Microtubules as Sensors Molecular Mechanisms

Although a sensory function of microtubules in the sensing of abiotic stimuli is supported by a large number of observations from different organisms, the molecular base of this sensory function has remained enigmatic so far. Principally, there are two possible routes and at the present (Fig. 1), limited, state of knowledge it is not possible to rule out any of those. And this may not be necessary, because these routes are not mutually exclusive:

1. Microtubules as Susceptors for Mechanosensitive Ion Channels.

In the first model, the actual perception of the mechanic stimulus occurs through mechanosensitive ion channels (Fig. 1A). The primary input are minute deformations of the membrane with energies that are, in

embrane

ion channel ion channel

Fig. 1 Models for the role of microtubules in mechanosensing. A Microtubules acting as mechanosusceptors in sensu (Bjorkman 1988). Membrane deformations are collected and focussed by a microtubule lever system towards a mechanosensitive ion channel such that the input energy exceeds thermal noise. B Microtubules acting as mechanoreceptors in sensu strictu. Microtubules constrict the opening of ion channels and disassemble upon mechanic load. Note that, in this model, the ion channel acts as a transducer, not as a receptor [in contrast to the model depicted in (A)]

Fig. 1 Models for the role of microtubules in mechanosensing. A Microtubules acting as mechanosusceptors in sensu (Bjorkman 1988). Membrane deformations are collected and focussed by a microtubule lever system towards a mechanosensitive ion channel such that the input energy exceeds thermal noise. B Microtubules acting as mechanoreceptors in sensu strictu. Microtubules constrict the opening of ion channels and disassemble upon mechanic load. Note that, in this model, the ion channel acts as a transducer, not as a receptor [in contrast to the model depicted in (A)]

most cases, below the fluctuations due to thermal noise. In order to obtain a sensible signal, these primary deformations have to be focussed by a lever system. The role of microtubules in this model would be that of a (mechanic) susceptor in sensu (Bjorkman 1988). Elimination of microtubules has been repeatedly found to activate calcium channels. Antimicrotubular compounds such as oryzalin, ethyl-N-

phenylcarbamate, or colchicine induce a six- to tenfold increase in the activity of calcium channels (Ding and Pickard 1993; Thion et al. 1996, 1998). Moreover, cold-induced calcium fluxes are amplified conspicuously by these drugs in tobacco cells (Mazars et al. 1997). However, the characterization of a channel activity as mechanosensitive is usually based on patch-clamp experiments and does not prove any physiological function in mechanosensing.

These pharmacological findings from plant cells are supported by the results from genetic screens for touch-sensitive ion channels in Caenorhab-ditis elegans. Using a system, where the phobic response to a specific touch stimulus was screened, so-called mechanosensation defective (mec) mutants could be recovered (Chalfie and Au 1989; for review see Chalfie 1993). Some of the mutated genes encoded a novel class of transmembrane proteins, the so-called degerins, that might represent components of a touch-sensitive ion channel. However, two of these mutants, mec7 and mec12 were affected in a unique set of microtubules consisting of 15 protofilaments that were confined to the axons of the touch-sensitive neurons responsible for the phobic response (Chalfie and Thomson 1982). MEC12 and MEC7 were later shown to encode specific isotypes of a-tubulin (Savage et al. 1989) and ^-tubulin (Fukushige et al. 1999), respectively. In both mutants, the loss of mechanosensation was correlated by specific changes in the organization of these 15-protofilament microtubules. This led to a model, where the microtubules act through specific linker proteins as a kind of lever system that amplifies minute deformations of the perceptive membrane into a strong aperture of the putative channels (Chalfie 1993). A similar set-up, where specialized mi-crotubules are able, via an intermediate protein to induce a functional spatial arrangement of receptors or ion channels has been proposed for the clustering of glycine receptors in rat spinal cord synapses, where the microtubule-associated protein gephyrin plays the role of the intermediate linker (Kirsch et al. 1993).

2. Microtubules as Primary Deformation Sensors.

In the model sketched above, microtubules act as signal amplifiers for mechanosensitive ion channels. However, microtubules might be the sensors themselves (Fig. 1B). Even in vitro the assembly of microtubules from soluble tubulin heterodimers could be shown to depend on vectorial forces. The preferential direction of the assembled microtubules could be modulated by centrifugal force (Tabony and Job 1992) although one should note that the forces acting here are orders of magnitude above the minute inputs that trigger the perception of abiotic stimuli in a physiological context. The same argument is true for experiments, where the microtubules could be reoriented by bending of maize coleoptiles by application of defined weights (Zandomeni and Schopfer 1994). The forces acting in those experiments approached the limits of membrane integrity and thus are far beyond the physiologically relevant range. This emphasizes the necessity of efficient signal amplification in deformation sensing. As will be explained in more detail later, due to their nonlinear dynamics microtubules themselves should be able to amplify small mechanic stimuli into clear net outputs that can be processed by downstream signalling cascades. It should also be kept in mind that microtubules can generate force not only through microtubule-motors such as kinesins or dyneins. In addition, microtubule disassembly could be recently shown to generate a force that is quite considerable and even exceeds the forces produced by motor proteins (Grishchuk et al. 2005).

Summarizing, both models for the sensory role of microtubules are compatible with our (admittedly still limited) knowledge on the molecular base of abiotic sensing. Both models rely on positive feedback circuits that are able to amplify the minute inputs (small deformations of the perceptive membranes in the first model or changes in the dynamic equilibrium between assembly and disassembly of microtubules themselves in the second model) into clear and nearly qualitative outputs that can then be processed by downstream signalling cascades. The distinction between the two models described above was introduced for the sake of conceptual clarity, it might be not as pronounced in the biological context of a cell, where both mechanisms could act in a complementary fashion. It will be a challenge for the next years not only to identify the molecular elements acting in the perception and modulation of abiotic stimuli, but to understand their interaction and systemic properties. This will require integration of molecular data with cell biological and physiological analysis and even mathematical modelling.

Was this article helpful?

0 0

Post a comment