Molecular Imaging of Halocynthia papillosa Cellulose William Helbert,*Yoshiharu Nishiyama,† Takeshi Okano,† and Junji Sugiyama* ,1 *Wood Research Institute, Kyoto University, Uji, Kyoto 611-0011, Japan; and †Graduate School of Agricultural and Life Science, The University of Tokyo, Yayoi, Tokyo 113-8256, Japan Received May 21, 1998, and in revised form September 17, 1998 The molecular organization of cellulose I  microfi- brils in the tunic of Halocynthia papillosa was analyzed by high-resolution cryoelectron micros- copy on ultrathin cross sections of artificially highly oriented microfibrils. The arrangement of cellulose chains intersected by the 0.6-, 0.53-, and 0.39-nm equatorial lattice planes was clearly imaged over the whole area of a parallelogram-shaped cross section of a microfibril. One edge of the parallelo- gram was parallel to the 0.6-nm lattice plane, while the other did not correspond to a crystallographic plane. Such organization is distinct from previous findings on algal cellulose I  -rich microfibrils, which have an almost square cross section bounded by both 0.6- and 0.53-nm crystallographic planes. A tentative model for microfibril formation is pro- posed by introducing a two-step biocrystallization mechanism: the formation of molecular sheets spaced by 0.53 nm between adjacent molecules, followed by self-deposition of these sheets by hydro- gen bonding between them.  1998 Academic Press Key Words: biogenesis; cellulose I  ; cellulose micro- fibril; Halocynthia papillosa; lattice imaging; cryo- microscopy. INTRODUCTION Cellulose is a macromolecule widely distributed in the living world. Its occurrence is not limited to algae and higher plants, but it can be also found in a few species of slime mold (Mu ¨hlethaler, 1956), fungi (Aronson et al., 1967; Bartinicki-Garcia, 1968), and animals (Rånby, 1952). To date, in all the organisms that have been investigated, cellulose appears natu- rally as crystalline microfibrils where the polymer chains have the same polarity (Hieta et al., 1984; Chanzy and Henrissat, 1984), and this form is called cellulose I. Another polymorph, most likely with anti-parallel packing, cellulose II, is also found on limited occasions: as a cell wall component of Halycis- tis (Sisson, 1938), as a product of mutant strains of bacteria (Kuga et al., 1993), or from a bacterial product after a low-temperature treatment (Hirai et al., 1997), all of which can be easily distinguished by their diffraction diagrams as well as a lack of microfi- brillar morphology. Although the chain polarity and microfibrillar nature enable native cellulose I to be identified, the chain packing in a microfibril leads to two possible allomorphs: cellulose I  and I  (Atalla and VanderHart, 1984), with the corresponding unit cells characterized as one-chain triclinic and two- chain monoclinic, respectively (Sugiyama et al., 1991). The biosynthesis of cellulose microfibrils through- out the living world seems to follow a similar mecha- nism. Cellulose synthases, named terminal com- plexes (TCs), polymerize unidirectional cellulose chains which crystallize into almost pure I  allo- morph in the case of some tunicate species, e.g., Halocynthia spp. (Belton et al., 1989; Larsson et al., 1995), or more generally into a I  /I  mixture in other species. On the basis of the I  /I  ratio, some species are considered I  rich (e.g., bacterial cellulose, algae in Cladophoraceae, Siphonocladaceae, Rhodophy- ceae, and Xantophyceae) or I  rich (e.g., wood cellu- lose, algae in Characeae, Zygnemataleceae) (Koyama et al., 1997a). In addition to such crystalline allomor- phism, the higher order of synthase architecture, i.e., the TC, is considered to play an important role in determining the other characteristics of a cellulose microfibril, such as size, shape, and organization of the lattice within (Brown, 1996). However, the exact role of TCs and the crystallization mechnism are still less clear. If we consider that two main events, polymeriza- tion and crystallization, take place during microfi- bril biosynthesis, Halocynthia papillosa cellulose seems to be a good model for study for several reasons. First, this tunicate cellulose is almost en- tirely composed of the I  allomorph (Belton et al., 1 To whom correspondence should be addressed. Fax: 81-774-38- 3635. JOURNAL OF STRUCTURAL BIOLOGY 124, 42–50 (1998) ARTICLE NO. SB984045 42 1047-8477/98 $25.00 Copyright  1998 by Academic Press All rights of reproduction in any form reserved.