A crescent-shaped ALIX dimer targets ESCRT-III CHMP4 filaments

Abstract

ALIX recruits ESCRT-III CHMP4 and is involved in membrane remodeling during endosomal receptor sorting, budding of some enveloped viruses, and cytokinesis. We show that ALIX dimerizes via the middle domain (ALIX(-V)) in solution. Structural modeling based on small angle X-ray scattering (SAXS) data reveals an elongated crescent-shaped conformation for dimeric ALIX lacking the proline-rich domain (ALIX(BRO1-V)). Mutations at the dimerization interface prevent dimerization and induce an open elongated monomeric conformation of ALIX(-V) as determined by SAXS modeling. ALIX dimerizes in vivo and dimeric ALIX colocalizes with CHMP4B upon coexpression. We show further that ALIX dimerization affects HIV-1 budding. C-terminally truncated activated CHMP4B retaining the ALIX binding site forms linear, circular, and helical filaments in vitro, which can be bridged by ALIX. Our data suggest that dimeric ALIX represents the active form that interacts with ESCRT-III CHMP4 polymers and functions as a scaffolding protein during membrane remodeling processes.

Figure 1. ALIX forms monomers and dimers in solution

(A) ALIX_Bro1-V and (B) ALIX_V, analyzed by MALLS indicates molecular weights of ~80 and ~160 kDa for ALIX_Bro1-V and ~40 and ~80 kDa for ALIX_V; left y-axis, refractive index; the molecular weight is blotted on the right axis in a logarithmic scale; x-axis shows the elution volume of the peaks.

Figure 2. Small angle X-ray scattering analysis of ALIX_Bro1-V.

(A) Experimental scattering intensity patterns obtained for monomeric (dark grey) and dimeric (light grey) ALIX_Bro1-V are shown as a function of resolution and after averaging and subtraction of solvent scattering. The scattering intensity patterns calculated from the ALIX_Bro1-V monomer and dimer SAXS model with the lowest χ value are shown as black lines. (B) P(r) function of both monomeric (dark grey) and dimeric (light grey) ALIX_Bro1-V (both curves have been adjusted to the same height). The ab initio model envelope of monomeric ALIX_Bro1-V is shown together with the manually docked ALIX_Bro1-V structure (Fisher et al., 2007) as inset. (C) Ab initio modeling of dimeric ALIX_Bro1-V reveals a crescent shape model spanning ~ 270 Å; two orientations of the bead model including the molecular envelope are shown.

Figure 3. Mapping of the conformational flexibility of ALIX_Bro1-V

Hydrogen/deuterium (H/D) exchange labeling coupled to mass spectrometry analysis was used to determine conformational differences between monomeric and dimeric ALIX_Bro1-V. (A) Example of local H/D exchange kinetic for the peptide sequence 637–648. The average mass (M_avg) is plotted against D₂O incubation time. This reveals a difference in H/D exchange within this region where dimeric ALIX_Bro1-V shows less deuterium exchange compared to monomeric ALIX_Bro1-V over a period of 30 min. (B) The H/D exchange results have been plotted onto the structure of ALIX_Bro1-V (Fisher et al., 2007) and the differences in deuteration between monomer and dimer (ΔM_avg) are scored according to the following color code: green regions are more protected from H/D exchange in dimeric ALIX_Bro1-V while regions shown in red < orange < yellow are less accessible to H/D exchange within monomeric ALIX_Bro1-V. Regions in blue show similar accessibility between monomeric and dimeric ALIX_Bro1-V and regions in grey were not covered by peptide mapping of the deuterated forms.

Figure 4. Mutagenesis of the ALIX_V dimer hinge region generates elongated ALIX_V monomers

(A) SEC reveals that ALIX_Vmut1 produces only one peak as compared to wild type ALIX_V and ALIX_Vmut2. (B) MALLS analyses show that ALIX_Vmut1 is monomeric and ALIX_Vmut2 exists in monomer and dimer conformations. Both refractive index and molecular weight analyses (logarithmic scale) are plotted against the elution volume. The molecular weight distribution across each peak is indicated by dots.

Figure 5. SAXS model of ALIX_Vmut1.

(A) X-ray scattering curves of wild type ALIX_V (red) and ALIX_Vmut1 (blue) is shown after averaging and subtraction of solvent scattering. The inset shows the corresponding Guinier plots. The theoretical scattering profiles calculated from the ab initio models with the lowest χ values are shown as black lines. (B) P(r) functions of both wild type and mutant ALIX_Vmut1. (C) Ab initio calculated model of ALIX_Vmut1 reveals an elongated 170Å long rod-like structure; two orientations of the bead models plus envelopes (grey) rotated by 90° are shown. (D) The three helical bundle structure (red ribbon) of one arm of ALIX_V fits into the molecular envelope of ALIXV_mut1 and the second arm (blue) would fit after a rotation of ~ 140° along the ALIX_V hinge region. (E) Model for dimeric ALIX. Dimerzation requires that ALIX_V opens and allows anti-parallel interaction of the ALIX V-domains arms.

Figure 6. ALIX_Bro1-V dimerizes in vivo

ALIX_Bro1-V was expressed using a split YFP expression system. (A) Co-expression of both ALIX_Bro1-V YFP fusion protein halves in HEK293 cells resulted in complementation (right panel). ALIX_Bro1-V expression thus induces ALIX_Bro1-V dimerization in vivo. (B) Expression of CHMP4B_ΔC-ALIX revealed inclusions partly localized along the plasma membrane (left panel). (C) Co-expression of ALIX_Bro1-V YFP fusion constructs and CHMP4B_ΔC-ALIX; red fluorescence CHMP4B_ΔC-ALIX (left panel), green fluorescence dimeric ALIX (middle panel) and the lower panel show the co-localization: CHMP4B_ΔC-ALIX recruits ALIX_Bro1-V dimers in vivo into large inclusions. Control staining with DAPI is shown in all panels.

Figure 7. ALIX dimerization is important for HIV-1 budding

HIV-1 particle release detected upon expression of RFP-ALIX_V-PRD wild type and the dimerization mutants. (A) Expression of RFP-ALIX_Vmut1-PRD reverses the dominant negative effect observed for expression of wild type RFP-ALIX_V-PRD. Lane 1, vector control; lane 2, wild type RFP-ALIX_V-PRD; lane 3, RFP-ALIX_Vmut1-PRD. Detection of extracellular Gag levels (virions, left panel) and intracellular Gag levels (right panels) are shown. (B) Expression of RFP-ALIX_Vmut2-PRD does not influence the dominant negative effect observed for expression of wild type RFP-ALIX_V-PRD. Lane 1, vector control; lane 2, wild type RFP-ALIX_V-PRD; lane 3, RFP-ALIX_Vmut2-PRD. (C) Expression levels of RFP-ALIX_V-PRD are shown for wild type (lane 2) and RFP-ALIX_Vmut1-PRD (lane 3) expression (lane 1, RFP vector expression control).

Figure 8. ALIX_Bro1-V interacts with CHMP4 polymers in vitro

Sucrose gradient centrifugation analysis of monomeric (A) and dimeric (B) ALIX_Bro1-V, (C) MBP-CHMP4B_ΔC, (D) MBP-CHMP4B_ΔC-ALIX, (E) CHMP4B_ΔC-ALIX, (F) MBP-CHMP4B_ΔC-ALIX and monomeric ALIX_Bro1-V, (G) MBP-CHMP4B_ΔC-ALIX and dimeric ALIX_Bro1-V, (H) Western blot detection of both MBP-CHMP4B_ΔC-ALIX and ALIX_Bro1-V present in the bottom fraction of gradients F and G; (I) CHMP4B_ΔC-ALIX and monomeric ALIX_Bro1-V, (J) CHMP4B_ΔC-ALIX and dimeric ALIX_Bro1-V. (Protein bands migrating below the 50kDa marker protein correspond to MBP* (*; C, D, E, F, G, I, J)).

Figure 9. Electron microscopy analyses of CHMP4B_ΔC-ALIX and ALIX_Bro1-V complexes

Negative staining EM images of (A) MBP-CHMP4B_ΔC, (B) MBP-CHMP4B_ΔC-ALIX, (C) CHMP4B_ΔC-ALIX, (D) CHMP4B_ΔC-ALIX - ALIX_Bro1-V (monomer) complexes, (E) CHMP4B_ΔC-ALIX- ALIX_Bro1-V (dimer) complexes. (F) Cryo-electron microscopy image of CHMP4B_ΔC-ALIX. (G) Cryo EM images of selected ring structures of CHMP4B_ΔC-ALIX; (H) Ring-like structures revealing the potential repeating unit, marked by arrows (left panel, negative staining, right panel cryo image). (I) 5 panels of cryo EM images showing CHMP4B_ΔC-ALIX - ALIX_Bro1-V (dimer) complexes. The length of one rung and the distance between rungs are indicated schematically in panel 2. The scale bars are 50 nm (A), 100 nm (B-F, I) and 30 nm (G, H).