A close temporal and spatial correlation between cell growth,cell wall synthesis and the activity of enzymes of mannan synthesis in Acetabularia mediterranea

The synthesis of cell wall mannan and the activities of guanosine-diphosphate-mannose-pyrophosphorylase (EC2.7.7.13) and mannan synthetase were studied during the development of nucleate and enucleated cells of the alga Acetabularia mediterranea. The activities of both enzymes are relatively high as long as the cells grow and synthesize mannans. With termination of growth and mannan synthesis, the activities of both enzymes, but especially of mannan synthetase, drop to a low value. Furthermore, the activities of both enzymes are distributed in the cell along an apical-basal gradient. High activities are present in the apical regions of the cell where growth and mannan synthesis mainly occur, whereas in the basal region, growth, mannan synthesis and the activity of the two enzymes are slight. Since the in vitro activity of GDP-Man-pyr is at least 100 times higher than that of mannan synthetase, it was concluded that mannan synthetase activity is the limiting factor in mannan synthesis. This conclusion is supported by the determined pool sizes of Fru 6-P, Man 6-P, Man 1-P and GDP-Man during the development of the cells. The control of mannan synthesis and with it cell wall formation and growth through the regulation of mannan synthetase activity is discussed.Abbreviations DD dark-dark regime - Fru 6-P fructose-6-phosphate - GDP-Man guanosine-diphosphate-mannose - GDP-Manpyr GDP-diphosphate-mannose-pyrophosphorylase - GTP guanosine-triphosphate - LD light-dark regime - Man 1-P mannose-1-phosphate - Man 6-P mannose-6-phosphate - TCA trichloracetic acid     

The polysaccharide structure of potato cell walls: Chemical fractionation

Cell walls of potato tubers were fractionated by successive extraction with various reagents. A slightly degraded pectic fraction with 77% galacturonic acid was extracted in hot, oxalate-citrate buffer at pH 4. A further, major pectic fraction with 38% galacturonic acid was extracted in cold 0.1 M Na₂CO₃ with little apparent degradation. These two pectic fractions together made up 52% of the cell wall. Most of the oxalate-citrate fraction could alternatively be extracted with cold acetate-N,N,N-tetracetic acid (CDTA) buffer, a non-degradative extractant which nevertheless removed essentially all the calcium ions. This fraction was therefore probably held only by calcium binding, and the remainder of the pectins by covalent bonds. Electrophoresis showed that both pectic fractions contained a range of molecular types differing in composition, with a high arabinose: galactose ratio as well as much galacturonic acid in the most extractable fractions. From methylation data, the main side-chains were 1,4-linked galactans and 1,5-linked arabinans, with smaller quantities of covalently attached xyloglucan. Extraction with NaOH-borate removed a small hemicellulose fraction and some cellulose. The main hemicelluloses were apparently a galactoxyloglucan, a mannan or glucomannan and an arabinogalactan.Abbreviations GLC gas-liquid chromatography - MS mass spectrometry - V₀ void volume - MW weight-average molecular weight - DMSO dimethylsulphoxide - EDTA ethylenediamine tetraacetic acid - TFA trifluoroacetic acid - CDTA N,N,N-tetraacetic acid     

Molecular weight distribution of cellulose in primary cell walls

The distribution pattern of the degree of polymerization (DP) of cellulose present in the cell walls of mesophyll- and suspension-cultured cells of tobacco was compared to that of newly synthesized ¹⁴C-labeled cellulose from regenerating tobacco protoplasts and suspension-cultured cells. The cellulose was nitrated, and, after fractionation according to differences in solubility in acetone/water, the DP pattern of labeled or unlabeled cellulose nitrate was determined by viscosity measurements. A low (DP<500) and high DP-fraction (DP>2500) of cellulose were predominant in the cell walls of protoplasts, suspension — cultured cells, and mesophyll cells. The average DP of the high molecular weight fraction of cellulose in the cell walls of mesophyll was higher (DP4,000) than in protoplasts or suspension — cultured cells (DP 2,500-3,000). In all cell walls tested, minor amounts of cellulose molecules with a broad spectrum of a medium DP were present. Pulse — chase experiments with either protoplasts or suspension —cultured cells showed that a large proportion of the low and medium DP-cellulose are a separate class of structural components of the cellulose network. The results are discussed in relation to the organization of cellulose in the primary cell wall.Abbreviations DP degree of polymerisation - 2,4-D 2,4-dichlorophenoxyacetic acid - IAA indole-3-acetic acid     

Control of thickness of collenchyma cell walls by pectins

Near-isotropic stresses were generated within collenchyma cell walls of celery (Apium graveolens L.) by exchanging K⁺ for Ca²⁺ ions, varying the ionic strength and de-esterifying the pectic carboxyl groups, treatments that changed the free-charge density of the pectic polysaccharides. The collenchyma strands swelled radially with increasing free-charge density but there was very little longitudinal swelling. Depolymerising the pectins by -elimination also induced much more radial than longitudinal swelling. Supported by earlier work on Nitella, these results indicate that pectins control the interlamellar spacing in cell walls and hold them together across their thickness, particularly against turgor stresses tending to delaminate the walls at the cell corners.The author thanks J.S.G. Reid (Department of Biological Sciences, University of Stirling, UK) and M. Demarty (SCUEOR, University of Rouen, France) for critical comments.     

A proton nuclear magnetic resonance study of the physical changes in growing plant cell walls

Changes in broadline proton nuclear magnetic resonance parameters of cell walls during growth of etiolated hypocotyls of bean (Phaseolus vulgaris L.) indicate that cell wall structure becomes more rigid during development. Most of the changes are completed in the first 6 cm below cotyledon insertion and are correlated with increased restriction of proton movements in regions of dense polymer packing.Abbreviations FID free induction decay - M₂ second moment - M_2interpair interpair second moment - NMR nuclear magnetic resonance - T_1D dipolar relaxation time - T₂ spin-spin relaxation time This work was supported by grants from Natural Sciences and Engineering Research Council of Canada to A.L.M., I.E.P.T. and M. Bloom.     

Identification of a highly conserved hydroxyproline-rich glycoprotein in the cell walls of Chlamydomonas reinhardtii and two other Volvocales

The unicellular alga Chlamydomonas reinhardtii Dang, has a cell wall made entirely from hydroxyproline-rich glycoproteins (HRGPs). We recently employed a quantiative in vitro reconstitution system (Adair et al. 1987, J. Cell Biol. 105, 2373–2382) to assign outer-wall HRGPs of C. reinhardtii to specific sublayers, and describe the major interactions responsible for their assembly. Some of these interactions appear to involve relatively conserved HRGP domains, as evidenced by interspecific cell-wall reconstitution between C. reinhardtii and two multicellular Volvocales (Volvoxcarteri lyengar and Gonium pectorale Müller). In the present report we provide biochemical and immunological evidence that the outer cell-walls of V. carteri and G. pectorale both contain prominent HRGPs closely related to C. reinhardtii GP2. Identification of conserved GP2 homologues indicates a molecular basis for interspecific reconstitution and provides a useful avenue for characterization of HRGP domains mediating cell-wall formation in these algae.Abbreviations GP1, 2, 3 outer-cell wall glycoproteins 1, 2, and 3 - GP2_dg deglycosylated GP2 - HRGP hydroxyprolinerich glycoprotein - SDS-PAGE sodium docecyl sulfate polyacrylamide gel electrophoresis     

Characterization of long-term extension of isolated cell walls from growing cucumber hypocotyls

Walls from frozen-thawed cucumber (Cucumis sativus L.) hypocotyls extend for many hours when placed in tension under acidic conditions. This study examined whether such creep is a purely physical process dependent on wall viscoelasticity alone or whether enzymatic activities are needed to maintain wall extension. Chemical denaturants inhibited wall creep, some acting reversibly and others irreversibly. Brief (15 s) boiling in water irreversibly inhibited creep, as did pre-incubation with proteases. Creep exhibited a high Q₁₀ (3.8) between 20° and 30°C, with slow inactivation at higher temperatures, whereas the viscous flow of pectin solutions exhibited a much lower Q₁₀ (1.35). On the basis of its temperature sensitivity, involvement of pectic gel-sol transitions was judged to be of little importance in creep. Pre-incubation of walls in neutral pH irreversibly inactivated their ability to creep, with a half-time of about 40 min. At 1 mM, Cu²⁺, Hg²⁺ and Al³⁺ were strongly inhibitory whereas most other cations, including Ca²⁺, had little effect. Sulfhydryl-reducing agents strongly stimulated creep, apparently by stabilizing wall enzyme(s). The physical effects of these treatments on polymer interactions were examined by Instron and stress-relaxation analyses. Some treatments, such as pH and Cu²⁺, had significant effects on wall viscoelasticity, but others had little or no apparent effect, thus implicating an enzymatic creep mechanism. The results indicate that creep depends on relatively rugged enzymes that are firmly attached to or entangled in the wall. The sensitivity of creep to SH-reducing agents indicates that thiol reduction of wall enzymes might provide a control mechanism for endogenous cell growth.Abbreviations DTT dithiothreitol - EDTA ethylenediaminetetraacetic acid - EGTA ethyleneglycol-bis-(-aminoethylether)-N,N,N,N-tetraacetic acid - Hepes N-2-hydroxyethylpiperazine-N-2-ethansulfonic acid     

Glycoprotein conformation in plant cell walls

The major structural glycoprotein of the cell wall of Chlamydomonas reinhardii has a protein core, at least 50% of which is in the unusual polyproline II conformation. This has been demonstrated by examining the circular dichroism of the cell wall, its constituent glycoproteins, and thermolysin released wall glycopeptides. One of these glycopeptides, T2, has a high hydroxyproline and sugar content, and possesses upward of 85% polyproline II structure. The main extracellular matrix glycoprotein therefore has a rigid, rod-like structure and the significance of this and its relation to higher plant cell wall glycoproteins is discussed. The unusual conformation appears to confer great stability on the glycoprotein as it is unchanged either by certain denaturing agents or during the transition from protomer to assembled cell wall.Abbreviations CD circular dichroism - HP 4-hydroxy-L-proline - PP poly-L-proline - SDS sodium dodecylsulphate This is the eight paper in a series entitled Structure, Composition and Morphogenesis of the Cell Wall of Chlamydomonas reinhardii. The last paper in this series was Catt et al. (1978)     

Pectin esterification is spatially regulated both within cell walls and between developing tissues of root apices

Monoclonal antibodies recognizing un-esterified (JIM5) and methyl-esterified (JIM7) epitopes of pectin have been used to locate these epitopes by indirect immunofluorescence and immunogold electron microscopy in the root apex of carrot (Daucus carota L.). Both antibodies labelled the walls of cells in all tissues of the developing root apex. Immunogold labelling observed at the level of the electron microscope indicated differential location of the pectin epitopes within the cell walls. The un-esterified epitope was located to the inner surface of the primary cell walls adjacent to the plasma membrane, in the middle lamella and abundantly to the outer surface at intercellular spaces. In contrast, the epitope containing methyl-esterified pectin was located evenly throughout the cell wall. In root apices of certain other species the JIM5 and JIM7 epitopes were found to be restricted to distinct tissues of the developing roots. In the root apex of oat (Avena sativa L.), JIM5 was most abundantly reactive with cell walls at the region of intercellular spaces of the cortical cells. JIM7 was reactive with cells of the cortex and the stele. Neither epitope occurred in walls of the epidermal or root-cap cells. These pattern of expression were observed to derive from the very earliest stages of the development of these tissues in the oat root meristem and were maintained in the mature root. In the coleoptile and leaf tissues of oat seedlings, JIM5 labelled all cells abundantly whereas JIM7 was unreactive. Other members of the Gramineae and also the Chenopodiaceae are shown to express similar restricted spatial patterns of distribution of these pectin epitopes in root apices.Abbreviations CDTA 1,2-diaminocyclohexane tetraacetic acid - RG rhamnogalacturonan J.P.K. was supported by the Agricultural and Food Research Council Cell Signalling and Recognition Programme. We thank J. Cooke and N. Stacey for technical assistance, H.A. Schols, Drs. P. Albersheim and A. Darvill for pectic polysaccharides, and Dr. R.R. Selvendran and M. McCann for useful discussions.     

Composition of the cell walls of Nicotiana alata Link et Otto pollen tubes

Cell walls isolated from pollen of Nicotiana alata germinated in vitro contain glucose and arabinose as the predominant monosaccharides. Methylation analysis and cytochemical studies are consistent with the major polysaccharides being a (13)--D-glucan (callose) and an arabinan together with small amounts of cellulose. The cell walls contain 2.8% uronic acids. Alcian blue stains the pollen-tube walls intensely at the tip, indicating that acidic polysaccharides are concentrated in the tip. Synthetic aniline-blue fluorochrome is specific primarily for (13)--D-glucans and stains the pollen-tube walls, except at the tip. Protein (1.5%), containing hydroxyproline (2.4%), is present in the cell wall.     

Use of antisera to localize callose,xylan and arabinogalactan in the cell-plate,primary and secondary walls of plant cells

Antibodies to cellobiose, L-arabinopyranose, L-arabinofuranose, D-galactose, oligosaccharides containing 14 xylose, oligosaccharides containing 14 glucose, and oligosaccharides containing 13 glucose have been raised in rabbits. The antisera have been characterized to show the specificity of binding to particular polysaccharides. They have been used for immunocytology using the electron microscope to locate the polymers in dividing and differentiating cells of bean (Phaseolus vulgaris L.) root, bean callus tissue and cells of Zinnia elegans L. in vitro. Arabinogalactans have been shown to be present in the cell-plate and primary walls but not in secondary thickening. Xylan as distinct from xyloglucan was found in the primary walls but not in the cell-plate. It was present in large amounts in the secondary thickening. Callose was found in the cell plate and also in the young growing wall. In the wall it was specifically located at the plasmodesmata. The use of the antibody against L-arabinofuranose enabled a specific organelle to be detected which was membranous and which occurred within the cytoplasm and also within the vacuole of the cells. Membranes carrying polymers containing L-arabinofuranose were also found in layers just under the plasmamembrane.Abbreviations L-Araf L-arabinofuranose - L-Arap L-arabinopyranose - BSA bovine serum albumin - Gal galactose - D-Galp D-galactopyranose - Glc glucose - Xyl xylose     

Investigations of mannan and glucan in the cell wall ofCandida boidinii cultivated on methanol

The polysaccharide components (mannan and glucan) in the cell wall ofCandida boidinii M 363 grown on methanol and glucose as control were investigated using electron microscopy, cytochemical and biochemical methods. An ultrastructural rearrangement of the polymers in the cell wall of yeasts cultivated on methanol in comparison to those cultivated on glucose was established. The morphological changes correlate to the quantitative changes in the polysaccharide constituents of the cell wall. The forming and the role of thiosemihydrocarbazide (TSHC) — negative zones in theCandida boidinii cell wall cultivated on methanol media are discussed.     

Structural analysis of the cell walls regenerated by carrot protoplasts

A procedure was developed to isolate protoplasts rapidly from carrot (Daucus carota L. cv. Danvers) cells in liquid culture. High purity of cell-wall-degrading enzymes and ease of isolation each contributed to maintenance of viability and initiation of regeneration of the cell wall by a great majority of the protoplasts. We used this system to re-evaluate the chemical structure and physical properties of the incipient cell wall. Contrary to other reports, callose, a (1 3)-d-glucan whose synthesis is associated with wounding, was not a component of the incipient wall of carrot protoplasts. Intentional wounding by rapid shaking or treatment with dimethyl sulfoxide initiated synthesis of callose, detected both by Aniline blue and Cellufluor fluorescence of dying cells and by an increase in (1 3)-linked glucan quantified in methylation analyses. Linkage analyses by gas-liquid chromatography of partially methylated alditol-acetate derivatives of polysaccharides of the incipient wall of protoplasts and various fractions of the cell walls of parent cells showed that protoplasts quickly initiated synthesis of the same pectic and hemicellulosic polymers as normal cells, but acid-resistant cellulose was formed slowly. Complete formation of the wall required 3 d in culture, and at least 5 d were required before the wall could withstand turgor. Pectic substances synthesized by protoplasts were less anionic than those of parent cells, and became more highly charged during wall regeneration. We propose that de-esterification of the carboxyl groups of pectin uronic-acid units permits formation of a gel that envelops the protoplast, and the rigid cellulose-hemicellulose frame-work forms along with this gel matrix.Abbreviations DEAE Diethylaminoethyl - DMSO dimethyl sulfoxide - ECP extracellular polymers - EDTA ethylenediaminetetraacetic acid - HGA nomogalacturonan - RG rhamnogalacturonan - Tes N-tris(hydroxymethyl)methyl-2-amino-ethanesufonic acid - TFA trifluoroacetic acid Journal paper No. 11,776 of the Purdue University Agriculture Experiment Station     

The proportion of calcium-bound pectin in plant cell walls

The amount of pectin held in cell walls by ionic bonds only was determined by extraction with cyclohexanediamine tetraacetic acid (CDTA) at room temperature, to remove calcium ions without degrading the galacturonan chains. Enzymic degradation was avoided by extracting the cell walls with phenol-acetic acid-water during preparation. From cell walls of celery petioles, cress hypocotyls and tomato and cucumber pericarp CDTA extracted 64–100 mg g^-1 pectin, leaving 80–167 mg g^-1 uronic acid in the residue. An additional extraction at high ionic strength was used to make the galacturonan chains more flexible and thus detach any pectins held by steric interactions, but the amount released in this way was small. Most of the residual uronic acid polymers could be extracted by cold alkali and remained soluble on neutralisation, showing that it was not water-insolubility that prevented their extraction with CDTA. Covalent bonding was thought more likely.     

Isolation and characterization of cell walls from the mesocarp of mature grape berries (Vitis vinifera)

Cell walls have been isolated from the mesocarp of mature grape (Vitis vinifera L.) berries. Tissue homogenates were suspended in 80% (v/v) ethanol to minimise the loss of water-soluble wall components and wet-sieved on nylon mesh to remove cytoplasmic material. The cell wall fragments retained on the sieve were subsequently treated with buffered phenol at pH 7.0, to inactivate any wall-bound enzymes and to dislodge small amounts of cytoplasmic proteins that adhered to the walls. Finally, the wall preparation was washed with chloroform/methanol (1:1, v/v) to remove lipids and dried by solvent exchange. Scanning electron microscopy showed that the wall preparation was essentially free of vascular tissue and adventitious protein of cytoplasmic origin. Compositional analysis showed that the walls consisted of approximately 90% by weight of polysaccharide and less than 10% protein. The protein component of the walls was shown to be rich in arginine and hydroxyproline residues. Cellulose and polygalacturonans were the major constituents, and each accounted for 30–40% by weight of the polysaccharide component of the walls. Substantial varietal differences were observed in the relative abundance of these two polysaccharides. Xyloglucans constituted approximately 10% of the polysaccharide fraction and the remainder was made up of smaller amounts of mannans, heteroxylans, arabinans and galactans. Received: 26 November 1996 / Accepted: 30 January 1997     

Cellulose-microfibril-orienting mechanisms in plant cells walls

A brief review is given of the changing views over the years, as knowledge of wall structure has developed, concerning the mechanism whereby cellulose chains may be oriented. This leads to an examination of current concepts, particularly those concerning microtubules. It is shown that none of the mechanisms suggested whereby microtubules might cause orientation of cellulose microfibrils is consistent with the known range of molecular architectures found in plant cell walls. It is further concluded that any mechanism which necessitates an indissoluble link between the plasmalemma and the cellulose-synthesising complex at the tip of a microfibril is unacceptable. A new proposal is presented in which it is speculated that both microtubules and microfibrils are oriented by a mechanism separate from both. It is shown that if two vectors are contemplated, one parallel to cell length and one at right angles, and a sensor exists on the plasmalemma surface which responds to changes in the vectors, then all known wall structures may be explained. The possible nature of the vectors and the sensor are considered.     

In vivo and in vitro swelling of cell walls during fruit ripening

Swelling properties of the cell walls of nine temperate fruit species, selected for their different ripening and textural characteristics, were studied during ripening. Cell wall swelling was examined in intact fruit using microscopy techniques and in vitro, using cell wall material isolated from fruit tissue. In fruit which ripened to a soft melting texture (persimmon, avocado, blackberry, strawberry, plum), wall swelling was pronounced, particularly in vitro. In-vivo swelling was marked only in avocado and blackberry. Fruit which ripened to a crisp, fracturable texture [apple (two cultivars), nashi pear, watermelon] did not show either in-vivo or in-vitro swelling of the cell wall. There was a correlation between swelling and the degree of pectin solubilisation, suggesting that wall swelling occurred as a result of changes to the viscoelastic properties of the cell wall during pectin solubilisation. Chemical and enzymatic removal of pectin from kiwifruit cell wall material supported the idea that swelling is associated with movement of water into voids left in the cellulose-hemicellulose network by the solubilised pectin. However, the results also suggested that swelling in vivo was more complex than this, and that the physicochemical changes which led to swelling included other elements of cell wall modification involving the site and mechanism of pectin solubilisation and-or the cellulose-xyloglucan complex. Received: 28 January 1997 / accepted: 11 March 1997     

Multinucleate cyst formation in Acetabularia mediterranea after x-irradiation

Cells of Acetabularia mediterranea were irradiated with increasing doses of X-rays (64.5–258·10^-4 kC kg^-1). The cells are radioresistant up to 193.5·10^-4 kC kg^-1 in terms of growth and progression through he life cycle but the morphogenesis of whorls, caps, and cysts is accompanied by morphological alterations. Microscopical examination of cyst bearing caps in irradiated cells has shown the presence of giant cysts neighboring particularly small ones. Photographic recording of cyst development showed that the multinucleate cap cytoplasm partitions into multinucleate portions rather than uninucleate ones as in the control cells. After complete cleavage a cyst wall is deposited onto the multinucleate cytoplasm. In contrast to uninucleate cysts with one lid the wall contains multiple lids. Their number appears to correspond to the number of nuclei in the cytoplasm compartment during cleavage. The results indicate that X-rays preferentially inhibit the synthesis of a factor which plays a role in establishing the normal spatial morphogenetic pattern necessary for cyst formation.     

The relationship between xyloglucan endotransglycosylase and in-vitro cell wall extension in cucumber hypocotyls

It has been proposed that cell wall loosening during plant cell growth may be mediated by the endotransglycosylation of load-bearing polymers, specifically of xyloglucans, within the cell wall. A xyloglucan endotransglycosylase (XET) with such activity has recently been identified in several plant species. Two cell wall proteins capable of inducing the extension of plant cell walls have also recently been identified in cucumber hypocotyls. In this report we examine three questions: (1) Does XET induce the extension of isolated cell walls? (2) Do the extension-inducing proteins possess XET activity? (3) Is the activity of the extension-inducing proteins modulated by a xyloglucan nonasaccharide (Glc₄-Xyl₃-Gal₂)? We found that the soluble proteins from growing cucumber (cucumis sativum L.) hypocotyls contained high XET activity but did not induce wall extension. Highly purified wall-protein fractions from the same tissue had high extension-inducing activity but little or no XET activity. The XET activity was higher at pH 5.5 than at pH 4.5, while extension activity showed the opposite sensitivity to pH. Reconstituted wall extension was unaffected by the presence of a xyloglucan nonasaccharide (Glc₄-Xyl₃-Gal₂), an oligosaccharide previously shown to accelerate growth in pea stems and hypothesized to facilitate growth through an effect on XET-induced cell wall loosening. We conclude that XET activity alone is neither sufficient nor necessary for extension of isolated walls from cucumber hypocotyls.     

Total epidermal cell walls of pea stems respond differently to auxin than does the outer epidermal wall alone

The effect of auxin on cell wall mass in the epidermis of third internodes of Pisum sativum L. cv. Alaska grown in dim red light was investigated using epidermal peels, to determine whether epidermal peels reflect the behavior of the outer epidermal cell wall. In contrast to the outer epidermal wall itself, where auxin caused thinning in proportion to growth (M.S. Bret-Harte et al, 1991, Planta 185, 462–471), auxin promoted an increase in wall mass in epidermal peels from treated internode segments in the absence of exogenously supplied sugar. The percentage gain in mass was smaller than the percentage elongation, however, so mass per unit length decreased in peels from auxin-treated segments. Epidermal peels from auxin-treated segments gained more wall mass than control peels even when adhering internal tissue at the basal end of the peel was removed. Epidermal peels also had a gross composition different from that of the outer wall alone (M.S. Bret-Harte and L.D. Talbott, 1993, Planta 190, 369–378). These discrepancies can be explained by the observation that the outer wall makes up only 30% of the mass of the epidermal peel. It appears that the inner walls of the epidermis, and walls of the outer layer of cortical cells that remain attached to the epidermis during peeling, nearly maintain their thickness by biosynthesis while the outer wall loses mass as previously described (Bret-Harte et al. 1991). These results indicate that epidermal peels may not be a good system for examining the biochemical and physiological properties of the outer epidermal cell wall.I would like to thank Dr. Peter M. Ray, of Stanford University, for the use of experimental facilities, helpful discussions, and technical and editorial assistance, Dr. Winslow R. Briggs, of the Carnegie Institute of Washington, for helpful discussions and for the use of experimental facilities, Dr. Paul B. Green, of Stanford University, for financial support, and Dr. Wendy K. Silk, of the Department of Land, Air, and Water Resources, University of California, Davis, for financial support. This work was supported by a National Science Foundation Graduate Fellowship, National Science Foundation grant DCB8801493 to Paul B. Green, and the generosity of Wendy K. Silk in the final writing.