Structural basis of katanin p60:p80 complex formation

Abstract

Interactions between microtubule (MT) interacting and trafficking (MIT) domains and their binding proteins are important for the accurate progression of many cellular processes that require the AAA+ ATPase machinery. Therefore, knowledge on the structural basis of MIT domain interactions is crucial for understanding the molecular mechanisms underlying AAA+ ATPase function. Katanin is a MT-severing AAA+ ATPase that consists of p60 and p80 subunits. Although, the hexameric p60 subunit is active alone, its association with the p80 subunit greatly enhances both the MT-binding and -severing activities of katanin. However, the molecular mechanism of how the p80 subunit contributes to katanin function is currently unknown. Here, we structurally and functionally characterized the interaction between the two katanin subunits that is mediated by the p60-MIT domain and the p80 C-terminal domain (p80-CTD). We show that p60-MIT and p80-CTD form a tight heterodimeric complex, whose high-resolution structure we determined by X-ray crystallography. Based on the crystal structure, we identified two conserved charged residues that are important for p60-MIT:p80-CTD complex formation and katanin function. Moreover, p60-MIT was compared with other MIT domain structures and similarities are discussed.

Figure 1

Biophysical and structural characterization of the p60-MIT:p80-CTD complex by AUC and X-ray crystallography. (A) Cartoon representation of the crystal structure of the p60-MIT L40P:p80-CTD complex. Blue, p60-MIT; red, p80-CTD. Microcephaly-associated mutations S538L, L543R and G581D, are highlighted as grey sticks. The intradomain hydrogen bond formed between S538 and R508 is indicated by a black dotted line. (B) c(S) distribution analysis of the p60-MIT:p80-CTD complex.

Figure 2

Zoom-in stereo view of a ribbon representation of the interaction interface between p60-MIT and p80-CTD with important residues highlighted in stick representation.

Figure 3

Importance of the C terminus of the p60-MIT domain for the p60-MIT:p80-CTD complex formation. (A) Superimposition of p60-MIT of the p60-MIT L40P:p80-CTD complex structure (in blue) with the p60-MIT solution NMR structure (PDB ID 2RPA, in grey), and the p60-MIT domain from the katanin:ASPM complex (PDB ID 5LB7, in red). The L40P mutation of p60-MIT that resulted in crystals suitable for structure determination, is highlighted in black. (B) c(S) distribution showing that the minimal version of p60-MIT (aa 1–72 of mouse p60) used to determine its NMR structure does not bind to p80-CTD. For technical reasons, these measurements were carried out with both proteins fused to thioredoxin. All proteins were used at a concentration of 20 µM. (C) Cartoon representation showing the interactions within the C-terminal part of the third helix of p60 that is crucial for p60-MIT:p80-CTD complex formation. Blue, p60-MIT; red, p80-CTD; black, hydrogen bonds. Interacting residues are highlighted as sticks.

Figure 4

Significance of p60-MIT:p80-CTD interface residues for katanin heterodimerization and function. (A) Streptavidin pull-down assay with lysates of HEK293T cells co-transfected with indicated GFP-p80 and GFP-p60 mutants and biotin ligase BirA. (B) Maximum projection of 10 s time lapse movies of indicated GFP-p60/p80 mutants co-expressed with EB3-TagRFP (a marker of MT plus ends) in HeLa cells. Scale bar, 2 μm. (C) Quantification of p60/p80 intensity on MT plus ends. The intensity is normalized to the average intensity of the wild-type protein. Data represent mean ± SD. N = 10 cells per condition.

Figure 5

Comparison of katanin p60-MIT interactions with those of other AAA+ ATPses. (A–E) The structures of different MIT domains are highly similar to each other and they all form an asymmetric three-helix bundle. Superimposition of available complex structures demonstrates an overlap of the binding motifs, although the molecular contacts between these complexes are in most cases not conserved. (A) Spastin-CHMP1B structure (PDB ID 3EAB), (B) ATG1-ATG13 structure (PDB ID 4P1N), (C) Vps4-Vps2 structure (PDB ID 2V6X), (D) Vps4-CHMP6 structure (PDB ID 2K3W), (E) VPS4-Vfa1 structure (PDB ID 5FVK).