Chemistry And Function Of Microtubules

Wolff, Jan

Abstract

Membrane Tubulin Microtubules make many contacts with plasma membranes and these same membranes contain firmly embedded tubulin (not due to preparative contamination) whose properties differ in some respect from cytoplasmic tubulin. The interaction between microtubules and the membrane has been studied by two approaches: 1. attachment via a linker protein; 2. attachment to lipid modified membrane tubulin. Interruption of prenylation reactions with lovastatin has shown that microtubules retract from the plasma membrane Since tubulin cannot be prenylated, we searched for a linker protein that might mediate microtubule attachment. From 3H-mevalonate labeling, protein size and antibody reaction, we identified 2?, 3?-cyclic nucleotide phospho-diesterase (CNP) as the linker which attaches microtubules to membranes via both prenylation and palmitoylation. Tubulin and CNP form a complex that is stable through a variety of manipulations. CNP promotes microtubule assembly and behaves like a conventional microtubule-associated protein (MAP) whose activity is abolished by phosphorylation. The tubulin-binding domain resides in the C-terminal tridecapeptide, which itself has activity whereas the truncated protein does not. The two proteins co-localize at and near the plasma membrane. Transfection of COS cells with truncated CNP leads to microtubule retraction as seen after lovastatin Thus CNP is an effective linker. In addition, tubulin is directly embedded in the plasma membrane due to palmitoylation of a number of the cysteine residues of tubulin (2, 3). Identification of the cysteines that are substrates for palmitoylation proved difficult because of the hydrophobicity and lability of the thioester-linked tryptic peptides. We resolved this dilemma by several different methods: 1. Identification of the more reactive SH groups toward alkylation; these are alphaC347, alpha C315-316, alphaC376, betaC241(239) and betaC356(354). The basis for SH reactivity is stabilization of the thiolate anion by a positive electrostatic environment. Positive edges of aromatic rings contribute significantly. A negative environment inhibits reactivity even of surface-exposed cysteines. 2. In another approach reactive cysteines are modified, then the remaining cysteines are treated with NEM followed by removal of the labile groups, followed by labeling of the newly exposed SH groups. This approach has been highly successful for disulfide bond formation and looks promising for thioester bond formation(palmitoylation). 3. Direct mass spectrometry of palmitoylated tubulin peptides has just become possible and we are using that approach. 4. Finally, although yeast tubulin shows substantial differences from mammalian tubulin (e.g. lack of response to colchicines, vinblastine etc, and a very low concentration of cytoplasmic microtubules), we have initiated a search for tubulin in membranes from S. cerevisiae with the hope of studying the effect of cysteine mutations on membrane tubulin localization.