322 SHLOMO HESTRIN 



If the diffusible entity which enters into the extracellular phase and 

 there serves as a precursor of the cellulose fibre is indeed itself chemically 

 identical with the cellulose in fibre, it could be anticipated that when cells 

 are incubated with glucose the point of time at which fibre appears might 

 be preceded by an interval during which a relatively large fraction of 

 cellulose molecules in the extracellular phase is still in a relatively dis- 

 organized state. Miihlethaler [ii] has shown that, when A. xylinum grows 

 in a complex medium, cellulose fibres arise in the vicinity of the cell within 

 an amorphous "slime". Studies recently conducted in our laboratory by 

 Mr. I. Ohad in collaboration with Dr. D. Danon at the Weizmann Institute 

 have indicated that when washed cells of this organism are suspended in 

 radioactive glucose solution the synthesis of cellulose proceeds linearly in 

 time without any observable induction phase, but that deposition of 

 fibre becomes apparent only after an initial interval during which electron- 

 microscopically discernible material is accumulated in the extracellular 

 phase within relatively amorphous formations which consist of an alkali- 

 insoluble, radioactive macromolecular compound — presumably cellulose 

 itself. Within a few minutes after contact of the cells with glucose, almost 

 all cellulose in the extracellular phase assumed a well-defined crystalline 

 habitat. Ribbons and ropes of cellulose arose by side-to-side aggregation 

 and intertwisting of an element which presented a remarkably constant 

 morphology — the cellulose " ultrastrand " The thickness of this element 

 was estimated by means of measurements of the length of the shadow which 

 it casts. On this basis, the thickness was shown to be in the range 15 A. 

 Width estimates were relatively more diflicult to arrive at in view of the 

 distortion imposed by metal-shadowing. Allowing about 80 A for con- 

 tribution made by metal to apparent width (cf. Hall [12]), the net width 

 of the ultrastrand can be estimated to have been about two or three times 

 the thickness. This would imply that the basic morphological element in 

 bacterial cellulose fibre is a bundle which comprises about twelve glucose 

 chains in its cross-section. 



Celluloses from widely different sources including preparations of 

 bacterial origin, have been generally supposed, on the basis of electron- 

 microscope studies, to consist of fibrils ^ 60 A in diameter [13]. However, 

 correction was not made in any of these earlier studies for the effect of 

 metal on the apparent fibril width, nor was the thickness estimated on the 

 basis of measurements of length of shadow. In view of the present findings 

 on bacterial cellulose, a re-examination of the value which has been assigned 

 to the thickness dimension of the cellulose fibril in celluloses of different 

 sources may be desirable. 



