RAB proteins are members of the RAS superfamily of small GTPases, with which they share sequence and structural homology. RAS isoforms were first identified in cancer-causing viruses and subsequently as oncogenes, prompting substantial interest. Subfamilies of the RAS GTPases include the RAB, RHO, ARF, RAP, and RAN proteins of which RABs are the largest group (Takai et al. 2001). Early work in yeast identified essential roles for the RABs Ypt1p and Sec4p in pre- and post-Golgi membrane trafficking (Salminen and Novick 1987; Segev et al. 1988). Efforts to clone other members of the RAB gene family quickly established that this family had undergone significant expansion in higher eukaryotes. As each new RAB was discovered, it also became clear that different RAB proteins could adopt specific subcellular localizations, associating with particular membrane compartments and regulating the functions of these organelles. RAB18 was partially cloned in 1992 and then fully cloned in 1993 (Chavrier et al. 1992; Yu et al. 1993).
In 2011, mutations in RAB18 were shown to cause Warburg Micro syndrome, an autosomal recessive neurodevelopmental disorder also associated with mutations in RAB3GAP1, RAB3GAP2, and TBC1D20 (Aligianis et al. 2005; Bem et al. 2011; Borck et al. 2011; Liegel et al. 2013). Subsequently, a binary complex of RAB3GAP1 and RAB3GAP2 has been found to function as a guanine nucleotide exchange factor (GEF) for RAB18, whereas TBC1D20 functions as a RAB18 GTPase activating protein (GAP) (Gerondopoulos et al. 2014; Handley et al. 2015). RAB18 is a highly conserved and ubiquitously expressed RAB protein. It has most frequently been reported to localize to lipid droplets (LDs) as well as to the Endoplasmic Reticulum (ER) and the cis-Golgi. A role for RAB18 in regulating LDs, lipolysis and lipogenesis, has been established, and a role for RAB18 in regulating ER structure and Golgi-ER membrane trafficking is emerging (Martin et al. 2005; Ozeki et al. 2005; Dejgaard et al. 2008; Pulido et al. 2011; Gerondopoulos et al. 2014; Handley et al. 2015). Clear goals for future work are to define how functional loss of RAB18 contributes to disease and to define its cellular role in terms of the molecular mechanisms underlying its functions.
Structure and Conservation
In common with other small GTPases, RAB proteins can bind to the nucleotides guanosine triphosphate (GTP) and guanosine diphosphate (GDP). The “active site” of each protein contains conserved residues that interact with nucleotide phosphate groups, Mg2+, and the guanosine base. Regions denoted “Switch I” and “Switch II” surround the γ-phosphate of GTP when this nucleotide is bound (Takai et al. 2001). These regions are associated with conformational differences according to whether the RAB is GTP- or GDP-bound and with conformational change when GTP is hydrolyzed to GDP. The distinct nucleotide-bound RAB conformations are, in turn, associated with altered protein-binding characteristics and altered interactions with regulatory proteins and with effectors, the mediators of their downstream cellular functions. The altered properties of RABs when bound to different nucleotides mean that they are frequently referred to as “molecular switches.”
Most RAB proteins have a C-terminal di-cysteine motif that is subject to modification by two geranylgeranyl lipid groups. In difference, RAB18, like RAB8, RAB13, and RAB23, has a C-terminal CAAX motif like that of RAS and RHO proteins. This motif is monoprenylated and then sequentially cleaved and carboxymethylated by ER-resident enzymes (Leung et al. 2007). In each case, the lipid modifications mediate the association of RABs with cellular membranes, when not sequestered by a general RAB regulatory protein called GDP-dissociation inhibitor (GDI). However, the carboxymethylation of RAB18 may affect its interaction with GDI and therefore its membrane-association. Since carboxymethylation is a potentially reversible modification, this may represent an additional layer of RAB18’s regulation.
Sequence and structural similarities between RAB protein isoforms mean that they interact with some common regulatory proteins. However, the differences between individual RAB isoforms confer specificity to other interactions and contribute to the distinct roles of specific RABs in specific membrane-trafficking steps. Several attempts have been made to classify RAB18 by phylogeny and in terms of sequence motifs, and RAB18 has been grouped together with RAB1 and RAB8 isoforms (Klopper et al. 2012). RAB18 is a highly conserved ancestral RAB protein, present in both animal and plant lineages, but is not present in some eukaryotes including some Yeast. It is therefore possible that RAB18 function is redundant in some organisms or that aspects of this function have been adopted by another RAB isoform(s) in these cases.
Regulator and Effector Proteins
Following synthesis and posttranslational lipid modification, the membrane-association and activity of RAB proteins is regulated by four classes of protein: GDIs, GDI displacement factors (GDFs), GEFs, and GAPs (Takai et al. 2001; Barr 2013). GDIs and GDFs are considered general regulators of multiple RABs, while GEFs and GAPs are thought to show more specificity for particular RAB proteins. GDIs can sequester GDP-bound RABs in the cytosol by binding to their hydrophobic prenyl groups, but also coordinate with GDFs in the delivery of RABs to membranes. GEFs catalyze the exchange of bound GDP for GTP and so can serve to concentrate specific RABs on specific membrane compartments. RAB proteins are not susceptible to GDI-mediated membrane extraction when GTP bound, and the GTP-bound conformation is usually that which associates with effectors. Thus, GEFs can be seen as mediating RAB activation. In opposition to this, RAB-GAPs stimulate their intrinsic GTP-hydrolysis activity, mediating conversion of bound GTP to GDP. This can both abrogate their effector interactions and render them susceptible to membrane extraction. GAPs have therefore been associated with RAB inactivation. One model for RAB protein function suggests that GEF-mediated activation at a “donor” membrane compartment precedes effector-mediated coordination of the trafficking of transport vesicles to a “target” compartment at which GAP-mediated inactivation promotes recycling to the donor compartment. However, it is also possible that the coordination of the trafficking between donor and target compartments can require repeated rounds of activation and inactivation.
With the discovery that biallelic loss-of-function mutations in RAB18 cause Warburg Micro syndrome (Micro syndrome), the function of RAB18 was genetically linked to that of RAB3GAP1 and RAB3GAP2, mutations in which cause the same disease (Aligianis et al. 2005; Bem et al. 2011; Borck et al. 2011). RAB3GAP1 and RAB3GAP2 were first characterized as forming a binary complex with RAB-GAP activity towards RAB3 isoforms (Fukui et al. 1997; Nagano et al. 1998). Subsequent work has now shown that this complex is also a RAB18-GEF (Gerondopoulos et al. 2014). The RAB3GAP complex has specific RAB18-GEF activity in vitro and is capable of ectopic recruitment of RAB18 to cellular membranes in vivo. Additionally, disease-associated nonsynonymous mutations affecting the RAB3GAP subunits have been shown to disrupt biochemical RAB18-GEF activity while leaving RAB3-GAP activity intact, further supporting a link between RAB18 activation and disease.
The gene most recently associated with Micro syndrome is TBC1D20 (Liegel et al. 2013). TBC1D20 is a member of a family of RAB-GAPs sharing a conserved domain structure and had been reported to regulate RAB1 and RAB2 isoforms. Again because of the genetic link, the relationship between TBC1D20 and RAB18 was investigated (Handley et al. 2015). Consistent with the suggestion that TBC1D20 functions as a RAB18-GAP in vivo, the absence of the protein was associated with stabilization of RAB18-membrane-association and redistribution of RAB18 away from the cis-Golgi in human patient fibroblasts. Since RAB-GAPs potentiate GDI-mediated dissociation of RAB proteins from membranes and recycling to the donor membrane compartment, these findings also implicate RAB18 in trafficking from the cis-Golgi. Further, because Micro syndrome is clinically indistinguishable whether caused by loss-of-function mutations in RAB18, RAB3GAP1, RAB3GAP2, or TBC1D20, they suggest that both RAB18 activation by RAB3GAP and its inactivation by TBC1D20 are equally essential for its disease-relevant physiological role.
The identification of RAB18 effectors by conventional affinity purification approaches has been hampered by difficulties in expressing functional recombinant mammalian RAB18 in bacteria. However, a recent study in which interacting proteins were co-purified with the Drosophila RAB18 ortholog expressed in Drosophila S2 cells has yielded a number of important findings (Gillingham et al. 2014). A strength of this study was that binding-partners of multiple RABs were identified and compared, providing a good indication of isoform-specific interactors. The most prominent RAB18 interactors identified were components of the Dsl1/NRZ complex, a complex involved in the tethering and fusion of transport vesicles with the ER (Tagaya et al. 2014). Also identified were proteins encoded by disease-associated genes including orthologs of Spartin and LRRK2. Spartin is linked to SPG20/Troyer syndrome, a form of hereditary spastic paraplegia that shares some similarities with Micro syndrome (Patel et al. 2002). LRRK2 is linked to Parkinson’s disease (Paisan-Ruiz et al. 2004). Interestingly, LRRK2 has recently been shown to regulate a subset of RAB proteins by phosphorylating them and altering their binding to their regulators (Steger et al. 2016). RAB18 was not characterized among its substrates and so may be an as yet uncharacterized substrate, or potentially involved in directing the kinase towards other RABs. Similarly, a putative interaction between RAB18 and TBC1D5 may suggest that RAB18 is a TBC1D5 substrate, or potentially an effector directed to another RAB(s) by RAB18. The established RAB3-GAP activity of RAB3GAP1 (Fukui et al. 1997) may indicate that RAB18 activation is coupled to RAB3 inactivation in particular cells and in particular cellular locations.
Subcellular Localization and Function
In most cell types, RAB18 localizes to the Golgi and the ER, and several reports have suggested that it is involved in Golgi-ER trafficking (Martin et al. 2005; Dejgaard et al. 2008; Handley et al. 2015). The concentration of the RAB3GAP/RAB18-GEF complex at the cis-Golgi is consistent with the activation of RAB18 here, whereas TBC1D20/RAB18-GAP is an intrinsic ER protein, indicating that RAB18 is inactivated at this organelle (Handley et al. 2015). Further, the putative interaction between RAB18 and the Dsl1/NRZ complex strongly suggests that RAB18 is involved in the process by which carrier vesicles undergo tethering and fusion at the ER (Gillingham et al. 2014). It has also been shown that RAB18 regulates ER structure, with the loss of RAB18 or its regulators associated with disruption in the dynamics of tubular ER and a spreading of ER sheets into the cell periphery (Gerondopoulos et al. 2014). More speculatively, altered ER structure could also underlie the finding that RAB18 knockdown may affect the nuclear import and export of protein since the nuclear envelope is an ER subdomain (Dopie et al. 2015).
The localization of RAB18 to LDs, and its concentration at ER membrane enveloping them, is well established, and RAB18 has also been linked to regulation of lipogenesis and lipolysis (Martin et al. 2005; Ozeki et al. 2005; Pulido et al. 2011). Further, LD biogenesis is similarly altered in cellular models deficient in RAB3GAP and TBC1D20 as well as those deficient in RAB18, supporting a requirement for functional RAB18 in normal lipid handling (Liegel et al. 2013; Carpanini et al. 2014; Gerondopoulos et al. 2014). The putative RAB18-Spartin interaction may be involved in its LD regulation, since LD size and number are reduced in Spartin-depleted cells (Eastman et al. 2009). However, it still remains to be seen whether the molecular basis for LD regulation by RAB18 is a discrete one or whether it is secondary to its roles in regulating ER structure (Gerondopoulos et al. 2014).
Both the RAB3GAP complex, and more recently TBC1D20, have been associated with the regulation of autophagy (Spang et al. 2014; Sidjanin et al. 2016). It is tempting to speculate that their involvement relates to their regulation of RAB18, particularly given a putative interaction between RAB18 and Atg14 (Gillingham et al. 2014). Disruption of RAB18 function has been associated with accelerated N-cadherin degradation in neurons (Wu et al. 2016). Further, loss of either RAB3GAP or TBC1D20 leads to the posttranscriptional accumulation of RAB18 in cell lines, perhaps suggesting that RAB18 is degraded as part of its normal function (Handley et al. 2015). RAB3GAP has been associated with autophagosome biogenesis, while TBC1D20 has been associated with autophagosome maturation (Spang et al. 2014; Sidjanin et al. 2016). Potentially, therefore, RAB18 could link these processes.
In addition to more general roles for RAB18 in membrane trafficking, there is evidence that it can adopt specialized roles in specific cell types and also that its expression can be induced and its localization altered by cellular stimulation. In endocrine cell lines and in pituitary melanotropes, it has been found associated with secretory granules and is suggested to function in modulation of the secretory response (Vazquez-Martinez et al. 2007, 2008). In PC12 and AtT20 cells, stimulation with KCl led to redistribution of RAB18 from the cytosol to a subpopulation of secretory granules. Similarly, in adipocytes, recruitment of RAB18 to lipid droplets was promoted by treatment with insulin, which stimulates lipogenesis (Pulido et al. 2011), or by the β-adrenoceptor agonist isoproterenol, which stimulates lipolysis (Martin et al. 2005; Pulido et al. 2011). Although RAB18 expression is ubiquitous, it is found to be induced in differentiating adipocytes (Pulido et al. 2011) and in endothelial cells stimulated with histamine (Schafer et al. 2000).
RAB18 in Disease
Micro syndrome is a clinically distinctive developmental disorder with ocular, neurological, and endocrine features (Aligianis et al. 2005; Bem et al. 2011; Handley et al. 2013; Liegel et al. 2013). Affected children have congenital cataracts, microphthalmia, and atonic pupils that do not respond to light or mydriatic agents. Even with early cataract surgery, vision is limited by severe cortical visual impairment characterized by normal electroretinogram (ERG), but virtually absent visually evoked potentials (VEPs). The affected individuals have global developmental delay, postnatal microcephaly, and brain malformations including polymicrogyria and hypoplasia or agenesis of the corpus callosum. They have hypothalamic hypogonadism, and also hypotonia, going on to develop ascending spastic paraplegia and contractures. They do not learn to sit independently, walk, or talk.
The pathogenesis of Micro syndrome indicates that RAB18 function is important in eye and lens development, brain development, neurotransmission, neuronal migration, and homeostasis. Further, the associated global developmental delay may indicate a more general cellular deficit. It is not clear at present whether symptoms arise because cells of the eye and nervous system are more sensitive to loss of a ubiquitous RAB18 function, or through loss of cell-type-specific RAB18 activity. Nevertheless, animal models of Micro syndrome have been useful in the characterization of cell-type-specific phenotypes. Two RAB18-null mice have been reported (Carpanini et al. 2014; Cheng et al. 2015). The null mice have a reduced body weight, recapitulate ocular features of Micro syndrome including cataracts, microphthalmia, and atonic pupils and also show hypogonadism. Importantly, the mice show progressive motor deterioration and this has been linked to abnormal accumulation of cytoskeletal components including neurofilament and microtubules at neuromuscular junctions (NMJs) and also cytoskeletal disorganization within motor axons (Carpanini et al. 2014). Consistent with a conserved role for TBC1D20 in regulating RAB18, TBC1D20-null mice show a broadly similar phenotype, with cataracts, hypogonadism, and progressive motor deterioration (Liegel et al. 2013; Sidjanin et al. 2016). Characterization of these mice has focused on autophagy, identifying impairment in the process in eye and testes and accumulation of sequestosome 1 and ubiquitin in the brain (Sidjanin et al. 2016). Interestingly, the reported RAB3GAP1-null mouse model does not show a strong phenotype, although several breeds of dog with RAB3GAP1-mutations show features similar to those of Micro syndrome (Sakane et al. 2006; Mhlanga-Mutangadura et al. 2016a, b; Wiedmer et al. 2016). In a mouse knockdown model, reduced RAB18 and RAB3GAP each causes a similar impairment of neuronal migration (Wu et al. 2016). In Drosophila, loss of the RAB3GAP2 ortholog has been associated with a RAB3-dependent impairment of synaptic homeostasis (Muller et al. 2011).
It is important to note that RAB18 may have roles in disease in addition to those in Micro syndrome. Dysregulation of RAB18 expression has been observed in a number of different cancers and has been extensively studied in the context of pituitary tumors (Vazquez-Martinez et al. 2008). More recently, it has been identified as a target of miRNAs in cancers including lung and gastric cancer (Zhong et al. 2014; Liu et al. 2016). In a number of studies, both RAB18 and TBC1D20 have been associated with replication of Hepatitis C virus (Salloum et al. 2013; Cho et al. 2015). Individually, TBC1D20 has also been associated with Herpes simplex virus and HIV-1, and RAB18 has been associated with Dengue virus replication (Zenner et al. 2011; Nachmias et al. 2012; Tang et al. 2014).
To summarize, much has been learned about RAB18 in recent years, but much still remains to be discovered. A primary aim for future work should be to determine the molecular basis by which loss-of-functional RAB18 causes Micro syndrome. This work will be aided by the knowledge that its regulators RAB3GAP and TBC1D20 are essential for its pathologically relevant cellular function. RAB18 is known or likely to regulate multiple important physiological processes including Golgi-ER membrane trafficking, maintenance of normal ER structure, LD biogenesis and metabolism, autophagy, and exocytosis. Additionally, multiple cell-type-specific phenotypes have been described in animal models of Micro syndrome, including cytoskeletal disorganization in motor neurons, aberrant LD biogenesis, and dysregulated autophagy. A host of candidate RAB18 effectors has been identified, including several disease-associated proteins. Further, RAB18 function appears likely to involve reciprocal regulation of several other RABs. Careful examination of key interactions should provide important insights into cell biology.
- Carpanini SM, McKie L, Thomson D, Wright AK, Gordon SL, Roche SL, et al. A novel mouse model of Warburg Micro syndrome reveals roles for RAB18 in eye development and organisation of the neuronal cytoskeleton. Dis Model Mech. 2014;7:711–22. doi:10.1242/dmm.015222.CrossRefPubMedPubMedCentralGoogle Scholar
- Handley MT, Morris-Rosendahl DJ, Brown S, Macdonald F, Hardy C, Bem D, et al. Mutation spectrum in RAB3GAP1, RAB3GAP2, and RAB18 and genotype-phenotype correlations in Warburg micro syndrome and Martsolf syndrome. Hum Mutat. 2013;34:686–96. doi:10.1002/humu.22296.CrossRefPubMedPubMedCentralGoogle Scholar
- Liegel RP, Handley MT, Ronchetti A, Brown S, Langemeyer L, Linford A, et al. Loss-of-function mutations in TBC1D20 cause cataracts and male infertility in blind sterile mice and Warburg micro syndrome in humans. Am J Hum Genet. 2013;93:1001–14. doi:10.1016/j.ajhg.2013.10.011.CrossRefPubMedPubMedCentralGoogle Scholar
- Mhlanga-Mutangadura T, Johnson GS, Schnabel RD, Taylor JF, Johnson GC, Katz ML, et al. A mutation in the Warburg syndrome gene, RAB3GAP1, causes a similar syndrome with polyneuropathy and neuronal vacuolation in Black Russian Terrier dogs. Neurobiol Dis. 2016b;86:75–85. doi:10.1016/j.nbd.2015.11.016.CrossRefPubMedPubMedCentralGoogle Scholar
- Sakane A, Manabe S, Ishizaki H, Tanaka-Okamoto M, Kiyokage E, Toida K, et al. Rab3 GTPase-activating protein regulates synaptic transmission and plasticity through the inactivation of Rab3. Proc Natl Acad Sci U S A. 2006;103:10029–34. doi:10.1073/pnas.0600304103.CrossRefPubMedPubMedCentralGoogle Scholar
- Steger M, Tonelli F, Ito G, Davies P, Trost M, Vetter M, et al. Phosphoproteomics reveals that Parkinson’s disease kinase LRRK2 regulates a subset of Rab GTPases. eLife. 2016;5. doi:10.7554/eLife.12813.Google Scholar
- Vazquez-Martinez R, Martinez-Fuentes AJ, Pulido MR, Jimenez-Reina L, Quintero A, Leal-Cerro A, et al. Rab18 is reduced in pituitary tumors causing acromegaly and its overexpression reverts growth hormone hypersecretion. J Clin Endocrinol Metab. 2008;93:2269–76. doi:10.1210/jc.2007-1893.CrossRefPubMedPubMedCentralGoogle Scholar
- Wiedmer M, Oevermann A, Borer-Germann SE, Gorgas D, Shelton GD, Drogemuller M, et al. A RAB3GAP1 SINE insertion in Alaskan huskies with polyneuropathy, ocular abnormalities, and neuronal vacuolation (POANV) resembling human warburg micro syndrome 1 (WARBM1). G3. 2016;6:255–62. doi:10.1534/g3.115.022707.CrossRefGoogle Scholar