The mean or net preferential orientation of cellulose fibrils in plant cell walls is detected with polarization confocal laser scanning microscopy using the fluorescence dichroism of Congo Red. Single cells, arrays of cells in a tissue, or the epidermis of whole organs can be assayed in vivo. Aerial parts require an extra pectinase treatment because of the cuticle, which is impermeable to aqueous solutions. Peeling off the epidermis can be an elegant alternative, especially for leaves. With this method the net preferential fibril orientation can be related to the symmetry axis of the cell in quantitative terms. Data issuing from this approach are useful in current research on plant biomechanics.
The cell wall is often considered the plant equivalent of the extracellular matrix of animal cells. A very typical and plant-specific function of cell walls is the role they play in the mechanics of the build-up and the maintenance of the plant body. The tensile strength of the cell wall counterbalances the turgor pressure of the cells. The main load bearing elements in the wall are the cellulose fibrils, and their specific orientation confers mechanical isotropy or anisotropy to the cell wall. In this perspective the parallel or random orientation of cellulose fibrils in the growing wall is particularly important for determining the direction of expansion of cells. This has been well known in plant developmental biology for many years (Ziegenspeck, 1948; Frey-Wyssling & Mühlethaler, 1965).
In herbaceous plants and in all seedlings the outer epidermal cell wall is of the utmost importance, as it is the plant/environment phase boundary. It can control both the size of the plant body (girth of stems, surface area of leaves, etc.) and its shape (rate of anisotropy in growth). The relation between cellulose fibril orientation in the outer epidermal wall and morphogenesis of the plant body has been emphasized during the last decade (Green & Selker, 1991; Hernandez & Green, 1993; Green et al., 1996). In current treatises on the molecular basis of plant development this is still a matter of debate (Howell, 1998).
The classic way of determining the orientation of cellulose fibrils in the wall is by polarized light microscopy. This method is limited to very thin preparations containing only one cell layer and cannot be used for preparations containing many cell layers or whole plant organs.
Staining of the cell walls and measurement of absorption- or fluorescence dichroism offers an alternative method that has been applied to some cell systems (Ziegenspeck, 1948; Isings, 1966; Herth & Schnepf, 1980). In brief, fluorophores preferentially absorb light with an orientation that lies parallel to their dipole moment. The fluorophore Congo Red typically binds to β 1-4 glucans (Wood, 1980). Binding on cellulose fibrils occurs in such a manner that its average dipole moment lies parallel to the long axis of the fibril. Absorption of green light (and accordingly their red colour) is strongest in light that vibrates parallel to the longitudinal axis of the cellulose fibril. Therefore, the staining of cellulose fibrils in cell walls with Congo Red is strongly dichroic, provided of course that the cellulose has a preferred orientation (Roelofson, 1959). This principle of fluorescence dichroism of stained cell walls, combined with the technical possibilities of a confocal laser scanning microscope, led to the development of a method to study cellulose fibril orientation in walls of cultured plant cells (Verbelen & Stickens, 1995). We have further explored the limits of this approach using different tissues and different plant species and demonstrate the very broad area of applicability of the method.
Materials and methods
Different tissues of the following plant species were used for analysis: Arabidopsis thaliana L., Allium cepa L., Valerianella locusta L., Cucumis sativus L. and Zea mais L.
The method used to visualize the orientation of cellulose fibrils is based on the fluorescence dichroism of Congo Red. Isolated tissues or whole organs were incubated for 30 min in a 1% solution of Congo Red (Merck CI22120) in water. Because the epidermis of aerial parts of the plant in general does not readily stain, owing to the presence of a very hydrophobic cuticle, pieces of stem or leaf were treated with a pectinase (Serva) solution (0.4% w/v in water) for 20 min. This enzyme dissolves the pectin layer generally present between the cell wall and the cuticle. By shaking the tissue after the incubation, patches of epidermal cells were freed from their cuticle and could be stained with Congo Red. This was done for Arabidopsis stems. Leaf primordia of A. thaliana, dissected but still attached to the stem, can be stained without the use of pectinase as their cuticle is not yet very well developed. A second method consisted of just peeling off the epidermis and letting it float on the Congo Red solution, as was done for Valerionella and Allium. The dye entered the cell wall compartment via the inner periclinal wall and the anticlinal walls of the cells and finally also stained the outer periclinal wall.
The dyed samples were rinsed with water and studied with a confocal microscope consisting of a Biorad MRC600 mounted on a Zeiss Axioskop upright microscope equipped with a coaxial rotating table. The distribution of Congo Red fluorescence in the wall was validated on cells of the onion bulb scale epidermis using the X–Z scanning mode using a high magnification objective (40 ×, NA 0.9) and a very narrow pinhole to assure a good resolution along the Z-axis. For all X–Y scanning modes we opted to collect thick optical sections (about 10 μm). This was achieved by using a low magnification objective (10 ×, NA 0.3 or 20 ×, NA 0.4) and adjusting the pinhole adequately. For all observations the 514 nm line of the laser source was used. A polarizing filter was inserted into the exciting beam such that the vector of the exciting beam was vertical on the microscope monitor.
The fluorescence of a cell wall is at a maximum when the vector is parallel to the predominant cellulose fibril orientation; it is at a minimum with the vector perpendicular to the fibrils. These orientations could be determined by rotating the microscope table. Walls with a totally random orientation of the cellulose fibrils have no predominant orientation and their fluorescence intensity does not depend on their orientation with respect to the vector of the exciting laser beam. It turns out that in most elongated plant cells either the maximum or the minimum fluorescence position always coincides with the longitudinal axis of the cell. Therefore the preferential orientation of the cellulose fibrils in a cell wall can also be expressed in the form of the axiality ratio. This is the ratio between the fluorescence intensity of the cell wall with the vector of the excitation light parallel to the long axis of the cell and the same value registered with the vector perpendicular to the cell axis. The axiality ratio will be unity for walls with a random cellulose orientation, greater than unity for walls with a mean cellulose orientation parallel to the long axis, and less than unity for a mean cellulose orientation perpendicular to the long axis. The mean pixel intensity (standard output of the confocal microscope) is taken as a measure of fluorescence intensity. The microscope settings have to be set to exclude maximum pixel intensity values, and for a meaningful comparison should not be changed during the observation of samples. For a detailed description of the method see Verbelen & Stickens (1995).
First the approach was validated on cells of the onion bulb scale epidermis using the X–Z scanning mode (Fig. 1). Clearly, the fluorochromic signal is distributed throughout the whole thickness of the cell wall. As a consequence, the Congo Red signal picked up taking thick optical sections in the X–Y plane is the sum of the signals that are produced by all possible layers in the cell wall; it is to be considered the net preferential orientation of cellulose fibrils for the whole thickness of the cell wall. The results of the X–Y scans on the onion bulb scale epidermis are shown on the two insets in Fig. 1. The cells have a net orientation of the microfibrils perpendicular to the longitudinal axis of the cell (see below for further explanation). If whole organs are studied, our observations are limited to the outermost cell wall layer of the plant, the outer periclinal cell wall of the epidermis.
In all of the following images, the orientation of the polarization is indicated by the arrows. Fluorescence is false colour coded for intensity, as shown by the colour scale. Red and white show the highest fluorescence intensity and blue shows the lowest.
Leaf primordia of A. thaliana were stained without pectinase treatment. Figure 2(a) is a transmission micrograph of one such leaf primordium. Figures 2(b) and (c) are the related confocal images. The higher fluorescence intensity seen in Fig. 2(c) indicates that this dome-shaped organ has in its outer epidermal cell wall a net orientation perpendicular to the axis of future longitudinal growth.
Leaves of different age were sampled from lamb's lettuce plants (V. locusta). In this case the upper epidermis was peeled off and stained. Figure 3 shows an adaxial epidermal peel from a very small, young leaf and Fig. 4 that from an old (fully developed) leaf. Two types of cellular organization are present: first there are the cells covering the mesophyll. They have very sinuous anticlinal walls giving the cells a typical jigsaw puzzle shape. In every state of development these cells show no preferential orientation of the cellulose fibrils, allowing growth in a random manner. By contast, the cells covering the minor veins show a net transverse orientation of the cellulose fibrils with respect to the major axis of the cell in young leaves (highest fluorescence in Fig. 3b), but this orientation changes to longitudinal in old leaves (highest fluorescence in Fig. 4a). When translated in calculated axiality ratios, the leaf data demonstrate the potential of the method presented. The first type of cell (with sinuous border) always has an axiality ratio of about 1 (areas of measurement not shown). The cells covering the veins, however, have a ratio of 0.86 in young leaves (see boxes on Fig. 3) and of 1.85 in old leaves (see boxes on Fig. 4). In other words, the transverse net orientation of young cells is quantitatively weaker than the longitudinal net orientation in adult cells. The respective mean pixel densities for the boxed areas are shown in the figures.
The hypocotyl of Arabidopsis was treated with a pectinase solution to remove the cuticle. The result of this approach is illustrated in Fig. 5. Parts of the hypocotyl are stained. The microfibrils here clearly show a net longitudinal orientation with respect to the axis of growth.
Roots can be stained without any prior treatment. Whole roots of Arabidopsis and Z.mais seedlings were assayed. Typically the Arabidopsis root tip has a preferential orientation perpendicular to the long axis of the root (Fig. 6). This region is about 300 μm long, but in general this depends on different parameters like age, growth medium and growth rate. Older parts of the roots show a more random orientation. Roots of maize seedlings, by contrast, exhibit a longitudinal orientation over the whole length of the organ, from the very tip (Fig. 7). Root hairs show a longitudinal orientation except at the very tip, where no preferential orientation was detected (results not shown).
Internal tissues of the plant can also be assayed. The orientation of cellulose in a dissected spiral tracheary element of cucumber fruits is illustrated in Fig. 8. Microfibrils are parallel to the rings/spirals of the secondary wall.
Traditionally, polarization microscopy of thin sections or of isolated cell walls is used to study cellulose orientation. The possibilities for research of wall structure offered by fluorescence dichroism were already recognized by Ziegenspeck (1948) and confirmed more recently by Herth & Schnepf (1980). In a previous paper (Verbelen & Stickens, 1995) we reported how we could detect and define random orientation and parallel orientation of cellulose microfibrils with Congo Red and polarization confocal microscopy in cell walls of a very simple system: single cultured cells of tobacco. A detailed evaluation and discussion of the technique are found therein.
Here we have shown how the technique can be adapted for use on whole plant organs. The data on the leaf primordium fit perfectly with the data obtained by Green (1985) with polarization microscopy of surgically removed outer periclinal walls of primordium-epidermis cells. The present method is much less time consuming and in priciple works on whole plants, or at least intact organs. The data on hypocotyl of Arabidopsis underline these two chracteristics of the method and confirm results from more laborious methods on other stems (Niklas & Paolillo, 1998).
It is interesting that, in root tips, fundamental differences were found in the organization of the outer epidermal cell wall between maize (a monocotyledonous plant) on one hand and Arabidopsis (a dicotyledonous plant) on the other hand. Also, the fundamental difference in organization of the outer epidermal wall between cells covering the veins and cells covering the mesophyll in a dicotyledonous leaf is striking. Both these findings fit with our current research on the mechanics of specific parts of plant morphogenesis.
In our laboratory the method described here has already proved to be useful in the definition of Arabidopsis mutants (results not shown) and we are confident that many more uses will appear in the future, especially as whole plant organs can be probed. The set-up of conventional confocal microscopes fits this purpose perfectly: thick samples can easily be handled for observation and all manufacturers offer the necessary wavelengths of light in the standard outfit of the microscope.
The authors acknowledge the support of the Research Programme of the Fund for Scientific Research — Flanders (Project G0034.79) for the availability of the confocal microscope. S.K. is a recipient of a PhD grant from the Flemisch Institute for the Promotion of Scientific and Technological Research in Industry (IWT).