TBC1D18, a novel Rab5-GAP, coordinates endosome maturation together with Mon1

Endosome maturation is essential for efficient degradation of internalized extracellular molecules and plasma membrane proteins. Two Rab GTPases, Rab5 and Rab7, are known to regulate endosome maturation, and a Rab5-to-Rab7 conversion mediated by a Rab7 activator, Mon1–Ccz1, is essential for progression of the maturation process. However, the importance and mechanism of Rab5 inactivation during endosome maturation is poorly understood. Here we report a novel Rab5 inactivator (Rab5-GTPase activating protein [Rab5-GAP]), TBC1D18, which is associated with Mon1 and mediates endosome maturation. We found that Rab5 hyperactivation in addition to Rab7 inactivation occurs in the absence of Mon1. We present evidence showing that the severe defects in endosome maturation observed in Mon1-KO cells are attributable to Rab5 hyperactivation rather than to Rab7 inactivation. We then identified TBC1D18 as a Rab5-GAP by comprehensive screening of TBC-domain-containing Rab-GAPs. Expression of TBC1D18 in Mon1-KO cells rescued the defects in endosome maturation, whereas its depletion attenuated endosome formation and degradation of endocytosed cargos. Moreover, TBC1D18 was found to be able to interact with Mon1, and it localized in close proximity to lysosomes in a Mon1-dependent manner. Thus, TBC1D18 is a crucial regulator of endosome maturation that functions together with Mon1.

However, the importance and mechanism of Rab5 inactivation during endosome maturation is poorly understood. Here we report a novel Rab5 inactivator (Rab5-GTPase activating protein [Rab5-GAP]), TBC1D18, which is associated with Mon1 and mediates endosome maturation. We found that Rab5 hyperactivation in addition to Rab7 inactivation occurs in the absence of Mon1. We present evidence showing that the severe defects in endosome maturation observed in Mon1-KO cells are attributable to Rab5 hyperactivation rather than to Rab7 inactivation. We then identified TBC1D18 as a Rab5-GAP by comprehensive screening of TBC-domain-containing Rab-GAPs.
Expression of TBC1D18 in Mon1-KO cells rescued the defects in endosome maturation, whereas its depletion attenuated endosome formation and degradation of endocytosed cargos. Moreover, TBC1D18 was found to be able to interact with Mon1, and it localized in close proximity to lysosomes in a Mon1-dependent manner. Thus, TBC1D18 is a crucial regulator of endosome maturation that functions together with Mon1.

INTRODUCTION
Endocytosis is a pivotal membrane dynamic process, in which cells internalize extracellular molecules and plasma membrane proteins (Sorkin & Zastrow, 2009;Sigismund et al., 2021). Invagination and fission of the plasma membrane results in the formation of early endosomes, where sorting of endocytosed cargos occurs. Some cargos are recycled back to the plasma membrane through recycling endosomes, whereas others are destined to lysosomes for degradation through the endocytic pathway in the following manner. Early endosomes first mature into late endosomes (also known as multivesicular bodies), a process that is accompanied by luminal acidification and acquisition of lysosomal enzymes from the trans-Golgi network, and the late endosomes then fuse with lysosomes. Endocytosed cargos are eventually degraded in the lysosomes, and the products of degradation are transported out of the lysosomes for energy generation or reutilization in biosynthetic pathways.
Rab GTPases are widely thought to coordinate protein sorting and vesicle/membrane trafficking during the endosome maturation processes. Rabs function as a molecular switch by cycling between an inactive form (GDP-bound) and active form (GTP-bound). Although inactive Rabs are present in the cytosol, active Rabs are localized to specific membrane compartments (or organelles) via their Cterminal cysteine residue(s) modified by geranylgeranyl transferase and recruit a specific effector protein(s) that promotes vesicle/membrane trafficking (Stenmark, 2009;Hutagalung & Novick, 2011;Pfeffer, 2017;Homma et al., 2021). Rab activity is thought to be spatiotemporally regulated by activation and inactivation enzymes, i.e., guanine nucleotide exchange factors (GEFs) and GTPase-activating proteins (GAPs), 4 respectively (Lamber et al., 2019). Two Rab family members, early endosomal Rab5 (Ypt5 in yeasts) and late endosomal Rab7 (Ypt7 in yeasts) are key Rabs that coordinate endosome maturation (Huotari & Helenius, 2011;Borchers et al., 2021). After endocytosis, Rab5 is activated by a specific GEF, such as Rabex5 (Horiuchi et al., 1997), and it facilitates the formation of early endosomes and their homotypic fusion (Stenmark et al., 1994). The early endosomal Rab5 then directly recruits the Mon1-Ccz1 heterodimeric complex, a Rab7-GEF, to the same compartment (Nordmann et al., 2010;Kinchen & Ravichandran, 2011), and the recruited Mon1-Ccz1 activates Rab7, which regulates subsequent late endosome maturation. The Rab7 on late endosomes recruits the hexameric homotypic fusion and vacuole protein sorting (HOPS) tethering complex, which mediates fusion between late endosomes and lysosomes (Balderhaar & Ungermann, 2013). Finally, the late endosomal Rab7 is inactivated by a GAP, such as TBC1D15 (Zhang et al., 2005), and reused for the next round of endosome maturation.
The spatiotemporal regulation of Rab5 and Rab7 activities by their GEFs and GAPs is important, because the Rab5-to-Rab7 conversion described above is essential for endosome maturation. While involvement of a Rab5-GEF, Rab7-GEF, and Rab7-GAP in endosome maturation has been well studied, very little has been known about the importance of Rab5 inactivation or the involvement of a specific Rab5-GAP in the Rab5-

Distinct phenotypes of Mon1-KO cells and Rab7-KO cells in endosome maturation
To investigate the function of Mon1 in endosome maturation in greater detail, we first established Mon1-KO (Mon1a/b-DKO) MDCK cells (Fig. S1A) and examined their lysosomal morphology by immunostaining for lysosomal membrane-associated protein 2 (LAMP2). Although, consistent with our previous observation, the Rab7-KO cells contained slightly larger lysosomes than WT cells did (Kuchitsu et al., 2018)

Hyperactivation of Rab5, not inactivation of Rab7, in Mon1-KO cells induced formation of greatly enlarged endosomes and lysosomes
Because Mon1 is known to be a component of the Rab7-GEF complex (Kinchen & Ravichandran, 2010;Nordmann et al., 2010;Poteryaev et al., 2010), we first checked the activity of Rab7 in WT and Mon1-KO cells by performing effector pull-down assays with beads coupled to T7-tagged RILP, the best characterized Rab7 effector (Cantalupo et al., 2001). After incubating lysates of WT and Mon1-KO cells with the T7-RILP beads, the active Rab7 trapped by the beads was analyzed by immunoblotting (Fig. 1E). As has an inhibitory role in relation to Rab5, that is, inactivates Rab5.

Comprehensive screening for a candidate(s) Rab5-GAP that regulates endosome maturation
To identify a candidate Rab5-GAP(s) that regulates endosome maturation, we first focused on the protein family members that share a TBC (Tre-2/Bub2/Cdc16)-domain, most of which are known to function as Rab-GAP domains (Fukuda, 2011;Frasa et al., 2012). More than 40 TBC-domain-containing proteins (simply referred to as TBC proteins below) are present in mammals, and although some of them have been shown to exhibit GAP activity toward Rab5 in vitro (Xiao et al., 1997;Pei et al., 2002;Chamberlain et al., 2004;Haas et al., 2005;Li et al., 2009;Lachmann et al., 2012;Rao et al., 2021;Yan et al., 2021), no attempt has ever been made to comprehensively screen for Rab5- We also tried to generate TBC1D18-KO cells, but did not succeed: at least one allele of TBC1D18 was always retained in the sgRNA-expressing cells (data not shown), the same as we had observed in Rab5-KO cells in a previous study (Homma et al., 2019), suggesting that TBC1D18 is essential for cell survival and/or growth, the same as Rab5 is. We therefore selected TBC1D18 as the prime candidate for the Rab5-GAP in endosome maturation.

TBC1D18 functions as a Rab5-GAP both in vitro and in cultured cells
Because nothing was known about the function of TBC1D18 (also known as RabGAP1L) in terms of Rab5 regulation, we performed in vitro GAP assays with purified components to demonstrate inactivation of Rab5 by TBC1D18. As shown in Fig. 4A and 4B, the amount of GTP bound to Rab5A, which was monitored by luminescence, was found to significantly decrease in a time-dependent manner in the presence of TBC1D18 when compared to the control BSA. Importantly, TBC1D18 did not display any GAP activity toward Rab4A or Rab7 under our experimental conditions. In contrast to these findings in the present study, it has previously been reported that TBC1D18 acts as a GAP for Rab22A, another early endosomal Rab (Itoh et al., 2006;Qu et al., 2016) in this study as opposed to GST-Rab22A in the previous study (Itoh et al., 2006).
Actually, N-terminal tagging of EGFP to certain Rabs is now known to distort their functions (Homma and Fukuda, 2016;Oguchi et al., 2020), suggesting that using GST-Rab5A to perform GAP assays is inappropriate.
To further assess whether TBC1D18 has the ability to inactivate Rab5 (i.e., Rab5-GAP activity) even in cultured cells, we performed active Rab5 pull-down assays with GST-Appl1-N in WT cells stably expressing EGFP-TBC1D18, the same as shown in Fig. 2C. The results showed that the amount of active Rab5 in the EGFP-TBC1D18expressing cells was clearly reduced in comparison with the control EGFP-expressing cells (Fig. 4C), indicating that TBC1D18 actually acts as a Rab5-GAP in cultured cells.
Consistent with this finding, the dotted signals of Rab5A and EEA1, a known Rab5 effector, had almost completely disappeared in the WT cells expressing TBC1D18 ( Fig.   4D and 4E), suggesting suppression of Rab5 activation by TBC1D18.

TBC1D18 is involved in endosome maturation through association with Mon1
Since Rab5 is required for endosome fusion and/or maturation (Langemeyer et al., 2018), we further evaluated the impact of TBC1D18 depletion on endosome maturation by performing 488-EGF uptake and EGFR degradation assays. After incubating cells with 488-EGF for 20 min, we immunostained them for EEA1 and LAMP2. The results showed that the EGF-positive dots in the TBC1D18-depleted cells were not co-localized with EEA1 but that the colocalization between EGF-positive dots between LAMP2 was unaffected ( Fig. 5A and 5B). Moreover, the same as in Mon1-depleted cells (Fig. 1D), the TBC1D18-depleted cells exhibited attenuated EGFR degradation (Fig. 5C), indicating that TBC1D18 is involved in endosome maturation.
Finally, we attempted to determine the functional relationship between Mon1 and TBC1D18 during endosome maturation. One of the simplest relationships would be an association between these two molecules, and the results of co-immunoprecipitation assays indicated that TBC1D18 actually interacts with both Mon1a and Mon1b (Fig. 5D).
We then investigated the localization of TBC1D18 in WT and Mon1-KO cells. Since TBC1D18 mostly exhibited cytosolic localization (Fig. 3A), which precluded a specific organelle association, we permeabilized the cells with digitonin prior to fixation to remove cytosolic TBC1D18. As shown in Fig. 5E, dotted signals of EGFP-TBC1D18 were observed in close proximity to the LAMP2-positive dots in the WT cells, whereas virtually no EGFP-TBC1D18 signals were observed in the Mon1-KO cells, indicating that Mon1 is required for the organelle localization of TBC1D18.

DISCUSSION
In the present study, we discovered a novel Rab5-GAP, TBC1D18, that mediates endosome maturation in a Mon1-depenedent manner. Since TBC1D18 interacted with Mon1, a Rab7-GEF component, and its organelle localization (i.e., close proximity to LAMP2-positive compartments) was Mon1-dependedent ( Fig. 5D and 5E inactivates Rab5" (this study) and promotes dissociation of Rabex-5 from early endosomes (Poteryaev et al., 2010). (iv) The recruited TBC1D18 then inactivates Rab5, and the Mon1-Ccz1 complex activates Rab7 as a Rab7-GEF, both of which promote efficient conversion from early endosomes to late endosomes (v). Thus, Mon1 is likely to function as a hub for endosome maturation that coordinates the Rab5-GEF activity of Rabex-5, the Rab5-GAP activity of TBC1D18, and its own Rab7-GEF activity.
Although Mon1 plays dual roles in Rab5 inactivation by promoting Rabex-5 dissociation and TBC1D18 recruitment and Rab7 activation via its GEF activity, the Rab5 inactivation appears to be more important than the Rab7 activation in terms of endosome maturation. Actually, Mon1-KO cells showed severe defects in endosome maturation (i.e., enlarged lysosomes and attenuated EGFR degradation), whereas Rab7-KO cells showed rather mild phenotypes ( Fig. 1A and 1D). Moreover, the phenotypes of the Mon1-KO cells resembled those of Rab5A-QL-expressing cells (Fig. S2A). Since Mon1 is recruited to early endosomes via Rab5 even in Rab7-KO cells, Rab5 must be 14 inactivated by the Mon1-TBC1D18 axis, which would maintain endo-lysosomal morphology and function to some extent. Hyperactivation of Rab5 is known to facilitate homotypic fusion between early endosomes (Stenmark et al., 1994), and in the presence of Rab5 these enlarged early endosomes may not mature into late endosomes.
Nevertheless, the enlarged EEA1-positive structures in Mon1-KO cells and Rab5-QLexpressing cells were also positive for LAMP2, suggesting that enlarged early endosomes can fuse with lysosomes before fully maturing into late endosomes. Although these hybrid structures were LysoTracker-and Magic Red-positive, lysosomal degradation activity as monitored by EGFR degradation and DQ-BSA was clearly inhibited (Figs 5C and S1E). Thus, proper inactivation of Rab5 before fusion with lysosomes is necessary to achieve efficient degradation of endocytic cargos.
In addition to TBC1D18, another TBC protein, RUTBC3, has previously been shown to function as a Rab5-GAP (Haas et al., 2005). The results of our initial comprehensive screening in Mon1-KO cells also pointed to RUTBC3 as a possible Rab5-GAP (Fig. 3A). Unlike TBC1D18, however, RUTBC3 KO or knockdown did not affect endosome formation ( Fig. S3C and S3D). Although human and mouse TBC1D18 and The transition from Rab5 to Rab7 during endosome maturation is a mechanism 15 that is common to all eukaryotic cells, and the Mon1-Ccz1 complex has been retained during evolution (Borchers et al., 2021). TBC1D18, however, is conserved in vertebrates alone, implying that the coordinated regulation of endosome maturation by TBC1D18 and Mon1 may have been evolutionarily acquired in vertebrates.
Nevertheless, since TBC proteins are widely present in invertebrates (Fukuda, 2011), including in fungi and plants, it is still possible that Mon1-dependent Rab5 inactivation is the common mechanism and that other TBC proteins function during endosome maturation in invertebrates (i.e., convergent evolution in terms of Rab5 inactivation).
Further research will be necessary to elucidate whether Mon1-dependent Rab5 inactivation is important for endosome maturation in other invertebrate species.
In summary, we succeeded in identifying a novel Rab5-GAP, TBC1D18, that mediates endosome maturation, and we have proposed a new model in which Mon1 recruits TBC1D18 to endosomes/lysosomes, which accelerates a transition from Rab5 to Rab7 during endosome maturation by inactivating Rab5 and activating Rab7.
Intriguingly, Mon1/TBC1D18-mediated Rab5 inactivation is more important than Mon1mediated Rab7 activation during endosome maturation, because the defects in endosome maturation were more severe after Rab5 hyperactivation than after Rab7 inactivation.
Since both Mon1 and TBC1D18 appear to localize mostly in the cytosol, the spatiotemporal regulation of their activities and the precise mechanism of their targeting to Rab5-localzied endosomes during endosome maturation are important issues that need to be addressed next in the future studies.

Materials
Rabbit polyclonal antibodies against Rab5B/C, Rab7, and Rab22A were prepared as described previously (Mrozowska & Fukuda, 2016). cDNA cloning and plasmid constructions cDNA encoding mouse Mon1a was prepared as described previously (Yasuda et al., 2016), and the mouse Mon1b cDNA was amplified from the Marathon-Ready adult mouse brain and testis cDNAs (Clontech/Takara Bio, Shiga, Japan) by conventional PCR techniques.

(Thermo Fisher Scientific) and Lipofectamine RNAiMAX (Thermo Fisher
Scientific), respectively, each according to the manufacturer's instructions.

Retrovirus production and infection into MDCK cells
Retroviruses were produced in Plat-E cells as described previously (Homma et al., 2019).
The virus-containing medium was added to the culture medium of MDCK cells in the presence of 8 µg/mL polybrene, and after 24 h the transformants were selected with 2 µg/mL puromycin (Merck) for 24-48h.
The final supernatant containing 3×FLAG-TBC1D18 was transferred to a fresh tube.

Immunoblotting
Protein extracts obtained from cells that had been lysed with an SDS sample buffer were

Immunofluorescence analysis
Cells grown on coverslips were fixed with 4% paraformaldehyde for 10 min then washed them with PBS three times. The fixed cells were permeabilized for 5 min at room temperature with 50 µg/mL digitonin (Sigma-Aldrich) in PBS and then blocked for 30 min at room temperature with 3% BSA in PBS. To remove cytosolic components of EGFP-TBC1D18 in Fig. 5E, the cells were permeabilized with digitonin was performed 21 before fixation. The permeabilized cells were incubated for 1 h with a primary antibody, washed with PBS, and then incubated for 1 h with Alexa-labeled anti-mouse or rabbit IgG secondary antibody. After washing the cells with PBS, the coverslips were mounted on glass slides with Prolong Diamond (Thermo Fisher Scientific, #P36961) and with DAPI.
All procedures were carried out at room temperature. Confocal fluorescence images were obtained through a confocal fluorescence microscope (Fluoview 1000; Olympus, Tokyo, Japan) equipped with a Plan-Apochromat 100×/1.45 oil-immersion objective lens and an electron-multiplying charge-coupled device camera (C9100; Hamamatsu Photonics, Shizuoka, Japan).

Electron microscopy
Cells were cultured on cell-tight C-2 cell disks (Sumitomo Bakelite, Tokyo, Japan; #MS-

EGFR degradation assay
Cells grown on a 24-well plate were starved for 24 h in serum-free DMEM. After preincubation at 37˚C for 30 min in the medium containing 100 µg/mL cycloheximide, the cells were exposed to 200 ng/mL EGF for the times indicated in Figs. 1D and 5C. The total EGFR protein level after the EGF treatment was evaluated by immunoblotting.

GTP-Rab pull-down assays performed with Rab effector domains
For the GTP-Rab7 pull-down assays, T7-tagged RILP, a Rab7 effector, was expressed in COS-7 cells and immobilized to anti-T7 tag-antibody-conjugated agarose beads (EMD Millipore) (Matsui et al., 2012). MDCK-WT and Mon1-KO cells were lysed with lysis buffer #1 containing a protease inhibitor cocktail, and the supernatants were incubated with rotation for 1 h at 4˚C with beads coupled to T7-RILP. The beads were washed three times with the lysis buffer #1, boiled for 5 min with an SDS sample buffer, and then subjected to SDS-PAGE and immunoblotting analyses.
The beads were then washed three times with a buffer (50 mM HEPES-KOH, pH 7.2, 150 mM NaCl, and 10 mM MgCl2), boiled for 5 min with an SDS sample buffer, and then subjected to SDS-PAGE and immunoblotting analyses.

In vitro GAP assay
GAP-accelerated GTP hydrolysis of Rab was measured by using a GTPase-Glo TM assay kit (Promega, Madison, WI; #V7681) according to the manufacture's instruction. In brief, purified recombinant Rab4A, Rab5A, Rab7, or Rab22A (0.45 µg each) with or without 3×FLAG-TBC1D18 (0.033 µg) was suspended in the reaction buffer in the kit.
After addition of GTP to a final concentration of 5 µM, the reaction mixture was 23 incubated at room temperature for the times indicated in Fig. 4A and 4B, and in Fig. S4C.
The GTPase-Glo TM reagent was added to the reaction mixture to convert the remaining GTP to ATP. The amount of ATP was then detected with luciferase to produce bioluminescence, which was measured with a Victor Nivo Multimode microplate reader (PerkinElmer, Waltham, MA).

Assay for endocytic cargo uptake
Cells grown on coverslips were exposed to 25 µg/mL DQ TM Red BSA in serum-free DMEM for 6 h or to 2 µg/mL 488-EGF in serum-free DMEM for 20 min. After incubation, the cells were fixed with 4% paraformaldehyde and subjected to immunofluorescence analysis.

Co-immunoprecipitation assay
COS-7 cells transiently expressing recombinant proteins were lysed for 10 min on ice in lysis buffer #1 containing a phosphatase inhibitor cocktail and protease inhibitor cocktail.
After centrifugation at 20,380×g for 10 min, the supernatants were incubated with gentle rotation for 1 h at 4˚C with anti-T7 tag-antibody-conjugated agarose beads. The beads were washed three times with lysis buffer #1, boiled for 5 min with an SDS sample buffer, and then subjected to SDS-PAGE and immunoblotting analyses.