MAPping tubulin mutations

Abstract

Microtubules are filamentous structures that play a critical role in a diverse array of cellular functions including, mitosis, nuclear translocation, trafficking of organelles and cell shape. They are composed of α/β-tubulin heterodimers which are encoded by a large multigene family that has been implicated in an umbrella of disease states collectively known as the tubulinopathies. De novo mutations in different tubulin genes are known to cause lissencephaly, microcephaly, polymicrogyria, motor neuron disease, and female infertility. The diverse clinical features associated with these maladies have been attributed to the expression pattern of individual tubulin genes, as well as their distinct Functional repertoire. Recent studies, however, have highlighted the impact of tubulin mutations on microtubule-associated proteins (MAPs). MAPs can be classified according to their effect on microtubules and include polymer stabilizers (e.g., tau, MAP2, doublecortin), destabilizers (e.g., spastin, katanin), plus-end binding proteins (e.g., EB1-3, XMAP215, CLASPs) and motor proteins (e.g., dyneins, kinesins). In this review we analyse mutation-specific disease mechanisms that influence MAP binding and their phenotypic consequences, and discuss methods by which we can exploit genetic variation to identify novel MAPs.

Keywords: disease; dynein; kinesin; microtubule-associated protein; microtubules; tubulinopathies.

FIGURE 1

Mapping of pathogenic tubulin mutations across tubulin isotypes. Tubulin proteins can be divided into three regions: the N-terminal (green), Intermediate (pink), and C-terminal (blue) domains. The latter constitutes the major MAP-binding region of tubulin. Variants shown to affect MAP interaction are highlighted in red.

FIGURE 2

In silico homology model of an α/β-tubulin heterodimer (A) facing the exterior surface of a microtubule polymer and (B) a rotated side view of an individual microtubule protofilament (PDB 2XRP) (Fourniol et al., 2010). α-tubulin is represented by a pink ribbon, β-tubulin in silver. Unstructured, tubulin carboxy-terminal tails are not shown. Mutated residues shown to effect MAP binding are mapped onto α- and β-tubulin subunits (red).

FIGURE 3

(A) In silico homology model of a microtubule protofilament with bound kinesin. α-tubulins are represented by a pink ribbon, β-tubulins in silver and kinesin in gold (PDB 6ZPI) (Atherton et al., 2020). (B) A detailed view of the predicted kinesin-interacting domain of β-tubulin. Mutated residues shown to effect kinesin binding and/or processivity shown in red.

FIGURE 4

(A) In silico homology model of a microtubule protofilament with bound dynein. α-tubulins are represented by a pink ribbon, β-tubulins in silver and dynein microtubule-binding domain in gold (PDB 3J1T) (Redwine et al., 2012). (B) A detailed view of the dynein-interacting domain. The R402 residue of α-tubulin (red) is a hot-spot for pathogenic variants in TUBA1A, which have been shown to affect dynein interaction and processivity (Aiken et al., 2019; Leca et al., 2020).

FIGURE 5

(A) Schematic alignment and positional conservation of TOG domains in XMAP215 family microtubule polymerases and CLASPs [adapted from (Al-Bassam and Chang, 2011; Byrnes and Slep, 2017)]. (B) Schematic depiction of dynamic behaviour of (i) pure microtubules (grey lines), (ii) accelerated microtubule polymerisation in the presence of XMAP215 family polymerases (gold), and (iii) microtubule rescue mediated by fission yeast CLASP, Cls1p (blue), adapted from (Al-Bassam and Chang, 2011). (C) Homology model depiction of a unpolymerised tubulin heterodimer complexed with TOG1 domain of S. cerevisiae microtubule polymerase Stu2 (gold) (PDB 4FFB) and docked TOGL2 domain of Stu1 (S. cerevisiae CLASP; blue) (PDB 6COK) (Ayaz et al., 2012; Majumdar et al., 2018). α-tubulins are represented by a pink ribbon, β-tubulins in silver. Both TOGs bind preferentially to curved tubulin heterodimers, hence α- and β-tubulin subunits are tilted 13 to form this complex (Ayaz et al., 2012). (D) A detailed view of the predicted TOG1 binding interface of tubulin heterodimers. The α-tubulin residue valine 409 (410 in yeast) (red) is directly involved in TOG1 binding, with substitutions affecting its binding and localisation to microtubule plus-ends (Hoff et al., 2022). (E) Detailed view of the predicted TOGL2 binding interface of tubulin heterodimers. Mutating α-tubulin proline 263 and arginine 402 (P264 and R402 in yeast) have been shown to reduce binding affinity with human CLASPs 1 & 2 (Yu et al., 2016).

FIGURE 6

(A) In silico homology model of a microtubule protofilament with bound Calponin Homology domain of S. pombe EB1 homologue, Mal3 (labelled EB1). α-tubulins are represented by a pink ribbon, β-tubulins in silver and EB1 in gold (PDB 4ABO) (Maurer et al., 2012). (B) A detailed view of the EB1-tubulin interface. The F265 residue (red), associated with neurodevelopmental disease and shown to affect EB1 binding in yeast (Jaglin et al., 2009; Denarier et al., 2019), is not predicted to interact directly with this plus-end MAP, suggesting it may act via allosteric reconfiguration. (C). TUBB5 Q15 and Y222 (red) are located within the β-tubulin GTP (orange)-binding site and variants affecting these residues might to affect nucleotide interaction and/or hydrolysis which, in turn, could influence EB2 binding.