Transcellular Strands in Sieve Tubes; What Are They?

We show that sieve elements of Nymphoides peltata (S. G. Gmel.) O. Kuntze contain strands which are bundles of P-protein filaments. We observe the strands under the light microscope (differential-interference contrast), and in the scanning electron microscope which shows some of them to be arranged as a parietal network. We find bundles of filaments which correspond to these strands in sections of embedded sieve elements in the transmission electron microscope, and also in freeze-fracture replicas of sieve elements in vascular bundles frozen intact while translocating carbon-14. Not all the strands are necessarily transcellular; some may end in the parietal layer just to the inside of the plasmalemma where they appear to come in contact with membranes, possibly of endoplasmic reticulum. The filaments in the strands have the same banded appearance as filaments in the sieve pores. We are unable to find any membrane or other special boundary round the strands; we propose they should be called ‘filamentous strands’. We suggest that the filaments are aggregated into strands by the Bernoulli effect when fluid flows through sieve elements. We suggest that the strands may be formed by flow during translocation as well as by flow due to injury.