Plant Physiol, October 2001, Vol. 127, pp. 381-385

SCIENTIFIC CORRESPONDENCE

Cell Walls at the Plant Surface Behave Mechanically Like Fiber-Reinforced Composite Materials¹

Department of Biology, University of Antwerp U.I.A., Universiteitsplein 1, B-2610 Wilrijk, Belgium (S.K., J.-P.V.); and Department of Physics, Laboratory of Biomedical Physics, University of Antwerp RUCA, Groenenborgerlaan 171, B-2020 Antwerp, Belgium (W.F.D.)

	ARTICLE

TOP ARTICLE LITERATURE CITED

Using extensiometry and polarization confocal microscopy, we provide here the first confirming evidence that the mean orientation of cellulose fibrils in the composite wall determines its mechanical properties. When there is a preferred orientation of the cellulose fibrils, the cell wall is reinforced parallel with the fibrils.

Composite materials consist of stiff strong fibers, plates, or particles placed in a relatively compliant matrix (Vincent, 1992). The mechanical properties of fiber-based composites depend on many variables such as fiber types, orientations, and general architecture. Plant cell walls can be considered as composite materials (Roland et al., 1989). They consist of stiff semicrystalline cellulose fibrils, cross-linked by hemicellulose polymers, embedded in a gel-like matrix of pectins (Cosgrove, 1997; Fujino et al., 2000). Cellulose fibrils have a history of being considered as the load-bearing elements in plant cell walls (Hofmeister, 1859; Niklas, 1992). The epidermis, the covering cell layer of plants, has typical patterns of mean cellulose orientation (Verbelen and Kerstens, 2000; Verbelen et al., 2001) and its mechanical properties are of utmost importance for the control of size and shape of the plant body (Green, 1980; Hernandez and Green, 1993; Niklas and Paolillo, 1997). Factual support for this axiome, however, is absent.

We addressed the experimental question on two well-defined single-cell layer models: epidermal peels from the adaxial side of the onion (Allium sativum) bulb scale and from the abaxial side of a Kalanchoe (Kalanchoe blossfeldiana Poelln.) leaf. The outer periclinal wall of these tissues forms the boundary between plant and environment.

The orientation of the cellulose fibrils was revealed by Congo Red staining and polarization confocal microscopy (Stickens and Verbelen, 1995; Verbelen and Kerstens, 2000). Congo Red specifically binds to -1,4-linked glucan polymers, thereby aligning the chromophoric groups. If cellulose fibrils are themselves aligned in parallel, the fluorescence intensity of the wall is at a maximum when the vector of the polarized exciting laser beam is parallel to the predominant cellulose orientation. It is at a minimum with the vector perpendicular to the fibrils. Walls with a totally random orientation of the cellulose fibrils have no predominant excitation orientation and their fluorescence intensity does not depend on the orientation of the vector of the exciting beam. In onion, the wall has an anisotropic architecture. Cellulose fibrils have a mean orientation parallel with the longitudinal axis of the cell, as is illustrated by the high intensity of Congo Red fluorescence when the polarization vector of the laser light is parallel to this axis (Fig. 1, a and b). On the contrary, the wall of Kalanchoe epidermal cells is isotropic: it has a random orientation of the cellulose fibrils. There is no difference in fluorescence intensity depending on the orientation of the laser beam (Fig. 1, c and d). In both figures, the arrows indicate the orientation of the electrical vector of the laser beam and the inserted wedge illustrates the color coding of the fluorescence intensity. Low, intermediate, and high fluorescence intensity are indexed as blue, green, and red, respectively.

View larger version (200K): [in this window] [in a new window]   Figure 1.   Polarization confocal micrographs of onion (a and b) and Kalanchoe (c and d) epidermal peels after Congo Red staining. Scale bar is 100 µm. The electrical vector of the polarized laser light is indicated by the arrows. The intensity of fluorescence is color coded from low (blue) to high (red) and is represented by the wedge included in b. The onion epidermis has a mean orientation of the cellulose fibrils parallel with the longitudinal axis of the cell (highest fluorescence intensity in a). For the Kalanchoe epidermal peels, there is no difference in fluorescence intensity and hence cellulose fibril orientation is random.

Is the mean orientation of cellulose fibrils related to the mechanical properties of the epidermis? To answer this question, strips of the two model tissues were subjected to uniaxial tension experiments in two directions, in the plane of the outer (periclinal) wall. For the anisotropic onion epidermis these directions were parallel and perpendicular to the mean cellulose orientation. For the isotropic Kalanchoe epidermis, we chose arbitrary directions: parallel and perpendicular to the leaf axis. For extensiometry strips of living tissues (5-mm width) were fixed between clamps and put in a bath of tapwater. The upper clamp was attached to a vibration exciter (type 4809, Bruel and Kjær, Naerum, Denmark) to extend the tissue. The lower clamp was attached to an isometric force transducer (model 31, Sensotec, Columbus, OH) to measure the generated force. The whole device was coupled to a personal computer with an input/ouput data acquisition board (PCI-6024E, National Instruments, Austin, TX) and controlled in Matlab. The tissue was pre-extended to +4%. After 1 min of relaxation, the tissue was cyclically (sinusoidally) extended between +3.5% and +6.5% (around 5%) at a frequency of 1 Hz. A function of the form:
is fitted to the measured force. The value of B₀ corresponds to the force of the point (5%) around which the sinus oscillates. The amplitude (A₁) corresponds to the elastic response of the tissue. There is a phase difference (₁) between the measured force and the extension, which corresponds to the viscous behavior of the cell wall (Ferry, 1970). The second harmonic sin(2x), with his own amplitude and phase difference A₂ and ₂, is a measure for the nonlinear behavior of the tissue. This empirical function fits well the measured force, generated by extending soft biological materials (Decraemer, 1977). Forces generated during repetitive extensions of the tissue were measured and divided by the average surface area of the transverse section of the samples to obtain tensions (MPa).

The relation between extension and tension for the first loop is represented in Figure 2. Each loop is from another sample. The onion epidermis is much stiffer parallel with the mean orientation of the cellulose fibrils in the wall (much higher tension is generated for the same extension), than transverse to this mean orientation (Fig. 2a). The tissue is thus mechanically anisotropic. It is also striking that the area of the loops for extension parallel to the cellulose orientation is substantially larger than for extension in the other direction. For the Kalanchoe epidermis, which has a random orientation of the cellulose fibrils in its cell wall, there is no difference in mechanical properties between both extension directions (Fig. 2b). For its surface characteristics, this tissue is mechanically isotropic.

View larger version (22K): [in this window] [in a new window]   Figure 2.   Stress-strain loops for unidirectional tensile experiments on onion and Kalanchoe epidermal peels. Every single ellipsoid curve refers to the first extension of another sample. a, Onion epidermal peels are stronger parallel (red) with the mean cellulose fibril (CF) orientation (higher tension is generated) than transverse to this orientation (blue). b, There is no orientation-linked difference in mechanical properties of the Kalanchoe epidermal peel (red and blue). Because there is no preferential cellulose orientation, the main vein of the leaf (MV) was used as a reference.

When the subsequent extension cycles of individual samples are considered, preconditioning becomes obvious (Fig. 3). This happens in most viscoelastic materials. It is clear that the area of the first loop is much larger than the area of subsequent loops in an onion sample when pulled parallel with the cellulose fibrils (Fig. 3a). A plastic extension occurs, seen as a notch in the experimental curve, but not seen on this fitted curve. This is not the case when the pulling force is exerted transverse to the cellulose fibrils. The Kalanchoe epidermis behaves in the two directions of applied extension like the onion epidermis in the longitudinal position but to a lesser extend (Fig. 3b). Except for the onion samples pulled transverse to the mean orientation of the cellulose fibrils, the material is thus in a different state after the first cycle. During the first extension, shear forces build up in the composite between the fibers and the matrix, due to their different modulus of elasticity (Piggott, 1980). Above a certain yield point, the matrix undergoes plastic deformation and the fibrils can slide along each other (Spatz et al., 1999). Whether this is a true plastic deformation or a long-term viscoelastic effect is a matter of discussion (Nolte and Schopfer, 1997). In the second and subsequent cycles, only viscoelastic behavior is seen, with little changes in plasticity. The onion epidermis pulled transverse to the mean orientation of the cellulose fibrils behaves purely viscoelastic from the first cycle on. These orientation-dependent differences in viscoelastic behavior are highlighted by the values of the variables in the formula mentioned above.

View larger version (21K): [in this window] [in a new window]   Figure 3.   Evolution of the stress-strain loops during nine successive extensions of onion and Kalanchoe peels. For each direction, only one sample is shown. Preconditioning becomes obvious after several cycles. a, For the onion epidermis pulled parallel with the cellulose fibrils, the first loop is of a totally different shape from the subsequent loops (red). This is not the case when the epidermis is pulled perpendicular (blue) to the mean orientation of cellulose fibrils (for clarity only the first, second, and ninth loops are shown). b, The Kalanchoe epidermis has the same behavior in both directions during the subsequent cycles (red and blue).

For the onion epidermis, there is a drastic drop in ₁ (phase difference related to viscous behavior) after the first extension cycle in the parallel orientation but not in the transverse orientation (Fig. 4a). The value for A₁ (amplitude related to elastic behavior) diminishes during subsequent extension cycles due to relaxation effects (Fig. 4c). In the direction perpendicular to the orientation of cellulose fibrils, a first extension of 6.5% correlates with an amplitude of 0.04 MPa (blue), but a first extension of 6.5% parallel with the cellulose fibrils correlates with an amplitude of 0.33 MPa (red). In other words, the wall is 8.25× less elastic in the direction parallel with its mean cellulose orientation. The course of the ₁ and A₁ values for the Kalanchoe epidermis during the subsequent extension cycles show that the tissue has the same characteristics in the two directions (Fig. 4, b and d). In a cell wall with a random cellulose orientation, there are always some fibrils parallel with the direction of extension. The changes in ₁ and A₁ explain for the greater part the differences in shape and size of the curves in Figure 2.

View larger version (23K): [in this window] [in a new window]   Figure 4.   The evolution of the phase difference 1 (a and b) and the amplitude A1 (c and d) during the successive loops. a, Substantial change in phase difference is seen between the first and the next cycles in the onion epidermis pulled parallel with the cellulose fibril orientation (red), whereas only a very smooth and gradual change in phase difference occurs when the epidermis is pulled perpendicular to the cellulose fibril orientation (blue). b, The Kalanchoe samples behave in both directions like the onion epidermis pulled parallel with the cellulose fibril orientation. c, During subsequent cycles in the onion, the amplitude A1 reveals stress relaxation in both directions. d, Kalanchoe peels have the same elasticity characteristics in both directions (red and blue).

These quantitative data demonstrate that the orientation of cellulose fibrils is a key factor in determining the mechanical properties of the plant cell wall; hence, plant cell walls can be regarded as fiber reinforced composite materials. Under tension, an epidermis with a mean parallel orientation of its cellulose fibrils in the wall will easily expand in the direction transverse to the mean orientation of cellulose fibrils; the material is very elastic and no plastic deformation happens. In the direction parallel with the mean orientation of cellulose fibrils, however, the wall is very stiff and during the first extension parallel with the cellulose fibrils a plastic deformation occurs. An epidermis with a random orientation of cellulose fibrils in the wall exhibits plastic deformation and a limited elasticity in all directions in the plane of the epidermis. In such walls, there is always a fraction of the cellulose fibrils parallel with the pulling force.

Experimental approaches on the viscoelastic behavior of cell walls have traditionally been done on whole organs or samples with high tissue complexity (Cleland, 1984; Kutschera, 1996). They seldom focused on the orientation of cellulose fibrils in the wall (Niklas and Paolillo, 1998; Lichtenegger et al., 2000). In this report, we relate wall viscoelastic properties to cellulose architecture in a single layer tissue. Considering the data, we can conclude that the cell wall behaves as a composite material with a matrix and a fiber phase and that the orientation of fibrils is important for the reinforcement efficiency just as it is in synthetic materials (Callister, 1991). In all walls, the reinforcement efficiency parallel with the fibers and perpendicular to the fibers is not in a ratio of 1:0 because the matrix as well as the fibers have a complex nature, with many cross-links. With fibers randomly distributed within a specific plane, the cell wall is reinforced to the same extend in any direction in the plane of the fibers, as is also the case in synthetic composite materials.

In the current models on cell wall architecture and expansion, hemicelluloses cross-link the cellulose fibrils (Carpita and Gibeaut, 1993; Pauly et al., 1999; Cosgrove, 2000). The results presented in this article are a new step in the research toward the unraveling of the cell wall biomechanics at the molecular level. Important data can be expected when applying our approach to actively growing tissue layers.

	ACKNOWLEDGMENT

We thank Professor Karl J. Niklas for critically reading the manuscript.

	FOOTNOTES

Received May 9, 2001; accepted June 13, 2001.

¹ This work was supported by the Fund for Scientific Research- Flanders for the confocal setup (grant nos. 3.0028.90 and 2.0049.93). S.K. is a recipient of a PhD grant from the Flemish Institute for the Promotion of Scientific and Technological Research in Industry (grant no. 981256).

^* Corresponding author; e-mail verbelen{at}uia.ua.ac.be; fax 32-3-820-2271.

www.plantphysiol.org/cgi/doi/10.1104/pp.010423.

	LITERATURE CITED

TOP ARTICLE LITERATURE CITED

• Callister WD Jr (1991) Materials Science and Engineering. John Wiley & Sons Inc., New York
• Carpita NC, Gibeaut DM (1993) Plant J 3: 1-30[CrossRef][ISI][Medline]
• Cleland RE (1984) Planta 160: 514-520[CrossRef]
• Cosgrove DJ (1997) Plant Cell 9: 1031-1041[CrossRef][ISI][Medline]
• Cosgrove DJ (2000) Nature 407: 321-326[CrossRef][Medline]
• Decraemer WF (1977) PhD thesis. University of Antwerp, Wilrijk, Belgium
• Ferry JD (1970) Viscoelastic Properties of Polymers. John Wiley & Sons Inc., New York
• Fujino T, Sone Y, Mitsuishi Y, Itoh T (2000) Plant Cell Physiol 41: 486-494
• Green PB (1980) Ann Rev Plant Physiol 31: 51-82
• Hernandez LF, Green PB (1993) Plant Cell 5: 1725-1738[Abstract]
• Hofmeister W (1859) Berichte über die Verhandlungen der Königlichen Sächsischen Gezellschaft der Wisschenhaften zu Leipzig (Mathematisch-physische Klasse)11: 175-204
• Kutschera U (1996) J Exp Bot 302: 1387-1394
• Lichtenegger H, Reiterer A, Tschegg SE, Müller M, Riekel C, Paris O, Fratzl P (2000) In H-C Spatz, T Speck, eds, Plant Biomechanics 2000. Thieme Verlag, Stuttgart, pp 436-442
• Niklas KJ (1992) Plant Biomechanics: an Engineering Approach to Plant Form and Function. University of Chicago Press, Chicago
• Niklas KJ, Paolillo DJ Jr (1997) Am J Bot 84: 735-744[Abstract]
• Niklas KJ, Paolillo DJ Jr (1998) Am J Bot 85: 1068-1081[Abstract]
• Nolte T, Schopfer P (1997) J Exp Bot 48: 2103-2107
• Pauly M, Albersheim P, Darvill A, York WS (1999) Plant J 20: 629-639[CrossRef][ISI][Medline]
• Piggott MR (1980) Loadbearing Fiber Composites. Pergamon Press, Oxford
• Roland JC, Reis D, Vian B, Roy S (1989) Biol Cell 67: 209-220[CrossRef]
• Spatz H-C, Köhler L, Niklas KJ (1999) J Exp Bot 202: 3269-3272
• Stickens D, Verbelen J-P (1995) J Microsc 177: 1-6
• Verbelen J-P, Kerstens S (2000) J Microsc 198: 101-107[Medline]
• Verbelen J-P, Vissenberg K, Kerstens S, Le J (2001) J Plant Physiol 158: 537-543[CrossRef]
• Vincent JFV (1992) In JFV Vincent, ed, Biomechanics. Oxford University Press, Oxford, 57-74