Tubulins were purified from the brain tissues of three Antarctic fishes, Notothenia gibberifrons, Notothenia coriiceps neglecta, and Chaenocephalus aceratus, by ion-exchange chromatography and one cycle of temperature-dependent microtubule assembly and disassembly in vitro, and the functional properties of the protein were examined. The preparations contained the alpha- and beta-tubulins and were free of microtubule-associated proteins. At temperatures between 0 and 24 degrees C, the purified tubulins polymerized readily and reversibly to yield both microtubules and microtubule polymorphs (e.g., "hooked" microtubules and protofilament sheets). Critical concentrations for polymerization of the tubulins ranged from 0.87 mg/mL at 0 degrees C to 0.02 mg/mL at 18 degrees C. The van't Hoff plot of the apparent equilibrium constant for microtubule elongation at temperatures between 0 and 18 degrees C was linear and gave a standard enthalpy change (delta H degree) of +26.9 kcal/mol and a standard entropy change (delta S degree) of +123 eu. At 10 degrees C, tubulin from N. gibberifrons polymerized efficiently at high ionic strength; the critical concentration increased monotonically from 0.041 to 0.34 mg/mL as the concentration of NaCl added to the assembly buffer was increased from 0 to 0.4 M. Together, the results indicate that the polymerization of tubulins from the Antarctic fishes is entropically driven and suggest that an increased reliance on hydrophobic interactions underlies the energetics of microtubule formation at low temperatures. Thus, evolutionary modification to increase the proportion of hydrophobic interactions (relative to other bond types) at sites of interdimer contact may be one adaptive mechanism that enables the tubulins of cold-living poikilotherms to polymerize efficiently at low temperatures.

Kinesin, a microtubule-dependent ATPase, is believed to be involved in anterograde axonal transport. The kinesin head, which contains both microtubule and ATP binding sites, has the necessary components for the generation of force and motility. We have used saturation binding and electron microscopy to examine the interaction of the kinesin motor domain with the microtubule surface and found that binding saturated at one kinesin head per tubulin heterodimer. Both negative staining and cryo-electron microscopy revealed a regular pattern of kinesin bound to the microtubule surface, with an axial repeat of 8 nm. Optical diffraction analysis of decorated microtubules showed a strong layer-line at this spacing, confirming that one kinesin head binds per tubulin heterodimer. The addition of Mg-ATP to the microtubule-kinesin complex resulted in the complete dissociation of kinesin from the microtubule surface.

Using atomic force microscopy (AFM) it has been possible to detect actin filaments that are beneath the cell membrane of living cells despite the fact that the AFM tip is applied to the surface of the cell. To determine whether the AFM tip actually penetrates or deforms the cell membrane we determined whether an intracellularly trapped fluorescent indicator was lost from cells during AFM. Using epifluorescence illumination to monitor the presence of fluo-3 in the cell, we found that AFM did not cause dye leakage from the cell. Further, force-distance curves indicated that standard tips did not penetrate the membrane while sharper Supertips TM did. In addition, the physiology of cells was found to be unaffected by AFM with standard tips since volume regulatory signal transduction mechanisms were intact in such studies. Thus, traditional AFM tips deform the cell membrane in order to reveal the presence of subcellular structures.

The cooling efficiencies of various fluids at low temperature were compared by measuring the temperature decay in 3 p1 water samples plunged into them. A simple model of cooling was used in order to discuss the results. Liquid ethane was found to produce a cooling rate of 660 K s-1, about twice that of liquid propane, while ethanol was almost as effective as ethane between 273 to 223 K. The heat-transfer coefficient of liquid ethane was estimated to be between 1500 and 5000 W m-2 K-l, depending on the physical state assumed for the water sample. Samples of flagellated organisms, after being frozen rapidly in the above way, were freezesubstituted by the method of Barlow & Sleigh (1979). Although this fixation did not give good definition of the microtubules of the flagellar axoneme, it exhibited reasonable tissue preservation in thin sections of the cell body. The fixation method resulted in preserved flagellar wave shapes, which were observed under the light microscope and in critical-point dried cells examined by scanning and conventional electron microscopy. It was concluded (a) that methods for preserving the wave shape of the flagellum and for preserving its internal structure may not be compatible, and (b) that although the present cooling method (with ethane) approaches the speed required to arrest a flagellar wave, further improvements in the speed of the method are desirable.