Skip to main content
NCBI home page
eLife logoLink to eLife
. 2014 Feb 11;3:e01623. doi: 10.7554/eLife.01623

Diversity and plasticity in Rab GTPase nucleotide release mechanism has consequences for Rab activation and inactivation

Lars Langemeyer 1, Ricardo Nunes Bastos 1, Yiying Cai 2, Aymelt Itzen 3, Karin M Reinisch 2, Francis A Barr 1,*
Editor: Suzanne R Pfeffer4
PMCID: PMC3919270  PMID: 24520163

Abstract

Ras superfamily GTPase activation and inactivation occur by canonical nucleotide exchange and GTP hydrolysis mechanisms. Despite conservation of active-site residues, the Ras-related Rab GTPase activation pathway differs from Ras and between different Rabs. Analysis of DENND1-Rab35, Rabex-Rab5, TRAPP-Rab1 and DrrA-Rab1 suggests Rabs have the potential for activation by distinct GDP-release pathways. Conserved active-site residues in the Rab switch II region stabilising the nucleotide-free form differentiate these pathways. For DENND1-Rab35 and DrrA-Rab1 the Rab active-site glutamine, often mutated to create constitutively active forms, is involved in GEF mediated GDP-release. By contrast, in Rab5 the switch II aspartate is required for Rabex mediated GDP-release. Furthermore, Rab1 switch II glutamine mutants refractory to activation by DrrA can be activated by TRAPP, showing that a single Rab can be activated by more than one mechanistically distinct GDP-release pathway. These findings highlight plasticity in the activation mechanisms of closely related Rab GTPases.

DOI: http://dx.doi.org/10.7554/eLife.01623.001

Research organism: human

eLife digest

The 70 or so members of the Rab subfamily of proteins perform a wide range of important tasks inside cells. A Rab protein is always bound to another molecule, which determines whether it is inactive or active. Binding to a molecule called GDP makes the Rab protein inactive, while binding to GTP makes it active. Proteins called guanine nucleotide exchange factors, or GEFs for short, activate the Rab protein by promoting the release of GDP and the binding of GTP. Other proteins—known as GAPs—lead to the inactivation of the Rab protein. Together these proteins form a molecular switch that can be turned on and off.

The Rab subfamily of proteins is part of the large Ras superfamily, and all members of this superfamily are activated and inactivated in a similar way, with the binding and unbinding of GDP and GTP taking place at a structure called the G-domain. The fact that the detailed structure of this domain (at the level of individual amino acids) has been conserved over evolution is often taken as an indication that its mechanism has also been conserved. Langemeyer et al. have now tested this assumption with four different types of GEFs—three from humans and one from the bacteria that cause Listeria—and found that the story is more complicated than expected.

The experiments showed that different amino acids in the active site of the Rab protein are involved when the GEFs mediate the release of the GDP during the activation process. For example, the amino acid glutamine is involved when the Listeria GEF and one of the human GEFs activate the protein, whereas a different amino acid—aspartate—is involved when one of the other human GEFs is responsible for the activation. Using this information, Langemeyer et al. create a human Rab protein that cannot be activated by the GEF from the bacteria that cause Listeria, but can still be activated by its normal human GEF.

By showing that different Rab proteins are activated by different mechanisms, and that a single Rab protein can be activated by more than one mechanism, the work of Langemeyer et al. clearly illustrates the on-going ability of evolution to surprise researchers.

DOI: http://dx.doi.org/10.7554/eLife.01623.002

Introduction

Rabs form an important and highly conserved subfamily of Ras-related GTPases that play essential roles in controlling membrane trafficking between the organelles of eukaryotic cells (Zerial and McBride, 2001; Pfeffer and Aivazian, 2004). Specific regulators controlling nucleotide exchange and hydrolysis promote kinetic proofreading of vesicle and target organelle membrane surfaces by Rab GTPases, and therefore permit accumulation of active Rabs only at the required sites (Barr and Lambright, 2010; Barr, 2013). The mechanistic details of how such regulators control Rab activation is therefore important for understanding the regulation of membrane identity and vesicle transport. Activation of Ras superfamily GTPases is thought to proceed by a general nucleotide exchange mechanism (Bos et al., 2007). In the Ras-SOS GEF (guanine nucleotide exchange factor) complex the Ras P-loop lysine interacts with a conserved glutamate intrinsic to the Ras active site switch II region, thereby stabilising the GEF bound nucleotide-free form of the GTPase (Boriack-Sjodin et al., 1998). Mutation of this glutamate therefore reduces GEF-stimulated GDP-release, and compromises Ras activation (Gasper et al., 2008). Because of the high degree of sequence conservation in the Ras superfamily and Rab subfamily (Klopper et al., 2012; Rojas et al., 2012), this mechanism might be expected to be the same in the Rab subfamily of GTPases. However, at odds with this simple idea mutation of the conserved Rab switch II glutamate residue to alanine has little effect on the rate of GEF-mediated nucleotide exchange (Gasper et al., 2008).

In addition to these potential differences between Ras and Rab activation, the mechanism of Rab inactivation by GTPase activating proteins (GAPs) diverges in key details from Ras. In Ras the conserved switch II glutamine 61 and an arginine residue contributed into the Ras active site by the GAP act together to promote GTP hydrolysis (Ahmadian et al., 1997; Scheffzek et al., 1997). Mutation of either residue therefore prevents GTP hydrolysis. This has great biological relevance since the Ras switch II glutamine is frequently mutated in cancers creating a constitutively active oncogenic form of the protein. In Rabs, although the switch II glutamine is conserved, crystal structures of Rab33 and Rab1 with TBC domain Rab GAPs, Gyp1p and TBC1D20, respectively, reveal that it does not play a direct role in GTP hydrolysis (Pan et al., 2006; Gavriljuk et al., 2012). Instead, the GAP contributes both arginine and glutamine residues important for catalysis to the Rab activate site (Pan et al., 2006; Gavriljuk et al., 2012).

Therefore, despite the high level of sequence conservation in the key switch regions of Ras and Rab family members, the mechanisms of activation and inactivation may differ between Ras and Rabs. In other words shared sequence cannot be assumed to imply shared mechanism. We therefore investigated the role of conserved Rab switch II active site residues in GEF-mediated activation to obtain insight into the function and reasons for their conservation. This analysis revealed that Rab activation diverges from the canonical Ras pathway in key details. Furthermore, we find that even within the Rab family different activation mechanisms are used by different Rabs.

Results

Differences in switch II interactions in Rab-GEF complexes

Inspection of the Ras-SOS GEF complex shows that the Ras P-loop lysine interacts with a conserved glutamate intrinsic to the active site switch II region (Figure 1A). Scrutiny of DENND1-Rab35 (Wu et al., 2011) and Rab1-DrrA (Schoebel et al., 2009; Suh et al., 2010) Rab-GEF complex crystal structures suggests divergence from the Ras activation pathway. In both DENND1 and DrrA complexes, the Rab P-loop lysine 21 interacts with the switch II glutamine 67 and aspartate 63 of the target GTPase (Figure 1B,C). Examination of other Rab GEF complexes provides further support for the view that Rab activation differs from Ras. In the case of Rab1 bound to TRAPP there is no interaction of the Rab P-loop lysine with switch II residues (Figure 1D). Instead, a glutamate at position 192 is contributed to the Rab active site by the C-terminal extension of the TRAPP Bet3 subunit (Cai et al., 2008; Chin et al., 2009; Figure 1D). This could be considered mimicry of the Rab switch II region glutamate by the GEF. For the Rab5 family GTPase Rab21 in complex with the Vps9 domain of the Rabex exchange factor, a conserved aspartate 74 in the switch II region of the Rab interacts with the P-loop lysine 32 (Delprato et al., 2004; Delprato and Lambright, 2007; Figure 1E). Analogous to TRAPP an acidic residue is also contributed by Rabex, in this case an aspartate at position 313 (Figure 1E). These different modes of Rab P-loop interaction with the switch II region or residues from the GEF suggest distinct GDP-release pathways exist for Rab GTPases.

Figure 1. Role of switch II residues in Rab GEF complexes.

Figure 1.

Open in a new tab

(A) The crystal structure of Ras with its exchange factor SOS highlighting the interaction of the Ras P-loop lysine 16 with the Ras switch II glutamate 62. Dotted yellow lines indicate potential Ras P-loop lysine interactions. (B) Portions of Rab structures from Rab–RabGEF complex crystal structures are shown for Rab35-DENND1 and (C) Rab1-DrrA. (D) Ypt1 (budding yeast Rab1) TRAPP and (E) Rab21-Vps9/Rabex complexes are shown. The Rab is indicated in grey while the GEF is depicted in green. Switch II residues are coloured according to their position for ease of reference.

DOI: http://dx.doi.org/10.7554/eLife.01623.003

Distinct roles for switch II residues in Rab activation

To test this idea a series of GEFs acting on Rab GTPases were analysed. First, the Rab35 GEF DENND1 was compared to the Rab1 GEF DrrA. Rab35 switch II glutamine 67 mutation to alanine (Q67A) greatly reduced DENND1-stimulated nucleotide exchange towards Rab35 (Figure 2A). Catalytic efficiency (kcat/Km) was reduced from 2.3 × 104 M−1s−1 similar to previous measurements using wild type Rab35 (Wu et al., 2011) to ∼7.5 × 102 M−1s−1 for the Q67A mutant. Importantly, the glutamine to alanine mutation had little effect on basal GEF-independent nucleotide exchange. In the case of Rab35 there is a specific requirement for glutamine at this position since alteration of the switch II glutamine to glutamate resulted in a form of Rab35 Q67E that was not activated by DENND1 (Figure 2A). In addition, mutation of the switch II glutamate 68 (E68A) required for nucleotide exchange in the canonical Ras activation pathway had little effect on DENND1-stimulated nucleotide exchange by Rab35 (Figure 2A). Similarly, the Q67A switch II mutation greatly reduced DrrA-mediated nucleotide exchange towards Rab1 from a kcat/Km of 6.0 × 104 to 5.3 × 103 M−1s−1 (Figure 2B). Mutation of the switch II glutamate (E68A) in Rab1 had little effect on DrrA-mediated nucleotide exchange, kcat/Km 6.0 × 104 M−1s−1 (Figure 2B). Removal of the switch II aspartate 63 contacting the P-loop lysine residue resulted in increased basal release for both Rab35 and Rab1 (Figure 2A,B), consistent with its role in nucleotide binding. For DENND1 this substitution resulted in a small reduction in kcat/Km to 1.7 × 104 M−1s−1 (Figure 2A), while for DrrA kcat/Km increased over threefold to 2 × 105 M−1s−1 (Figure 2B). These results suggest that the switch II glutamine in these Rabs plays a crucial and previously unsuspected role in GDP-release during Rab activation. By contrast, the switch II aspartate contributes to nucleotide binding and therefore limits GDP-release (Pai et al., 1989; John et al., 1993).

Figure 2. Distinct roles for switch II residues in GEF-mediated nucleotide exchange independent of the basal nucleotide release pathway.

Figure 2.

Open in a new tab

(A) Initial rates of nucleotide exchange as a function of GEF concentration are plotted for Rab35-DENND1, (B) Rab1-DrrA, (C) Rab5-Rabex and (D) Rab1-TRAPP. Wild type and mutant Rabs were used as indicated; curves are colour coded as in Figure 1 according to the position in the switch II region predicted to be important for GEF-mediated nucleotide release. Wild type full-length GEFs were used for DENND1, DrrA, Rabex and TRAPP, as well as the Rabex D313A mutant, and the TRAPP Bet3 E192A/D193A mutant.

DOI: http://dx.doi.org/10.7554/eLife.01623.004

The structure of the Rab5 family member Rab21 with its Vps9 domain GEF Rabex suggested that in this case the conserved glutamine is not in a position to promote nucleotide exchange (Figure 1E). Instead, the structure suggests an alternative pathway where the Rab P-loop lysine interacts with the Rab switch II aspartate and an aspartate finger residue D313 in Rabex. As reported previously, mutation of the Rabex aspartate to alanine D313A abolished the activity of the GEF (Figure 2C; Delprato et al., 2004). In support of the role of the Rab switch II aspartate in this alternative exchange pathway, mutation of the Rab5 conserved aspartate D76A greatly reduced Rabex-mediated nucleotide exchange (Figure 2C). Because the numbering of Rab5c is increased by two amino acids with respect to the Rab21, the same residue is D74 in the Rab21 structure (Figure 1E). Catalytic efficiency was reduced from 2.5×104 M−1s−1 in agreement with previous measurements using wild type Rab5 (Delprato et al., 2004; Delprato and Lambright, 2007) to ∼2.5 × 102 M−1s−1 for the D76A mutant. As expected from the crystal structure of the Rab21-Rabex complex (Figure 1E), the Q80A mutation (Q78 in Figure 1E) had little effect on the activity of Rabex towards Rab5 (Figure 2C). Interestingly, mutation of Rab5 switch II glutamate 81 (Rab21 E79 in Figure 1E) resulted in an increase in the rate of Rabex-stimulated GDP-release from a kcat/Km of 2.5 × 104 to 4.6 × 104 M−1s−1 (Figure 2C). This increase is possibly due to removal of the negative charge on the switch II favouring entry of Rabex aspartate 313 into the vicinity of the P-loop lysine. Rab5 activation by Rabex therefore proceeds via an alternative pathway in which the conserved switch II aspartate interacts with the P-loop lysine to promote GDP-release. This is different to Rab35-DENND1 or Ras-SOS where the switch II glutamine or glutamate fulfil an equivalent role.

Analysis of TRAPP and DrrA reveals plasticity in Rab1 activation

Both TRAPP and DrrA can activate Rab1 family GTPases, however the crystal structures of these Rab-GEF complexes reveal differences in the interaction network around the P-loop lysine. For Rab1-DrrA complexes, the Rab1 P-loop lysine contacts the switch II aspartate 63 and glutamine 67 residues (Figure 1C). This is different to the structure of Rab1 with its longin domain GEF TRAPP where there are no obvious contacts between switch II and the P-loop lysine 21 (Figure 1D). Instead, the structure suggests an alternative pathway where the Rab1 P-loop lysine interacts with an acidic glutamate finger residue E192 provided by the Bet3 subunit of the TRAPP GEF complex. The Rab1 switch II glutamine mutation is therefore predicted to have no effect on TRAPP mediated nucleotide exchange. In agreement with this idea, the switch II glutamine mutation Q67A that greatly reduced DrrA-mediated nucleotide exchange towards Rab1 (Figure 2B) had little effect on the activity of budding yeast TRAPP towards Rab1/Ypt1 (Figure 2D). As for DrrA, removal of the Rab1 switch II aspartate 63 contacting the P-loop lysine residue resulted in increased basal and TRAPP-stimulated GDP-release (Figure 2D). Interestingly, mutation of switch II glutamate 68 increased the rate of TRAPP-stimulated GDP-release over 10-fold from a kcat/Km of 1.3 × 105 M−1s−1. This was possibly due to removal of the negative charge on the switch II favouring entry of Bet3 glutamate 192 into the vicinity of the P-loop lysine. To confirm the requirement for the Bet3 acidic C-terminal region, TRAPP complexes containing a Bet3 E192A/D193A mutant were produced. These showed a fourfold reduction in GEF activity towards Rab1 (Figure 2C). This is similar to the result obtained with Rabex, which also inserts an acidic residue to coordinate the P-loop lysine during GDP-release. Rab1 activation by TRAPP therefore proceeds via a pathway in which the P-loop lysine does not interact with switch II, but is stabilised by an acidic residue from the GEF thereby promoting GDP-release.

Together, these results define previously unrecognised pathways for activation of a Ras superfamily GTPase that are discrete from the switch II glutamate dependent pathway followed by Ras. Notably, the results obtained with Rab1 also reveal that a single Ras superfamily GTPase has the potential to follow more than one activation pathway depending on the GEF it is coupling with.

Differential requirements for the switch II glutamine in GTP hydrolysis

Since the switch II glutamine appears to be an important element in the activation pathway of Rab proteins by GEFs, we also wanted to analyse the significance of its involvement in GAP-mediated Rab-inactivation. Therefore, the role of the switch II glutamine in the GTP hydrolysis reaction leading to Rab inactivation was examined. Crystal structures of Rab33 and Rab1 with TBC domain Rab GAPs, Gyp1p and TBC1D20, respectively, show that the GAP rather than the cognate GTPase contributes the glutamine residue important for nucleotide hydrolysis (Pan et al., 2006; Gavriljuk et al., 2012). In these cases the only contribution of the switch II glutamine appears to be in contacting the peptide backbone of the GAP. Therefore it may stabilise the Rab-GAP complex rather than playing a direct role in catalysis. The TBC domain Rab GAPs acting on Rab35, Rab5 and Rab1 were then analysed (Haas et al., 2005, 2007; Fuchs et al., 2007). Switch II glutamine mutation of Rab35 and Rab1 greatly reduced GTP hydrolysis stimulated by TBC1D10A and TBC1D20, respectively, but had less than ∼1.5-fold effect on basal GTP hydrolysis (Figure 3A,B). By contrast, the switch II glutamine mutation had no effect on the activity of RUTBC3 towards Rab5, but caused a greater than fivefold reduction in basal GTP hydrolysis (Figure 3C). Consequently, depending on the Rab, the switch II glutamine is important for Rab inactivation by GAP-stimulated or basal hydrolysis reactions.

Figure 3. Differential requirements for the Rab switch II glutamine in GAP-stimulated GTP hydrolysis.

Figure 3.

Open in a new tab

(A) Rab GTP hydrolysis as a function of GAP concentration are plotted for Rab35-TBC1D10, (B), and Rab1-TBC1D20 (C) Rab5-RUTBC3. Both wild type and switch II glutamine mutant Rabs were used. Basal GTP hydrolysis of purified Rabs in the absence of the cognate GAP is shown in the bar graph insets (mean +/− the deviation from the mean, n = 2).

DOI: http://dx.doi.org/10.7554/eLife.01623.005

A Rab35 switch II glutamine mutant is inactive not dominant active

Rab switch II glutamine mutants have been widely used as dominant active forms. This is best understood for Rab5 where such mutants promote endosome fusion (Stenmark et al., 1994; Rybin et al., 1996). The results presented here indicate that this may not apply to other Rabs, since the same mutation can interfere with GEF stimulated GDP-release and hence Rab activation. This suggests that in some Rabs switch II glutamine mutations will be uncoupled from their GEF and hence inactive rather than dominant active when expressed in cells. This idea was tested using Rab35 and the plasma membrane to Golgi trafficking pathway with Shiga toxin B as cargo protein. These experiments confirmed that wild type Rab35 supported delivery of cargo to the Golgi (Figure 4A, 90 min timepoint), as previously reported (Fuchs et al., 2007; Yoshimura et al., 2010). Under the same conditions the switch II mutant Rab35Q67A did not actively interfere with Shiga toxin transport (Figure 4A, 90 min timepoint). This indicates it does not act as a dominant negative inhibitor of transport. However, in the absence of the endogenous protein, Rab35Q67A failed to support efficient Shiga toxin transport to the Golgi (Figure 4B). Instead, Shiga toxin accumulated in small punctate structures and was not efficiently delivered to the Golgi even after 90 min (Figure 4B). This reduced accumulation in the Golgi was not due to altered binding of Shiga toxin to the cell surface, since this was equivalent in all the conditions used (Figure 4A,B, 0 min timepoint). Quantitation showed transport efficiency was over 80% for wild type Rab35, whereas Rab35Q67A did not rescue transport above the level seen in the absence of Rab35 (Figure 4C). Therefore Rab35 switch II glutamine mutants are inactive rather than dominant active.

Figure 4. The glutamine switch II mutant Rab35 fails to support Shiga toxin transport from the cell surface to the Golgi.

Figure 4.

Open in a new tab

(A) Shiga toxin B (STxB) uptake assays were performed for 0 and 90 min in HeLa cells expressing either Rab35, or (B) the Rab35Q67A mutant. Endogenous Rab35 was depleted using siRNA directed to the 3′-UTR or a mock depletion performed using a non-targeting control duplex. Cells were stained with a GM130 antibody to mark the Golgi. Scale bar is 10 µm. Cells outlined in yellow dotted lines in the 90 min timepoint express GFP-Rab35 or Rab35Q67A, and asterisks mark non-transfected cells. Arrowheads mark those cells shown in the enlarged panels to the right. (C) Delivery of Shiga toxin into the Golgi was scored and is plotted in the bar graph (mean +/− deviation from the mean, n = 2). (D) FRAP experiments were performed on cells expressing wild type and Switch II glutamine mutant Rab35 or (E) Rab5 (mean +/− standard deviation from the mean, n = 12). Western blots show the distribution of GFP-Rab35 or GFP-Rab5 in the membrane and cytosol fractions marked by the Golgi membrane protein GM130 and tubulin, respectively.

DOI: http://dx.doi.org/10.7554/eLife.01623.006

To test if this loss of function is associated with reduced membrane recruitment, the rate of wild type Rab35 and Q67 switch II mutant recruitment to organelle membranes was investigated using fluorescence recovery after photobleaching (FRAP). In agreement with the GEF activity data, Rab35 was rapidly recruited to membranes while the Rab35 Q67A or Q67L mutants showed only slow recovery (Figure 4D). For Rab5 the Q80A/L switch II mutation (Q78 in Rab21, Figure 1E) did not alter activation by the GEF Rabex (Figure 2C), and accordingly had no effect on the rate of recovery (Figure 4E), suggesting it is rapidly recruited to membranes like wild type Rab5. Western blot analysis confirmed that at steady state the membrane bound fractions of wild type and switch II glutamine mutant Rab35 and Rab5 were unaltered (Figure 4D,E, inset blots). This suggests that membrane anchoring of the Rabs was not affected by the mutations. Together, these results show that switch II glutamine mutant Rab35Q67A is defective for rapid membrane recruitment, and fails to support Shiga toxin trafficking to the Golgi.

Discussion

Divergence of the Ras and Rab activation cycles

The structural and biochemical analysis presented in this study identifies distinct pathways for Rab activation used by different RabGEF families. These can be divided into two main groups depending on the role played by the switch II region. For the first class of Rab GEFs, exemplified by DrrA and the DENN family member DENND1, GDP-release is promoted by interaction of the P-loop lysine with the conserved glutamine residue intrinsic to switch II (Figure 5A). Some details of this mechanism remain unresolved, particularly the question of how charge neutralization of the Rab P-loop lysine occurs remains unclear. The Rab–RabGEF complex structures suggest this is possibly due to the negatively charged switch II aspartate, however mutation of this residue does not reduce GEF mediated GDP-release. Future work, analysing structures of such mutants will therefore be necessary. For the second class of GEFs represented by the Vps9 family member Rabex and TRAPP, the Rab P-loop lysine interacts with a negatively charged residue, either aspartate or glutamate, provided by the GEF and extrinsic to the Rab (Figure 5A). This latter group can be further subdivided, since stabilisation of the P-loop lysine in Rab1 bound to TRAPP appears not to involve direct interaction with any residues in switch II. All these cases are different to the general nucleotide exchange mechanism proposed for the Ras family, in which formation of the nucleotide-free form of the GTPase is promoted by interaction of a glutamate intrinsic to the switch II region with the P-loop lysine (Boriack-Sjodin et al., 1998; Bos et al., 2007; Gasper et al., 2008). For Ras, this results in discrete requirements for conserved glutamate and glutamine residues in switch II for GEF-mediated activation and GAP-stimulated inactivation, respectively. However, for DENND1 and DrrA, the switch II glutamine of the target Rabs interacts with the P-loop lysine to promote conversion of the Rab GDP form to a nucleotide-free intermediate that then binds GTP (Figure 5B). Mutation of this glutamine therefore reduces the rate of GDP release by the Rab. In addition, the same glutamine interacts with the GAP and aides the GTP hydrolysis reaction converting the GTP to a GDP bound Rab (Figure 5B), thereby inactivating the Rab. This results in activation and inactivation processes with a shared requirement for the switch II glutamine residue. The glutamine side chain swings from an ‘in’ conformation where it contacts the P-loop lysine during nucleotide exchange (activation), to an ‘out’ conformation where it contacts the GAP during GTP hydrolysis (Figure 5C). DENND1, the Rab35 GEF studied here, is a member of the DENN and DENN-related proteins that form the largest family of Rab nucleotide exchange factors (Yoshimura et al., 2010; Barr, 2013). Therefore, these findings are likely to be broadly applicable to many other Rab GTPases. As a consequence they have wide significance for future studies of membrane trafficking and GTPase regulation in other systems.

Figure 5. Diversity and plasticity in Rab GTPase nucleotide release mechanism.

Figure 5.

Open in a new tab

(A) A schematic depicting the three different P-loop lysine interactions with the Rab switch II region and GEF, or Ras and the GEF SOS. Circled residues are required for GEF-mediated GDP release. (B) The switch II glutamine is required for both Rab GAP stimulated GTP hydrolysis, and DrrA or DENN GEF mediated nucleotide exchange reactions. GEF interaction with the GTPase results in distortion of switch I and II regions, and reduced affinity for both the guanosine and terminal phosphate of bound GDP. The switch II glutamine interacts with the P-loop lysine to displace the β-phosphate. This does not occur for switch II Q-mutant Rabs and GDP-release therefore fails. Adapted from Figure 5 of Thomas and Wittinghofer 2007 (Thomas et al., 2007). (C) A revised Rab GTPase cycle in which GEF-stimulated activation and GAP-mediated inactivation share a common determinant with respect to switch II.

DOI: http://dx.doi.org/10.7554/eLife.01623.007

Switch II function in GTP hydrolysis

The findings presented here also indicate that switch II glutamine mutant Rabs will behave differently depending on the nature of their GEF activator and GAP inactivator. For DENN GEF targets such as Rab35, because both GEF-stimulated activation and GAP-mediated inactivation are compromised (Figures 2A and 3A), the Rab does not fulfil its cellular purpose (Figure 4A–C). For Rab1/Ypt1 activated by TRAPP, switch II mutants slow but do not prevent the GAP-stimulated GTPase reaction (Figure 3B), possibly explaining the subcritical reduction in transport efficiency reported previously (De Antoni et al., 2002). In the case of Rab5, switch II glutamine Q79A mutation prolongs the lifetime of the active state by reducing the intrinsic or spontaneous rate of GTP hydrolysis (Figure 3C), and as a result promotes endosome fusion (Stenmark et al., 1994; Rybin et al., 1996).

Rab5 has served a paradigm for the function of Rab GTPases in membrane traffic (Zerial and McBride, 2001), yet as shown here switch II glutamine mutations in other Rabs may not behave in the same fashion as they do in Rab5. For Rab5 the switch II glutamine mutation results in a form that displays near wild type properties in terms of GAP-stimulated GTP hydrolysis, yet strongly compromised basal GTP hydrolysis. Results obtained previously for Rab33B and its GAP RUTBC1 (Nottingham et al., 2011) show this is true for at least one other Rab. However, for Rab1 and Rab35 tested here the same mutation greatly reduces GAP-stimulated GTP hydrolysis. These results have implications for the interpretation of published work on Rab5 regulation. While removal of the Rab5 GAP RUTBC3 results in moderately enlarged endosomes and a delay in endocytic trafficking, expression of Rab5 glutamine mutants causes formation of grossly enlarged vacuolar early endosomes and blocks the pathway at this stage (Stenmark et al., 1994; Haas et al., 2005). Together these findings indicate that in cells basal hydrolysis is an important determinant of Rab5 lifetime and turnover. This agrees with the idea that the major regulatory determinants of Rab5 levels at endosomal membranes are inputs driving GEF activity such as ubiquitylated cargo and PI-lipids rather than GAP activity (Del Conte-Zerial et al., 2008; Barr, 2013). By contrast, GAP-stimulated Rab5 turnover may occur in response to specific signals during cell migration and cell adhesion (Palamidessi et al., 2013).

Switch II glutamine mutants: a trap for the unwary

Switch II glutamine mutants have typically been used to trap Rab GTPases in the active state. As shown here this may generate a GTPase refractory to GEF-mediated activation and GAP-stimulated inactivation. The results reported here therefore have implications for the interpretation of many prior studies of Rab GTPase function that have relied on the use of switch II point mutants for individual Rabs or as part of libraries or Rab toolkits (Ullrich et al., 1996; Richardson et al., 1998; Clark et al., 2011; Dambournet et al., 2011; Gallegos et al., 2012; Ishida et al., 2012; Xiong et al., 2012). Depending on the nature of the GEF for the GTPase in question, then, the mutation may have additional consequences for GEF stimulated nucleotide exchange. Where this is not known the results of such experiments must be treated with great caution.

Multiple switch II configurations reveal plasticity in Rab activation

In analysing the role of switch II residues, this work has uncovered plasticity in terms of the different pathways leading to Rab activation. Although all Rabs switch between common GDP and GTP bound states, the intermediates of the nucleotide exchange reaction differ in crucial features. In some cases a single Rab, shown here for Rab1, can follow a different activation pathway depending on the nature of its cognate GEF. Importantly, switch II mutations that compromise one pathway do not necessarily affect other activation routes. Thus, Rab1 switch II glutamine mutants can be activated by TRAPP, yet are refractory to activation by DrrA. These findings raise the possibility that small molecule inhibitors could be developed to target these mechanistically discrete pathways and associated switch II conformations.

Materials and methods

Reagents and antibodies

General laboratory chemicals were obtained from Sigma–Aldrich, UK and Fisher Scientific, UK. Commercially available antibodies were used to GM130 (mouse clone 35; BD Biosciences, UK). Secondary antibodies raised in donkey to mouse, rabbit, sheep/goat, and human conjugated to HRP, Alexa-488, Alexa-555, Alexa-568, and Alexa-647 were obtained from Molecular Probes/Life Technologies, UK and Jackson ImmunoResearch Laboratories Inc, West Grove, PA.

Molecular biology and protein purification

The libraries of hexahistidine-GST in pFAT2 and eGFP-tagged Rab GTPases and human GEF coding sequences have been described previously (Yoshimura et al., 2010). Mutagenesis was performed using the Quickchange method according to the protocol (Agilent Technologies, UK). Rab proteins in pFAT2 were expressed in BL21 (DE3) pRIL at 18°C for 12–14 hr, then purified using Ni-NTA agarose as described previously (Fuchs et al., 2007). In brief, cell pellets were lysed in 20 ml IMAC20 (20 mM Tris–HCl, pH 8.0, 300 mM NaCl, 20 mM imidazole, and protease inhibitor cocktail; Roche Diagnostics, UK) using an Emulsiflex C-5 system (AVESTIN, Germany). Lysates were clarified by centrifugation at 16,000×g rpm in a JA-17 rotor for 30 min. To purify the tagged protein, 0.5 ml of nickel-charged NTA-agarose (QIAGEN, UK) was added to the clarified lysate and rotated for 2 hr. The agarose was washed three times with IMAC20 and the bound proteins eluted in IMAC200 (IMAC20 with 200 mM imidazole) collecting 1.5 ml fractions. All manipulations were performed on ice or in an 8°C cold room. Hexahistidine-tagged Rabex5, DENND1B-S, in pQE32 were expressed in JM109 at 18°C for 12–14 hr, then purified using nickel-charged NTA agarose using the same procedure as the Rabs. RabGAPs RUTBC3, TBC1D10A and TBC1D20 were purified as described previously (Fuchs et al., 2007). Purified proteins were dialyzed against TBS (50 mM Tris–HCl, pH 7.4, and 150 mM NaCl) and then snap frozen in liquid nitrogen for storage at −80°C. Protein concentration was measured using the Bradford assay. TRAPPI-complex and DrrA were purified as described elsewhere (Cai et al., 2008; Schoebel et al., 2009).

Nucleotide binding and Rab GEF assays

First 10 nmol of hexahistidine-GST-Rab was loaded with 2′-(3′)-bis-O-(N-methylanthraniloyl)-GDP (Mant-GDP) (Jena Bioscience, Germany) in 20 mM HEPES, pH 6.8, 1 mg/ml BSA, 20 mM EDTA, pH 8.0, 40 mM Mant-GDP at 30°C for 30 min. After loading 25 mmol MgCl2 was added and the sample was exchanged into reaction buffer (20 mM HEPES, pH 6.8, 1 mg/ml BSA, 150 mM NaCl, 1 mM MgCl2) using Zeba spin columns (Fisher Scientific). This step removes the free Mant-GDP leaving only Rab bound nucleotide. Nucleotide exchange was then measured using 1 nmol of the loaded Rab and the amount of GEF specified in the figure legends in a final volume of 100 µl reaction buffer by monitoring the quenching of fluorescence after release of Mant-GDP using a Tristar LB 941 plate reader (Berthold Technologies, UK) under control of MikroWin Software. Samples were excited at 350 nm and emission monitored at 440 nm. GTP was added to a final concentration of 0.1 mM to start the exchange reaction at 30°C. Curve fitting and extraction of pseudo first order rate constants (kobs) was carried out as described previously (Delprato et al., 2004; Delprato and Lambright, 2007). Since kobs = (kcat/Km)[GEF] + kbasal where kbasal is the rate constant measured in the absence of GEF, catalytic efficiency (kcat/Km) can be obtained.

Rab GAP assays

For GTP-loading reactions, 10 µl assay buffer (20 mM HEPES, pH 6.8, 1 mg/ml BSA), 73 µl H2O, 10 µl of 10 mM EDTA, pH 8.0, 5 µl of 1 mM GTP, 2 µl γ-[32P]-GTP (6000 Ci/mmol 10 mCi/ml Perkin Elmer, UK), and 100 pmol Rab-GTPase were mixed on ice. After 30 min of incubation at 30°C, loaded GTPases were stored on ice. GAP reactions were started by the addition of GAP to 1 µM Rab in 50 µl final volume as specified in the figures. A 2.5 μl aliquot of the assay mix was scintillation counted to measure the specific activity in counts per minute per picomole of GTP. Reactions were then incubated at 30°C for 0 to 60 min and then split into two equal aliquots. 5 µl of each aliquot was immediately added to 795 µl of ice-cold 5% (wt/vol) activated charcoal slurry in 50 mM NaH2PO4, left for 1 hr on ice, and centrifuged at 20,000×g to pellet the charcoal. A 400 µl aliquot of the supernatant was scintillation counted, and the amount of GTP hydrolyzed was calculated from the specific activity of the reaction mixture.

Shiga toxin uptake assays

HeLa cells were cultured on No. 1.5 glass coverslips (Menzel-Gläser, Fisher Scientific) in DMEM containing 10% bovine calf serum (Invitrogen) at 37°C and 5% CO2. Endogenous Rab35 was depleted using siRNA duplexes obtained from Qiagen directed against the 3′-UTR (Hs_RAB35_4 SI00092638), target sequence 5′-CCTGGGAAGAACCGAGTTTAA-3′ transfected using Oligofectamine (Life Technologies) for 72 hr. Cells were then transfected with eGFP-Rab35 or Rab35Q67A for 18 hr using Mirus LT1 (Mirus Bio LLC, Madison, WI). Shiga toxin assays were then carried out as described previously (Fuchs et al., 2007). For imaging samples were washed twice with 2 ml of PBS, and fixed with 2 ml of 3% (wt/vol) paraformaldehyde in PBS for 15 min. Fixative was removed and the cells quenched with 2 ml of 50 mM NH4Cl in PBS for 10 min. Coverslips were washed three times in 2 ml PBS before permeabilization in 0.2% (vol/vol) Triton-X 100 for 5 min. Cells were then stained with GM130 antibodies. Primary and secondary antibody staining was carried out in PBS for 60 min at room temperature. Coverslips were mounted in Mowiol 4-88 mounting medium (Merck Millipore, UK). Fixed samples on glass slides were imaged using a 60x NA1.35 oil immersion objective on an Olympus BX61 upright microscope with filter sets for DAPI, GFP/Alexa-488, Alexa-555, Alexa-568, and Alexa-647 (Chroma Technology Corp., Bellows Falls, VT), a CoolSNAP HQ2 camera (Roper Scientific), and Metamorph 7.5 imaging software (Molecular Dynamics Inc., Sunnyvale, CA).

FRAP and membrane fractionation assays for Rab recruitment

For live cell imaging using spinning disk confocal microscopy, cells were plated in 35 mm dishes with a 14 mm No. 1.5 thickness coverglass window in the bottom (MatTek Corporation, Ashland, MA). Cells were left for 24 hr then transfected with eGFP-Rab constructs for a further 16 hr. For imaging, the dishes were placed in a 37°C and 5% CO2 environment chamber (Tokai Hit CO., Ltd, Japan) on the microscope stage. Imaging was performed at 37°C in 5% CO2 using an Olympus IX81 inverted microscope with a 60x 1.42NA oil immersion objective coupled to an Ultraview Vox spinning disk confocal system (Perkin Elmer) fitted with a C9100-13 EM-CCD camera (Hamamatsu Photonics Limited, UK). For FRAP, five image stacks of four planes with 0.2 μm spacing were acquired at 1 s intervals during the pre-bleach period. Bleach of the eGFP-Rab signal was performed using an UltraVIEW PK Device with the 488 nm laser set at 10% with the following settings: cycles 5, step size 1, spot period 10, stop period 10, spot cycles 1, small spot size and no attenuation. Recovery images were acquired for 30 time points every 2 s using 50 ms exposures at 4% laser power and then a further 30 time points every 4 s. Quantification and analysis of the FRAP data were performed using ImageJ. For membrane fractionation the cells were washed from the dish in PBS containing 1 mM EDTA, then homogenized using 20 passes through an 18-gauge needle in 50 mM HEPES-NaOH pH7.4, 200 mM sucrose. Unbroken cells were removed by centrifugation at 1000×g for 10 min in a microfuge. A membrane pellet and cytosol were prepared from this post-nuclear supernatant by centrifugation at 100,000×g for 60 min in a TLA-100 rotor. Equivalent proportions of the membrane pellet and cytosol were analysed by western blotting.

Funding Statement

The funder had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Funding Information

This paper was supported by the following grant:

  • Wellcome Trust 097769/Z/11/Z to Francis A Barr.

Additional information

Competing interests

The authors declare that no competing interests exist.

Author contributions

LL, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article.

RNB, Acquisition of data, Analysis and interpretation of data, Contributed unpublished essential data or reagents.

YC, Analysis and interpretation of data, Drafting or revising the article, Contributed unpublished essential data or reagents.

AI, Analysis and interpretation of data, Drafting or revising the article, Contributed unpublished essential data or reagents.

KR, Analysis and interpretation of data, Drafting or revising the article, Contributed unpublished essential data or reagents.

FAB, Conception and design, Analysis and interpretation of data, Drafting or revising the article.

References

  1. Ahmadian MR, Stege P, Scheffzek K, Wittinghofer A. 1997. Confirmation of the arginine-finger hypothesis for the GAP-stimulated GTP-hydrolysis reaction of Ras. Nature Structural Biology 4:686–689. 10.1038/nsb0997-686 [DOI] [PubMed] [Google Scholar]
  2. Barr F, Lambright DG. 2010. Rab GEFs and GAPs. Current Opinion Cell Biology 22:461–470. 10.1016/j.ceb.2010.04.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Barr FA. 2013. Rab GTPases and membrane identity: causal or inconsequential? The Journal of Cell Biology 202:191–199. 10.1083/jcb.201306010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Boriack-Sjodin PA, Margarit SM, Bar-Sagi D, Kuriyan J. 1998. The structural basis of the activation of Ras by Sos. Nature 394:337–343. 10.1038/28548 [DOI] [PubMed] [Google Scholar]
  5. Bos JL, Rehmann H, Wittinghofer A. 2007. GEFs and GAPs: critical elements in the control of small G proteins. Cell 129:865–877. 10.1016/j.cell.2007.05.018 [DOI] [PubMed] [Google Scholar]
  6. Cai Y, Chin HF, Lazarova D, Menon S, Fu C, Cai H, Sclafani A, Rodgers DW, De La Cruz EM, Ferro-Novick S, Reinisch KM. 2008. The structural basis for activation of the Rab Ypt1p by the TRAPP membrane-tethering complexes. Cell 133:1202–1213. 10.1016/j.cell.2008.04.049 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chin HF, Cai Y, Menon S, Ferro-Novick S, Reinisch KM, De La Cruz EM. 2009. Kinetic analysis of the guanine nucleotide exchange activity of TRAPP, a multimeric Ypt1p exchange factor. Journal of Molecular Biology 389:275–288. 10.1016/j.jmb.2009.03.068 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Clark BS, Winter M, Cohen AR, Link BA. 2011. Generation of Rab-based transgenic lines for in vivo studies of endosome biology in zebrafish. Developmental Dynamics 240:2452–2465. 10.1002/dvdy.22758 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Dambournet D, Machicoane M, Chesneau L, Sachse M, Rocancourt M, El Marjou A, Formstecher E, Salomon R, Goud B, Echard A. 2011. Rab35 GTPase and OCRL phosphatase remodel lipids and F-actin for successful cytokinesis. Nature Cell Biology 13:981–988. 10.1038/ncb2279 [DOI] [PubMed] [Google Scholar]
  10. De Antoni A, Schmitzova J, Trepte HH, Gallwitz D, Albert S. 2002. Significance of GTP hydrolysis in Ypt1p-regulated endoplasmic reticulum to Golgi transport revealed by the analysis of two novel Ypt1-GAPs. The Journal of Biological Chemistry 277:41023–41031. 10.1074/jbc.M205783200 [DOI] [PubMed] [Google Scholar]
  11. Del Conte-Zerial P, Brusch L, Rink JC, Collinet C, Kalaidzidis Y, Zerial M, Deutsch A. 2008. Membrane identity and GTPase cascades regulated by toggle and cut-out switches. Molecular Systems Biology 4:206. 10.1038/msb.2008.45 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Delprato A, Lambright DG. 2007. Structural basis for Rab GTPase activation by VPS9 domain exchange factors. Nature Structural & Molecular Biology 14:406–412. 10.1038/nsmb1232 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Delprato A, Merithew E, Lambright DG. 2004. Structure, exchange determinants, and family-wide rab specificity of the tandem helical bundle and Vps9 domains of Rabex-5. Cell 118:607–617. 10.1016/j.cell.2004.08.009 [DOI] [PubMed] [Google Scholar]
  14. Fuchs E, Haas AK, Spooner RA, Yoshimura S, Lord JM, Barr FA. 2007. Specific Rab GTPase-activating proteins define the Shiga toxin and epidermal growth factor uptake pathways. The Journal of Cell Biology 177:1133–1143. 10.1083/jcb.200612068 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Gallegos ME, Balakrishnan S, Chandramouli P, Arora S, Azameera A, Babushekar A, Bargoma E, Bokhari A, Chava SK, Das P, Desai M, Decena D, Saramma SD, Dey B, Doss AL, Gor N, Gudiputi L, Guo C, Hande S, Jensen M, Jones S, Jones N, Jorgens D, Karamchedu P, Kamrani K, Kolora LD, Kristensen L, Kwan K, Lau H, Maharaj P, Mander N, Mangipudi K, Menakuru H, Mody V, Mohanty S, Mukkamala S, Mundra SA, Nagaraju S, Narayanaswamy R, Ndungu-Case C, Noorbakhsh M, Patel J, Patel P, Pendem SV, Ponakala A, Rath M, Robles MC, Rokkam D, Roth C, Sasidharan P, Shah S, Tandon S, Suprai J, Truong TQ, Uthayaruban R, Varma A, Ved U, Wang Z, Yu Z. 2012. The C. elegans rab family: identification, classification and toolkit construction. PLOS ONE 7:e49387. 10.1371/journal.pone.0049387 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Gasper R, Thomas C, Ahmadian MR, Wittinghofer A. 2008. The role of the conserved switch II glutamate in guanine nucleotide exchange factor-mediated nucleotide exchange of GTP-binding proteins. Journal of Molecular Biology 379:51–63. 10.1016/j.jmb.2008.03.011 [DOI] [PubMed] [Google Scholar]
  17. Gavriljuk K, Gazdag EM, Itzen A, Kotting C, Goody RS, Gerwert K. 2012. Catalytic mechanism of a mammalian Rab.RabGAP complex in atomic detail. Proceedings of the National Academy of Sciences of the United States of America 109:21348–21353. 10.1073/pnas.1214431110 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Haas AK, Fuchs E, Kopajtich R, Barr FA. 2005. A GTPase-activating protein controls Rab5 function in endocytic trafficking. Nature Cell Biology 7:887–893. 10.1038/ncb1290 [DOI] [PubMed] [Google Scholar]
  19. Haas AK, Yoshimura S, Stephens DJ, Preisinger C, Fuchs E, Barr FA. 2007. Analysis of GTPase-activating proteins: Rab1 and Rab43 are key Rabs required to maintain a functional Golgi complex in human cells. Journal of Cell Science 120:2997–3010. 10.1242/jcs.014225 [DOI] [PubMed] [Google Scholar]
  20. Ishida M, Ohbayashi N, Maruta Y, Ebata Y, Fukuda M. 2012. Functional involvement of Rab1A in microtubule-dependent anterograde melanosome transport in melanocytes. Journal of Cell Science 125:5177–5187. 10.1242/jcs.109314 [DOI] [PubMed] [Google Scholar]
  21. John J, Rensland H, Schlichting I, Vetter I, Borasio GD, Goody RS, Wittinghofer A. 1993. Kinetic and structural analysis of the Mg(2+)-binding site of the guanine nucleotide-binding protein p21H-ras. The Journal of Biological Chemistry 268:923–929 [PubMed] [Google Scholar]
  22. Klopper TH, Kienle N, Fasshauer D, Munro S. 2012. Untangling the evolution of Rab G proteins: implications of a comprehensive genomic analysis. BMC Biology 10:71. 10.1186/1741-7007-10-71 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Nottingham RM, Ganley IG, Barr FA, Lambright DG, Pfeffer SR. 2011. RUTBC1 protein, a Rab9A effector that activates GTP hydrolysis by Rab32 and Rab33B proteins. The Journal of Biological Chemistry 286:33213–33222. 10.1074/jbc.M111.261115 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Pai EF, Kabsch W, Krengel U, Holmes KC, John J, Wittinghofer A. 1989. Structure of the guanine-nucleotide-binding domain of the Ha-ras oncogene product p21 in the triphosphate conformation. Nature 341:209–214. 10.1038/341209a0 [DOI] [PubMed] [Google Scholar]
  25. Palamidessi A, Frittoli E, Ducano N, Offenhauser N, Sigismund S, Kajiho H, Parazzoli D, Oldani A, Gobbi M, Serini G, Di Fiore PP, Scita G, Lanzetti L. 2013. The GTPase-activating protein RN-tre controls focal adhesion turnover and cell migration. Current biology: CB 23:2355–2364. 10.1016/j.cub.2013.09.060 [DOI] [PubMed] [Google Scholar]
  26. Pan X, Eathiraj S, Munson M, Lambright DG. 2006. TBC-domain GAPs for Rab GTPases accelerate GTP hydrolysis by a dual-finger mechanism. Nature 442:303–306. 10.1038/nature04847 [DOI] [PubMed] [Google Scholar]
  27. Pfeffer S, Aivazian D. 2004. Targeting Rab GTPases to distinct membrane compartments. Nature Reviews Molecular Cell Biology 5:886–896. 10.1038/nrm1500 [DOI] [PubMed] [Google Scholar]
  28. Richardson CJ, Jones S, Litt RJ, Segev N. 1998. GTP hydrolysis is not important for Ypt1 GTPase function in vesicular transport. Molecular and Cellular Biology 18:827–838 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Rojas AM, Fuentes G, Rausell A, Valencia A. 2012. The Ras protein superfamily: evolutionary tree and role of conserved amino acids. The Journal of Cell Biology 196:189–201. 10.1083/jcb.201103008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Rybin V, Ullrich O, Rubino M, Alexandrov K, Simon I, Seabra MC, Goody R, Zerial M. 1996. GTPase activity of Rab5 acts as a timer for endocytic membrane fusion. Nature 383:266–269. 10.1038/383266a0 [DOI] [PubMed] [Google Scholar]
  31. Scheffzek K, Ahmadian MR, Kabsch W, Wiesmuller L, Lautwein A, Schmitz F, Wittinghofer A. 1997. The Ras-RasGAP complex: structural basis for GTPase activation and its loss in oncogenic Ras mutants. Science 277:333–338. 10.1126/science.277.5324.333 [DOI] [PubMed] [Google Scholar]
  32. Schoebel S, Oesterlin LK, Blankenfeldt W, Goody RS, Itzen A. 2009. RabGDI displacement by DrrA from Legionella is a consequence of its guanine nucleotide exchange activity. Molecular Cell 36:1060–1072. 10.1016/j.molcel.2009.11.014 [DOI] [PubMed] [Google Scholar]
  33. Stenmark H, Parton RG, Steele-Mortimer O, Lutcke A, Gruenberg J, Zerial M. 1994. Inhibition of rab5 GTPase activity stimulates membrane fusion in endocytosis. The EMBO Journal 13:1287–1296 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Suh HY, Lee DW, Lee KH, Ku B, Choi SJ, Woo JS, Kim YG, Oh BH. 2010. Structural insights into the dual nucleotide exchange and GDI displacement activity of SidM/DrrA. The EMBO Journal 29:496–504. 10.1038/emboj.2009.347 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Thomas C, Fricke I, Scrima A, Berken A, Wittinghofer A. 2007. Structural evidence for a common intermediate in small G protein-GEF reactions. Molecular Cell 25:141–149. 10.1016/j.molcel.2006.11.023 [DOI] [PubMed] [Google Scholar]
  36. Ullrich O, Reinsch S, Urbe S, Zerial M, Parton RG. 1996. Rab11 regulates recycling through the pericentriolar recycling endosome. The Journal of Cell Biology 135:913–924. 10.1083/jcb.135.4.913 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Wu X, Bradley MJ, Cai Y, Kummel D, De La Cruz EM, Barr FA, Reinisch KM. 2011. Insights regarding guanine nucleotide exchange from the structure of a DENN-domain protein complexed with its Rab GTPase substrate. Proceedings of the National Academy of Sciences of the United States of America 108:18672–18677. 10.1073/pnas.1110415108 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Xiong B, Bayat V, Jaiswal M, Zhang K, Sandoval H, Charng WL, Li T, David G, Duraine L, Lin YQ, Neely GG, Yamamoto S, Bellen HJ. 2012. Crag is a GEF for Rab11 required for rhodopsin trafficking and maintenance of adult photoreceptor cells. PLOS Biology 10:e1001438. 10.1371/journal.pbio.1001438 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Yoshimura S, Gerondopoulos A, Linford A, Rigden DJ, Barr FA. 2010. Family-wide characterization of the DENN domain Rab GDP-GTP exchange factors. The Journal of Cell Biology 191:367–381. 10.1083/jcb.201008051 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Zerial M, McBride H. 2001. Rab proteins as membrane organizers. Nature Reviews Molecular Cell Biology 2:107–117. 10.1038/35052055 [DOI] [PubMed] [Google Scholar]
eLife. 2014 Feb 11;3:e01623. doi: 10.7554/eLife.01623.008

Decision letter

Editor: Suzanne R Pfeffer1

eLife posts the editorial decision letter and author response on a selection of the published articles (subject to the approval of the authors). An edited version of the letter sent to the authors after peer review is shown, indicating the substantive concerns or comments; minor concerns are not usually shown. Reviewers have the opportunity to discuss the decision before the letter is sent (see review process). Similarly, the author response typically shows only responses to the major concerns raised by the reviewers.

Thank you for sending your work entitled “Plasticity in nucleotide release mechanism results in coupling of Rab GTPase activation and inactivation” for consideration at eLife. Your article has been evaluated by a Senior editor and 3 reviewers, one of whom, Suzanne Pfeffer, is a member of our Board of Reviewing Editors.

The Reviewing editor and the other reviewers discussed their comments before we reached this decision, and the Reviewing editor has assembled the following comments to help you prepare a completely revised submission.

This is an important and interesting paper that shows that Rab GTPase GEFs and GAPs act by distinct mechanisms, even when a single Rab is the substrate. The work has the potential be of broad interest to cell biologists interested in molecular mechanisms. However, the manuscript needs significant reworking to clarify the story for the non-cognoscenti. The authors are encouraged to step back and rewrite the paper to guide the reader through the structures carefully, one figure at a time and keeping all of the referee comments in mind when crafting their presentation.

The reviewers felt strongly that the use of the word coupling is not correct and would be confusing to cell biologists. The concept of coupling implies coupled reactions, which is not the case since the GEFs and GAPs presumably act in different locations in cells. In addition, two expert enzymologists feel that classifying families based on one Q or D/E residue is not a sound practice. You need to convince them that this is helpful for the field if you believe otherwise. Finally, one wrote, “...The authors need to re-draw their conclusions in a way that maintains a high level of scholarly integrity.” I think they mean tone down the idea that this is a major conceptual advance – it is, nevertheless, a surprising and interesting result that these enzymes use different mechanisms. This reviewer felt a change in the manuscript title would also help.

Because rewriting is required I attach here all the referee comments. Normally, a complete rewriting could be cause for rejection but the referees believe that a very interesting finding can be found in the core of this manuscript and encourage a second round of evaluation.

Reviewer #1:

1) The figure legends need to describe the figures much more clearly and accurately. 1b has no legend; Figure 3 legend is out of order; Figure 4 b,d,f,h are either reverse labeled or need better clarification in terms of what is presented.

2) Figure 2, panel b, bottom. Please show the Rab1-GDP with the same projection as Rab1-DrrA and/or move it for comparison with Rab1-GTP to indicate residue orientation for comparison. Is there a Q-GEF for Rab1 or is that a microbial adaptation and approach to perfection? This could be tested – does Rab1-Q mutant rescue a Rab1 phenotype.

3) Figure 3. Panel d should be moved to Figure 7b. Please compare Rab1-GTP with Rab1-GDP and show key residues for comparison. Bottom text “Divergent in Q/E positons” is misplaced.

4) Figure 4. Please include Bet3 E192 mutant for completeness. Can you include an S-GEF example? Please relabel b,d,f,h so that the reader understands what is shown.

5) Figure 5 basal hydrolysis needs to be shown as a rate, not a one hour time point. This is important since GTP hydrolysis rates vary significantly between Rabs. Would inclusion of the Rab5Q or Rab5GTP structure be useful to show here to understand why Rab5Q mutant is so poor in intrinsic hydrolysis?

6) How many cells were counted in 6c? The images would be easier to understand if cells expressing the particular Rab were indicated – it seems some cells express more of the Rab and the phenotype does not seem uniform. Please present this more clearly. Also, the FRAP in d, e may be the result of differences in capacity to interact with GDI and without monitoring relative prenylation and cytosol/membrane fraction cannot really be interpreted. Please remove or add additional data so that the FRAP can be understood molecularly.

Please include the fact that RUTBC1 was shown to activate a catalytically inhibited Rab33B mutant (Q92A), in support of a dual finger mechanism for RUTBC1 action (Nottingham & Barr et al. 2011).

Reviewer #2:

The manuscript by Langemeyer et al investigates the mechanisms by which different Rab GTPases are activated by the nucleotide exchange activity of their respective GEFs and, in some cases, whether this activation is 'coupled' to inactivation by the GAPs. Starting from a survey of available crystal structures and sequence information, the authors noted that different GEFs may use distinct residues (Q, E/D, or S, sometimes in the Rab Switch II loop and sometimes supplied by the GEF) for activation of Rab, and the same Rab (e.g., Rab1) may use distinct residues for activation by different GEFs. This is experimentally tested in the cases of Rab35 (Q), Rab5 (D), and Rab1 (Q- or E-) for GEF-mediated nucleotide exchange and GAP-mediated GTPase activation. These experiments also show that the same glutamine in Rab35 and Rab1 is important for both GEF-mediated activation and GAP-mediated inactivation, hence the notion of 'coupled' 'pathways' for activation and inactivation. The importance of the glutamine for activation of Rab35 in vivo is further verified in a Shiga toxin B uptake assay.

Overall, this is more of a 'hypothesis' paper (4 model/literature review figures vs 3 data figures) that (i) suggests a new way of classifying the Rab GEFs based on the residue used in Rab activation; and (ii) proposes that activation and inactivation of some Rabs that use Q-GEFs are coupled. The experiments are done well and made the points summarized above, but by themselves they are not highly novel if not placed in the context of the authors' hypotheses.

Regarding (i), the question that immediately arises is whether it is justified to classify Rab GEFs based on a single residue that participates in nucleotide exchange. I have a hard time doing so. The idea of a single residue being responsible for, and is hence the 'key' for activation, is incompatible with the principles of macromolecular activity, which is almost always the result of a collective network of interacting residues. For example, will Q67 explain GEF-induced displacement of Switch I loop and active site Mg2+? Are all other mechanisms of disrupting the Rab-GDP interaction the same for GEFs? If the authors cared to mutate other residues around the Rab or GEF active site, will they identify other mutations that disrupt GEF-mediated activation? (The answer is almost certainly yes.) Given this, what justifies placing paramount importance on the Q or D/E in Switch II of Rab? Second, such a classification offers no predictive power, as both Q and E residues are conserved in the Rab Switch II loop, but only one of them are used, and sometimes neither is used. Most researchers will have to begin with a null hypothesis with or without knowing that GEFs can be Q, D/E or S family. Thirdly, the available data are not sufficiently comprehensive to support, for example, the presence of the 'S-family' or that the Ypt1is a D/E family member.

Regarding (ii), while I agree that the same glutamine is used by Rab35 and Rab1 for GEF-mediated exchange and GAP-mediated hydrolysis, I have a hard time considering this as 'coupled' activation and inactivation. One can certainly identify many aspects of the mechanisms of these two reactions that are highly different, and hence 'uncoupled'. 'Coupled' in this paper has changed meaning, and refers to a requirement for the Switch II glutamine. To what extent is this concept valuable? Only when this glutamine is mutated, and one thus disrupt both sets of Rab regulation. Pragmatically, I agree with the authors that this gives researchers in the field a warning not to assume that mutating the Switch II glutamine only disrupts GAP activity and to make wrong interpretations, but that seems all there is – I can not come up with more examples in which this 'coupled' concept would be useful. I will be happy to hear more if the authors can explain.

[Minor comments not shown]

The title (Plasticity in nucleotide release mechanism results in coupling of Rab GTPase activation and inactivation) implies a causal relationship that I have a hard time digesting. Plasticity in this paper refers to Rabs being able to use Q, D/E, or S for activation by GEF. “Coupling of activation and inactivation” refers to the observation that for a Rab that uses a Q-GEF, the same Q can also be used in GAP-mediated hydrolysis. Do the authors see a strong connection between the two that I don't?

Reviewer #3:

The starting off point for this work are the structures of complexes between various Rab proteins and their cognate GEFs, in particular with reference to the fate of an essential lysine residue in the P-loop of the GTPases after displacement of GDP by a GEF. It had been suggested earlier that a common feauture of such GTPase:GEF complexes is an interaction of this lysine with a glutamate residue in the switch II region immediately following the catalytic glutamine. Comparing the structures discussed in the manuscript, it is clear that this does not apply generally, in particular here for the Rab proteins. The authors define 4 classes of GEFs, which they call the D, E, Q or S GEFs, where the single letter refers to the type of residue that interacts with the P-loop lysine in the Rab:GEF structure.

These residues can be in the switch II region, elsewhere in the Rab sequence or in the GEF. Of special interest, and this is the origin of the title of the manuscript and of several statements made throughout the manuscript to the effect that there is coupling between GTPase activation and inactivation, are the Q GEFs. The origin of this statement is the fact that in the Q GEFs, the lysine interacts with the conserved catalytic glutamine in the Rab:GEF complexes for DENND1 and DrrA, GEFs for Rab35 and Rab1, respectively. This is an interesting observation, and mutational analysis (referred to as alanine scanning analysis, which it doesn't appear to be) confirms that in these cases, this glutamine is important not only for GTPase activity, as already known, but also for GEF activity. However, and this is my main criticism of the manuscript, this has nothing to do with coupling of activation and inactivation of Rabs, whatever this is supposed to mean. The fact that the same glutamine residue is important for the GEF reaction and the GTPase reaction does not imply coupling in any manner. Thus, after displacement of GDP and generation of the Rab:GEF complex, there will be immediate binding of GTP and dissociation of the GEF, but what happens next, or better, the time scale of what happens next depends on the rate of the intrinsic (basal) GTPase reaction and the availability and activity of a cognate GAP, but this has nothing to do with the question of whether the P-loop lysine interacted with a glutamine, a glutamate, an aspartate or a serine in the nucleotide-free Rab:GEF complex, and where this residue is in which sequence.

The emphasis on the (in my opinion incorrect) idea of coupling of activation and inactivation means that the manuscript cannot be published in this form. The observations are interesting in terms of basic principles, since they show that there are apparently several ways of stabilizing Rab:GEF complexes via interactions with the P-loop lysine (in one case 2 different ways for the same Rab), but they are probably most interesting with respect to the use of mutations of the essential glutamine for cell biological studies. This approach to generating stable GTPase:GTP complexes for cell biological studies is known to be flawed in the case of Rabs because the effect on GAP activation is much less than for other GTPases because TBC domain containing GAPs supply a glutamine to take over the role that the switch II glutamine plays in other classes of GTPases. The work presented here demonstrates that the approach is flawed for a further reason, i.e., because in the case of the Q Rabs activation will be inhibited. This is important information for scientists working in this area, but is only relevant to the use of these mutants, and not to the physiological situation.

The authors need to reconsider what the main message of this paper should be. It cannot be that RabGTPase activation and inactivation are coupled for some classes of GEFs, or better Rab-GEF combinations.

On examining the structures discussed in a little more detail, there are several more points to be made. If the P-loop lysine interacts with a neutral sidechain (glutamine, serine), the question arises as to charge neutralization of the protonated lysine amino group. Looking at the Rab1:DrrA structure, it is clear that there is a strong interaction with D63 from the Rab molecule (reported in the Schoebel et al. paper in Mol.Cell, 2009), in fact probably much stronger than the one with with Q67 discussed in this manuscript, where there is a bond length of 3.6 Angstroms. D63is the highly conserved aspartate in the WDTAGQE sequence, and is in fact the equivalent residue to D74 in Rab5 that was identified in this manuscript (and earlier, of course) as the Rab residue interacting with the P-loop lysine in the complex with Vps9. Looking at the DENND1:Rab35 structure, we see the same constellation (i.e., interactions of the lysine with D63 and Q67). So should DrrA and DENND1 be called D/Q Rabs? And Rab5 a D/E Rab? In the case of Sec2:Sec4, examination of one structure in the pdb does indeed show an interaction of the lysine with Ser161. A quick look for an acidic residue for charge neutralization did not reveal an obvious partner, although D101 of Sec4 is quite near. The GEF molecule is too far away to interact. In another Sec2:Sec4 structure in the PDB, the lysine is far removed from Ser161, but interacts with an inorganic phosphate group bound to the complex. These are all points that need discussion in the manuscript.

The reader is left with the distinct impression that an attempt has been made to make a bigger story out of the results than is justified. This oversell is unworthy of the intelligence and reputation of the authors. Please rewrite with an emphasis on interpretations that are justified by the arguments.

[Minor comments not shown]

[Editors’ note: further clarifications were requested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled “Diversity and plasticity in Rab GTPase nucleotide release mechanism has consequences for Rab activation and inactivation” for further consideration at eLife. Your revised article has been favorably evaluated by a Senior editor and me. All the reviewers agree that the manuscript has been improved significantly but there are some remaining issues that need to be addressed before acceptance, as outlined below:

Reviewer 2 wrote:

“Overall the clarity of the presentation is significantly improved from the previous manuscript. The attempt of the authors to downplay the 'classification' of RabGEFs into different families is also appreciated. The distinction between RabGEFs that stabilize the P-loop by directly coordinating the P-loop lysine or repositioning the switch II loop of Rab to coordinate this lysine may provide a conceptually sounder line of classification, although this still awaits to be seen.

Much more emphasis in this revision appears to be placed on the contribution of the Switch II glutamine to the intrinsic GTP hydrolysis rate of the Rabs. This raises questions about the quality of GTP hydrolysis measurements in Figure 3. I am puzzled by the data. First, the units reported denote mass (pmol), whereas a hydrolysis rate or rate constant should be reported in time units (s-1 or min-1). Secondly, the basal GTP hydrolysis activity is presumably the y-intercept in the main panels, but these values do not match those in the inset that reports 'basal GTP hydrolysis'. In Figure 3c the basal GTP hydrolysis activity is in fact higher than the GAP-activated hydrolysis. Thirdly, I would assume that a proper 'GTP hydrolysis' measurement involves a time course (as all the arguments are about kinetics of hydrolysis), but re-reading the Methods section, I am less sure this time how these hydrolysis rate constants were measured and analyzed. PLEASE CLARIFY. Even if all the measurements are accurate, the author's statement “ in cells basal hydrolysis is a more significant determinant of Rab5 lifetime and turnover than GAP-mediated turnover” is highly speculative. The qualifier for this sentence “Together these findings indicate that...” needs to be significantly softened.”

[Minor comments not shown]

Roger Goody wrote:

“I think there is still more to be done on the role of the conserved aspartate in the Rab1-DrrA and Rab35-DENND1 interaction. As I alluded to in previous comments, mutation of this aspartate is problematic, since it has a negative effect on nucleotide binding, making the interpretation of the results difficult. Thus, the question arises as to whether we should be comparing kcat/Km values for wt and mutant, or perhaps the acceleration of the GDP release rate at saturating GEF concentrations. From Fig. 2b, it looks as if the rate constant for uncatalyzed GDP release from Rab1b with the aspartate removed is ca. 0.001/s, about 100 times faster than for wt-Rab1b. Thus, what is interpreted here as an increase of efficiency of GDP release catalyzed by DrrA on removing the aspartate might turn out to be a reduction if the acceleration ratio is considered. In addition, there is still the question of how charge neutralization of the P-loop lysine occurs without a negatively charged interaction partner, even if, as the authors point out, it is not clear whether the lysine is protonated or not. To my knowledge, without having checked exhaustively, there is always such an interaction in binary GTPase-GEF complexes, either in cis or trans (apart from the relatively poorly defined Sec2:Sec4 structure that has fortunately now been removed from the discussion in the present paper). I wonder whether there might be a change of mechanism with the aspartate mutant? It would be interesting to see which residues the lysine interacts with in the binary complex when the conserved aspartate has been removed. However, I am not suggesting that this structure should be solved before acceptance of this paper.”

* The Reviewing editor would be satisfied with a careful discussion of this point without a need for additional work. *

[Minor comments not shown]

eLife. 2014 Feb 11;3:e01623. doi: 10.7554/eLife.01623.009

Author response

[Editors’ note: the author responses to the first round of peer review follow.]

Reviewer #1:

1) The figure legends need to describe the figures much more clearly and accurately. 1b has no legend; Figure 3 legend is out of order; Figure 4 b,d,f,h are either reverse labeled or need better clarification in terms of what is presented.

We have thoroughly revised the manuscript to more clearly present the data and explain the key findings. All figure legends and the main text have been revised. Figures and panels are now labeled strictly in the order of appearance.

2) Figure 2, panel b, bottom. Please show the Rab1-GDP with the same projection as Rab1-DrrA and/or move it for comparison with Rab1-GTP to indicate residue orientation for comparison. Is there a Q-GEF for Rab1 or is that a microbial adaptation and approach to perfection? This could be tested – does Rab1-Q mutant rescue a Rab1 phenotype.

Structures are now are presented in the new Figure 1. We have spent some time trialing different figure layouts and hope this is clearer than the previous representation. The Rab1-GDP structure was removed as part of these changes in response to the comment of Reviewer 2 that some aspects of the figures were too much like a literature review.

3) Figure 3. Panel d should be moved to Figure 7b. Please compare Rab1-GTP with Rab1-GDP and show key residues for comparison. Bottom text “Divergent in Q/E positons” is misplaced.

This figure panel was removed in the revision.

4) Figure 4. Please include Bet3 E192 mutant for completeness. Can you include an S-GEF example? Please relabel b,d,f,h so that the reader understands what is shown.

The effects of TRAPP and Rabex acidic finger mutations are now tested, and additional mutations in the Rab switch II region are included as requested by the other referees. Together these data provide additional support the major conclusions of the work.

Because the manuscript has been extensively restructured the figure numbering has now changed. This data is now presented more simply in Figure 2.

5) Figure 5 basal hydrolysis needs to be shown as a rate, not a one hour time point. This is important since GTP hydrolysis rates vary significantly between Rabs. Would inclusion of the Rab5Q or Rab5GTP structure be useful to show here to understand why Rab5Q mutant is so poor in intrinsic hydrolysis?

The values shown are the amount of GTP hydrolysis per hour under conditions that allow the effect of the GAP to be tested. This is within the experimentally validated linear range of the reaction and the value is therefore a rate (pmol/h).

6) How many cells were counted in 6c? The images would be easier to understand if cells expressing the particular Rab were indicated – it seems some cells express more of the Rab and the phenotype does not seem uniform. Please present this more clearly.

Over 300 cells for each of two independent experiments were counted for the Shiga toxin transport assays. The wild type rescue and mutant phenotypes were highly reproducible, and non-transfected cells depleted of Rab35 did not transport Shiga toxin into the Golgi apparatus. Cells rescued with the switch II mutant Rab35 did not support Shiga toxin traffic into the Golgi, and it is spread throughout the cytoplasm in what we think are small vesicles. Cells transfected with the Rab are outlined to focus the attention of the reader, and we thank the referee for this suggestion.

Also, the FRAP in d, e may be the result of differences in capacity to interact with GDI and without monitoring relative prenylation and cytosol/membrane fraction cannot really be interpreted. Please remove or add additional data so that the FRAP can be understood molecularly.

This is an important control and we have added a Western blot analysis of Rab cytosol/membrane distribution as suggested to address the issue of whether prenylation is altered by the mutations. No differences were seen, supporting the original interpretation of the FRAP data.

Please include the fact that RUTBC1 was shown to activate a catalytically inhibited Rab33B mutant (Q92A), in support of a dual finger mechanism for RUTBC1 action (Nottingham & Barr et al. 2011).

This reference is now added and we apologize for this omission.

Reviewer #2:

[…] The idea of a single residue being responsible for, and is hence the 'key' for activation, is incompatible with the principles of macromolecular activity, which is almost always the result of a collective network of interacting residues. For example, will Q67 explain GEF-induced displacement of Switch I loop and active site Mg2+? Are all other mechanisms of disrupting the Rab-GDP interaction the same for GEFs? If the authors cared to mutate other residues around the Rab or GEF active site, will they identify other mutations that disrupt GEF-mediated activation? (The answer is almost certainly yes.) Given this, what justifies placing paramount importance on the Q or D/E in Switch II of Rab?

We thank the referee for the comments that the experiments are well done, but don’t completely agree with the view that this is only a question of GEF classification. Both this reviewer and Reviewer 3 acknowledge that questions remain about the mechanisms used by Rab GEFs. We show experimental data confirming that there is plasticity in terms of the P-loop switch II conformations leading to GDP-release. It is of course a conformation induced by GEF binding, and this is perhaps the crux of the matter. A single Rab is potentially permissive for multiple different conformations, and these are in fact used as the analysis of Rab1 with TRAPP and DrrA shows. This is an important piece of information for anyone studying GTPases and their function.

A further point is that we have identified mutations that discriminate nucleotide exchange from nucleotide binding. Previous work has shown that conserved aromatic residues in switch I are important for nucleotide exchange; however this is because they contribute to nucleotide binding as the referee notes. Mutation of these aromatic residues, for example, increases basal nucleotide release. The switch II glutamine mutations characterized here for DENN and DrrA GEFs show unaltered nucleotide binding properties, and reduce GEF-stimulated nucleotide release. We don’t disagree with the view that a complex network of interactions promotes nucleotide release, however we feel this work identifies a key node in this network – the switch II P-loop lysine interaction.

Second, such a classification offers no predictive power, as both Q and E residues are conserved in the Rab Switch II loop, but only one of them are used, and sometimes neither is used. Most researchers will have to begin with a null hypothesis with or without knowing that GEFs can be Q, D/E or S family.

Our analysis shows that DENND1 a member of the large (27) group of DENN GEFs acts via a switch II glutamine (Q) dependent mechanism. Therefore, the switch II Q mutation should probably be treated with caution for this family. Although the intrinsic and GAP-mediated GTP-hydrolysis rates are likely to be decreased by this oft-used mutation, the Q67A (and homologous) substitutions will also affect the activation of Rabs by DENN/DrrA GEFs. Consequently, the activations state of such Rab mutants will be ill defined under physiological conditions, as shown by our DENND1-Rab35 data in vitro and in vivo. Different to Ras, the switch II glutamate (E) is never used in Rabs so far as we can tell. The switch II aspartate (D) is used for Rabex, a member of the large group (∼11 in humans) of Vps9 domain GEFs. This isn’t something applying to only a few Rabs, since the data we provide is relevant for 30-40 human Rab GEFs and their target GTPases. That implies it has general value for future research.

Thirdly, the available data are not sufficiently comprehensive to support, for example, the presence of the 'S-family' or that the Ypt1is a D/E family member.

The original idea was to treat the Rab activation pathway as a journey, with the route referred to by a key waypoint along the journey. In the naming system we adopted this waypoint or intermediate was named according to the switch II interaction with the P-loop. In response to the comments from two referees who felt that this was a bad idea, we have revised the main text and figures to remove this naming convention, eliminate the model figures, and focus on the new data.

We accept the reservations about the electron density in the P-loop of the Sec2-Sec4 structure expressed by the referee; however we feel that this must still fall into a different class in terms of its activation mechanism. To test this we have measured GEF activity of human Rabin8 towards Rab8 and find catalytic efficiency of 4.2x104 M-1 sec-1 higher than the published value of 2.6x104 M-1 sec-1 (Author response image 1; Guo et al. JBC 288:32466-32474).

Author response image 1.

Author response image 1.

Open in a new tab

Here we have used full-length Rab8 and Rabin8 rather than partly truncated proteins, which may explain this difference. Mutation of the switch II glutamine to alanine increased this further to 7.5x104 M-1 sec-1, while mutation of T150 (equivalent to S161 in Sec4) resulted in loss of GEF activation. We have provided them to the referees to indicate that we have investigated this issue further and not simply ignored the comment.

Additional data has been added for all Rabs showing the effects of mutating the most highly conserved residues in switch II (see revised Figure 2).

Regarding (ii), while I agree that the same glutamine is used by Rab35 and Rab1 for GEF-mediated exchange and GAP-mediated hydrolysis, I have a hard time considering this as 'coupled' activation and inactivation. One can certainly identify many aspects of the mechanisms of these two reactions that are highly different, and hence 'uncoupled'. 'Coupled' in this paper has changed meaning, and refers to a requirement for the Switch II glutamine. To what extent is this concept valuable? Only when this glutamine is mutated, and one thus disrupt both sets of Rab regulation. Pragmatically, I agree with the authors that this gives researchers in the field a warning not to assume that mutating the Switch II glutamine only disrupts GAP activity and to make wrong interpretations, but that seems all there is – I can not come up with more examples in which this 'coupled' concept would be useful. I will be happy to hear more if the authors can explain.

The finding that Rab GTPase activation (GDP-GTP exchange) and inactivation (GTP hydrolysis) are linked in the way described is a previously unreported finding. Its value is that it adds to our detailed understanding of Rab regulation, and as the referee notes also highlights precautions that should be taken when mutating the switch II region. We also show that a single Rab can undergo activation via intermediates with different P-loop-switch II conformations. GEFs binding to different sites at the surface of the target GTPase do not necessarily have to induce the same conformational changes in switch I/II to promote nucleotide exchange. As we show, more than one P-loop-switch II conformation is permitted for a single Rab GTPase. This plasticity in the Rab family is surely a finding likely to have general relevance for the Ras superfamily, and should be valuable for future work.

[Minor comments and responses not shown]

The title (Plasticity in nucleotide release mechanism results in coupling of Rab GTPase activation and inactivation) implies a causal relationship that I have a hard time digesting. Plasticity in this paper refers to Rabs being able to use Q, D/E, or S for activation by GEF. “Coupling of activation and inactivation” refers to the observation that for a Rab that uses a Q-GEF, the same Q can also be used in GAP-mediated hydrolysis. Do the authors see a strong connection between the two that I don't?

We have altered the title to address this point. We show that plasticity in Rab GTPase nucleotide release mechanism has consequences for both activation and inactivation pathways, and the cellular function of Rabs.

Reviewer #3:

[…] These residues can be in the switch II region, elsewhere in the Rab sequence or in the GEF. Of special interest, and this is the origin of the title of the manuscript and of several statements made throughout the manuscript to the effect that there is coupling between GTPase activation and inactivation, are the Q GEFs. The origin of this statement is the fact that in the Q GEFs, the lysine interacts with the conserved catalytic glutamine in the Rab:GEF complexes for DENND1 and DrrA, GEFs for Rab35 and Rab1, respectively. This is an interesting observation, and mutational analysis (referred to as alanine scanning analysis, which it doesn't appear to be) confirms that in these cases, this glutamine is important not only for GTPase activity, as already known, but also for GEF activity. However, and this is my main criticism of the manuscript, this has nothing to do with coupling of activation and inactivation of Rabs, whatever this is supposed to mean. The fact that the same glutamine residue is important for the GEF reaction and the GTPase reaction does not imply coupling in any manner. Thus, after displacement of GDP and generation of the Rab:GEF complex, there will be immediate binding of GTP and dissociation of the GEF, but what happens next, or better, the time scale of what happens next depends on the rate of the intrinsic (basal) GTPase reaction and the availability and activity of a cognate GAP, but this has nothing to do with the question of whether the P-loop lysine interacted with a glutamine, a glutamate, an aspartate or a serine in the nucleotide-free Rab:GEF complex, and where this residue is in which sequence.

We have revised the figures and manuscript text to remove the two issues identified by the referee. Rather than refer to D/E/Q/S GEFs we simply describe the results for the different Rab and GEF combinations and discuss these to develop a more coherent picture that we hope will inform future work. The notion of coupling has been removed entirely and we now note that Rab activation and inactivation can share a common determinant with regard to the role of the switch II glutamine. This is likely to apply to the large (27 member in humans) family of DENN GEFs, and therefore has relevance for many trafficking pathways.

The emphasis on the (in my opinion incorrect) idea of coupling of activation and inactivation means that the manuscript cannot be published in this form. The observations are interesting in terms of basic principles, since they show that there are apparently several ways of stabilizing Rab:GEF complexes via interactions with the P-loop lysine (in one case 2 different ways for the same Rab), but they are probably most interesting with respect to the use of mutations of the essential glutamine for cell biological studies. This approach to generating stable GTPase:GTP complexes for cell biological studies is known to be flawed in the case of Rabs because the effect on GAP activation is much less than for other GTPases because TBC domain containing GAPs supply a glutamine to take over the role that the switch II glutamine plays in other classes of GTPases. The work presented here demonstrates that the approach is flawed for a further reason, i.e., because in the case of the Q Rabs activation will be inhibited. This is important information for scientists working in this area, but is only relevant to the use of these mutants, and not to the physiological situation.

Mechanistic details and detailed X-ray structures do count for the physiological situation, and are not only relevant for designing mutations to use in cell biological studies. This study is part of a larger body of work by many groups that aims to explain Rab GTPase regulation and function. We hope the referee will agree that without such details our understanding remains little better than cartoon models.

The authors need to reconsider what the main message of this paper should be. It cannot be that RabGTPase activation and inactivation are coupled for some classes of GEFs, or better Rab-GEF combinations.

We have thought carefully about the criticisms of all the referees, and have focused and simplified the manuscript as a consequence. The main message of the paper is focused on the role of the switch II region in Rab activation by different GEFs, and depending on the nature of the cognate GEF, the plasticity that emerges from this.

On examining the structures discussed in a little more detail, there are several more points to be made. If the P-loop lysine interacts with a neutral sidechain (glutamine, serine), the question arises as to charge neutralization of the protonated lysine amino group. Looking at the Rab1:DrrA structure, it is clear that there is a strong interaction with D63 from the Rab molecule (reported in the Schoebel et al. paper in Mol.Cell, 2009), in fact probably much stronger than the one with with Q67 discussed in this manuscript, where there is a bond length of 3.6 Angstroms. D63is the highly conserved aspartate in the WDTAGQE sequence, and is in fact the equivalent residue to D74 in Rab5 that was identified in this manuscript (and earlier, of course) as the Rab residue interacting with the P-loop lysine in the complex with Vps9. Looking at the DENND1:Rab35 structure, we see the same constellation (i.e., interactions of the lysine with D63 and Q67). So should DrrA and DENND1 be called D/Q Rabs? And Rab5 a D/E Rab? In the case of Sec2:Sec4, examination of one structure in the pdb does indeed show an interaction of the lysine with Ser161. A quick look for an acidic residue for charge neutralization did not reveal an obvious partner, although D101 of Sec4 is quite near. The GEF molecule is too far away to interact. In another Sec2:Sec4 structure in the PDB, the lysine is far removed from Ser161, but interacts with an inorganic phosphate group bound to the complex. These are all points that need discussion in the manuscript.

As the referee suggests, the ionic aspartate-lysine interaction may be stronger than the H-bridge between lysine and glutamine, but nobody knows whether the lysine in the Rab-GEF-complex is protonated or not. The protonation state will be important to estimate any strength of interaction. However, the crystal structures do not provide any data on this issue and it is therefore hard to come to any firm conclusion about the strength of interactions of the participating amino acid side chains. Still, the glutamine is important because the experiments we present clearly show this for DENN and DrrA GEFs. The glutamine is therefore involved either in stabilizing the lysine directly or in organizing its stabilization. The discrete distances in the various complexes don't resolve this issue unambiguously since they only provide a static snapshot. The argument we put forward is that the glutamine is located at the heart of the nucleotide-Rab-interaction in the case of DrrA- and DENND1-complexes and therefore contributes to the exchange reaction. It has been shown in other instances that the lysine requires stabilization during the transition through the nucleotide-free GTPase-GEF complex and therefore amino acid residues contributing to this stabilization must be important. We have also tested the role of the switch II aspartate in GDP-release for all the Rab-GEF pairs tested (Figure 2). Only in the case of Rabex is it required for GEF stimulated GDP-release. For the other Rabs it either results in an increase in basal release, consistent with the role in nucleotide binding. This suggests that the aspartate does not play an essential role in lysine stabilization during the GEF stimulated GDP-release reaction.

The reader is left with the distinct impression that an attempt has been made to make a bigger story out of the results than is justified. This oversell is unworthy of the intelligence and reputation of the authors. Please rewrite with an emphasis on interpretations that are justified by the arguments.

We have carried out a careful analysis of how Rab activation works, and generated data that we were very excited about. The referee has explained why they do not agree with all our ideas and the way they were presented. We have thought about this carefully and done our best to address their comments and revise some of our own conclusions in light of the valuable comments provided.

[Editors’ note: the author responses to the re-review follow.]

Reviewer 2 wrote:

“Overall the clarity of the presentation is significantly improved from the previous manuscript. The attempt of the authors to downplay the 'classification' of RabGEFs into different families is also appreciated. The distinction between RabGEFs that stabilize the P-loop by directly coordinating the P-loop lysine or repositioning the switch II loop of Rab to coordinate this lysine may provide a conceptually sounder line of classification, although this still awaits to be seen.

We agree with the comment of the reviewer, and would like to add that far more structures of Rab-RabGEF pairs are needed before we can make any definitive statements on GEF classification. It is clear that major questions about Rab regulation and function still need to be addressed.

Much more emphasis in this revision appears to be placed on the contribution of the Switch II glutamine to the intrinsic GTP hydrolysis rate of the Rabs. This raises questions about the quality of GTP hydrolysis measurements in Figure 3. I am puzzled by the data. First, the units reported denote mass (pmol), whereas a hydrolysis rate or rate constant should be reported in time units (s-1 or min-1). Secondly, the basal GTP hydrolysis activity is presumably the y-intercept in the main panels, but these values do not match those in the inset that reports 'basal GTP hydrolysis'. In Figure 3c the basal GTP hydrolysis activity is in fact higher than the GAP-activated hydrolysis. Thirdly, I would assume that a proper 'GTP hydrolysis' measurement involves a time course (as all the arguments are about kinetics of hydrolysis), but re-reading the Methods section, I am less sure this time how these hydrolysis rate constants were measured and analyzed. PLEASE CLARIFY. Even if all the measurements are accurate, the author's statement “ in cells basal hydrolysis is a more significant determinant of Rab5 lifetime and turnover than GAP-mediated turnover” is highly speculative. The qualifier for this sentence “Together these findings indicate that...” needs to be significantly softened.”

The referee asks an important question about the data plotted in Figure 3. GAP-stimulated and basal nucleotide hydrolysis for three Rab-RabGAP pairs are shown separately in this figure. This allows a simple comparison of the effects of the Rab switch II mutations on GAP-stimulated and basal hydrolysis. Basal hydrolysis is plotted in the bar graph in pmol/h. GAP-stimulated hydrolysis is plotted as a function of GAP concentration for wild type and switch II glutamine mutant Rabs. These values were measured using the protocol in the methods, and are plotted in the line graph. Hydrolysis is in pmol/h, and this is now correctly indicated in the figure. All GAP-stimulated hydrolysis values are corrected for basal GTP hydrolysis in the absence of GAP, so the y-intercept is not the basal hydrolysis value.

In response to this comment the text has been edited to soften the conclusions, and part of the text was moved to the discussion section. All the data shown was in the original submission, and relatively minor changes were made to the text describing these data. It is therefore unclear to us why the referee feels that more emphasis is placed on the contribution of the Switch II glutamine to the intrinsic GTP hydrolysis rate of the Rabs. We hope the changes we have made address the concern of the referee.

[Minor comments and responses not shown]

Roger Goody wrote:

“I think there is still more to be done on the role of the conserved aspartate in the Rab1-DrrA and Rab35-DENND1 interaction. As I alluded to in previous comments, mutation of this aspartate is problematic, since it has a negative effect on nucleotide binding, making the interpretation of the results difficult. Thus, the question arises as to whether we should be comparing kcat/Km values for wt and mutant, or perhaps the acceleration of the GDP release rate at saturating GEF concentrations. From Fig. 2b, it looks as if the rate constant for uncatalyzed GDP release from Rab1b with the aspartate removed is ca. 0.001/s, about 100 times faster than for wt-Rab1b. Thus, what is interpreted here as an increase of efficiency of GDP release catalyzed by DrrA on removing the aspartate might turn out to be a reduction if the acceleration ratio is considered. In addition, there is still the question of how charge neutralization of the P-loop lysine occurs without a negatively charged interaction partner, even if, as the authors point out, it is not clear whether the lysine is protonated or not. To my knowledge, without having checked exhaustively, there is always such an interaction in binary GTPase-GEF complexes, either in cis or trans (apart from the relatively poorly defined Sec2:Sec4 structure that has fortunately now been removed from the discussion in the present paper). I wonder whether there might be a change of mechanism with the aspartate mutant? It would be interesting to see which residues the lysine interacts with in the binary complex when the conserved aspartate has been removed. However, I am not suggesting that this structure should be solved before acceptance of this paper.”

We briefly explain that the conserved aspartate in switch II has a role in nucleotide binding. This is indirect and the aspartate does not make direct contact with the bound metal ion. For example, in Rab1 the P-loop serine (S22) and threonine (T40) in the (TIGVD motif) make direct contact with the bound metal ion. In addition, the beta- and gamma-phosphates of the bound GTP contact the metal ion. All these interactions fall in one plane around the magnesium ion. Finally, there are two water molecules positioned above and below the metal ion. The aspartate (D63 in Rab1) may influence the environment and possibly contact of one of these water molecules as well as the P-loop serine (S22). The same water molecule is also predicted to interaction with the carbonyl oxygen of the polypeptide back at threonine 65 in the switch II region. Many studies of Ras superfamily GTPases use a serine to asparagine (S22N) mutation in the P-loop since this prevents magnesium binding, and hence stable GDP or GTP binding. Removal of the aspartate as we have done will reduce the affinity for Mg:GDP or GTP, but does not abolish binding in the same way.

We agree with the referee that the aspartate may play a role in charge neutralization of the P-loop lysine in some cases (DENND1-Rab35 and DrrA-Rab1). The discussion has been extended to mention this point.

We disagree with the use of an acceleration ratio as suggested by the referee. The biological activity of a GEF is only fulfilled if it produces a sufficient number of active Rab molecules in a given time. It is therefore the absolute amount of activity (Rab GTP created), not a ratio that is the relevant parameter. We were careful in wording the text to avoid making the statement made by the referee: “what is interpreted here as an increase of efficiency of GDP release catalyzed by DrrA”. Our text states that GDP release was increased and we don’t say this was because the GEF is more efficient. In fact, we suggest that this because removal of the aspartate contributes to nucleotide binding (via the coordination of the metal ion), and therefore limits GDP-release.

[Minor comments and responses not shown]

Articles from eLife are provided here courtesy of eLife Sciences Publications, Ltd

RESOURCES