with either HIP1R-Halo, HIP1R-Halo-Rabex5, or HIP1R-Halo-Rabex5(D313A), and then imaged at the bottom surface using TIRF microscopy. Experiments were repeated three (b, g) or four times (c–f) with similar results. Halo-tagged constructs in (b–g) were stained with the JFX₆₅₀-HaloTag ligand. Data are shown as mean ± SEM in (b), (c), and (e). Statistical analysis was performed using the ordinary one-way ANOVA, followed by Dunnett’s multiple comparisons test with the vector-expressed group; ****P < 0.0001; ns, not significant. Scale bars, 5 μm in (b), (c), (f), and (g), or 10 μm in (d). Source numerical data are available in source data.

We chose HIP1R and Epsin1, two clathrin-binding proteins recruited during clathrin-coated pit formation^33,34, to direct the Vps9 domain to clathrin-coated pits. We used GAK to direct the Vps9 domain to clathrin-coated vesicles (Fig. 4a). The Vps9 domain of Rabex5 was fused to HIP1R, Epsin1, or GAK to create the HIP1R-Halo-Vps9, Halo-Epsin1-Vps9, or Halo-GAK-Vps9 chimeras. In clathrin-TagRFP^+/+ cells, HIP1R-Halo-Vps9 and Halo-Epsin1-Vps9 were effectively recruited to the assembling clathrin-coated pits and Halo-GAK-Vps9 was recruited in a transient burst to clathrin-coated vesicles (Supplementary Fig. 7a–c). However, in Rabex5/hRME6-DKO cells, although HIP1R-Halo-Vps9, Halo-Epsin1-Vps9, or Halo-GAK-Vps9 were similarly recruited to clathrin-coated pits or vesicles, they were unable to recruit Rab5 to these structures (Fig. 4b) and failed to promote Rab5c-positive early endosome formation (Fig. 4d, e). Similarly, overexpression of tag-free Rabex5, but not of tag-free Rab5c, Rab5c(Q80L), HIP1R-Vps9, or Epsin1-Vps9, rescued Rab5c-positive early endosome formation in DKO cells (Supplementary Fig. 7d).

We utilized OCRL and Sac2, two proteins recruited to uncoated endocytic carriers and endosomes^35,36, to direct the Vps9 domain to uncoated endocytic carriers. Similar to OCRL, the OCRL-Vps9 and OCRL-Vps9(D313A) chimeras were recruited to uncoated endocytic carriers (Supplementary Fig. 7e). Remarkably, Halo-tagged OCRL-Vps9, but not OCRL-Vps9(D313A), effectively restored the recruitment of EGFP-Rab5c to uncoated endocytic carriers (Fig. 4c) and the formation of early endosomes in the DKO cells (Fig. 4d, e). The Vps9 domain alone exhibited a cytosolic distribution and could not rescue Rab5c recruitment or early endosome formation (Figs. 3d, 4e). Similarly, Sac2-Vps9 and Sac2-Vps9(D313A) chimeras were recruited to uncoated endocytic carriers (Supplementary Fig. 7f). Expression of Sac2-Vps9, but not Sac2-Vps9(D313A), restored Rab5c recruitment to uncoated endocytic carriers (Supplementary Fig. 7g) and early endosome formation in the DKO cells (Fig. 4d, e). These unexpected results collectively demonstrate that targeting the Vps9 domain to clathrin-coated pits is insufficient to induce Rab5 recruitment.

In cells overexpressing Rabex5, a small amount of ectopically expressed full-length Rabex5 was recruited to clathrin-coated pits. However, endogenous Rab5 remained absent from clathrin-coated pits containing ectopically expressed Rabex5 (Supplementary Fig. 8a,b). Furthermore, targeting more ectopically expressed Rabex5 to clathrin-coated pits using the HIP1R-Halo-Rabex5 chimera still failed to recruit endogenous Rab5c to these sites (Fig. 4f). In contrast, endogenous Rabaptin5 was efficiently recruited to clathrin-coated pits by HIP1R-Halo-Rabex5 (Fig. 4g). In addition to localizing at clathrin-coated pits, HIP1R-Halo-Rabex5 was also found at intracellular carriers, where it restored Rab5-positive early endosome formation in DKO cells (Fig. 4f). These findings indicate that, similar to its Vps9 domain, full-length Rabex5 is not sufficient to induce Rab5 recruitment to clathrin-coated pits.

Targeting Rabex5 or its Vps9 domain to various organelle membranes, but not the plasma membrane, induces Rab5 recruitment

The above results suggest that Rabex5 or its Vps9 domain may not be the sole determinant for the spatial targeting of Rab5. To further investigate this hypothesis, we generated a series of chimeras to redirect the Vps9 domain or full-length Rabex5 to mitochondria, the ER, the Golgi, or the plasma membrane. Consistent with previous observations²⁰, we found that directing Rabex5 or its Vps9 domain to mitochondria, as well as to the ER and Golgi, effectively recruited endogenous Rab5c to the membrane of these organelles in cells with or without Rabex5/hRME6-DKO (Fig. 5a–c and Supplementary Fig. 8c–e).

Fig. 5 — a–c The WT or Rabex5/hMRE6-DKO EGFP-Rab5c^+/+ clathrin-TagRFP^+/+ cells were transfected with the specified Halo-tagged constructs, and then imaged by spinning-disk confocal microscopy. d The WT or Rabex5/hMRE6-DKO EGFP-Rab5c^+/+ clathrin-TagRFP^+/+ cells were transfected with Vps9-Halo-CAAX, and then imaged by spinning-disk confocal microscopy. Left: images showing the lack of recruitment of EGFP-Rab5c to the plasma membrane by Vps9-Halo-CAAX. Center: plots showing the fluorescence intensities of EGFP-Rab5c and Vps9-Halo-CAAX along the line on the merged image (PM, plasma membrane; Cyto, cytoplasm). Right: plots showing the averaged relative fluorescence intensities (n = 34 and 39 cells). CAAX represents the membrane targeting sequence (including the palmitoylation sites and CAAX motif) of H-Ras. e The WT or Rabex5/hMRE6-DKO EGFP-Rab5c^+/+ clathrin-TagRFP^+/+ cells were transfected with Rabex5-Halo-CAAX and then imaged by spinning-disk confocal microscopy. The plots show the averaged relative fluorescence intensities (n = 20 and 36 cells). f Cells were transfected with the specified mEGFP-tagged Rab5c variants and Halo-tagged constructs, and then imaged by spinning-disk confocal microscopy. Images show the distribution of mEGFP-tagged Rab5 variants and Halo-tagged constructs at the middle plane of the cells. Plots show the relative fluorescence intensities of Rab5c variants and Halo-tagged constructs around the plasma membrane (28–34 cells). g The ratio of mean fluorescence intensities of mEGFP-tagged Rab5 variants around the plasma membrane and in the cytoplasm (n = 35, 29, 34, 30, and 28 cells). h Cells were transfected with the specified constructs and the interaction of Vps9-mEGFP or Vps9(D313A)-mEGFP with Halo-tagged different Rab5c variants was analyzed by co-immunoprecipitation. Experiments were repeated three times with similar results. Halo-tagged constructs in (a–f) were stained with the JFX₆₅₀-HaloTag ligand. Data are shown as mean ± SD in (d–f) or mean ± SEM in (g). Statistical analysis was performed using the ordinary one-way ANOVA with Tukey’s multiple comparisons test; ****P < 0.0001; ns, not significant. Scale bars, 10 μm. Source numerical data are available in source data.

Unexpectedly, redirecting the Vps9 domain of Rabex5 to the plasma membrane using the C-terminal membrane targeting sequence of H-Ras (containing two cysteine palmitoylation sites and one CAAX motif, abbreviated as CAAX)³⁷, did not lead to the recruitment of endogenous Rab5 to the plasma membrane (Fig. 5d). Similarly, targeting full-length Rabex5 to the plasma membrane by CAAX also failed to recruit Rab5 to the plasma membrane (Fig. 5e). Vps9-Halo-CAAX and Rabex5-Halo-CAAX were also localized to intracellular endosomal carriers, where they induced the recruitment of Rab5 in the DKO cells (Fig. 5d, e). Collectively, these results demonstrate that while the Vps9 domain or full-length Rabex5 can effectively target Rab5 to various intracellular organelles, they are insufficient for recruiting Rab5 to the plasma membrane.

The Vps9 domain of Rabex5 interacts with and recruits GDP-bound Rab5

It is established that the GDP-to-GTP switch enables Rab5 to interact with cellular membranes^6,7. Similar to wild-type Rab5, the constitutively GTP-bound Rab5 was not recruited to the plasma membrane by the Vps9 domain of Rabex5 (Fig. 5f, g). Surprisingly, although the constitutively GDP-bound Rab5 cannot associate with the plasma membrane, it was effectively recruited to the plasma membrane by the mistargeted Vps9 domain (Fig. 5f, g). The Vps9(D313A) mutant could also weakly recruit these GDP-bound Rab5 to the plasma membrane (Fig. 5f, g). Consistent with the observations with the Vps9 domain, mistargeting full-length Rabex5 to the plasma membrane resulted in the recruitment of constitutively GDP-bound but not GTP-bound Rab5 (Supplementary Fig. 8f).

Additionally, although mistargeting the Vps9 domain or Rabex5 to clathrin-coated pits failed to recruit Rab5 to clathrin-coated pits (Fig. 4b, f and Supplementary Fig. 8a), mistargeting the Vps9 domain resulted in the recruitment of constitutively GDP-bound but not GTP-bound Rab5 to clathrin-coated pits (Supplementary Fig. 8g). These results suggest that, compared to the wild-type or constitutively GTP-bound Rab5, the constitutively GDP-bound Rab5 shows a preferential association with the Vps9- or Rabex5-mistargeted plasma membrane. Indeed, immunoprecipitation analysis confirmed that the Vps9 domain interacts with Rab5, exhibiting markedly stronger binding with the constitutively GDP-bound Rab5 (Fig. 5h). The Vps9(D313A) mutant also showed weak binding with the constitutively GDP-bound Rab5 (Fig. 5h).

PI(4,5)P₂ inhibits the recruitment of Rab5 to the plasma membrane in cells

The unexpected finding that Rabex5 or its Vps9 domain was unable to recruit endogenous Rab5 or the constitutively GTP-bound Rab5 to the plasma membrane prompted us to investigate additional factors involved in regulating the membrane association of Rab5. Such factor(s) are likely specific to the plasma membrane and may interfere with Rab5 binding or activation. GTPase activating proteins (GAPs) catalyze the conversion of Rab5 from its active GTP-bound form to its inactive GDP-bound form⁷. Of the four putative Rab5 GAPs identified in previous studies (RabGAP5, RN-tre, TBC1D17, and TBC1D18), only RN-tre was localized to the plasma membrane, where it was enriched at focal adhesions but absent from clathrin-coated pits^38–41 (Supplementary Fig. 9a, b). Knockdown of RN-tre did not lead to Rab5 recruitment to the Vps9- or Rabex5-mistargeted plasma membrane (Supplementary Fig. 9c). Therefore, the absence of Rab5 recruitment to the Vps9- or Rabex5-mistargeted plasma membrane is less likely related to GAP proteins.

In addition to Rab GTPases, phosphoinositides—a group of signaling phospholipids selectively distributed in various cellular membranes—are another key factor in determining organelle identity and regulating membrane trafficking⁴². Among them, PI(4,5)P₂ is the most abundant species that is primarily located in the inner leaflet of the plasma membrane⁴³. We hypothesize that PI(4,5)P₂ may inhibit Rab5 recruitment to the plasma membrane, because it is known to be depleted following the budding of endocytic vesicles, a process mediated by the activities of inositol 5-phosphatases^15,44–46. To test this hypothesis, we first verified that Rab5 recruitment occurred following PI(4,5)P₂ depletion during clathrin-mediated endocytosis (Supplementary Fig. 10a). We then assessed whether endogenous Rab5 could be recruited to the Vps9-mistargeted plasma membrane after acute PI(4,5)P₂ depletion, which was achieved by rapamycin-induced recruitment of the inositol 5-phosphatase domain of OCRL from the cytosol to the plasma membrane⁴⁷. Remarkably, endogenous Rab5c was rapidly recruited to the Vps9-targeted plasma membrane following PI(4,5)P₂ depletion (Fig. 6a and Supplementary Movie 5). In contrast, recruitment of the inositol 5-phosphatase domain of the OCRL mutant OCRL(D523G)—which lacks phosphatase activity—had no effect on Rab5 recruitment (Fig. 6b). Cells expressing Vps9(D313A)-Halo-CAAX (Fig. 6c) or Halo-CAAX (Fig. 6d) did not support Rab5 recruitment to the PI(4,5)P₂-depleted plasma membrane. Furthermore, we confirmed Rab5 recruitment to the PI(4,5)P₂-depleted plasma membrane by the mistargeted full-length Rabex5 (Supplementary Fig. 10b). These results indicate that the recruitment of Rab5c to the plasma membrane depends on both the presence of an active GEF domain and the depletion of PI(4,5)P₂.

Fig. 6 — a EGFP-Rab5c^+/+ cells co-expressing LYN11-FRB-ECFP, mCherry-FKBP-OCRL, and Vps9-Halo-CAAX were imaged at 15-s intervals by spinning-disk confocal microscopy. Rapamycin was added (set as 0 s) during continuous imaging to trigger acute depletion of PI(4,5)P₂ by recruiting mCherry-FKBP-OCRL from the cytosol to the plasma membrane. The change in lipid composition is illustrated on the left. Kymographs (right panels) were generated along the line in the “0 s” image. b–f EGFP-Rab5c^+/+ cells co-expressing LYN11-FRB-ECFP together with the specified mCherry-tagged and Halo-tagged plasmids were treated, imaged, and presented as in (a). g Left: Diagram of the GEF assay. Vps9-associated liposomes, without or with 2% PI(4,5)P₂, PI(3,4)P₂, or PI4P, were pre-incubated with MANT-GDP-loaded Rab5c (high fluorescence). The addition of GTPγS initiated the reaction, leading to the release of MANT-GDP from Rab5c and a subsequent decrease in fluorescence. Right: Time courses of MANT-GDP release from Rab5c measured by incubating Rab5c with Vps9-liposomes containing different lipid compositions (n = 3 independent experiments; mean ± SEM). The control experiment (no Vps9) lacked Vps9-liposomes. h Top: Diagram of the liposome flotation assay. Reaction mixtures from the GEF assays were combined with sucrose and subjected to the liposome flotation assay. Rab5c bound to Vps9-liposomes with different lipid compositions was floated to the top of the sucrose gradient. Five fractions were collected and analyzed by Western blot. Bottom: Western blot analysis of Rab5c in the five collected fractions. Experiments were repeated three (e–h) or four times (a–d) with similar results. Halo-tagged constructs in (a–f) were stained with the JFX₆₅₀-HaloTag ligand. Scale bars, 10 μm. Source numerical data are available in source data.

The dephosphorylation of PI(4,5)P₂ by inositol 5-phosphatases generates PI4P. To confirm that it is the depletion PI(4,5)P₂, rather than the generation of PI4P, that regulates Rab5 recruitment, we depleted PI4P using the catalytic domain of the S. cerevisiae phosphatase Sac1, and simultaneously depleted both PI(4,5)P₂ and PI4P using an OCRL-Sac1 chimera⁴⁸. As observed with the depletion of PI(4,5)P₂ alone, the simultaneous depletion of PI(4,5)P₂ and PI4P resulted in the recruitment of Rab5 to the Vps9-mistargeted plasma membrane (Supplementary Fig. 10c). In contrast, acute depletion of PI4P alone did not affect Rab5 recruitment (Fig. 6e). Additionally, acute depletion of PI(3,4)P₂ and PI(3,4,5)P₃, two other phosphoinositide species present at the plasma membrane^42,49, did not result in Rab5 recruitment to the Vps9-mistargeted plasma membrane (Fig. 6f and Supplementary Fig. 10d). These findings suggest that PI(4,5)P₂ specifically modulates Rab5 recruitment. Furthermore, we confirmed that PI(4,5)P₂ depletion did not lead to the recruitment of Rabex5 to the plasma membrane (Supplementary Fig. 10e), indicating that PI(4,5)P₂ does not regulate Rabex5 recruitment.

PI(4,5)P₂ inhibits Rab5 recruitment to membranes of PI(4,5)P₂-containing liposomes

The inhibitory effect of PI(4,5)P₂ on Rab5 recruitment may be direct or through the inhibition of the GEF activity of Rabex5. To distinguish between these possibilities, we devised an in vitro assay.

To mimic the cellular conditions under which Rabex5 is targeted to cellular membrane, the purified Vps9 domain of Rabex5 was conjugated to liposomes containing various phosphoinositide compositions. The prenylated Rab5c was purified from membrane fraction of HEK293S GnTI⁻ cells. The GEF activity of Vps9 was examined by incubating it with GDP-loaded Rab5c and GTPγS. After the reaction, the samples were subjected to a liposome flotation assay to measure Rab5 binding to liposomes. Our results showed that the liposome-bound Vps9 exhibited GEF activity, efficiently promoting the release of GDP from Rab5c (Fig. 6g). Notably, the presence of PI(4,5)P₂, PI(3,4)P₂, or PI4P did not affect the GEF activity of the Vps9 domain (Fig. 6g). However, while the incorporation of PI(4,5)P₂ into liposomes did not affect the flotation of Vps9-containing liposomes, it markedly reduced the binding of Rab5c to the floated liposome membranes (Fig. 6h and Supplementary Fig. 10f). In contrast, the addition of PI(3,4)P₂ or PI4P did not affect Rab5 binding to liposome membranes (Fig. 6h). Therefore, using this in vitro system with minimal components, we demonstrated that PI(4,5)P₂ negatively regulates the membrane association of Rab5.

Defective PI(4,5)P₂ hydrolysis reduces early endosome formation in Lowe syndrome cells

OCRL is an inositol 5-phosphatase that is recruited to nascent endocytic carriers and early endosomes³⁵. Mutations in the OCRL gene cause Lowe syndrome, a rare X-linked multisystemic disorder⁵⁰. OCRL deficiency impairs PI(4,5)P₂ hydrolysis, leading to its abnormal accumulation on early endosomes^51,52. We therefore hypothesized that the intracellular accumulation of PI(4,5)P₂ may hinder Rab5 recruitment, thereby reducing the formation of EEA1-positive early endosomes. Due to the lack of reliable antibodies for Rab5 immunostaining, we stained for endogenous EEA1 in de-identified Lowe syndrome patient fibroblasts and control fibroblasts. We observed a significant reduction in the number of EEA1-positive early endosomes in Lowe syndrome cells (Fig. 7a), while the overall EEA1 protein expression remained unchanged (Fig. 7b). Moreover, the transient overexpression of OCRL increased the number of EEA1-positive early endosomes in Lowe syndrome cells (Supplementary Fig. 10g). These results indicate that timely depletion of PI(4,5)P₂ from nascent endocytic carriers is important for early endosome formation.

Fig. 7 — a De-identified Lowe syndrome patient fibroblasts and control fibroblasts were stained for endogenous EEA1 and then imaged by using spinning-disk confocal microscopy. Numbers of identified EEA1-labeled endosomes in 41 control and 40 Lowe syndrome cells are shown on the right (mean ± SEM). Statistical analysis was performed using the two-tailed unpaired Student’s t test; ****P < 0.0001. b Expression of the indicated proteins in the control and Lowe syndrome cells was analyzed by Western blot. CHC, clathrin heavy chain. c Schematic summary of the dynamics and regulation of Rab5 during early endosome formation. Clathrin-coated pits form at the PI(4,5)P₂-enriched plasma membrane. After scission, endocytic vesicles undergo rapid lipid conversion, including depletion of PI(4,5)P₂. Auxilin1/GAK is then recruited in a burst-like manner to the budded coated vesicles to generate the uncoated endocytic carriers. The two GEF molecules, Rabex5 and hRME6, are recruited continuously or transiently following clathrin uncoating. Unlike uncoated endocytic carriers or endosomes, which lack PI(4,5)P₂, the mistargeting of Rabex5 or its Vps9 domain to the plasma membrane or coated vesicles fails to recruit Rab5. Thus, PI(4,5)P₂ depletion and Rabex5/hRME6 recruitment together determine Rab5 recruitment and activation during the initial stage of early endosome formation. The upper part of the schematic summary was modified from¹⁵. Experiments were repeated three times (a) or twice (b) with similar results. Scale bars, 10 μm. Source numerical data are available in source data.

Discussion

In this study, by leveraging the advantages of genome-editing and single-molecule imaging tools, we resolved the dynamics and molecular mechanism underlying Rab5 recruitment during the first step of early endosome formation (Fig. 7c). We identified that Rabex5 and hRME6, the two GEF molecules recruited continuously or transiently after clathrin uncoating, together regulate the recruitment and activation of Rab5. The depletion of PI(4,5)P₂ at uncoated endocytic carriers further promotes Rab5 recruitment. Our findings reveal the dual regulation of Rab5 recruitment/activation during early endosome formation through both the spatial localization of GEFs and changes in lipid composition.

Identifying the specific GEF(s) that determine Rab5 recruitment/activation is critical for understanding the molecular mechanism underlying early endosome biogenesis. Our previous study¹⁵ and the current study collectively demonstrate that all three Rab5 isoforms are recruited to clathrin-derived endocytic carriers after uncoating, indicating the existence of specific GEF molecules during this early stage of early endosome formation. Rabex5 is the first identified and most extensively studied GEF for Rab5⁵³. In HeLa cells, knockout of Rabex5 did not eliminate EEA1-labeled early endosomes but caused the redistribution of EEA1-labeled clusters of early endosomes to the cell periphery²⁸. Similarly, in C. elegans lacking Rabex5, the Rab5-labeled structures were reduced and distributed mostly at the cell periphery²⁶. These observations indicate that Rabex5 primarily regulates the formation of early endosomes near the perinuclear region, suggesting that it may not be the sole GEF controlling early endosome formation. Indeed, our current study uncovers that simultaneous knockout of Rabex5 and hRME6 is necessary and sufficient to fully prevent the recruitment of Rab5 to uncoated endocytic carriers and the formation of Rab5-positive early endosomes. Therefore, we propose that Rabex5 and hRME6 jointly control Rab5 recruitment/activation and early endosome biogenesis in mammalian cells. However, the precise mechanism governing the initial recruitment of Rabex5 and hRME6 to uncoated endocytic carriers remains unclear. It may involve their interaction with specific lipid species or proteins that emerge following the scission and uncoating of endocytic carriers, necessitating further investigation in future studies.

Previous studies have suggested that the subcellular localization of Rabex5 determines the membrane targeting of Rab5²⁰. However, we uncover that artificially targeting Rabex5 or its Vps9 domain to clathrin-coated pits or the plasma membrane is insufficient to recruit Rab5, suggesting that GEFs are not the sole determinants of Rab5 recruitment. Studies have shown that purified Rab5a displayed varying binding affinity to lipid vesicles containing different amounts or combinations of DOPC, DMPC, or DOPE⁵⁴. Purified prenylated Rab5a was also found to be preferentially associated with liquid-disordered membrane domains⁵⁵. Additionally, molecular dynamics simulations have suggested that certain lipid species, including cholesterol and PI3P, may influence the interaction between the hypervariable region of geranylgeranylated Rab5 and membranes⁵⁶. These observations raise the possibility that locally enriched lipid species, or their associated effector proteins, may play a role in regulating Rab5 recruitment in coordination with the active GEF domain. It is estimated that ∼20,000 PI(4,5)P₂ molecules are present per μm² in the inner plasma membrane, a concentration that far exceeds the abundance of even the most abundant proteins in cells⁴³. This positions PI(4,5)P₂ as a potential modulator of Rab5 membrane association. During clathrin-mediated endocytosis, membrane scission triggers the rapid depletion of PI(4,5)P₂ from uncoated endocytic carriers^15,17, which coincides with the recruitment timing of Rabex5, hRME6, and Rab5. Remarkably, we found that depletion of PI(4,5)P₂ facilitates the recruitment of Rab5 to the plasma membrane by mistargeted Rabex5 or its Vps9 domain, uncovering an unexpected negative role of PI(4,5)P₂ in Rab5 recruitment. The precise mechanism by which PI(4,5)P₂ negatively regulates Rab5’s membrane association remains unclear. Our in vitro assays suggest that PI(4,5)P₂ may directly interfere with Rab5’s membrane association, potentially by creating a lipid environment unfavorable for the insertion of Rab5’s lipid tail. In cellular contexts, additional mechanisms might be involved. For example, PI(4,5)P₂ may inhibit Rab5’s membrane association by preventing the GDP-Rab5/GDI complex from approaching the plasma membrane, exerting negative regulation by its product DAG, or recruiting effector proteins that repel Rab5. These possibilities require further investigation.

Previous studies have demonstrated a cascade of phospholipid conversion during clathrin-mediated endocytosis and early endosome formation^15,44,45. Abnormal accumulation of PI(4,5)P₂ on endocytic carriers impairs actin dynamics and endocytic trafficking in Lowe syndrome^46,51,52. Consistent with our findings, OCRL deficiency has been shown to cause persistent perinuclear positioning of early endosomes in Lowe syndrome cells⁵⁷. Additionally, we observed a significant reduction in the number of EEA1-positive early endosomes in those cells, in agreement with previous observations in OCRL1-deficient zebrafish⁵⁸. During macropinocytosis, hydrolysis of PI(4,5)P₂ is required for macropinosome closure⁵⁹. Therefore, the recruitment and activation of Rab5 are likely regulated by both lipid molecules and GEF proteins, providing a dual safeguard mechanism to ensure that endocytic vesicles can mature into early endosomes only after scission and uncoating.

Methods

Cell culture and chemical treatment

SUM159 cells⁶⁰ were cultured at 37 °C and 5% CO₂ in DMEM/F12 (Gibco), supplemented with 5% FBS (Gibco), 100 U/mL penicillin and streptomycin (Corning), 5 μg/mL insulin (Sigma-Aldrich), and 1 μg/mL hydrocortisone (Sigma-Aldrich), pH 7.4. HeLa cells (CCL-2, ATCC), de-identified Lowe syndrome patient fibroblasts and control fibroblasts⁴⁶ were cultured at 37 °C and 5% CO₂ in DMEM (Corning), supplemented with 10% FBS, and 100 U/mL penicillin and streptomycin. Cells were verified to be mycoplasma-free using the TransDetect PCR Mycoplasma Detection Kit (TransGen Biotech). Rapamycin (Merck Millipore, 553210-100UG, 0.5 or 1 μM) was added to cells during continuous imaging using a syringe pump (Harvard Apparatus).

Plasmids and transfection

The DNA sequences encoding human Rab5a, Rab5b, or Rab5c were amplified by PCR from cDNA clones and inserted into a vector containing mEGFP to generate the plasmids mEGFP-Rab5a, mEGFP-Rab5b, or mEGFP-Rab5c using the Gibson assembly method (pEASY-Uni Seamless Cloning and Assembly Kit, TransGen Biotech). The DNA sequences encoding human Rabex5 (NM_014504.3) or hRME6 (NM_015635.4) were amplified by PCR from the cDNA of HEK293 cells and inserted into a vector containing mEGFP to generate the plasmids Rabex5-mEGFP or mEGFP-hRME6. Rabex5-Halo or Halo-hRME6 were generated by replacing the coding sequence of mEGFP in plasmid Rabex5-mEGFP or mEGFP-hRME6 with the coding sequence of HaloTag. The DNA sequences encoding different domains of Rabex5 or hRME6 were amplified by PCR from Rabex5-mEGFP or mEGFP-hRME6 and then inserted into a vector containing mEGFP to generate the mEGFP-tagged truncated Rabex5 or hRME6. A flexible (GGS)₃ or (GGS)₅ linker was inserted between mEGFP/HaloTag and the protein-coding cDNA fragments. The DNA sequences encoding HIP1R, OCRL, Epsin1, Sac2, the PTEN and clathrin-binding domains of GAK¹⁷, Sec61β, Golgi84, or the yeast mitochondrial targeting sequence (Mito)⁶¹ were amplified by PCR from the corresponding expression vectors to construct the Halo-tagged constructs using the Gibson assembly method.

The plasmids LYN11-FRB-ECFP (#38003), mCherry-FKBP-INPP4B (#116864), NES-EGFP-cPHx3 (#116855), and mCherry-FKBP-PTEN (#116866) were obtained from Addgene. mCherry-FKBP-ptaseOCRL, mCherry-FKBP-ptaseOCRL(D523G), mCherry-FKBP-Sac1, and mCherry-FKBP-Sac1(C392S) were generated by replacing PTEN in mCherry-FKBP-PTEN (#116866) with the 5-phosphatase domain of OCRL from mCherry-CRY2-5ptaseOCRL (Addgene, #66836), the D523G mutant of the 5-phosphatase domain of OCRL from mCherry-CRY2-5ptaseOCRL(D523G) (Addgene, #66837), Sac1 from PJ-Sac (Addgene, #38000), or Sac1(C392S) from PJ-DEAD (Addgene, #38002). Plasmid transfection was performed using Lipofectamine 3000 (Invitrogen) according to the manufacturer’s instructions.

Generation of knock-in cells using the CRISPR/Cas9 approach

Cells were gene-edited to incorporate EGFP at the N-terminus of Rab5b or Rab5c, mEGFP at the N-terminus of hRME6 or Rabaptin5, and mEGFP at the C-terminus of Rabex5 using the CRISPR/Cas9 approach as described^15,62. Donor constructs used for homologous recombination were generated by cloning into the pUC19 vector with two ∼600-800-nucleotide fragments of genomic DNA upstream and downstream of the start codon of human RAB5B, RAB5C, GAPVD1, and RABEP1, or the stop codon of RABGEF1, and the open reading frame of EGFP/mEGFP/mScarlet-I using the pEASY-Uni Seamless Cloning and Assembly Kit (TransGen Biotech). A flexible (GGS)₃ linker was inserted between the start or stop codon of the gene and the open reading frame of EGFP/mEGFP/mScarlet-I.

Single-guide RNA (sgRNA) targeting human RAB5B (5’-CCTTTCCCATTCTGATAATC-3’), human RAB5C (5’-TGGACGGGCAATGGCGGGTC-3’), human GAPVD1 (encoding hRME6) (5’-AGCCTTTCCCACATTGAAGA-3’), human RABEP1 (encoding Rabaptin5) (5’-CCGCCTGGTCATGGCGCAGC-3’), or human RABGEF1 (encoding Rabex5) (5’-GATCATCCTGCATAAACTTG-3’) was delivered as a PCR amplicon containing a PCR-amplified U6-driven sgRNA expression cassette⁶². Cells were transfected with the donor plasmid and the PCR amplicon using Lipofectamine 3000 (Invitrogen) according to the manufacturer’s instructions. Cells expressing EGFP/mEGFP were enriched by fluorescence-activated cell sorting (FACS) 5–7 days after transfection (FACSAria II or FACSAria Fusion, BD Biosciences). The enriched cells were expanded and then subjected to single-cell sorting into 96-well plates. The genome-edited clonal cells were characterized by PCR, imaging, and western blot analysis.

Generation of knockout cells using the CRISPR/Cas9 approach

Knockout of Rabex5, hRME6, or Rabaptin5 expression was performed using the CRISPR/Cas9 approach as described⁶². The sgRNA targeting human RABGEF1 (encoding Rabex5) (5’-AGCCTTAAGTCTGAACGCCG-3’), human GAPVD1 (encoding hRME6) (5’-GAAACTAGATATTCATACTC-3’), or human RABEP1 (encoding Rabaptin5) (5’-TCAGCAACGGGTAGCAGAAT-3‘) was cloned into the lentiviral vector (LentiCRISPR) encoding Cas9 and the sgRNA targeting sequence⁶³. The cells were infected with the lentivirus and then incubated with puromycin (2 μg/ml) for 2–3 days. The surviving cells were plated into 96-well plates and cultured for another 2–3 weeks. The monoclonal cell populations with mutations in both alleles of Rabex5, hRME6, or Rabaptin5 were identified by sequencing, and the loss of Rabex5, hRME6, or Rabaptin5 protein expression was confirmed by western blot.

To generate the Rabex5 and hRME6 double knockout cell lines, the sgRNA targeting human RABGEF1 or human GAPVD1 was cloned into pSpCas9(BB)−2A-Halo. The Rabex5-KO cells were transfected with the constructed plasmid for 24–36 h, stained with JFX₆₅₀-HaloTag ligand, and then subjected to single-cell sorting into 96-well plates (FACSAria Fusion, BD Biosciences). The monoclonal cell populations with hRME6 knockout were identified by sequencing and western blot.

Knockdown of Rabex5, hRME6, and Rabaptin5 using siRNA

The siRNA sequences used to knock down the expression of Rabex5, hRME6, and Rabaptin5 were 5’-AAATTAAGCCTCCGAATCA-3’, 5’-TTAGAATAGTGGTGCGTTA-3’, and 5’-GTAGTATGCTGTATGAATA-3’ respectively. Rabex5, hRME6, and Rabaptin5 knockdown was achieved through two sequential transfections using Lipofectamine RNAiMAX (Invitrogen) as described¹⁵. The first transfection, second transfection, and imaging were performed on day one (after overnight plating), day three, and day five, respectively.

Knockdown of RN-tre using shRNA

Lentivirus shRNA containing the target sequence 5’-GTAGATAGTCCCGTGAGATAT-3’ was utilized to knock down the expression of RN-tre. The shRNA was generated by cloning the target sequence into the pLKO.1-TRC cloning vector (Addgene, #10878). The scramble shRNA (Addgene, #1864) was used as a control. shRNA expressing lentivirus was produced in HEK293FT cells. The supernatant containing lentivirus was harvested 48 hours after transfection and concentrated 5-fold (Biogeek, #BG20101L) prior to infection of EGFP-Rab5c^+/+ cells. After 24 h, the cells were replaced with fresh medium containing 2 μg/mL puromycin (InvivoGen). After another 96 h, the cells were subjected to live-cell imaging and q-PCR analysis.

Live-cell TIRF imaging and imaging analysis

SUM159 cells plated overnight on single-well, 4-well, or 8-well confocal dishes (Cellvis) were imaged in phenol-free DMEM/F12 (Corning) containing 5% FBS and 20 mM HEPES. HeLa cells were imaged in FluoroBrite DMEM (Thermo Fisher Scientific) containing 10% FBS and 20 mM HEPES. The TIRF microscopes used for live-cell imaging were built on a Nikon Ti2-E microscope or a Nikon TIE microscope as described⁶⁴. The Nikon Ti2-E microscope was equipped with a manual TIRF Illuminator Unit (Nikon), a CFI Apochromat TIRF 100X objective (1.49 NA, Nikon), a Perfect Focus Unit (Nikon), a UNO Stage Top Incubator (Okolab), OBIS CellX lasers (405, 488, 561 and 637 nm, Coherent), the W-VIEW GEMINI-2C Image Splitting Optics (Hamamatsu), and two EMCCD cameras (Evolve 512 Delta, Photometrics). The 2× tube lens was applied to achieve the final pixel size corresponding to 80 nm of the image. Images were acquired using Micro-Manager 1.4⁶⁵. The Nikon TiE microscope was equipped with a motorized TIRF Illuminator Unit (Nikon), a CFI Apochromat TIRF 100X objective (1.49 NA, Nikon), a Perfect Focus Unit (Nikon), a Motorized XY stage (Prior Scientific), a fully enclosed and environmentally controlled cage incubator (Okolab), OBIS 488, 561 and 647 nm lasers (Coherent), W-VIEW GEMINI Image splitting optics (Hamamatsu), and an EMCCD camera (iXon Life 888, Andor Technology). The 1.5× tube lens was applied to achieve the final pixel size corresponding to 86.7 nm of the image. Images were acquired using Micro-Manager 2.0⁶⁵. For single-molecular imaging of EGFP/mEGFP-tagged proteins, the 488 nm laser power measured at the objective exit in epi-illumination mode was 1.5 mW. The exposure time was 100 ms per channel, and the EM gain of the EMCCD camera was set to 280 (Evolve 512 Delta) or 300 (iXon Life 888). The imaging interval and duration for each experiment are reported in the respective figure legends. Montages and merged images were generated using Fiji⁶⁶. Kymographs were created using the Multi Kymograph plugin in Fiji.

The automated detection and tracking of the recruitment of Rab5 and other proteins during clathrin-mediated endocytosis were performed using the cmeAnalysis software package⁶⁷ (https://github.com/DanuserLab/cmeAnalysis) in MATLAB (MathWorks), with clathrin as the “master” and other proteins as the “slave” channel. The minimum and maximum tracking search radii were 1 and 3 pixels, and the maximum gap length in a trajectory was 2 frames. The start buffer and end buffer for the valid tracks were 5 and 10 frames. Valid tracks from single clathrin-coated pits were used for further analysis. Persistent tracks and large clathrin structures (‘plaques’)^68,69 were excluded from this analysis. The intensity-lifetime cohorts were generated using the cmeAnalysis software package.

To compare the relative timing of GAK and Rab5a/b/c recruitment during clathrin uncoating, the tracks with more than three positive frames in the slave channels were analyzed. The timing of GAK signal appearance was used as the reference point for aligning different tracks. The 10 frames before and after GAK appearance from multiple tracks were averaged and then plotted as heatmaps. To compare the relative appearance time of GAK and Rab5a/b/c in each endocytic event, the appearance times of GAK or Rab5 were divided by the length of the track and then plotted as density scatter plots. The length of the track was determined by the lifetime of clathrin plus the end buffer (10 frames). The timing of GAK or Rab5 appearance was determined by the frame at which the fluorescent intensity of GAK or Rab5 was statistically higher than the local background by the cmeAnalysis software package.

Live-cell spinning-disk confocal imaging and imaging analysis

The spinning-disk confocal microscope was built on the Nikon TiE microscope described above. The microscope was equipped with a CSU-X1 spinning-disk confocal unit (Yokogawa) and an EMCCD camera (iXon Ultra 897, Andor Technology), positioned on the left side port of the microscope.

To quantify the number of Rab5-positive early endosomes in individual cells, the EGFP-Rab5c^+/+ clathrin-TagRFP^+/+ cells (with or without Rabex5 or/and hRME6 expression) were cultured overnight in confocal dishes (Cellvis), and then imaged from the bottom to the top surfaces of the cells (planes spaced at 0.35 μm) using the spinning-disk confocal microscope. The cell boundaries were manually segmented based on cell fluorescence in Fiji. A maximum-intensity Z projection from four sequential optical sections (starting from the bottom surface) of each cell was generated in Fiji. The projected and binary cell boundary images were loaded into CellProfiler 4⁷⁰ (http://www.cellprofiler.org) and the number of spots within the cell boundaries was identified automatically. The results were then exported from Cellprofiler and plotted in GraphPad Prism 9.

To generate the intensity profile of EGFP-Rab5 at and around the plasma membrane, a straight line (31-pixel width) perpendicular to the local plasma membrane was drawn manually. The intensity profiles of both EGFP-Rab5 and Halo-tagged CAAX along this line were measured using Fiji. The position of the cell membrane was defined by the maximum fluorescent signal of CAAX. To determine the background signals for Rab5 and CAAX channels, the average intensity of 21 continuous pixels along this line, positioned 15 pixels outside the cell and away from the cell membrane, was computed. Subsequently, the fluorescence intensity of Rab5 and CAAX for each cell was adjusted by subtracting their corresponding background values and then normalized to their respective maximum intensity. The background-subtracted and normalized intensity profiles from each cell were aligned based on the maximum fluorescent signal of CAAX (set as 0 μm) and then plotted.

To measure the relative enrichment of EGFP-Rab5 at the plasma membrane, the intensity of EGFP-Rab5 at the plasma membrane was determined by averaging the 3 pixels (before, on, after the maximum signal of CAAX) of EGFP-Rab5 (F_membrane). The intensity of EGFP-Rab5 in the cytosol was determined by averaging 21 continuous pixels inside the cell, positioned 15 pixels away from the cell membrane (F_cytoplasm). Subsequently, the intensity ratio (F_membrane / F_cytoplasm) was calculated for each cell and then plotted.

Immunofluorescence

Human fibroblasts were cultured overnight on coverslips (NEST, 801010) precoated with poly-D-Lysine (Sigma, P1149-10MG). Subsequently, cells were fixed using 4% paraformaldehyde (Electron Microscopy Sciences, 157-8) and permeabilized with 0.5% Triton X-100 in PBS. After incubation in blocking buffer (1% BSA in PBS) for 1 h at room temperature, the samples were incubated with primary antibodies against EEA1 (BD Biosciences, 610456, 1:400 in 1% BSA) overnight at 4˚C. Following five washes with PBS, the samples were incubated with Alexa Fluor 488-conjugated secondary antibodies or Alexa Fluor 555-conjugated secondary antibodies (Thermo Fisher Scientific, 1:1000 in 1% BSA) for 1 h at room temperature. Then the samples were imaged using the spinning-disk confocal microscope described above. The EEA1-positive endosomes were identified using the TrackMate plugin in Fiji.

Assessment of transferrin and EGF uptake by flow cytometry

Transferrin and EGF uptake were assessed as described^17,71. In brief, SUM159 cells were cultured overnight in 12-well plates, starved for 1 h in DMEM/F12 (Corning), and then incubated with 5 µg/mL Alexa Fluor 647–conjugated transferrin or 100 ng/mL Alexa Fluor 555–conjugated EGF (Life Technologies) in pre-warmed α-MEM (Corning) at 37 °C for 10 min. Then the plates were either washed with ice-chilled PBS and transferred to ice or washed three times with pre-warmed DMEM/F12. Plates washed with DMEM/F12 were replenished with pre-warmed complete culture medium and placed in the incubator for an additional 0, 10, 20, or 30 min. Following the incubation, the cells were washed with ice-chilled PBS. Surface-bound Alexa Fluor 647–conjugated transferrin or Alexa Fluor 555–conjugated EGF was removed through two brief incubations with acid wash medium (150 mM NaCl, 0.125 mM CaCl₂, 1 mM MgCl₂, and 0.1 M glycine, pH 2.5). The cells were then detached using 0.25% Trypsin and fixed with 4% paraformaldehyde for 30 min at room temperature. The cells were subsequently washed and suspended in PBS. The amount of internalized Alexa Fluor 647–conjugated transferrin or Alexa Fluor 555–conjugated EGF was determined by flow cytometry (FACSAria Fusion, BD Biosciences).

Expression and purification of proteins

To purify prenylated Rab5c, the DNA sequence encoding human Rab5c was cloned into the pEG BacMam vector, with a 10× His tag, EGFP tag, and a PreScission protease recognition site fused to the N-terminus of Rab5c. After transfection, the HEK293S GnTI^– cells were cultured at 37 °C for 10 ~ 12 h, followed by an additional 48 h at 30 °C in the presence of 10 mM sodium butyrate. Cells were harvested by centrifugation at 9,280 g for 20 min and stored at –80 °C. The cell pellets were resuspended in Buffer A (50 mM HEPES, 150 mM NaCl, 15% glycerol, 2 μg/mL DNase, 1 mM DTT, 1 mM MgCl₂, 1 mM GTP, and a cocktail of protease inhibitors, pH 7.64), sonicated on ice, and centrifuged at 20,000 g for 80 min at 4 °C. The pellets were then suspended in Buffer A with the addition of 1% n-dodecyl-β-D-maltoside (DDM) and 0.1% cholesteryl hemisuccinate (CHS) for 2 h at 4 °C. Cell debris was removed by centrifugation at 38,900 g for 50 min at 4 °C. The supernatants were incubated with anti-GFP nanobody-coupled beads for 2 h at 4 °C. The beads were washed with 50 column volumes of Buffer B (25 mM HEPES, 150 mM NaCl, 1 mM DTT, 1 mM MgCl₂, and 0.02% DDM-0.002% CHS, pH 7.64) and then incubated with PreScission protease (4:1 w/w ratio) at 4 °C overnight. The released Rab5c was mixed with GST beads at 4 °C for 1 h to remove the PreScission protease, and then eluted with 4 column volumes of Buffer B. The protein was further purified by size exclusion chromatography using a Superose 6 Increase 10/300 GL column (GE Healthcare, 29-0915-96).

To purify 6x His tagged Vps9, the DNA sequence encoding the Vps9 domain of human Rabex5 was cloned into a pET-28a vector, with an N-terminal 6x His tag. The plasmid was transformed into E. coli strain BL21 (TransGen Biotech, #CD601). Protein expression was induced with 0.3 mM isopropyl β-D-1-thiogalactoside when the culture reached an OD₆₀₀ of 0.6, followed by overnight incubation at 16 °C. The bacterial pellet was resuspended in 25 mL of Buffer D (50 mM HEPES, 150 mM NaCl, 5 mM MgCl₂, 5 mM β-mercaptoethanol, pH 8.0) supplemented with 10 mM imidazole and 1 mM PMSF, lysed by sonication for 10 min, and centrifuged at 18,000 g for 30 min at 4 °C. The supernatant was incubated with 500 µL of nickel-NTA agarose beads (Sigma, #H0537) for 2 h at 4 °C. The beads were washed three times with 30 mL of Buffer D containing 30 mM imidazole. His-Vps9 was eluted in 2.5 mL of Buffer D containing 250 mM imidazole and then exchanged into 3.5 mL of storage buffer (20 mM HEPES, 150 mM NaCl, 5 mM MgCl₂, pH 7.2) using PD-10 desalting columns (Cytiva, #17085101).

GEF assay

For GEF assay and liposome flotation assay, POPC (#850457), POPS (#840034), DGS-NTA(Ni) (#790404), PI(4,5)P₂ (#850155), PI(3,4)P₂ (#850153), and PI4P (#850151) were purchased from Avanti Polar Lipids. POPC and POPS were used to mimic the lipid composition of the plasma membrane⁷². All lipids were stored in chloroform at −20 °C. The lipids were mixed in a composition of 73% POPC, 15% POPS, 10% DGS-NTA(Ni), and different phosphoinositides. The lipid mixture was evaporated under nitrogen gas and then placed in a SpeedVac for an additional hour. The dried lipids were resuspended in 100 µL of GEF buffer (20 mM HEPES, 150 mM NaCl, 5 mM MgCl₂, pH 7.5) to a final concentration of 5 mM and subjected to 11 freeze-thaw cycles. Small unilamellar liposomes were generated by extruding the mixture through 100 nm polycarbonate membranes (Avanti Polar Lipids, #610005) 11 times. The liposomes were aliquoted, snap-frozen in liquid nitrogen, and stored at −80 °C.

The GEF activity of Vps9 was assessed using N-Methylanthraniloyl-GDP (MANT-GDP) (Sigma, #69244). To load MANT-GDP onto Rab5c, 70 µL of purified Rab5c (1.8 µg/µL) was exchanged into 70 µL of low MgCl₂ buffer (20 mM HEPES, 150 mM NaCl, 0.5 mM MgCl₂, pH 7.5) using spin desalting columns (Thermo Fisher Scientific, #89883). This was then mixed with an equal volume of 2×MANT-GDP loading buffer (20 mM HEPES, 150 mM NaCl, 0.5 mM MgCl₂, 10 mM EDTA, 2 mM DTT, and a 20-fold molar excess of MANT-GDP, pH 7.5) and incubated at 20 °C for 90 min in the dark. The reaction was terminated by adding 1 M MgCl₂ to a final concentration of 10 mM, followed by further incubation for 30 min. Subsequently, 130 µL of MANT-GDP-loaded Rab5c was exchanged into 130 µL of GEF buffer using spin desalting columns.

To prepare Vps9-liposomes, 0.25 µM of His-Vps9 was conjugated to 2.5 mM liposomes via His–Ni²⁺-NTA interaction in GEF buffer at 4 °C for 1 h. A mixture containing MANT-GDP-loaded Rab5c and Vps9-liposomes in GEF buffer was transferred into a 96-well microplate (Thermo Fisher Scientific, #237108). The concentrations in the 50 µL reaction were 5 µM for MANT-GDP-loaded Rab5c, 1 mM for liposomes, and 100 nM for His-Vps9. Fluorescence measurements were performed using a microplate spectrometer (TECAN), with an excitation wavelength of 360 nm and an emission wavelength of 440 nm. The reaction was initiated by adding 3 µL of 8.3 mM GTPγS (final concentration of 500 µM) once the initial fluorescent signal stabilized. Measurements were recorded every 15 s at room temperature. The fluorescence intensity over time was normalized by the mean intensity before GTPγS addition.

Liposome flotation assay

Following the GEF assay, the reaction samples were subjected to the liposome flotation assay. Briefly, 30 µL of the sample was transferred to a polycarbonate tube (Beckman, #343775), mixed with 100 μL of 1.9 M sucrose at the bottom of the tube, and overlaid sequentially with 100 μL of 1.25 M sucrose and 20 μL of GEF buffer. Liposomes floated to the top after ultracentrifugation at 174,000 g for 1 h. Five 50 µL fractions were collected from the top, mixed with SDS loading buffer, and analyzed by Western blot using antibodies against Rab5c (Proteintech, 27219-1-AP, 1:10000).

Western blot analysis

Western blot analysis was performed as described¹⁵. Briefly, cells were lysed at 4˚C for 20 min in RIPA lysis buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS) containing a protease inhibitor cocktail (Thermo Fisher Scientific, 78442), and then pelleted at 12,000 g for 15 min at 4˚C. The supernatant was mixed with 5× sample buffer (GenScript, MB01015), heated to 100˚C for 8-10 min, fractionated by SDS-PAGE, and then transferred to nitrocellulose membranes (PALL, 66485). The membranes were incubated in TBST buffer containing 5% skim milk for 1 h at room temperature, followed by overnight incubation at 4˚C with specific primary antibodies. After three washes in TBST (5 min each), the membranes were incubated with the appropriate HRP-conjugated secondary antibody (Beyotime, A0208 or A0216, 1:1,000) for 1 h at room temperature. The membranes were incubated with the SignalFireTM ECL Reagent (Cell Signaling, 6883S) or BeyoECL Moon (Beyotime, P0018FS) and imaged using the Tanon-5200 chemiluminescent imaging system (Tanon) or MiniChemi 610 chemiluminescent imaging system (Sage Creation Science). The primary antibodies used in this study were: GFP (Invitrogen, 14-6674-82, 1:1,000), EEA1 (BD Biosciences, 610456, 1:1,000), GAPDH (Proteintech, 60004-1-Ig, 1:50,000), Rabex5 (Santa Cruz, sc-166611, 1:500), RME6/Gapex-5 (Abcam, ab86033, 1:1,000), Rab5a (Proteintech, 11947-1-AP, 1:1,000), Rab5b (Proteintech, 27403-1-AP, 1:1,000), Rab5c (Proteintech, 27219-1-AP, 1:1,000), Rabaptin5 (Proteintech, 14350-1-AP, 1:1,000), CHC (Abcam, ab21679, 1:1,000), ALS2CL (Santa Cruz, sc-377278, 1:500), OCRL (Proteintech, 17695-1-AP, 1:1,000), and His (MBL Life science, D291-3S, 1:1000).

Co-immunoprecipitation

SUM159 cells were transfected with the specified mEGFP- and Halo-tagged plasmids for 24 h, and then lysed in lysis buffer supplemented with a cocktail of protease and phosphatase inhibitors (Thermo Fisher Scientific, 78442) on ice for 30 min. Then the cells were centrifuged at 20,000 g for 15 min at 4°C. The supernatants were collected and incubated with GFP-Nanoab-Agarose beads (Lablead, GNA-25-500) with gentle rotation for 2 h at 4°C. Then the beads were washed five times with lysis buffer, and the bound proteins were eluted using 2× SDS sample buffer, heated to 100˚C for 10 min. The samples were then subjected to immunoblot analysis as described above.

Statistical tests

Statistical analyses were performed using GraphPad Prism 10 (GraphPad Software). The statistical methods used were specified in the figure legends. *: P < 0.05; **: P < 0.01; ***: P < 0.001; ****: P < 0.0001; ns: no significant difference. P < 0.05 was considered statistically significant. No statistical methods were used to predetermine sample sizes.

Reporting summary

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

Supplementary information

Supplementary Information^{(26.7MB, pdf)}

41467_2026_70543_MOESM2_ESM.pdf^{(187.1KB, pdf)}

Descriptions of Additional Supplementary Files

Supplementary Movie 1^{(25.1MB, mp4)}

Supplementary Movie 2^{(28MB, mp4)}

Supplementary Movie 3^{(26.4MB, mp4)}

Supplementary Movie 4^{(25.2MB, mp4)}

Supplementary Movie 5^{(21.9MB, mp4)}

Reporting Summary^{(7MB, pdf)}

Transparent Peer Review file^{(20.7MB, pdf)}

Source data

Source Data zip file^{(24MB, zip)}

Acknowledgements

We thank Dr. Pietro De Camilli for the de-identified Lowe syndrome patient fibroblasts and control fibroblasts. We thank Dr. Martin Loose and Ivana Matijevic for their expertise and input. We thank Drs. Tom Kirchhausen, Joan Brugge, and Luke Lavis for generous gifts of reagents. This research was funded by the National Key R&D Program of China (2021YFA0804802 and 2022YFA1304500 to K.H., and 2021YFA1300301 to H.C.) and the National Natural Science Foundation of China (32525023, 32321004, 92354305, and 91957106 to K.H., and 32425018 to H.C.).

Author contributions

Y.D. and Y.G. performed imaging experiments; Y.D., Y.L., Z.A., and K.H. performed imaging analysis; Y.D., S.D., and Y.G. generated genome-edited cell lines; X.M., H.C., X.B., Y.D., and Z.Z. purified proteins and performed in vitro assays; Y.D. and Z.W. performed co-IP experiments; Y.D., X.Z. and Y.Y. performed cargo uptake experiments; K.H. and H.C. supervised the work.

Peer review

Peer review information

Nature Communications thanks Guangpu Li, Reinhard Jahn and the other anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Data availability

The data that support the findings of the current study are included in Supplementary Figs. 1–10 and Supplementary Movies 1–5. The source imaging data presented in this paper are available from the corresponding author upon reasonable request. The full gel source data for western blot and PCR have been provided in Source data are provided with this paper.

Code availability

The download links for published codes were provided in the Methods. Other custom MATLAB routines are available upon request from the corresponding author.

Competing interests

The authors declare no competing interests.

Footnotes

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

These authors contributed equally: Yongtao Du, Xilin Miao.

Contributor Information

Huaqing Cai, Email: huaqingcai@ibp.ac.cn.

Kangmin He, Email: kmhe@genetics.ac.cn.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-026-70543-8.

References

1.Schmid, S. L., Sorkin, A. & Zerial, M. Endocytosis: Past, present, and future. Cold Spring Harb. Perspect. Biol.6, a022509 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Cullen, P. J. & Steinberg, F. To degrade or not to degrade: mechanisms and significance of endocytic recycling. Nat. Rev. Mol. Cell Biol.19, 679–696 (2018). [DOI] [PubMed] [Google Scholar]
3.Solinger, J. A. & Spang, A. Sorting of cargo in the tubular endosomal network. Bioessays44, e2200158 (2022). [DOI] [PubMed] [Google Scholar]
4.Maxfield, F. R. Role of endosomes and lysosomes in human disease. Cold Spring Harb. Perspect. Biol.6, a016931 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Villasenor, R., Kalaidzidis, Y. & Zerial, M. Signal processing by the endosomal system. Curr. Opin. Cell Biol.39, 53–60 (2016). [DOI] [PubMed] [Google Scholar]
6.Stenmark, H. Rab GTPases as coordinators of vesicle traffic. Nat. Rev. Mol. Cell Biol.10, 513–525 (2009). [DOI] [PubMed] [Google Scholar]
7.Wandinger-Ness, A. & Zerial, M. Rab proteins and the compartmentalization of the endosomal system. Cold Spring Harb. Perspect. Biol.6, a022616 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Edler, E. & Stein, M. Recognition and stabilization of geranylgeranylated human Rab5 by the GDP Dissociation Inhibitor (GDI). Small GTPases10, 227–242 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Bucci, C. et al. The small GTPase rab5 functions as a regulatory factor in the early endocytic pathway. Cell70, 715–728 (1992). [DOI] [PubMed] [Google Scholar]
10.Stenmark, H. et al. Inhibition of rab5 GTPase activity stimulates membrane fusion in endocytosis. EMBO J.13, 1287–1296 (1994). [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Chen, P. I., Kong, C., Su, X. & Stahl, P. D. Rab5 isoforms differentially regulate the trafficking and degradation of epidermal growth factor receptors. J. Biol. Chem.284, 30328–30338 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Gurkan, C. et al. Large-scale profiling of Rab GTPase trafficking networks: the membrome. Mol. Biol. Cell16, 3847–3864 (2005). [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Hatoyama, Y., Homma, Y., Hiragi, S. & Fukuda, M. Establishment and analysis of conditional Rab1- and Rab5-knockout cells using the auxin-inducible degron system. J. Cell Sci.134, jcs259184 (2021). [DOI] [PubMed]
14.Homma, Y. et al. Comprehensive knockout analysis of the Rab family GTPases in epithelial cells. J. Cell Biol.218, 2035–2050 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
15.He, K. et al. Dynamics of phosphoinositide conversion in clathrin-mediated endocytic traffic. Nature552, 410–414 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Kirchhausen, T., Owen, D. & Harrison, S. C. Molecular structure, function, and dynamics of clathrin-mediated membrane traffic. Cold Spring Harb. Perspect. Biol.6, a016725 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
17.He, K. et al. Dynamics of Auxilin 1 and GAK in clathrin-mediated traffic. J. Cell Biol.219, e201908142 (2020). [DOI] [PMC free article] [PubMed]
18.Massol, R. H., Boll, W., Griffin, A. M. & Kirchhausen, T. A burst of auxilin recruitment determines the onset of clathrin-coated vesicle uncoating. Proc. Natl. Acad. Sci. USA103, 10265–10270 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Oliinyk, O. S., Shemetov, A. A., Pletnev, S., Shcherbakova, D. M. & Verkhusha, V. V. Smallest near-infrared fluorescent protein evolved from cyanobacteriochrome as versatile tag for spectral multiplexing. Nat. Commun.10, 279 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Blumer, J. et al. RabGEFs are a major determinant for specific Rab membrane targeting. J. Cell Biol.200, 287–300 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Hadano, S. et al. ALS2CL, the novel protein highly homologous to the carboxy-terminal half of ALS2, binds to Rab5 and modulates endosome dynamics. FEBS Lett.575, 64–70 (2004). [DOI] [PubMed] [Google Scholar]
22.Zhu, H. et al. Rabaptin-5-independent membrane targeting and Rab5 activation by Rabex-5 in the cell. Mol. Biol. Cell18, 4119–4128 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Mattera, R. & Bonifacino, J. S. Ubiquitin binding and conjugation regulate the recruitment of Rabex-5 to early endosomes. EMBO J.27, 2484–2494 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Zhu, H., Qian, H. & Li, G. Delayed onset of positive feedback activation of Rab5 by Rabex-5 and Rabaptin-5 in endocytosis. PLoS ONE5, e9226 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Sato, M. et al. Caenorhabditis elegans RME-6 is a novel regulator of RAB-5 at the clathrin-coated pit. Nat. Cell Biol.7, 559–569 (2005). [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Zhang, W. et al. LET-502/ROCK Regulates Endocytic Recycling by Promoting Activation of RAB-5 in a Distinct Subpopulation of Sorting Endosomes. Cell Rep.32, 108173 (2020). [DOI] [PubMed] [Google Scholar]
27.Semerdjieva, S. et al. Coordinated regulation of AP2 uncoating from clathrin-coated vesicles by rab5 and hRME-6. J. Cell Biol.183, 499–511 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Lauer, J. et al. Auto-regulation of Rab5 GEF activity in Rabex5 by allosteric structural changes, catalytic core dynamics and ubiquitin binding. Elife8, e46302 (2019). [DOI] [PMC free article] [PubMed]
29.Zhang, Z. et al. Molecular mechanism for Rabex-5 GEF activation by Rabaptin-5. Elife3, e02687 (2014). [DOI] [PMC free article] [PubMed]
30.Delprato, A., Merithew, E. & Lambright, D. G. Structure, exchange determinants, and family-wide rab specificity of the tandem helical bundle and Vps9 domains of Rabex-5. Cell118, 607–617 (2004). [DOI] [PubMed] [Google Scholar]
31.Lippe, R., Miaczynska, M., Rybin, V., Runge, A. & Zerial, M. Functional synergy between Rab5 effector Rabaptin-5 and exchange factor Rabex-5 when physically associated in a complex. Mol. Biol. Cell12, 2219–2228 (2001). [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Lodhi, I. J. et al. Gapex-5, a Rab31 guanine nucleotide exchange factor that regulates Glut4 trafficking in adipocytes. Cell Metab.5, 59–72 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Messa, M. et al. Epsin deficiency impairs endocytosis by stalling the actin-dependent invagination of endocytic clathrin-coated pits. Elife3, e03311 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Engqvist-Goldstein, A. E. et al. The actin-binding protein Hip1R associates with clathrin during early stages of endocytosis and promotes clathrin assembly in vitro. J. Cell Biol.154, 1209–1223 (2001). [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Erdmann, K. S. et al. A role of the Lowe syndrome protein OCRL in early steps of the endocytic pathway. Dev. Cell13, 377–390 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Nakatsu, F. et al. Sac2/INPP5F is an inositol 4-phosphatase that functions in the endocytic pathway. J. Cell Biol.209, 85–95 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Cadwallader, K. A., Paterson, H., Macdonald, S. G. & Hancock, J. F. N-terminally myristoylated Ras proteins require palmitoylation or a polybasic domain for plasma membrane localization. Mol. Cell Biol.14, 4722–4730 (1994). [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Palamidessi, A. et al. The GTPase-activating protein RN-tre controls focal adhesion turnover and cell migration. Curr. Biol.23, 2355–2364 (2013). [DOI] [PubMed] [Google Scholar]
39.Haas, A. K., Fuchs, E., Kopajtich, R. & Barr, F. A. A GTPase-activating protein controls Rab5 function in endocytic trafficking. Nat. Cell Biol.7, 887–893 (2005). [DOI] [PubMed] [Google Scholar]
40.Hiragi, S., Matsui, T., Sakamaki, Y. & Fukuda, M. A.-O. TBC1D18 is a Rab5-GAP that coordinates endosome maturation together with Mon1. J. Cell Biol.221, e202201114 (2022). [DOI] [PMC free article] [PubMed]
41.Rao, X. S. et al. AMPK-mediated phosphorylation enhances the auto-inhibition of TBC1D17 to promote Rab5-dependent glucose uptake. Cell Death Differ.28, 3214–3234 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Balla, T. Phosphoinositides: tiny lipids with giant impact on cell regulation. Physiol. Rev.93, 1019–1137 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Wills, R. C. & Hammond, G. R. V. PI(4,5)P2: signaling the plasma membrane. Biochem J.479, 2311–2325 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Shin, H. W. et al. An enzymatic cascade of Rab5 effectors regulates phosphoinositide turnover in the endocytic pathway. J. Cell Biol.170, 607–618 (2005). [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Schmid, S. L. & Mettlen, M. Cell biology: lipid switches and traffic control. Nature499, 161–162 (2013). [DOI] [PubMed] [Google Scholar]
46.Nandez, R. et al. A role of OCRL in clathrin-coated pit dynamics and uncoating revealed by studies of Lowe syndrome cells. Elife3, e02975 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Idevall-Hagren, O. & De Camilli, P. Detection and manipulation of phosphoinositides. Biochim Biophys. Acta1851, 736–745 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Hammond, G. R. et al. PI4P and PI(4,5)P2 are essential but independent lipid determinants of membrane identity. Science337, 727–730 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Wang, J. et al. Spatial organization of PI3K-PI(3,4,5)P(3)-AKT signaling by focal adhesions. Mol. Cell84, 4401–4418.e4409 (2024). [DOI] [PubMed] [Google Scholar]
50.Mehta, Z. B., Pietka, G. & Lowe, M. The cellular and physiological functions of the Lowe syndrome protein OCRL1. Traffic15, 471–487 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Vicinanza, M. et al. OCRL controls trafficking through early endosomes via PtdIns4,5P(2)-dependent regulation of endosomal actin. EMBO J.30, 4970–4985 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Festa, B. P. et al. OCRL deficiency impairs endolysosomal function in a humanized mouse model for Lowe syndrome and Dent disease. Hum. Mol. Genet.28, 1931–1946 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Horiuchi, H. et al. A novel Rab5 GDP/GTP exchange factor complexed to Rabaptin-5 links nucleotide exchange to effector recruitment and function. Cell90, 1149–1159 (1997). [DOI] [PubMed] [Google Scholar]
54.Kirsten, M. L., Baron, R. A., Seabra, M. C. & Ces, O. Rab1a and Rab5a preferentially bind to binary lipid compositions with higher stored curvature elastic energy. Mol. Membr. Biol.30, 303–314 (2013). [DOI] [PubMed] [Google Scholar]
55.Kulakowski, G. et al. Lipid packing defects and membrane charge control RAB GTPase recruitment. Traffic19, 536–545 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Edler, E., Schulze, E. & Stein, M. Membrane localization and dynamics of geranylgeranylated Rab5 hypervariable region. Biochim. Biophys. Acta Biomembr.1859, 1335–1349 (2017). [DOI] [PubMed] [Google Scholar]
57.Wang, B. A.-O. et al. OCRL regulates lysosome positioning and mTORC1 activity through SSX2IP-mediated microtubule anchoring. EMBO Rep.22, e52173 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Oltrabella, F. et al. The Lowe syndrome protein OCRL1 is required for endocytosis in the zebrafish pronephric tubule. PLoS Genet.11, e1005058 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
59.Maxson, M. E., Sarantis, H., Volchuk, A., Brumell, J. H. & Grinstein, S. Rab5 regulates macropinocytosis by recruiting the inositol 5-phosphatases OCRL and Inpp5b that hydrolyse PtdIns(4,5)P2. J. Cell Sci.134, jcs252411 (2021). [DOI] [PubMed]
60.Forozan, F. et al. Molecular cytogenetic analysis of 11 new breast cancer cell lines. Br. J. Cancer81, 1328–1334 (1999). [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Robinson, M. S., Sahlender, D. A. & Foster, S. D. Rapid inactivation of proteins by rapamycin-induced rerouting to mitochondria. Dev. Cell18, 324–331 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
62.Ran, F. A. et al. Genome engineering using the CRISPR-Cas9 system. Nat. Protoc.8, 2281–2308 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
63.Shalem, O. et al. Genome-scale CRISPR-Cas9 knockout screening in human cells. Science343, 84–87 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Bai, X. et al. Structure and dynamics of the EGFR/HER2 heterodimer. Cell Discov.9, 18 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
65.Edelstein, A., Amodaj, N., Hoover, K., Vale, R. & Stuurman, N. Computer control of microscopes using µManager. Curr. Protoc. Mol. Biol.92, 14–20 (2010). [DOI] [PMC free article] [PubMed]
66.Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis. Nat. Methods9, 676–682 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
67.Aguet, F., Antonescu, C. N., Mettlen, M., Schmid, S. L. & Danuser, G. Advances in analysis of low signal-to-noise images link dynamin and AP2 to the functions of an endocytic checkpoint. Dev. Cell26, 279–291 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
68.Lampe, M., Vassilopoulos, S. & Merrifield, C. Clathrin coated pits, plaques and adhesion. J. Struct. Biol.196, 48–56 (2016). [DOI] [PubMed] [Google Scholar]
69.Saffarian, S., Cocucci, E. & Kirchhausen, T. Distinct dynamics of endocytic clathrin-coated pits and coated plaques. PLoS Biol.7, e1000191 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
70.Stirling, D. R. et al. CellProfiler 4: improvements in speed, utility and usability. BMC Bioinform.22, 433 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
71.Kalaidzidis, I. et al. APPL endosomes are not obligatory endocytic intermediates but act as stable cargo-sorting compartments. J. Cell Biol.211, 123–144 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
72.Lin, X., Nair, V., Zhou, Y. & Gorfe, A. A. Membrane potential and dynamics in a ternary lipid mixture: insights from molecular dynamics simulations. Phys. Chem. Chem. Phys.20, 15841–15851 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Supplementary Information^{(26.7MB, pdf)}

41467_2026_70543_MOESM2_ESM.pdf^{(187.1KB, pdf)}

Descriptions of Additional Supplementary Files

Supplementary Movie 1^{(25.1MB, mp4)}

Supplementary Movie 2^{(28MB, mp4)}

Supplementary Movie 3^{(26.4MB, mp4)}

Supplementary Movie 4^{(25.2MB, mp4)}

Supplementary Movie 5^{(21.9MB, mp4)}

Reporting Summary^{(7MB, pdf)}

Transparent Peer Review file^{(20.7MB, pdf)}

Source Data zip file^{(24MB, zip)}

Data Availability Statement

The download links for published codes were provided in the Methods. Other custom MATLAB routines are available upon request from the corresponding author.

[CR1] 1.Schmid, S. L., Sorkin, A. & Zerial, M. Endocytosis: Past, present, and future. Cold Spring Harb. Perspect. Biol.6, a022509 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR2] 2.Cullen, P. J. & Steinberg, F. To degrade or not to degrade: mechanisms and significance of endocytic recycling. Nat. Rev. Mol. Cell Biol.19, 679–696 (2018). [DOI] [PubMed] [Google Scholar]

[CR3] 3.Solinger, J. A. & Spang, A. Sorting of cargo in the tubular endosomal network. Bioessays44, e2200158 (2022). [DOI] [PubMed] [Google Scholar]

[CR4] 4.Maxfield, F. R. Role of endosomes and lysosomes in human disease. Cold Spring Harb. Perspect. Biol.6, a016931 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR5] 5.Villasenor, R., Kalaidzidis, Y. & Zerial, M. Signal processing by the endosomal system. Curr. Opin. Cell Biol.39, 53–60 (2016). [DOI] [PubMed] [Google Scholar]

[CR6] 6.Stenmark, H. Rab GTPases as coordinators of vesicle traffic. Nat. Rev. Mol. Cell Biol.10, 513–525 (2009). [DOI] [PubMed] [Google Scholar]

[CR7] 7.Wandinger-Ness, A. & Zerial, M. Rab proteins and the compartmentalization of the endosomal system. Cold Spring Harb. Perspect. Biol.6, a022616 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR8] 8.Edler, E. & Stein, M. Recognition and stabilization of geranylgeranylated human Rab5 by the GDP Dissociation Inhibitor (GDI). Small GTPases10, 227–242 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR9] 9.Bucci, C. et al. The small GTPase rab5 functions as a regulatory factor in the early endocytic pathway. Cell70, 715–728 (1992). [DOI] [PubMed] [Google Scholar]

[CR10] 10.Stenmark, H. et al. Inhibition of rab5 GTPase activity stimulates membrane fusion in endocytosis. EMBO J.13, 1287–1296 (1994). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR11] 11.Chen, P. I., Kong, C., Su, X. & Stahl, P. D. Rab5 isoforms differentially regulate the trafficking and degradation of epidermal growth factor receptors. J. Biol. Chem.284, 30328–30338 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR12] 12.Gurkan, C. et al. Large-scale profiling of Rab GTPase trafficking networks: the membrome. Mol. Biol. Cell16, 3847–3864 (2005). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR13] 13.Hatoyama, Y., Homma, Y., Hiragi, S. & Fukuda, M. Establishment and analysis of conditional Rab1- and Rab5-knockout cells using the auxin-inducible degron system. J. Cell Sci.134, jcs259184 (2021). [DOI] [PubMed]

[CR14] 14.Homma, Y. et al. Comprehensive knockout analysis of the Rab family GTPases in epithelial cells. J. Cell Biol.218, 2035–2050 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR15] 15.He, K. et al. Dynamics of phosphoinositide conversion in clathrin-mediated endocytic traffic. Nature552, 410–414 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR16] 16.Kirchhausen, T., Owen, D. & Harrison, S. C. Molecular structure, function, and dynamics of clathrin-mediated membrane traffic. Cold Spring Harb. Perspect. Biol.6, a016725 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR17] 17.He, K. et al. Dynamics of Auxilin 1 and GAK in clathrin-mediated traffic. J. Cell Biol.219, e201908142 (2020). [DOI] [PMC free article] [PubMed]

[CR18] 18.Massol, R. H., Boll, W., Griffin, A. M. & Kirchhausen, T. A burst of auxilin recruitment determines the onset of clathrin-coated vesicle uncoating. Proc. Natl. Acad. Sci. USA103, 10265–10270 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR19] 19.Oliinyk, O. S., Shemetov, A. A., Pletnev, S., Shcherbakova, D. M. & Verkhusha, V. V. Smallest near-infrared fluorescent protein evolved from cyanobacteriochrome as versatile tag for spectral multiplexing. Nat. Commun.10, 279 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR20] 20.Blumer, J. et al. RabGEFs are a major determinant for specific Rab membrane targeting. J. Cell Biol.200, 287–300 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR21] 21.Hadano, S. et al. ALS2CL, the novel protein highly homologous to the carboxy-terminal half of ALS2, binds to Rab5 and modulates endosome dynamics. FEBS Lett.575, 64–70 (2004). [DOI] [PubMed] [Google Scholar]

[CR22] 22.Zhu, H. et al. Rabaptin-5-independent membrane targeting and Rab5 activation by Rabex-5 in the cell. Mol. Biol. Cell18, 4119–4128 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR23] 23.Mattera, R. & Bonifacino, J. S. Ubiquitin binding and conjugation regulate the recruitment of Rabex-5 to early endosomes. EMBO J.27, 2484–2494 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR24] 24.Zhu, H., Qian, H. & Li, G. Delayed onset of positive feedback activation of Rab5 by Rabex-5 and Rabaptin-5 in endocytosis. PLoS ONE5, e9226 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR25] 25.Sato, M. et al. Caenorhabditis elegans RME-6 is a novel regulator of RAB-5 at the clathrin-coated pit. Nat. Cell Biol.7, 559–569 (2005). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR26] 26.Zhang, W. et al. LET-502/ROCK Regulates Endocytic Recycling by Promoting Activation of RAB-5 in a Distinct Subpopulation of Sorting Endosomes. Cell Rep.32, 108173 (2020). [DOI] [PubMed] [Google Scholar]

[CR27] 27.Semerdjieva, S. et al. Coordinated regulation of AP2 uncoating from clathrin-coated vesicles by rab5 and hRME-6. J. Cell Biol.183, 499–511 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR28] 28.Lauer, J. et al. Auto-regulation of Rab5 GEF activity in Rabex5 by allosteric structural changes, catalytic core dynamics and ubiquitin binding. Elife8, e46302 (2019). [DOI] [PMC free article] [PubMed]

[CR29] 29.Zhang, Z. et al. Molecular mechanism for Rabex-5 GEF activation by Rabaptin-5. Elife3, e02687 (2014). [DOI] [PMC free article] [PubMed]

[CR30] 30.Delprato, A., Merithew, E. & Lambright, D. G. Structure, exchange determinants, and family-wide rab specificity of the tandem helical bundle and Vps9 domains of Rabex-5. Cell118, 607–617 (2004). [DOI] [PubMed] [Google Scholar]

[CR31] 31.Lippe, R., Miaczynska, M., Rybin, V., Runge, A. & Zerial, M. Functional synergy between Rab5 effector Rabaptin-5 and exchange factor Rabex-5 when physically associated in a complex. Mol. Biol. Cell12, 2219–2228 (2001). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR32] 32.Lodhi, I. J. et al. Gapex-5, a Rab31 guanine nucleotide exchange factor that regulates Glut4 trafficking in adipocytes. Cell Metab.5, 59–72 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR33] 33.Messa, M. et al. Epsin deficiency impairs endocytosis by stalling the actin-dependent invagination of endocytic clathrin-coated pits. Elife3, e03311 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR34] 34.Engqvist-Goldstein, A. E. et al. The actin-binding protein Hip1R associates with clathrin during early stages of endocytosis and promotes clathrin assembly in vitro. J. Cell Biol.154, 1209–1223 (2001). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR35] 35.Erdmann, K. S. et al. A role of the Lowe syndrome protein OCRL in early steps of the endocytic pathway. Dev. Cell13, 377–390 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR36] 36.Nakatsu, F. et al. Sac2/INPP5F is an inositol 4-phosphatase that functions in the endocytic pathway. J. Cell Biol.209, 85–95 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR37] 37.Cadwallader, K. A., Paterson, H., Macdonald, S. G. & Hancock, J. F. N-terminally myristoylated Ras proteins require palmitoylation or a polybasic domain for plasma membrane localization. Mol. Cell Biol.14, 4722–4730 (1994). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR38] 38.Palamidessi, A. et al. The GTPase-activating protein RN-tre controls focal adhesion turnover and cell migration. Curr. Biol.23, 2355–2364 (2013). [DOI] [PubMed] [Google Scholar]

[CR39] 39.Haas, A. K., Fuchs, E., Kopajtich, R. & Barr, F. A. A GTPase-activating protein controls Rab5 function in endocytic trafficking. Nat. Cell Biol.7, 887–893 (2005). [DOI] [PubMed] [Google Scholar]

[CR40] 40.Hiragi, S., Matsui, T., Sakamaki, Y. & Fukuda, M. A.-O. TBC1D18 is a Rab5-GAP that coordinates endosome maturation together with Mon1. J. Cell Biol.221, e202201114 (2022). [DOI] [PMC free article] [PubMed]

[CR41] 41.Rao, X. S. et al. AMPK-mediated phosphorylation enhances the auto-inhibition of TBC1D17 to promote Rab5-dependent glucose uptake. Cell Death Differ.28, 3214–3234 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR42] 42.Balla, T. Phosphoinositides: tiny lipids with giant impact on cell regulation. Physiol. Rev.93, 1019–1137 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR43] 43.Wills, R. C. & Hammond, G. R. V. PI(4,5)P2: signaling the plasma membrane. Biochem J.479, 2311–2325 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR44] 44.Shin, H. W. et al. An enzymatic cascade of Rab5 effectors regulates phosphoinositide turnover in the endocytic pathway. J. Cell Biol.170, 607–618 (2005). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR45] 45.Schmid, S. L. & Mettlen, M. Cell biology: lipid switches and traffic control. Nature499, 161–162 (2013). [DOI] [PubMed] [Google Scholar]

[CR46] 46.Nandez, R. et al. A role of OCRL in clathrin-coated pit dynamics and uncoating revealed by studies of Lowe syndrome cells. Elife3, e02975 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR47] 47.Idevall-Hagren, O. & De Camilli, P. Detection and manipulation of phosphoinositides. Biochim Biophys. Acta1851, 736–745 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR48] 48.Hammond, G. R. et al. PI4P and PI(4,5)P2 are essential but independent lipid determinants of membrane identity. Science337, 727–730 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR49] 49.Wang, J. et al. Spatial organization of PI3K-PI(3,4,5)P(3)-AKT signaling by focal adhesions. Mol. Cell84, 4401–4418.e4409 (2024). [DOI] [PubMed] [Google Scholar]

[CR50] 50.Mehta, Z. B., Pietka, G. & Lowe, M. The cellular and physiological functions of the Lowe syndrome protein OCRL1. Traffic15, 471–487 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR51] 51.Vicinanza, M. et al. OCRL controls trafficking through early endosomes via PtdIns4,5P(2)-dependent regulation of endosomal actin. EMBO J.30, 4970–4985 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR52] 52.Festa, B. P. et al. OCRL deficiency impairs endolysosomal function in a humanized mouse model for Lowe syndrome and Dent disease. Hum. Mol. Genet.28, 1931–1946 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR53] 53.Horiuchi, H. et al. A novel Rab5 GDP/GTP exchange factor complexed to Rabaptin-5 links nucleotide exchange to effector recruitment and function. Cell90, 1149–1159 (1997). [DOI] [PubMed] [Google Scholar]

[CR54] 54.Kirsten, M. L., Baron, R. A., Seabra, M. C. & Ces, O. Rab1a and Rab5a preferentially bind to binary lipid compositions with higher stored curvature elastic energy. Mol. Membr. Biol.30, 303–314 (2013). [DOI] [PubMed] [Google Scholar]

[CR55] 55.Kulakowski, G. et al. Lipid packing defects and membrane charge control RAB GTPase recruitment. Traffic19, 536–545 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR56] 56.Edler, E., Schulze, E. & Stein, M. Membrane localization and dynamics of geranylgeranylated Rab5 hypervariable region. Biochim. Biophys. Acta Biomembr.1859, 1335–1349 (2017). [DOI] [PubMed] [Google Scholar]

[CR57] 57.Wang, B. A.-O. et al. OCRL regulates lysosome positioning and mTORC1 activity through SSX2IP-mediated microtubule anchoring. EMBO Rep.22, e52173 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR58] 58.Oltrabella, F. et al. The Lowe syndrome protein OCRL1 is required for endocytosis in the zebrafish pronephric tubule. PLoS Genet.11, e1005058 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR59] 59.Maxson, M. E., Sarantis, H., Volchuk, A., Brumell, J. H. & Grinstein, S. Rab5 regulates macropinocytosis by recruiting the inositol 5-phosphatases OCRL and Inpp5b that hydrolyse PtdIns(4,5)P2. J. Cell Sci.134, jcs252411 (2021). [DOI] [PubMed]

[CR60] 60.Forozan, F. et al. Molecular cytogenetic analysis of 11 new breast cancer cell lines. Br. J. Cancer81, 1328–1334 (1999). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR61] 61.Robinson, M. S., Sahlender, D. A. & Foster, S. D. Rapid inactivation of proteins by rapamycin-induced rerouting to mitochondria. Dev. Cell18, 324–331 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR62] 62.Ran, F. A. et al. Genome engineering using the CRISPR-Cas9 system. Nat. Protoc.8, 2281–2308 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR63] 63.Shalem, O. et al. Genome-scale CRISPR-Cas9 knockout screening in human cells. Science343, 84–87 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR64] 64.Bai, X. et al. Structure and dynamics of the EGFR/HER2 heterodimer. Cell Discov.9, 18 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR65] 65.Edelstein, A., Amodaj, N., Hoover, K., Vale, R. & Stuurman, N. Computer control of microscopes using µManager. Curr. Protoc. Mol. Biol.92, 14–20 (2010). [DOI] [PMC free article] [PubMed]

[CR66] 66.Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis. Nat. Methods9, 676–682 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR67] 67.Aguet, F., Antonescu, C. N., Mettlen, M., Schmid, S. L. & Danuser, G. Advances in analysis of low signal-to-noise images link dynamin and AP2 to the functions of an endocytic checkpoint. Dev. Cell26, 279–291 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR68] 68.Lampe, M., Vassilopoulos, S. & Merrifield, C. Clathrin coated pits, plaques and adhesion. J. Struct. Biol.196, 48–56 (2016). [DOI] [PubMed] [Google Scholar]

[CR69] 69.Saffarian, S., Cocucci, E. & Kirchhausen, T. Distinct dynamics of endocytic clathrin-coated pits and coated plaques. PLoS Biol.7, e1000191 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR70] 70.Stirling, D. R. et al. CellProfiler 4: improvements in speed, utility and usability. BMC Bioinform.22, 433 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR71] 71.Kalaidzidis, I. et al. APPL endosomes are not obligatory endocytic intermediates but act as stable cargo-sorting compartments. J. Cell Biol.211, 123–144 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR72] 72.Lin, X., Nair, V., Zhou, Y. & Gorfe, A. A. Membrane potential and dynamics in a ternary lipid mixture: insights from molecular dynamics simulations. Phys. Chem. Chem. Phys.20, 15841–15851 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Lipid turnover and GEF recruitment collectively determine Rab5 recruitment and activation during the first step of early endosome formation

Yongtao Du

Xilin Miao

Yu Gao

Yu Liang

Xue Bai

Song Dang

Zhongsheng Wu

Zhengyang An

Xinyi Zhang

Yaran Yang

Zhe Zhang

Huaqing Cai

Kangmin He

Abstract

Introduction

Results

Rab5 is recruited specifically to uncoated endocytic carriers in genome-edited cells

Fig. 1. Simultaneous knockout of Rabex5 and hRME6 expression eliminates Rab5 recruitment to uncoated endocytic carriers.

Simultaneous depletion of hRME6 and Rabex5 eliminates Rab5 recruitment to uncoated endocytic carriers and abolishes Rab5-positive early endosome formation

Distinct recruitment dynamics of endogenous Rabex5 and hRME6 to uncoated endocytic carriers

Fig. 2. Distinct recruitment dynamics of endogenous Rabex5 and hRME6 to uncoated endocytic carriers in genome-edited cells.

The Vps9 domain of Rabex5 is necessary for Rabex5 recruitment to uncoated endocytic carriers and Rab5-positive early endosome formation

Fig. 3. Influence of various Rabex5 or hRME6 constructs on Rab5 recruitment and Rab5-positive early endosome formation in cells with Rabex5/hRME6-DKO.

A burst of hRME6 recruitment triggers Rab5 recruitment to uncoated endocytic carriers and promotes Rab5-positive early endosome formation

Targeting the Vps9 domain of Rabex5 to clathrin-coated pits fails to induce Rab5 recruitment and Rab5-positive early endosome formation

Fig. 4. Targeting the Vps9 domain to uncoated endocytic carriers and various organelles, but not to clathrin-coated pits/vesicles or the plasma membrane, induces Rab5 recruitment.

Targeting Rabex5 or its Vps9 domain to various organelle membranes, but not the plasma membrane, induces Rab5 recruitment

Fig. 5. Mistargeted Vps9 domain and nucleotide binding status regulate Rab5 recruitment.

The Vps9 domain of Rabex5 interacts with and recruits GDP-bound Rab5

PI(4,5)P2 inhibits the recruitment of Rab5 to the plasma membrane in cells

Fig. 6. The presence of PI(4,5)P2 at the plasma membrane impedes the recruitment of Rab5.

PI(4,5)P2 inhibits Rab5 recruitment to membranes of PI(4,5)P2-containing liposomes

Defective PI(4,5)P2 hydrolysis reduces early endosome formation in Lowe syndrome cells

Fig. 7. PI(4,5)P2 regulates Rab5 recruitment and early endosome formation.

Discussion

Methods

Cell culture and chemical treatment

Plasmids and transfection

Generation of knock-in cells using the CRISPR/Cas9 approach

Generation of knockout cells using the CRISPR/Cas9 approach

Knockdown of Rabex5, hRME6, and Rabaptin5 using siRNA

Knockdown of RN-tre using shRNA

Live-cell TIRF imaging and imaging analysis

Live-cell spinning-disk confocal imaging and imaging analysis

Immunofluorescence

Assessment of transferrin and EGF uptake by flow cytometry

Expression and purification of proteins

GEF assay

Liposome flotation assay

Western blot analysis

Co-immunoprecipitation

Statistical tests

Reporting summary

Supplementary information

Source data

Acknowledgements

Author contributions

Peer review

Peer review information

Data availability

Code availability

Competing interests

Footnotes

Contributor Information

Supplementary information

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases

PI(4,5)P₂ inhibits the recruitment of Rab5 to the plasma membrane in cells

Fig. 6. The presence of PI(4,5)P₂ at the plasma membrane impedes the recruitment of Rab5.

PI(4,5)P₂ inhibits Rab5 recruitment to membranes of PI(4,5)P₂-containing liposomes

Defective PI(4,5)P₂ hydrolysis reduces early endosome formation in Lowe syndrome cells

Fig. 7. PI(4,5)P₂ regulates Rab5 recruitment and early endosome formation.