Identification of Rab30 as a novel regulator of tubular endosomes

Abstract

Tubular-shaped recycling endosomes, known as tubular endosomes, are present in certain types of cells, including HeLa cells, and they regulate the recycling of clathrin-independent endocytosed (CIE) cargo proteins to the plasma membrane. Several key regulators of tubular endosomes, including Rab small GTPases and related proteins, have been identified thus far, but the entire process of tubular endosome formation is not yet fully understood. We previously showed that expression of a Golgi-related Rab-GTPase-activating protein TBC1D22B in HeLa cells caused their tubular structures to disappear, suggesting a possible link between tubular endosome formation and a certain Golgi function(s). However, nothing is known about the target Rab(s) of TBC1D22B in tubular endosome formation or about the functional relationship between tubular endosomes and the Golgi apparatus. Here, we performed comprehensive Rab-knockdown screening in combination with dominant-negative Rab expression and succeeded in identifying 12 Rabs as regulators of tubular endosome formation. One of them, Rab30, a Golgi-resident Rab, is a novel target of TBC1D22B and involved in both tubular endosome formation and CIE cargo trafficking. We also showed that a Rab30–BICD2–KIF5B axis is likely to be involved in tubular endosome formation. Our findings suggest the importance of Rab30-mediated post-Golgi trafficking in tubular endosome formation.

Key words: Golgi, GTPase-activating protein (GAP), Rab30, siRNA screening, tubular endosome

Graphical Abstract

Introduction

Endocytosis is a process by which cells uptake nutrients or transmit extracellular signals to the inside of the cell via highly conserved pathways. The best characterized pathway is the clathrin-mediated endocytosis (CME) pathway (Kaksonen and Roux, 2018), but some cargo proteins are internalized even in the absence of CME (Damke et al., 1995; Lamaze et al., 2001). The CME-independent pathways are collectively referred to as clathrin-independent endocytosis (CIE) pathways (Mayor and Pagano, 2007), and many CIE cargo proteins have been reported thus far (Grant and Donaldson, 2009). Internalized cargo proteins are initially sorted into early endosomes and then delivered to their own destinations, i.e., to lysosomes for degradation, to the Golgi apparatus, or to the plasma membrane for recycling (Maxfield and McGraw, 2004). In certain types of cells, e.g., HeLa cells, tubular-shaped recycling endosomes (so-called tubular endosomes) are utilized to recycle some CIE cargo proteins to the plasma membrane. Although key regulators of tubular endosome formation, including Arf6, MICAL-L1, and EHD1, have gradually been identified (Radhakrishna and Donaldson, 1997; Caplan et al., 2002; Sharma et al., 2009), the significance of the tubular structure is not fully understood.

Rab small GTPases are key regulators of membrane traffic in all eukaryotes and together with appropriate Rab effectors regulate various steps or types of membrane traffic (Stenmark, 2009; Pfeffer, 2013; Homma et al., 2021). Rab activity, i.e., cycling between GTP-bound active Rab and GDP-bound inactive Rab, is controlled by a guanine nucleotide exchange factor (GEF) and a GTPase-activating protein (GAP) (Barr and Lambright, 2010). Previous studies have shown that several Rabs are involved in tubular endosome formation or CIE cargo trafficking (Weigert et al., 2004; Hattula et al., 2006; Sharma et al., 2009; Rahajeng et al., 2012; Delevoye et al., 2014; Del Olmo et al., 2019; Etoh and Fukuda, 2019), however, the investigations mainly focused on the localization of these Rabs on tubular endosomes. We recently screened for Rab-GAPs involved in tubular endosome formation and identified four TBC (Tre-2/Bub2/Cdc16)-domain-containing Rab-GAPs, whose GAP activity is required for tubular endosome formation (Nakashima and Fukuda, 2025). One of them, TBC1D22B, is a Golgi-related protein (Haas et al., 2007; Greninger et al., 2013) and has been shown to serve as a GAP for two Golgi-resident Rabs, Rab1B and Rab33B, in vitro (Pan et al., 2006; Martino et al., 2025). However, whether these Golgi-resident Rabs are actually involved in tubular endosome formation remains unknown, and almost nothing is known about the functional communication between tubular endosomes and the Golgi apparatus.

In this study, we comprehensively screened 58 Rabs that are conserved in mammals and found that 12 Rabs, including some non-tubular-endosome-localized Rabs, are involved in tubular endosome formation. One of them, Rab30, a Golgi-resident Rab, is a novel target of TBC1D22B both in vitro and in cultured cells. We also showed that Rab30-knockout (KO) in HeLa cells impairs both tubular endosome formation and CIE cargo trafficking. Moreover, we found that a Rab30-binding protein, BICD2, and KIF5B are required for tubular endosome formation. Our findings indicate the importance of the Golgi apparatus, presumably of post-Golgi trafficking, in tubular endosome formation.

Materials and Methods

Materials

The specific materials, including cell lines, antibodies (Mrozowska and Fukuda, 2016), primers, plasmids (Tsuboi and Fukuda, 2006; Ishibashi et al., 2009; Oguchi et al., 2017, 2020), and siRNAs, used in this study are summarized in Table S1. The KD efficiency of Rab siRNAs was previously evaluated by co-transfecting each Rab siRNA and its corresponding EGFP-tagged Rab (Aizawa and Fukuda, 2015). Unless otherwise specified, all other general reagents used in this study were of analytical grade or the highest grade commercially available.

Plasmid construction

The cDNAs used in this study were obtained by performing PCR with specific primers and subcloning into the appropriate vectors listed in Table S1. Plasmids were prepared by the standard molecular biology techniques.

Cell culture, transfections, infection, and establishment of Rab30-KO cells

Cells listed in Table S1 were grown at 37°C in Dulbecco’s modified Eagle’s medium (D-MEM), supplemented with 10% fetal bovine serum, 100 U/mL penicillin G, and 100 μg/mL streptomycin, in a 5% CO₂ incubator. Plasmids and siRNAs were transfected into these cultured cells by using Lipofectamine 2000 or 3000 and RNAiMAX (Thermo Fisher Scientific, Waltham, MA, USA), respectively, each according to the manufacturer’s instructions.

pMRX vectors and pLP/VSVG were cotransfected into PlatE cells, and medium containing retroviruses was prepared as described previously (Nakashima and Fukuda, 2025). To concentrate retroviruses, virus-containing medium was mixed with 4×polyethylene glycol (PEG) 6000 solution (32 w/v% PEG 6000, 400 mM NaCl, and 40 mM HEPES-NaOH, pH 7.4) and incubated at 4°C for 18 h. The mixture was then centrifuged at 2600 g at 4°C for 20 min. After removing the supernatant, the precipitated viruses were resuspended in the medium. HeLa cells were infected by using the retrovirus-containing medium and 8 μg/mL polybrene (Sigma-Aldrich, St. Louis, MO, USA), and uninfected cells were removed with a 2 μg/mL concentration of puromycin (Merck Millipore) or 5 μg/mL concentration of blastcidin S (FUJIFILM Wako Pure Chemical, Osaka, Japan).

Rab30-KO HeLa cells were established as described previously (Nakashima et al., 2023). A single-guide RNA (sgRNA) sequence was designed by using SYNTHEGO gRNA design tools (https://design.synthego.com/#/).

Immunoblotting and immunofluorescence analysis

SDS-PAGE and immunoblotting were performed as described previously (Nakashima and Fukuda, 2025). Immunoreactive bands were detected by using enhanced chemiluminescence and the ChemiDoc Touch imaging system (Bio-Rad, Hercules, CA, USA).

Immunofluorescence staining, including cell fixation with 4% paraformaldehyde, permeabilization with 50 μg/mL digitonin, blocking with 3% bovine serum albumin, and incubation with primary and secondary antibodies, was also performed essentially as described previously (Nakashima and Fukuda, 2025). To remove cytosolic components, cells were treated with 50 μg/mL digitonin for 30 sec and washed with PBS before fixation. All samples were examined with a confocal fluorescence microscope (Fluoview 1000; Evident/Olympus, Tokyo, Japan) equipped with a Plan-Apochromat 100×/1.45 NA oil-immersion objective lens or a Plan-Apochromat 60×/1.35 NA oil-immersion objective lens (Evident/Olympus). The images acquired were processed with ImageJ software (version 1.54r; https://imagej.nih.gov/ij/index.html).

Purification of recombinant proteins from COS-7 cell lysates

3×FLAG-tagged TBC1D22B (WT or RK) or FLAG-tagged Rab30 or Rab3B was transiently expressed in COS-7 cells, and the cells were lysed with a lysis buffer (50 mM HEPES-KOH, pH 7.2, 150 mM NaCl, 1% Triton X-100, and a protease inhibitor cocktail [Roche, Penzberg, Germany]) containing 1 mM EDTA. After centrifugation at 17,900 g at 4°C for 10 min, the supernatant was recovered, and the 3×FLAG-tagged or FLAG-tagged proteins in the supernatant were purified with anti-FLAG M2 magnetic beads (Sigma-Aldrich). In brief, the COS-7 cell lysates were incubated at 4°C for 1 h with the beads, and the beads were then washed three times with a wash buffer A (0.1% Triton X-100, 50 mM HEPES-KOH, pH 7.2, and 150 mM NaCl) and three times with a wash buffer B (50 mM HEPES-KOH, pH 7.2, and 150 mM NaCl). The beads were incubated, with agitation, at 4°C for 30 min in a 500 μg/mL 3×FLAG peptide (Sigma-Aldrich) solution. Finally, the supernatant containing the purified 3×FLAG-tagged or FLAG-tagged proteins was collected, and the purity of the recombinant proteins were assessed by 12% SDS-PAGE followed by staining with Coomassie Brilliant Blue Rapid Stain (BIOCRAFT, Tokyo, Japan).

In vitro GAP assay

The GAP activity of TBC1D22B toward Rab30 and Rab3B was measured by using a GTPase-Glo assay kit (Promega, Madison, WI, USA). For GTP loading, purified FLAG-tagged Rab30 (or Rab3B) was incubated with a loading buffer (50 mM HEPES-KOH, pH 7.2, 150 mM NaCl, and 5 mM EDTA) for 2 h at room temperature, and then 0.5 mM MgCl₂, 1 mM DTT, and 0.1 mM GTP were added to the solution. After incubation for 5 min on ice, excessive free GTP was removed by using a 10K ultrafiltration spin column (APRO Science, Naruto, Japan). The GTP-loaded Rab30 (or Rab3B) was mixed with 1 μM of 3×FLAG-TBC1D22B (WT or RK) (or BSA as a negative control in the presence of 3×FLAG-peptide), 100 μM of DTT, and 1 μM of GTP, and the cocktail was incubated at 30°C for 0 h or 1 h. GTP-ATP conversion and detection of ATP were performed according to the manufacture’s instruction by using a Victor Nivo Multimode microplate reader (PerkinElmer, Waltham, MA, USA). The data obtained were normalized by both 0 min input and the BSA control data (Fig. 3D and S2B).

Anti-CD147 recycling and transferrin (Tf) uptake assays

Anti-CD147 recycling and Alexa Fluor594-conjugated Tf (Alexa594-Tf; Thermo Fisher Scientific) uptake assays were performed as described previously (Nakashima et al., 2023).

Cell surface biotinylation assay

WT, Rab30-KO, and rescued HeLa cells were washed twice with ice-cold PBS (+) (PBS containing 0.1 mM CaCl₂ and 0.1 mM MgCl₂) and incubated for 30 min at 4°C with 0.5 mg/mL Sulfo-NHS-LC-biotin (Thermo Fisher Scientific) in PBS (+). The reaction was quenched by washing twice with a quenching buffer (100 mM glycine and 0.3% BSA in PBS (+)), and the dishes were subsequently washed twice with PBS (+). The cells were then lysed with the lysis buffer containing 1 mM EDTA described above. The supernatant was incubated for 1 h at 4°C with High Capacity Streptavidin Agarose beads (Thermo Fisher Scientific). The beads were then washed three times with the lysis buffer, twice with a high-salt wash buffer (50 mM Tris-HCl, pH 7.5, 500 mM NaCl, and 1% Triton X-100), and once with 50 mM Tris-HCl, pH 7.5. The samples were boiled with an SDS sample buffer, and then immunoblotted with the antibodies indicated in Fig. 4G.

Purification of GST-Rab30

GST-tagged Rab30 (Itoh et al., 2006) and GST were expressed in Escherichia coli JM109 and purified as described previously (Kuroda and Fukuda, 2005).

In vitro binding assay

3×FLAG-tagged BICD2-CC3 was transiently expressed in COS-7 cells, and the lysate was collected as described above. A 0.5 mM GTPγS or 1 mM GDP solution was loaded into 5 μg of GST-tagged Rab30 or GST alone as described above. The GST fusion proteins, the cell lysate, and anti-FLAG M2 magnetic beads were mixed and incubated for 1 h at 4°C. The samples were washed three times with wash buffer A, boiled with an SDS sample buffer, and then immunoblotted with the antibodies indicated in Fig. 5F.

Quantification and statistical analysis

The number of cells containing at least one >20 μm tubule was manually counted (the number of cells analyzed in each experiment is stated in the legends of Fig. 1B, 2B, D, 4C, 5C, I, S6B, and S7B). The Pearson correlation coefficient (PCC) between endogenous TGN46 (or GM130) and EGFP-tagged Rab was automatically measured by the ImageJ software (20 cells were analyzed in each experiment; Fig. 3B and S1C). The number of EGFP-Rab30-positive dots was automatically counted by ImageJ (15 cells were analyzed in each experiment; Fig. 3F and 5E). Total fluorescence intensity of CD147 and Alexa594-Tf was also measured by ImageJ (30 cells were analyzed in each experiment; Fig. 4E, F, and S5B). The intensity of surface and total CD147 bands was also automatically measured by ImageJ, and the values were normalized by total CD147 and WT sample (Fig. 4H). The experiments were independently repeated three times for each analysis, and quantitative data are expressed as the means ± s.e.m. One-way ANOVA and Tukey’s test or Dunnett’s test (for multiple comparisons) were performed with Prism4 (GraphPad software, version 4.0a), and the two-tailed unpaired Student’s t-test was performed with Excel software (Microsoft, Redmond, WA, USA, San Diego, CA, USA). The following criteria were used for statistical significance: *, p<0.05; **, p<0.01; and ***, p<0.001.

Fig. 1

Identification of candidate Rabs involved in tubular endosome formation in HeLa cells by comprehensive KD screening

(A) Typical images of MICAL-L1 in Rab10-KD (siRab10#1), Rab21-KD (siRab21#1), Rab22A-KD (siRab22A#1), Rab30-KD (siRab30#1), Rab35-KD (siRab35#1), and Control (siControl) HeLa cells. Note that Rab10-KD, Rab22A-KD, and Rab30-KD caused a reduction in MICAL-L1-positive tubular structures, whereas Rab21-KD and Rab35-KD had no effect. Scale bars: 20 μm or 2 μm (inset). (B) Quantitative analysis of MICAL-L1-positive tubular structures in Rab-KD cells shown in (A). The graph shows the percentage of cells containing at least one >20 μm tubule per 10 cells. Each data set was normalized by the siControl cell data. The data shown are means ± s.e.m. of the data obtained in three independent experiments. The broken red line indicates that 50% of the control, and the cyan bars represent candidate Rabs whose knockdown reduced the percentage of cells containing a >20 μm tubule to less than 50% in comparison with the siControl cells.

Fig. 2

Secondary screening for candidate Rabs involved in tubular endosome formation by using independent siRNAs and by overexpressing constitutively negative Rab mutants

(A) Typical images of MICAL-L1 in Rab1A-KD (siRab1A#2), Rab10-KD (siRab10#2), Rab30-KD (siRab30#2) and control (siControl) HeLa cells. Note that Rab10-KD and Rab30-KD caused a reduction in MICAL-L1-positive tubular structures, whereas Rab1A-KD had no effect. Scale bars: 20 μm or 2 μm (inset). (B) Quantitative analysis of MICAL-L1-positive tubular structures in Rab-KD cells shown in (A). The graph shows the percentage of cells containing at least one >20 μm tubule per 20 cells. Each data set was normalized by the siControl cell data. The data shown are means ± s.e.m. of the data obtained in three independent experiments. The broken red line indicates that 50% of the control, and the cyan bars represent candidate Rabs whose knockdown reduced the percentage of cells containing a >20 μm tubule to less than 50% in comparison with the siControl cells. (C) Typical images of MICAL-L1 (red) in HeLa cells transiently expressing EGFP-tagged Rab5A-SN, Rab27A-TN, and Rab30-TN (green). Note that Rab5A-SN and Rab30-TN overexpression caused a reduction in MICAL-L1-positive tubular structures, whereas Rab27A-TN overexpression had no effect. Scale bars: 20 μm or 2 μm (inset). (D) Quantitative analysis of MICAL-L1-positive tubular structures in cells overexpressing constitutively negative Rab mutants shown in (C). The graph shows the percentage of cells containing at least one >20 μm tubule per 20 cells. Each data set was normalized by the data of cells expressing EGFP alone (negative control). The data shown are means ± s.e.m. of the data obtained in three independent experiments. The broken red line indicates that 50% of the control, and the cyan bars represent candidate Rabs whose constitutively-negative-mutant-overexpression reduced the percentage of cells containing a >20 μm tubule to less than 50% in comparison with control EGFP-expressing cells. (E) Venn diagram of the results shown in (B) and (D).

Fig. 3

Rab30 is a novel target of TBC1D22B

(A) Typical images of TGN46 (red) in HeLa cells transiently expressing EGFP-tagged Rab10, Rab30, and control EGFP alone (green). Note that Rab30 colocalized well with TGN46-positive compartments, whereas Rab10 colocalized less. Scale bars: 20 μm or 2 μm (inset). (B) The PCCs between endogenous TGN46 and EGFP-tagged Rabs shown in (A). The data shown are means ± s.e.m. of the data obtained in three independent experiments. The broken red line indicates that a PCC value of 0.5. The cyan bars represent Rabs that colocalized well with TGN46. (C) Purified recombinant 3×FLAG-tagged TBC1D22B (WT and RK) and FLAG-tagged Rab30 proteins from COS-7 cells visualized by Coomassie Brilliant Blue (CBB) staining. Mr, molecular mass markers. (D) In vitro Rab30-GAP activity of TBC1D22B as assessed with the GTPase-Glo assay kit (see Materials and Methods for details). In brief, GTP-loaded FLAG-tagged Rab30 was incubated for 0 h or 1 h with either 3×FLAG-tagged TBC1D22B-WT, TBC1D22B-RK, or BSA (negative control), and GTP in the reaction mixtures was detected as luminescence with the GTPase-Glo assay kit. Note that luminescence (i.e., GTP-Rab30) was significantly reduced only in the presence of TBC1D22B-WT. Each value was normalized by both 0 min and the BSA control data. The data shown are means ± s.e.m. of the data obtained in three independent experiments. *, p<0.05; **, p<0.01; NS, not significant (one-way ANOVA and Tukey’s test). (E) Typical images of EGFP-tagged Rab30 (green) in HeLa cells transiently expressing mStr-tagged TBC1D22B (WT and RK) (red). Note that overexpression of WT TBC1D22B decreased the number of Rab30-positive dots outside the Golgi region, whereas its RK mutants had no effect. Insets showed magnified views of the boxed areas. Scale bars: 20 μm or 2 μm (insets). (F) Quantitative analysis of the Rab30-positive dots in the mStr-tagged TBC1D22B (WT or RK)-expressing cells shown in (E). The graph shows the number of Rab30-positive dots measuring less than 1 μm². The data shown are means ± s.e.m. of the data obtained in three independent experiments. (n = 15 cells in each experiment). *, p<0.05; **, p<0.01; NS, not significant (one-way ANOVA and Tukey’s test).

Fig. 4

Rab30 is required for both tubular endosome formation and CIE cargo trafficking.

(A) Expression of Rab30 protein in parental cells (WT), Rab30-KO cells (–), and rescued cells. The positions of the molecular mass markers (in kDa) are shown on the left. The arrowhead and asterisk indicate the positions of Rab30 and non-specific bands, respectively. (B) Typical images of MICAL-L1 in parental cells (WT), Rab30-KO cells (–), and rescued cells. Note that MICAL-L1-positive tubular structures were clearly diminished in the Rab30-KO cells and that tubular structures were completely recovered by re-expression of Rab30. Insets showed magnified views of the boxed areas. Scale bars: 20 μm or 2 μm (insets). (C) Quantitative analysis of MICAL-L1-positive tubular structures in the parental cells (WT), Rab30-KO cells (–), and rescued cells shown in (B). The graph shows the percentage of cells containing at least one >20 μm tubule per 40 cells. The data shown are means ± s.e.m. of the data obtained in three independent experiments. **, p<0.01; ***, p<0.001; NS, not significant (one-way ANOVA and Tukey’s test). (D) Typical images of internalized CD147 in parental cells (WT), Rab30-KO cells (–), and rescued cells at the times indicated. Note that the CD147-positive signals decreased in a time dependent manner in both the WT and rescued cells. Insets showed magnified views of the boxed areas. Scale bars: 20 μm or 2 μm (insets). (E) Quantitative analysis of the total signal intensity of CD147 in the parental cells (WT), Rab30-KO cells (–), and rescued cells shown in (D). The graphs show the average values of 30 cells and were normalized to the values at 0 min. Data shown are the means ± s.e.m. of three independent experiments. **, p<0.01; ***, p<0.001; NS, not significant (one-way ANOVA and Tukey’s test). (F) Quantitative analysis of the total signal intensity of CD147 in parental cells (WT), Rab30-KO cells (–), and rescued cells at 0 min shown in (D). The graphs show the average values of 30 cells and were normalized to the values by WT. The data shown are means ± s.e.m. of the data obtained in three independent experiments. ***, p<0.001; NS, not significant (one-way ANOVA and Tukey’s test). (G) Biotinylation of cell surface proteins in the parental cells (WT), Rab30-KO cells (–), and rescued cells. Biotinylated proteins were purified with streptavidin-conjugated beads followed by immunoblotting with the antibodies indicated. The positions of the molecular mass markers (in kDa) are shown on the left. Note that the reduction in cell surface CD147 in the Rab30-KO cells in comparison with the WT and rescued cells (compare lanes 5 and 4 or 6 in the upper panel). Input blots show 2% of the total lysate. (H) Quantitative analysis of biotinylated CD147 in the parental cells (WT), Rab30-KO cells (–), and rescued cells shown in (G). The values were normalized by input and the WT data in each experiment. The data shown are means ± s.e.m. of the data obtained in three independent experiments. *, p<0.05; NS, **, p<0.01; NS, not significant (one-way ANOVA and Tukey’s test).

Fig. 5

Both BICD2 and KIF5B are involved in tubular endosome formation

(A) The KD efficiency of Rab30, BICD2, and KIF5B (KHC) in their KD cells and control cells as revealed by immunoblotting. The positions of the molecular mass markers (in kDa) are shown on the left. The arrowheads and asterisks indicate the positions of the target proteins (Rab30 and KIF5B) and non-specific bands, respectively. (B) Typical images of MICAL-L1 in Rab30-KD (siRab30), BICD2-KD (siBICD2), KIF5B-KD (siKIF5B), and control (siControl) HeLa cells. Note that MICAL-L1-positive tubular structures were clearly diminished by the depletion of either Rab30, BICD2, or KIF5B. Insets showed magnified views of the boxed areas. Scale bars: 20 μm or 2 μm (insets). (C) Quantitative analysis of MICAL-L1-positive tubular structures in the Rab30-KD, BICD2-KD, and KIF5B-KD cells shown in (B). The graph shows the percentage of cells containing at least one >20 μm tubule per 30 cells. The data shown are means ± s.e.m. of the data obtained in three independent experiments. **, p<0.01 (one-way ANOVA and Dunnett’s test). (D) Typical images of EGFP-Rab30 in BICD2-KD (siBICD2), KIF5B-KD (siKIF5B), and control (siControl) HeLa cells. Note that EGFP-Rab30-positive dots were clearly diminished by the depletion of either Rab30, BICD2, or KIF5B. Insets showed magnified views of the boxed areas. Scale bars: 20 μm or 2 μm (insets). (E) Quantitative analysis of the Rab30-positive dots in the BICD2-KD, and KIF5B-KD cells shown in (D). The graph shows the number of Rab30-positive dots measuring less than 1 μm². The data shown are means ± s.e.m. of the data obtained in three independent experiments (n = 15 cells in each experiment). **, p<0.01; (one-way ANOVA and Dunnett’s test). (F) Interaction between 3×FLAG-tagged BICD2-CC3 and GST-tagged Rab30 (or GST alone). The interaction was analyzed by pulldown assays with anti-FLAG-tag-antibody conjugated beads followed by immunoblotting with the antibodies indicated. The positions of the molecular mass markers (in kDa) are shown on the left. Note that the interaction between BICD2-CC3 and Rab30 was stronger in the presence of GTPγS than in the presence of GDP (compare lanes 3 and 4 in the upper panel). Input blots show 1% of the total lysate. (G) The expression level of BICD2-WT, BICD2-ΔCC3 and their siRNA-resistant (SR) mutants in BICD2-KD cells as revealed by immunoblotting. The positions of the molecular mass markers (in kDa) are shown on the left. (H) Typical images of MICAL-L1 (red) in BICD2-KD (siBICD2) HeLa cells transiently expressing EGFP-tagged BICD2-WT^SR, BICD2-ΔCC3^SR, and EGFP alone (control). Note that MICAL-L1-positive tubular structures were not recovered by the BICD2-ΔCC3^SR expression. Insets showed magnified views of the boxed areas. Scale bars: 20 μm or 2 μm (insets). (I) Quantitative analysis of MICAL-L1-positive tubular structures in BICD2-KD (siBICD2) HeLa cells transiently expressing EGFP-tagged proteins shown in (H). The graph shows the percentage of cells containing at least one >20 μm tubule per 15 cells. The data shown are means ± s.e.m. of the data obtained in three independent experiments. ***, p<0.001 (one-way ANOVA and Tukey’s test).

Results

siRNA-mediated knockdown (KD) screening of candidate Rabs involved in tubular endosome formation

We previously performed localization-based screening and identified Rabs that mainly localize to tubular endosomes (tubular-endosome-resident Rabs, e.g., Rab10) in HeLa cells (Etoh and Fukuda, 2019); however, it was unclear whether non-tubular-endosome-resident Rabs other than Rab5, which is required for the formation of Rab22A-positive endosomes (Nakashima and Fukuda, 2025), are also involved in tubular endosome formation. To identify non-tubular-endosome-resident Rabs involved in tubular endosome formation, we used an effective human Rab siRNA library previously established (Aizawa and Fukuda, 2015) and performed comprehensive KD screening of Rabs that are commonly conserved in mammals (Homma et al., 2021). We used endogenous MICAL-L1 as a tubular endosome marker (Sharma et al., 2009) in each Rab-KD cell and analyzed the impact of Rab-KD on tubular endosome formation. The results showed that KD of about half of the Rabs reduced MICAL-L1-positive tubular structures (Fig. 1A and B, cyan bars). Consistent with the previous reports, KD of Rab11 or Rab22A reduced tubular structures (Weigert et al., 2004; Delevoye et al., 2014), whereas KD of Rab35 had no effect (Rahajeng et al., 2012), suggesting the validity of our KD screening.

Twelve of the candidate Rabs are involved in tubular endosome formation

KD screening occasionally misidentified candidates because of an off-target effect of the siRNAs. To exclude a possible off-target effect, we next performed two different types of second screening methods. The first second screening method consisted of using independent siRNA sites for the initial candidate Rabs identified in Fig. 1 to repeat KD screening (Fig. 2A, B). The second method consisted of overexpressing constitutively negative (CN) mutants (Tamura et al., 2009; Ishida et al., 2012) of the initial candidate Rabs (Fig. 2C, D), because CN mutants are thought to dominant-negatively interfere with the function of endogenous small GTPases by trapping their GEFs (Feig, 1999) and overexpression of certain Rab CN mutants has previously been shown to reduce tubular structures (Weigert et al., 2004; Patel et al., 2021). As a result of using these two second screening methods, we were able to identify 12 Rabs (Fig. 2E), including known tubular endosome regulators Rab10 and Rab22A, as strong candidates for involvement in tubular endosome formation, and 8 of them appeared not to localize to tubular endosomes based on the previous localization data (Homma et al., 2021).

Rab30 is a novel target of TBC1D22B

Rab1B and Rab33B have been reported to be substrates of TBC1D22B (Pan et al., 2006; Martino et al., 2025); however, neither of them was included among the final candidates identified by our screenings (Fig. 2E), suggesting the presence of an additional unidentified substrate(s) of TBC1D22B. Since TBC1D22B is a Golgi-related protein (Haas et al., 2007; Greninger et al., 2013), there was a strong possibility that the suspected additional substrate(s) would also be present at the Golgi. Actually, the results of the subcellular localization analysis of the final candidate Rabs indicated that several Rabs, including Rab22A, Rab24, and Rab30 well colocalized with TGN46 (trans-Golgi network marker) and/or GM130 (cis-Golgi marker) and that the two markers best colocalized with Rab30 (Fig. 3A, B, and S1). We therefore focused on Rab30 as a novel target of TBC1D22B. The results of in vitro GAP assays showed that TBC1D22B-WT significantly promoted Rab30 GTPase activity, whereas BSA (negative control) or a TBC1D22B-RK mutant (Pan et al., 2006) had no effect (Fig. 3C, D). By contrast, TBC1D22B was unable to promote Rab3B GTPase activity (Fig. S2). Intriguingly, the number of Rab30-positive dots in the cytoplasmic region outside the Golgi region was lower when TBC1D22B-WT was overexpressed, but not when TBC1D22B-RK was overexpressed (Fig. 3E, F). We speculate that Golgi-localized Rab30 even in the presence of TBC1D22B-WT corresponds to GDP-Rab30, because Rab30-TN, a constitutively negative mutant, was still localized at TGN46-positive compartments, but not to be present at cytoplasmic dots (Fig. S3). Cytoplasmic Rab30-positive dots are likely to be formed only when Rab30 is active in the GTP-bound form. Taken together, these results indicated that Rab30 is a novel candidate for a target of TBC1D22B.

Rab30 is involved in both tubular endosome formation and CIE cargo recycling

Rab30 is known to be a Golgi-related Rab, and several functions of Rab30, including regulation of the endosome-to-TGN retrograde pathway, autophagy, and transport of intracellular nanovesicles (Oda et al., 2016; Larocque et al., 2021; Zulkefli et al., 2021), have previously been reported; however, nothing was known about its involvement in tubular endosome formation. We first co-stained EGFP-Rab30-positive dots with MICAL-L1 and various known organelle markers, but no clear colocalization between Rab30-positive dots and MICAL-L1 (or other organelles) was observed (Fig. S4), suggesting that Rab30-positive dots are not directly involved in an initial step of tubular endosome formation. To further investigate whether the possible TBC1D22B target Rab30 is also involved in tubular endosome formation, we next established Rab30-KO HeLa cells and corresponding rescued cells (i.e., Rab30-KO cells stably expressing Rab30) (Fig. 4A) and immunostained the cells for MICAL-L1 (Fig. 4B). The results showed that MICAL-L1-positive tubules were significantly reduced by Rab30 depletion and that they were completely recovered by Rab30 re-expression (Fig. 4B, C). We then investigated the effect of Rab30-KO on CIE and CME cargo trafficking by using CD147 and transferrin (Tf), respectively, as model cargos. Consistent with the fact that tubular endosomes are involved in CIE cargo recycling (Maldonado-Báez et al., 2013; Higashi et al., 2022; Nakashima et al., 2023), the internalized CD147 intensity was significantly reduced in WT and rescued cells through CD147 recycling (Fig. 4D, E). Moreover, the amount of internalized CD147 intensity in the WT and rescued cells was greater than that in Rab30-KO cells (Fig. 4F), presumably because the amount of cell surface CD147 was significantly lower in the Rab30-KO cells (Fig. 4G, H). By contrast, the internalized Alexa594-Tf intensity was unaffected by Rab30 depletion (Fig. S5A, B), as reported previously (Zulkefli et al., 2021). In addition, Rab30-KO had no effect on Golgi morphology (Fig. S5C), consistent with the previous reports on Rab30-depleted cultured cells and tissues (Homma et al., 2019; Smith et al., 2024). We therefore concluded that, unlike TBC1D22B, Rab30 is specifically involved in the CIE pathway but does not regulate Golgi morphology.

BICD2, a Rab30-binding protein involved in tubular endosome formation

Finally, we investigated the possible involvement of a Rab30 effector in tubular endosome formation. To identify a candidate Rab30 effector(s), we referred to a list of Rab30-binding proteins previously reported based on proteome analyses in fly and human cultured cells in which a constitutively active Rab30 mutant was used as bait (Gillingham et al., 2014, 2019). We selected one of them, BICD2, as a candidate Rab30 effector, because BICD2 together with Rab6 has been reported to function in a Golgi–ER transport pathway in a dynein–dynactin-dependent manner (Matanis et al., 2002) and in a secretory pathway in a KIF5B-dependent manner (Grigoriev et al., 2007). Since KIF5B regulates secretory vesicle transport cooperatively with KIF13B (Serra-Marques et al., 2020), which is one of the key tubular endosome regulators (Etoh and Fukuda, 2019), we hypothesized that Rab30 regulates tubular endosome formation together with BICD2 and KIF5B. To test this hypothesis, we used specific siRNAs to knock down endogenous BICD2 and KIF5B in HeLa cells (Fig. 5A) and evaluated the impact of their KD on MICAL-L1-positive tubules (Fig. 5B). The results showed a significant reduction in tubular structures in BICD2-KD or KIF5B-KD cells were significantly reduced, the same as in Rab30-KD cells (Fig. 5B, C). In addition, the numbers of EGFP-Rab30-positive dots in both BICD2-KD cells and KIF5B-KD cells were also lower (Fig. 5D, E). Since BICD2 contains three coiled-coil (CC) domains and interacts with Rab6A through the CC3 domain (Matanis et al., 2002), we predicted that Rab30 also interacts with the BICD2 CC3 domain, and we confirmed their binding by performing pulldown assays. The results showed that BICD2-CC3 preferentially interacted with an active form of Rab30 (Fig. 5F). To directly assess the functional importance of the BICD2 CC3 domain in tubular endosome formation, we prepared an siRNA-resistance (SR) mutant of BICD2 (WT^SR and ΔCC3^SR, which lacks a CC3 domain; Fig. 5G) and performed a KD–rescue experiment. As expected, overexpression of BICD2-WT^SR rescued tubule deficiency of BICD2-KD cells, whereas the ΔCC3^SR mutant failed to restore tubular endosome formation (Fig. 5H, I). Furthermore, when we overexpressed BICD2-CC3, an active Rab30-binding domain, in HeLa cells, MICAL-L1-positive tubular structures were dramatically reduced by BICD2-CC3 expression (Fig. S6), which also supports our idea that Rab30–BICD2 interaction is important for tubular endosome formation. These results taken together suggested that Rab30 functions together with BICD2 and KIF5B in tubular endosome formation.

Discussion

In the present study, we have performed comprehensive Rab-KD screening and demonstrated that 12 Rabs, including several non-tubular-endosome-resident Rabs, e.g., Rab5A and Rab30, are involved in tubular endosome formation. We also identified Rab30 as a novel target of TBC1D22B, which we had previously identified as a Golgi-related Rab-GAP involved in tubular endosome formation (Nakashima and Fukuda, 2025), and Rab30 itself has been shown to be required for tubular endosome formation, CIE cargo recycling, and maintenance of the amount of cell surface CIE cargos. Moreover, Rab30-related molecules, BICD2 and KIF5B, were also found to be required for the formation of both tubular endosomes and cytoplasmic Rab30-positive dots outside the Golgi region. A recent study showed that the internalization of CIE cargo proteins is important for transient tubular endosome formation in H1650 non-small cell lung cancer cells (Misri et al., 2025), suggesting that the internalization and trafficking of CIE cargos in HeLa cells are also required for efficient tubular endosome formation. Since both BICD2 and KIF5B are involved in post-Golgi trafficking (Matanis et al., 2002) and disruption of the Golgi morphology by exposure to Brefeldin A (BFA) for 2 h caused a significant reduction in tubular structures (Fig. S7), Rab30 is involved in tubular endosome formation in a somewhat indirect fashion, presumably by trafficking of CIE cargos from the Golgi apparatus to the plasma membrane.

Our study did not attempt to unravel the precise mechanism of communication between tubular endosomes and the Golgi apparatus through Rab30. However, several recent reports have suggested important connections between tubular endosomes and other organelles. First, one of the tubular endosome markers, Rab10, has been shown to be a regulator of post-Golgi trafficking in neurons (Liu et al., 2013). Second, Rab10-positive tubules have been demonstrated to partially colocalize with TGN (Etoh and Fukuda, 2019). Third, some recycling endosomes are known to attach to TGN in a recycling pathway (Fujii et al., 2020). Fourth, KD of certain Golgi-related proteins has been shown to result in the formation of an excess of tubular endosomes in glioblastoma, which do not contain tubular endosomes under basal conditions (Joseph et al., 2023). Finally, BLTP2-positive tubular internal membranes continuous with the plasma membrane, which is in contact with ER, have been shown to be connected to Rab10-dependent tubular recycling endosomes (Dai et al., 2025). In order to answer the remaining questions, e.g., why certain types of cells can only tubulate recycling endosomes, in the future we would like to focus our research on communications between tubular endosomes and other organelles.

We did not investigate the function of any tubular-endosome-related Rabs newly identified in this study other than Rab30, but their roles in tubular endosome formation may be inferred from the functions previously reported. For example, Rab3B is involved in cross presentation of MHC-I, one of the CIE cargos (Eyster et al., 2009), in dendritic cells (Zou et al., 2009), and the Rab3 subfamily is involved in synaptic vesicle exocytosis at the presynaptic plasma membrane (Lledo et al., 1994). We therefore hypothesized that Rab3B is involved in the final step of CIE cargo recycling. Extensive research, including the establishment of Rab3B-KO cells and identification of a specific Rab3B effector, will be necessary to test our hypothesis in the future. Since there are a number of as yet uncharacterized tubular-endosome-related Rabs, a long, winding road still lies ahead before we are able to fully understand the Rab-mediated trafficking and/or signaling networks that are required for tubular endosome formation.

Author Declaration Statements

Funding

This work was supported in part by Grant-in-Aid for Scientific Research(B) and Grant-in-Aid for Challenging Exploratory Research from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan (grant numbers 24K22081 and 25K02267 to MF), and by the Japan Society for the Promotion of Science (24KJ0370 to SN).

Conflict of Interest Statement

The authors declare that they have no conflict of interest.

Data Availability Statement

Data that support the results reported here are available from the corresponding author, MF, upon reasonable request.

Author Contribution Statement

Conceptualization, investigation, analysis, funding acquisition, and writing-original draft: Shumei Nakashima. Conceptualization, funding acquisition, writing-review and editing, and supervision: Mitsunori Fukuda

Ethics Approval and Consent to Participate

Not applicable.

Patient Consent for Publication

Not applicable.

Acknowledgments

We thank Dr. Toshio Kitamura (The University of Tokyo, Tokyo, Japan) for kindly donating materials, and all members of the Fukuda laboratory for helpful discussions.

References

•  Aizawa,  M. and  Fukuda,  M. 2015. Small GTPase Rab2B and its specific binding protein Golgi-associated Rab2B interactor-like 4 (GARI-L4) regulate Golgi morphology. J. Biol. Chem., 290: 22250–22261.
•  Barr,  F. and  Lambright,  D.G. 2010. Rab GEFs and GAPs. Curr. Opin. Cell Biol., 22: 461–470.
•  Caplan,  S.,  Naslavsky,  N.,  Hartnell,  L.M.,  Lodge,  R.,  Polishchuk,  R.S.,  Donaldson,  J.G., and  Bonifacino,  J.S. 2002. A tubular EHD1-containing compartment involved in the recycling of major histocompatibility complex class I molecules to the plasma membrane. EMBO J., 21: 2557–2567.
•  Dai,  A.,  Xu,  P.,  Amos,  C.,  Fujise,  K.,  Wu,  Y.,  Yang,  H.,  Eisen,  J.N.,  Guillén-Samander,  A., and  De Camilli,  P. 2025. Multiple interactions recruit BLTP2 to ER-PM contacts to control plasma membrane dynamics. J. Cell Biol., 224: e202504027.
•  Damke,  H.,  Baba,  T.,  van der Bliek,  A.M., and  Schmid,  S.L. 1995. Clathrin-independent pinocytosis is induced in cells overexpressing a temperature-sensitive mutant of dynamin. J. Cell Biol., 131: 69–80.
•  Delevoye,  C.,  Miserey-Lenkei,  S.,  Montagnac,  G.,  Gilles-Marsens,  F.,  Paul-Gilloteaux,  P.,  Giordano,  F.,  Waharte,  F.,  Marks,  M.,  Goud,  B., and  Raposo,  G. 2014. Recycling endosome tubule morphogenesis from sorting endosomes requires the kinesin motor KIF13A. Cell Rep., 6: 445–454.
•  Del Olmo,  T.,  Lauzier,  A.,  Normandia,  C.,  Larcher,  R.,  Lecours,  M.,  Jean,  D.,  Lessard,  L.,  Steinberg,  F.,  Boisvert,  F.M., and  Jean,  S. 2019. APEX2-mediated RAB proximity labeling identifies a role for RAB21 in clathrin-independent cargo sorting. EMBO Rep., 20: e47192.
•  Etoh,  K. and  Fukuda,  M. 2019. Rab10 regulates tubular endosome formation through KIF13A and KIF13B motors. J. Cell Sci., 132: jcs226977.
•  Eyster,  C.A.,  Higginson,  J.D.,  Huebner,  R.,  Porat-Shliom,  N.,  Weigert,  R.,  Wu,  W.W.,  Shen,  R.F., and  Donaldson,  J.G. 2009. Discovery of new cargo proteins that enter cells through clathrin-independent endocytosis. Traffic, 10: 590–599.
•  Feig,  L.A. 1999. Tools of the trade: Use of dominant-inhibitory mutants of Ras-family GTPases. Nat. Cell Biol., 1: E25–E27.
•  Fujii,  S.,  Kurokawa,  K.,  Inaba,  R.,  Hiramatsu,  N.,  Tago,  T.,  Nakamura,  Y.,  Nakano,  A.,  Satoh,  T., and  Satoh,  A.K. 2020. Recycling endosomes attach to the trans-side of Golgi stacks in Drosophila and mammalian cells. J. Cell Sci., 133: jcs236935.
•  Gillingham,  A.K.,  Sinka,  R.,  Torres,  I.L.,  Lilley,  K.S., and  Munro,  S. 2014. Toward a comprehensive map of the effectors of Rab GTPases. Dev. Cell, 31: 358–373.
•  Gillingham,  A.K.,  Bertram,  J.,  Begum,  F., and  Munro,  S. 2019. In vivo identification of GTPase interactors by mitochondrial relocalization and proximity biotinylation. eLife, 8: e45916.
•  Grant,  B.D. and  Donaldson,  J.G. 2009. Pathways and mechanisms of endocytic recycling. Nat. Rev. Mol. Cell Biol., 10: 597–608.
•  Greninger,  A.,  Knudsen,  G.,  Betegon,  M.,  Burlingame,  A., and  DeRisi,  J. 2013. ACBD3 interaction with TBC1 domain 22 protein is differentially affected by enteroviral and kobuviral 3A protein binding. mBio, 4: e00098-13.
•  Grigoriev,  I.,  Splinter,  D.,  Keijzer,  N.,  Wulf,  P.S.,  Demmers,  J.,  Ohtsuka,  T.,  Modesti,  M.,  Maly,  I.V.,  Grosveld,  F.,  Hoogenraad,  C.C., and  Akhmanova,  A. 2007. Rab6 regulates transport and targeting of exocytotic carriers. Dev. Cell, 13: 305–314.
•  Haas,  A.K.,  Yoshimura,  S.,  Stephens,  D.J.,  Preisinger,  C.,  Fuchs,  E., and  Barr,  F.A. 2007. Analysis of GTPase-activating proteins: Rab1 and Rab43 are key Rabs required to maintain a functional Golgi complex in human cells. J. Cell Sci., 120: 2997–3010.
•  Hattula,  K.,  Furuhjelm,  J.,  Tikkanen,  J.,  Tanhuanpää,  K.,  Laakkonen,  P., and  Peränen,  J. 2006. Characterization of the Rab8-specific membrane traffic route linked to protrusion formation. J. Cell Sci., 119: 4866–4877.
•  Higashi,  S.,  Makiyama,  T.,  Sakane,  H.,  Nogami,  S., and  Shirataki,  H. 2022. Regulation of Hook1-mediated endosomal sorting of clathrin-independent cargo by γ-taxilin. J. Cell Sci., 135: jcs258849.
•  Homma,  Y.,  Kinoshita,  R.,  Kuchitsu,  Y.,  Wawro,  P.S.,  Marubashi,  S.,  Oguchi,  M.E.,  Ishida,  M.,  Fujita,  N., and  Fukuda,  M. 2019. Comprehensive knockout analysis of the Rab family GTPases in epithelial cells. J. Cell Biol., 218: 2035–2050.
•  Homma,  Y.,  Hiragi,  S., and  Fukuda,  M. 2021. Rab family of small GTPases: An updated view on their regulation and functions. FEBS J., 288: 36–55.
•  Ishibashi,  K.,  Kanno,  E.,  Itoh,  T., and  Fukuda,  M. 2009. Identification and characterization of a novel Tre-2/Bub2/Cdc16 (TBC) protein that possesses Rab3A-GAP activity. Genes Cells, 14: 41–52.
•  Ishida,  M.,  Ohbayashi,  N.,  Maruta,  Y.,  Ebata,  Y., and  Fukuda,  M. 2012. Functional involvement of Rab1A in microtubule-dependent anterograde melanosome transport in melanocytes. J. Cell Sci., 125: 5177–5187.
•  Itoh,  T.,  Satoh,  M.,  Kanno,  E., and  Fukuda,  M. 2006. Screening for target Rabs of TBC (Tre-2/Bub2/Cdc16) domain-containing proteins based on their Rab-binding activity. Genes Cells, 11: 1023–1037.
•  Joseph,  B.B.,  Naslavsky,  N.,  Binti,  S.,  Conquest,  S.,  Robison,  L.,  Bai,  G.,  Homer,  R.O.,  Grant,  B.D.,  Caplan,  S. and  Fay,  D.S. 2023. Conserved NIMA kinases regulate multiple steps of endocytic trafficking. PLoS Genet., 19: e1010741.
•  Kaksonen,  M. and  Roux,  A. 2018. Mechanisms of clathrin-mediated endocytosis. Nat. Rev. Mol. Cell Biol., 19: 313–326.
•  Kuroda,  T.S. and  Fukuda,  M. 2005. Identification and biochemical analysis of Slac2-c/MyRIP as a Rab27A-, myosin Va/VIIa-, and actin-binding protein. Methods Enzymol., 403: 431–444.
•  Lamaze,  C.,  Dujeancourt,  A.,  Baba,  T.,  Lo,  C.G.,  Benmerah,  A., and  Dautry-Varsat,  A. 2001. Interleukin 2 receptors and detergent-resistant membrane domains define a clathrin-independent endocytic pathway. Mol. Cell, 7: 661–671.
•  Larocque,  G.,  Moore,  D.J.,  Sittewelle,  M.,  Kuey,  C.,  Hetmanski,  J.H.R.,  La-Borde,  P.J.,  Wilson,  B.J.,  Clarke,  N.I.,  Caswell,  P.T., and  Royle,  S.J. 2021. Intracellular nanovesicles mediate α5β1 integrin trafficking during cell migration. J. Cell Biol., 220: e202009028.
•  Liu,  Y.,  Xu,  X.H.,  Chen,  Q.,  Wang,  T.,  Deng,  C.Y.,  Song,  B.L.,  Du,  J.L., and  Luo,  Z.G. 2013. Myosin Vb controls biogenesis of post-Golgi Rab10 carriers during axon development. Nat. Commun., 4: 2005.
•  Lledo,  P.-M.,  Johannes,  L.,  Vernier,  P.,  Zorec,  R.,  Darchen,  F.,  Vincent,  J.D.,  Henry,  J.P., and  Mason,  W.T. 1994. Rab3 proteins: Key players in the control of exocytosis. Trends Neurosci., 17: 426–432.
•  Maldonado-Báez,  L.,  Cole,  N.B.,  Kramer,  H., and  Donaldson,  J.G. 2013. Microtubule-dependent endosomal sorting of clathrin-independent cargo by Hook1. J. Cell Biol., 201: 233–247.
•  Martino,  F.,  Lupi,  M.,  Murabito,  A.,  Bedin,  F.,  Villari,  G.,  Andreoli,  L.,  Freddi,  S.,  Matoskova,  M.,  Pennisi,  R.,  Fontana,  S.,  Fardin,  A.,  Boncompain,  C.,  Perez,  F.,  Bussolino,  F.,  Cuomo,  A.,  Sigismund,  S., and  Lanzetti,  L. 2025. TBC1D22B regulates ER-to-Golgi trafficking via RAB1B inactivation and promotes oncogenic programs in breast cancer. Adv. Sci., 29: e02269.
•  Matanis,  T.,  Akhmanova,  A.,  Wulf,  P.,  Del Nery,  E.,  Weide,  T.,  Stepanova,  T.,  Galjart,  N.,  Grosveld,  F.,  Goud,  B.,  De Zeeuw,  C.I.,  Barnekow,  A., and  Hoogenraad,  C.C. 2002. Bicaudal-D regulates COPI-independent Golgi–ER transport by recruiting the dynein–dynactin motor complex. Nat. Cell Biol., 4: 986–992.
•  Maxfield,  F.R. and  McGraw,  T.E. 2004. Endocytic recycling. Nat. Rev. Mol. Cell Biol., 5: 121–132.
•  Mayor,  S. and  Pagano,  R.E. 2007. Pathways of clathrin-independent endocytosis. Nat. Rev. Mol. Cell Biol., 8: 603–612.
•  Misri,  G.,  Murakonda,  A.B.,  Naslavsky,  N., and  Caplan,  S. 2025. CIN85 and CD2AP are novel constituents of dynamic tubular recycling endosomes that regulate recycling upon recruitment by MICAL-L1. Traffic, 26: e70015.
•  Mrozowska,  P.S. and  Fukuda,  M. 2016. Regulation of podocalyxin trafficking by Rab small GTPases in 2D and 3D epithelial cell cultures. J. Cell Biol., 213: 355–369.
•  Nakashima,  S. and  Fukuda,  M. 2025. Identification of Rab GTPase-activating proteins required for tubular endosome formation. Traffic, 26: e70007.
•  Nakashima,  S.,  Matsui,  T., and  Fukuda,  M. 2023. Vps9d1 regulates tubular endosome formation through specific activation of Rab22A. J. Cell Sci., 136: jcs260522.
•  Oda,  S.,  Nozawa,  T.,  Nozawa-Minowa,  A.,  Tanaka,  M.,  Aikawa,  C.,  Harada,  H., and  Nakagawa,  I. 2016. Golgi-Resident GTPase Rab30 promotes the biogenesis of pathogen-containing autophagosomes. PLOS One, 11: e0147061.
•  Oguchi,  M.E.,  Noguchi,  K., and  Fukuda,  M. 2017. TBC1D12 is a novel Rab11-binding protein that modulates neurite outgrowth of PC12 cells. PLOS One, 12: e0174883.
•  Oguchi,  M.E.,  Okuyama,  K.,  Homma,  Y., and  Fukuda,  M. 2020. A comprehensive analysis of Rab GTPases reveals a role for Rab34 in serum starvation-induced primary ciliogenesis. J. Biol. Chem., 295: 12674–12685.
•  Pan,  X.,  Eathiraj,  S.,  Munson,  M., and  Lambright,  D.G. 2006. TBC-domain GAPs for Rab GTPases accelerate GTP hydrolysis by a dual-finger mechanism. Nature, 442: 303–306.
•  Patel,  N.M.,  Siva,  M.S.A.,  Kumari,  R.,  Shewale,  D.J.,  Rai,  A.,  Ritt,  M.,  Sharma,  P.,  Setty,  S.R.G.,  Sivaramakrishnan,  S., and  Soppina,  V. 2021. KIF13A motors are regulated by Rab22A to function as weak dimers inside the cell. Sci. Adv., 7: eabd2054.
•  Pfeffer,  S.R. 2013. Rab GTPase regulation of membrane identity. Curr. Opin. Cell Biol., 25: 414–419.
•  Radhakrishna,  H. and  Donaldson,  J.G. 1997. ADP-ribosylation factor 6 regulates a novel plasma membrane recycling pathway. J. Cell Biol., 139: 49–61.
•  Rahajeng,  J.,  Giridharan,  S.S.P.,  Cai,  B.,  Naslavsky,  N., and  Caplan,  S. 2012. MICAL-L1 is a tubular endosomal membrane hub that connects Rab35 and Arf6 with Rab8a. Traffic, 13: 82–93.
•  Serra-Marques,  A.,  Martin,  M.,  Katrukha,  E.A.,  Grigoriev,  I.,  Peeters,  C.A.,  Liu,  Q.,  Hooikaas,  P.J.,  Yao,  Y.,  Solianova,  V.,  Smal,  I.,  Pedersen,  L.B.,  Meijering,  E.,  Kapitein,  L.C., and  Akhmanova,  A. 2020. Concerted action of kinesins KIF5b and KIF13b promotes efficient secretory vesicle transport to microtubule plus ends. eLife, 9: e61302.
•  Sharma,  M.,  Giridharan,  S.S.P.,  Rahajeng,  J.,  Naslavsky,  N., and  Caplan,  S. 2009. MICAL-L1 links EHD1 to tubular recycling endosomes and regulates receptor recycling. Mol. Biol. Cell, 20: 5181–5194.
•  Smith,  D.M.,  Liu,  B.Y., and  Wolfgang,  M.J. 2024. Rab30 facilitates lipid homeostasis during fasting. Nat. Commun., 15: 4469.
•  Stenmark,  H. 2009. Rab GTPases as coordinators of vesicle traffic. Nat. Rev. Mol. Cell Biol., 10: 513–525.
•  Tamura,  K.,  Ohbayashi,  N.,  Maruta,  Y.,  Kanno,  E.,  Itoh,  T., and  Fukuda,  M. 2009. Varp is a novel Rab32/38-binding protein that regulates Tyrp1 trafficking in melanocytes. Mol. Biol. Cell, 20: 2900–2908.
•  Tsuboi,  T. and  Fukuda,  M. 2006. Rab3A and Rab27A cooperatively regulate the docking step of dense-core vesicle exocytosis in PC12 cells. J. Cell Sci., 119: 2196–2203.
•  Weigert,  R.,  Yeung,  A.C.,  Li,  J., and  Donaldson,  J.G. 2004. Rab22a regulates the recycling of membrane proteins internalized independently of clathrin. Mol. Biol. Cell, 15: 3758–3770.
•  Zou,  L.,  Zhou,  J.,  Zhang,  J.,  Li,  J.,  Liu,  N.,  Chai,  L.,  Li,  N.,  Liu,  T.,  Li,  L.,  Xie,  Z.,  Liu,  H.,  Wan,  Y., and  Wu,  Y. 2009. The GTPase Rab3b/3c-positive recycling vesicles are involved in cross-presentation in dendritic cells. Proc. Natl. Acad. Sci. USA, 106: 15801–15806.
•  Zulkefli,  K.L.,  Mahmoud,  I.S.,  Williamson,  N.A.,  Gosavi,  P.K.,  Houghton,  F.J., and  Gleeson,  P.A. 2021. A role for Rab30 in retrograde trafficking and maintenance of endosome-TGN organization. Exp. Cell Res., 399: 112442.