The Connecdenn DENN domain: a GEF for Rab35 mediating cargo-specific exit from early endosomes

Abstract

The DENN domain is an evolutionarily ancient protein module. Mutations in the DENN domain cause developmental defects in plants and human diseases, yet the function of this common module is unknown. We now demonstrate that the connecdenn/DENND1A DENN domain functions as a guanine nucleotide exchange factor (GEF) for Rab35 to regulate endosomal membrane trafficking. Loss of Rab35 activity causes an enlargement of early endosomes and inhibits MHC class I recycling. Moreover, it prevents early endosomal recruitment of EHD1, a common component of tubules involved in endosomal cargo recycling. Our data reveal an enzymatic activity for a DENN domain and demonstrate that distinct Rab GTPases can recruit a common protein machinery to various sites within the endosomal network to establish cargo-selective recycling pathways.

Figure 1. The connecdenn DENN domain binds lipids and is stably associated with membranes

A. Schematic representation of connecdenn (CD) indicating peptide motifs binding AP-2 (FXDXF, DPF, WXXF) and SH3 domain proteins (proline-rich) (Allaire et al., 2006). Borders of constructs used in the study are indicated B. Equal protein aliquots of homogenate (H) and CCVs isolated from HEK 293-T cells transfected with the indicated Flag-tagged connecdenn (CD) constructs were blotted with antibodies against clathrin heavy chain (CHC) and Flag. C. Equal amounts of CCVs isolated from cells expressing the indicated constructs were incubated with control buffer A or a Tris buffer to strip clathrin coats and the coat (supernatant, S) and vesicle (pellet, P) fractions generated by centrifugation were analyzed by Western blot, using CHC as control. D. Equal protein aliquots of homogenate (H) and CCVs isolated cells expressing Flag-tagged C-terminal region of NECAP 1 (Flag-N1-CT) or a chimeric protein, where the connecdenn DENN domain had been fused to the NECAP 1 C-terminal region (Flag-DENN/N1-CT) were probed by Western blot with Flag and CHC antibodies along with coat (S) and vesicle (P) fractions prepared as described in C. E. Purified GST or GST-DENN domain were incubated with liposome suspensions (1 mg/ml) containing 90 mol% of PtdChol (PC) and 10 mol% of the indicated phospholipids. The bars represent the mean percentage (±STDEV) of starting material retained on liposomes, N=5. Paired T-tests revealed significant differences for DENN domain binding to PtdIns(3)P compared to the other lipids as indicated (* p<0.05, ** p<0.01). See also Fig. S1.

Figure 2. Rab35 is associated with CCVs

A. Equal protein aliquots of various fractions obtained during the purification of CCVs from adult rat brain were analyzed by Western blot for the indicated proteins. B. Purified CCVs were separated on a continuous sucrose gradient and aliquots of the gradient fractions were analyzed by Western blot with the indicated antibodies. C. COS-7 cells were transfected with Flag-Rab35 and the localization of Rab35 was compared to endogenous clathrin light chains (CLCs). The higher magnification shows the area highlighted by the box. Bar: 20 μm (low mag.), 5 μm (high mag.).

Figure 3. The connecdenn DENN domain binds Rab35 and functions as a GEF

A. Purified GST or GST-Rab35 were used to affinity-select Flag-tagged full-length connecdenn (CD), the C-terminal region (CD-CT) or the isolated DENN domain. Binding was performed in HEPES buffer alone or in the presence of 10 mM EDTA (HEPES+EDTA), which renders GTPases nucleotide-free, and interactions were detected by Western blot. B. Purified GST and GST-DENN domain were used to affinity-select the indicated GFP-tagged Rab35 constructs. Binding was revealed by GFP Western blot. For A and B, starting material (SM) equals 10% of the material used in each condition. C. COS-7 cells were transfected with Flag-tagged connecdenn (Flag-CD) and myc-tagged Rab35 S22N or Q67L and processed for immunofluorescence using antibodies directed against the tags. The regions highlighted by boxes a-d are shown in higher magnification. Bar: 20 μm (low mag.), 5 μm (insets a-d). D. HEK 293-T cells were transfected with GFP-tagged Rab35 wt, S22N, and Q67L. Cell lysates were incubated with protein A-agarose alone (mock IP) or protein A-agarose with anti-GFP antibody (GFP IP). Specifically bound proteins were detected by blot with anti-GFP antibody or antibody recognizing endogenous connecdenn. The positions of the GFP-tagged Rab35 variants is indicated by the arrowheads. E. Purified GST and GST-DENN domain were incubated with lysates from HEK 293-T cells transfected with the indicated GFP-tagged Rabs and prepared under nucleotide-free conditions (HEPES+EDTA). Binding was revealed by Western blot using GFP antibodies. For D and E, starting material (SM) equals 10% of the material added to the beads. F/G. To measure GEF activity, 18.7 pmoles of GDP-loaded Rab3, Rab5, or Rab35 were incubated with 1.5 pmoles of connecdenn DENN domain, Rabex5 (aa1-399), or DH/PH domain of intersectin 1-l, a specific GEF for Cdc42, in the presence of 5 μM cold GTPγS and 0.2 mCi/mol ³⁵S-GTPγS. At the time points indicated, an aliquot of the reaction was analyzed for nucleotide exchange as described in the Experimental Procedures. pmoles of GTP exchanged are plotted over time (in seconds) and points represent mean (±STDEV), N=3. The curve was fit by nonlinear regression one phase association. The dotted line in F represents the total amount in pmole of Rab35 present in the reaction.

Figure 4. Connecdenn or Rab35 KD causes perinuclear clustering and enlargement of early endosomes

A. Equal aliquots of lysates from COS-7 cells transduced with control shRNAmiR or shRNAmiRs for KD of connecdenn (CD nt248 and CD nt275) or Rab35 (Rab35 nt63 and Rab35 nt419) were analyzed by Western Blot for the expression levels of the indicated proteins. For the bar graph, Western blots from four independent transductions were quantified by densitometry using ImageJ. Expression levels are represented in percent (±STDEV), setting the control shRNAmiR to 100%. B. Control and KD COS-7 cells were processed for immunofluorescence to reveal the localization of endogenous markers of early endosomes (EEA1, Rab5) and the recycling endosome (Rab11). Bar: 10 μm. See also Fig. S2.

Figure 5. Connecdenn and Rab35 do not control transferrin trafficking

A. Control and KD COS-7 cells were surface-labeled on ice with AlexaFluor647-transferrin, shifted to 37°C for the indicated times, and intracellular transferrin was measured by flow cytometry. The graph represent the mean of percentage of the initial surface label (N=4) and statistical analysis by Repeated Measure Two-Way ANOVA followed by Bonferroni posttests revealed no significant differences between control and KD cells. B/C. The data for the 1 min (B) and 2 min time points (C) from panel (A) were replotted as bar graphs to highlight the lack of any endocytic defect (note the overlap in error bars for the different conditions). D. Control and KD COS-7 cells were continuously labeled with AlexaFluor647-transferrin at 37°C for one hour, chased at 37°C for the indicated times, and intracellular transferrin was measured by flow cytometry. The graph represent the mean of percentage of the initial surface label (N=4) and statistical analysis by Repeated Measure Two-Way ANOVA followed by Bonferroni posttests revealed no significant differences between control and KD cells. See also Fig. S3.

Figure 6. Connecdenn and Rab35 KD affects MHCI trafficking and EHD1 recruitment

A. Control and KD COS-7 cells were incubated for 20 min with antibodies against MHCI and processed by immunofluorescence to reveal the localization of the internalized antibodies and endogenous EEA1. Bar: 20 μm. B. Control and KD COS-7 cells were incubated for 20 min with antibodies against MHCI, chased for 20 min, and intracellular MHCI antibodies were detected by flow cytometry. The graph represent the mean fluorescence after the 20 min chase (N=4), plotted as relative ratio (±SEM) with the control shRNAmiR set to 1. Statistical analysis by Repeated Measure One-Way ANOVA followed by Dunnett’s posttests revealed significant differences between control and KD cells (* p<0.05, ** p<0.01, *** p<0.001). C. COS-7 cells transduced with lentivirus for the expression of control shRNAmiR and transfected with myc-EHD1 were incubated with antibodies against MHCI for twenty minutes and processed by immunofluorescence. Bar: 20 μm. D. Control and KD COS-7 were transfected with myc-EHD1 and processed by immunofluorescence to reveal the localization of EHD1 and endogenous EEA1. Bar: 20 μm. E. COS-7 cells transfected with expression constructs for GFP-tagged Rab35 and myc-tagged EHD1 were processed by immunofluorescence to reveal the localization of Rab35, EHD1, and endogenous EEA1. The arrowheads indicate areas of co-localization of all three proteins and the area highlighted by the box is shown in higher magnification. Bar: 20 μm (low mag.), 10 μm (high mag.). See also Fig. S4/S5.

Figure 7. Loss of Rab35 activity causes the KD phenotypes

A. COS-7 cells were transfected with GFP-Rab5 wt or GFP-Rab5 Q67L alone, or co-transfected with GFP-Rab5 wt and myc-Rab35 S22N and processed by immunofluorescence to compare the effect of inactive Rab35 on the localization of Rab5. The arrowhead indicates the transfected cell when only the Rab5 signal is shown. Bar: 20 μm. B. Connecdenn KD cells (CD nt248) were transfected with GFP alone or GFP-tagged Rab35 Q67L and processed by immunofluorescence for endogenous EEA1 to reveal the localization and morphology of the early endosomal compartment. Arrowheads indicate the transfected cells when only the EEA1 signal is shown. Bar: 40 μm. C. Quantification of (B). The bar graph represents the mean percentage (±SEM, N=4) of transfected cells showing the EEA1 KD phenotype in connecdenn KD cells expressing GFP alone or GFP-Rab35 Q67L. Statistical analysis by two-tailed Mann Whitney test revealed a significant difference between GFP- and GFP-Rab35 Q67L-expressing cells (* p<0.05). D. COS-7 cells transfected with GFP alone or with GFP-tagged Rab5 Q79L were incubated for 20 min with antibodies against MHCI, chased for 20 min, and intracellular MHCI antibodies were detected by flow cytometry. The graph represent the mean fluorescence after the 20 min chase (N=4), plotted as relative ratio (±SEM) with GFP set to 1. Statistical analysis by two-tailed T-test revealed a significant difference between GFP and GFP-Rab5 Q79L-expressing cells (* p<0.05). E. COS-7 cells transfected with GFP or GFP-Rab5 alone or co-transfected with GFP-Rab5 and myc-tagged Rabex5 were processed as described for (D). The graph represent the mean fluorescence after the 20 min chase (N=4), plotted as relative ratio (±SEM) with GFP set to 1. Statistical analysis by Repeated Measure One-Way ANOVA followed by Dunnett’s posttests revealed significant differences between GFP and Rab5 or Rab5+Rabex5-expressing cells (* p<0.05).