Wheat tissues freeze-etched during exposure to extracellular freezing: distribution of ice

Pieces excised from leaf bases and laminae of seedlings of Triticum aestivum L. cv. Lennox were slowly frozen, using a specially designed apparatus, to temperatures between 2° and 14° C. These treatments ranged from non-damaging to damaging, based on ion-leakage tests to be found in the accompanying report (Pearce and Willison 1985, Planta 163, 304–316). The frozen tissue pieces were then freeze-fixed by rapidly cooling them, via melting Freon, to liquid-nitrogen temperature. The tissue was subsequently prepared for electron microscopy by freeze-etching. Ice crystals formed during slow freezing would tend to be much larger than those formed during subsequent freeze-fixation. Ice crystals surrounding the excised tissues were much larger in the frozen than in the control tissues (the latter rapidly freeze-fixed from room temperature). Large ice crystals were present between cells of frozen laminae and absent from controls. Intercellular spaces were infrequent in control leaf bases and no ice-filled intercellular spaces were found in frozen leaf bases. Intracellular ice crystals were smaller in frozen tissues than in controls. It is concluded that all ice formation before freeze-fixation was extracellular. This extracellular ice was either only extra-tissue (leaf bases), or extra-tissue and intercellular (laminae). Periplasmic ice was sometimes present, in control as well as slowly frozen tissues, and the crystals were always small; thus they were probably formed during freeze-fixation rather than during slow freezing. The plasma membrane sometimes showed imprints of cell-wall microfibrils. These were less abundant in leaf bases at 8° C than in controls, and were present on only a minority of plasma membranes from laminae. Therefore, extracellular ice probably did not compress the cells substantially, and changes in cell size and shape were possibly primarily a result of freezing-induced dehydration. Fine-scale distortions (wrinkles) in the plasma membrane, while absent from controls, were present, although only rarely, in both damaged and non-damaged tissues; they were therefore ice-induced but not directly related to the process of damage.