The cell wall is important biologically for four reasons: (1) Since it completely surrounds almost all plant cells, materials entering or leaving cells must pass through it. (2) As one of the main differentiating cellular elements it determines the morphology and to some extent the functions of the cell. (3) Since it is present early in the life of a cell and develops with it. it is possible that it carries with it some kind of developmental record. (4) Since it forms the limiting envelope of the cell it may be directly involved in regulating cell expansion. A study of the structure and properties of the cell wall, therefore, may give information about the nature of the growth process.

There are also technological reasons for knowing something about cell wall structure. The properties of timber, and of pulp and paper, are dependent to a very great extent on the molecular structure of the plant cell wall, and an understanding of wall structure might point the way to technological improvements.

As might be expected, a tremendous effort has been put into the study of the structure and growth of the cell wall over the years; but one of the surprising things that comes out of a consideration of the literature is how little is finally resolved about fundamental processes of wall formation and wall growth. In this paper, an attempt will be made to review the present state of our knowledge on these topics and to point to some of the fundamental problems which are so far unsolved.

Cell Wall Constituents

The principal constitutents of the cell are cellulose, hemicellulose, pectic substances, lignin and proteins. Waxes, together with cutin, suberin and sporopollenin are also found.

There is another class of substances (gums, tannins, colouring matter, etc.) whose presence becomes noticeable as sapwood passes into heartwood. They are not regarded as normal constituents of the cell wall, as they generally appear after the death of the protoplasm. Technologically they are important, however, because they render heartwood less susceptible to attack by fungi, and insects.

Some cell walls also contain mineral deposits, e.g. calcium carbonate, and silicates.

Cellulose is the skeletal substance of the cell wall, and is the most aboundant substance in the plant kingdom. It is a polymer of B- d- glucose residues joined in long chains by 1-4 links (Figs. 1a and b). Over certain parts of their length these chains lie parallel to each other and are very regularly spaced, so as to form long crystalline microfibrils. A microfibril may be thought of as having a highly crystalline core (70Å × 30Å in cross-section) surrounded by a para-crystalline region, which makes the overall dimensions about 100Å × 50Å. About 80 molecular chains occupy the cross-section of the core, the width and breadth of which is generally about 70Å × 30Å. In some species (Valonia and Cladophora, for example), the crystalline core is much bigger in section (200Å × 100Å). Around the outside of the crystalline core there are one or two layers of chain molecules, the crystallinity of which is upset by the incorporation of chain compounds other than glucans. A cross-section of a microfibril is shown in Fig. 2. The appearance of microfibrils on the inner wall of a Nitella cell is shown in Fig. 3 (a) and the appearance of individual microfibrils in Fig. 3 (b).

Fig. 1: Cellulose chain composed of glucose units.

The hemicelluloses are a class of substances which dissolve in alkali. but are not soluble (or are only slightly soluble) in water. On hydrolysis they yield mainly d-xylose. d-galactose, d-mannose, l-arabinose and l-rhamnose. In contrast to cellulose, the hemicelluloses are generally not crystalline in their natural condition, although they have been found in a crystalline state after extraction (Roelofsen. 1959).

The pectic substances are polymers of galacturonic acid which are found mainly in the middle lamella and in primary walls. They are probably amorphous in the concentrations in which they are usually found, although crystalline regions have been found in the walls of certain collenchyma cells which have a high pectin concentration (Roelofsen and Kreger, 1951).

Lignin is an amorphous substance which occurs as an incrustation between cellulose microfibrils. The concentration is highest in the page 117

Fig. 2: Diagrammatic representation of a cellulose microfibill in transvo:se section The oblique solid lines are projections of the glucose rings of the cellulose chains. The oblique broken lines are projections of rings of chains containing sugars other than glucose. The central latione represents the crystalline core which is about 70Å × 30Å. (after Preston, 1961).

Fig. 3: (a) Microfibrils on the inner surface of Nitella wall (X 18,000). Note the preferred orientation of the microfibriis in a direction which is parallel to the transverse axis of the cell. (b) Individual Valonia microfibrils (X 80,000).

page 118 middle lamella, and falls off towards the lumen. It is an important structural material in that its presence confers great strength on wood. Lignin is not a specific substance, but is a three-dimensional polymer consisting of various derivatives of phenylpropane. Its chemical structure is, however, not yet completely known. It does not occur alone in the wall but is found only in association with other wall constituents.

The outermost wall of the epidermis of most above-ground organs in land plants is completely covered by a water-repellent layer called the cuticle. It consists of cutin probably with an admixture of wax. Another closely related substance, suberin, occurs mainly in the periderm.

Protein: Although in some preparations much of the protein is undoubtedly derived from cytoplasmic debris, there is no doubt that some part of it occurs in the wall itself (Tripp et al., 1951). Preston (1960) has reviewed the role of polysaccharide-protein complexes in both plant and animal physiology, and points out that although there is no direct biochemical evidence of close association between cellulose and protein in plant cell walls, morphological and structural determinations, together with evidence from parallel conditions in animals (where there is an apparent association) is strongly suggestive that there may be an association in the plant cell. Ginsburg (1961) has investigated the factors which modify the action of chelating agents in dissolving the intercellular cement in plant tissue (pea root tips), and has reached the conclusion that the intercellular cement can be regarded as an oriented gel structure containing protein molecules cross-linked by two types of metallic ion. the metallic cross linkage being chelate in character.

The characteristic feature of cellulose which distinguishes it from the other components is that is normally occurs in the form of crystalline microfibrils, whereas the other components are not normally thought to be in this form. Its structure is such that it can only be elongated by stretching primary valence bonds and by opening valence angles. Treloar (1960) has estimated the modulus of elasticity of cellulose from the strength of the C-C linkages and the C-O-C linkages. The value of 10 × 10⁶ Ib./sq.in. so obtained is in good agreement with the highest value obtained on well-oriented native fibres (14.5 × 10⁶). For comparison, the corresponding figure for steel is 28 × 10⁶ Ib./sq.in. The fact that these cellulose microfibrils are such effective reinforcing cell wall components has great signifiance in any discussion on how the wall deforms during growth and, in particular, in any discussion on the way in which particular arrangements of microfibrils in the wall influence the shapes which cells ultimately acquire. It is fortunate, that because of its crystallinity, it can be studied in great detail by a number of techniques, and such studies have thrown a great deal of light on the nature of the growth process.

Cell Wall Organisation

In discussing cell wall organisation it is usual to distinguish between two types of wall. The wall which surrounds the growing cell while it is increasing in area is the ‘primary’ wall. About the time the cell ceases to increase in area, it starts to lay down a ‘secondary’ wall inside the primary wall. As a rough approximation, one can think of the primary wall as increasing in area, while remaining constant in thickness, and the secondary wall as remaining constant in area while increasing in thickness.

The primary and secondary walls are different in chemical composition and in fine structure, and therefore, have different physical properties. It is the thickened secondary walls which comprise the great bulk of the material we call wood.

The principal constituents of some selected cell walls are given in Table 1.

In the wood of Pinus radiata the percentages of cellulose and lignin are high while the pectic substances are present only in trace quantities, if at all. In the growing cells of the cambium, however, lignin is absent and the percentage of pectic substances is quite high.

A clear distinction between the composition of growing (extending) and mature walls is also shown in the case of cotton hairs.

Because of its ability to form gels, it has been thought in the past that pectin was a necessary component of rapidly expanding cell walls—e.g. the high pectin content of the cambium and of growing cotton hairs. The pectin content of the primary wall of Avena coleoptiles is very low, however, and it seems that in this case the special properties of these walls depend more on the presence of a high proportion of hemicelluloses than on the small amount of pectin. More will be said about this later.

The figures in Table 1 are on a dry weight basis, but the primary wall in particular, in its native state, is highly hydrated. Roelofsen (1959) obtained a figure of 60% for the water content of isolated walls of corn coleoptiles. Figures as high as 90% are often quoted.

The Cell Plate

The cell plate arises in the equatorial plane of the fibril spindle which connects the two daughter nuclei. It usually shows a gradual expansion, by which it finally reaches the cell wall and it is commonly assumed that this is due to new material being deposited along its circumferences. The cell plate is considered to consist partly, or perhaps mainly of pectin, but the chemical evidence for this view is, however, rather meagre. It is usually isotropic between crossed nicols but later it becomes positively birefringent. At this stage the cell plate consists of three layers; namely, the middle lamella and two primary walls. There are two different opinions with regard to the way in which the two primary walls join the wall of the mother cell. One view is that the daughter cells form new and completely continuous primary cells within the envelope of the mother cell. The other is that the primary walls formed on the cell plate become continuous with the existing wall by a process which involves breaking down the old wall to some extent and re-synthesising this local area of wall so that the new and old primary walls are fused. The weight of evidence seems to be in favour of the first view, but it is not proposed to consider this further here.

Organisation of the Primary Wall

(i) Multinet Structure

The arrangement of cellulose microfibrils in some young isodiametric cells is such that they run in all directions within the plane of the wall with no preferred direction; for example, in apical initials of onion roots (Scott et al., 1956) and in the first formed wall of Valonia (Steward and Muhlethaler (1953).

page 121

Fig. 4: Diagrammatic representation of ‘Multinet’ growth In the most recantly laid down layer microfibrils are laid down with an approximately transverse orientation. Microfibrils in layers which were laid down at an earlier stage of cell wall development have become disoriented by deformation of the amorphous matrix in which they are embedded. The microfibril density has also been reduced since the same number have been spread over a greater area of wall.

If the cell is elongated, however, one usually observes a net transverse orientation. Several types of elongated cells were studied by Roelofsen and Houwink (1953, 1954) who carried out an electron microscope examination of the cell wall of Phycomyces sporangiophores, and of growing hairs of Gossypium (cotton), Ceiba and Asclepias, Tradescantia staminal hairs, and root hairs of Zea mays. They were struck with a feature which was common to all of these cells, viz., that on the inner wall there was a compact transverse arrangement of microfibrils while, on the outside, there was a very loose texture with the microfibrils either oriented in an approximately axial direction, or in what appeared to be a completely random manner. They suggested that this difference in the arrangement of the microfibrils could be explained if one made the following assumptions:

(a)	New fibrils are continually secreted by the cytoplasm at the inner face of the wall (apposition growth) and they are oriented at the time of deposition in an approximately transverse direction.
(b)	Since the microfibrils are embedded in an amorphous matrix, microfibrils already present in the wall will be passively reoriented by the matrix, and will tend to assume a more axial orientation as the matrix deforms during cell elongation.

page 122

As the area of the wall increases and as deposition of new transverse fibrils at the inner surface keeps pace with it, any given group of microfibrils will: (i) tend to assume an axial orientation, (ii) appear to migrate towards the outside of the wall, and (ii) suffer a reduction in fibril density. These changes are shown in diagrammatic form in Fig. 4. The appearance of the Nitella inner surface is shown in Fig. 3a, and the outer surface in Fig. 5.

Because the outside of the wall looked like a loosely meshed fishing net they called this ‘multinet’ growth. It seems to provide a satisfactory explanation of the microfibrillar morphology of elongating walls in which transverse microfibrils are deposited. The theory applies in its simplest form to cells in which growth is uniform over the entire cell wall; for example, in Nitella (Green, 1954). There appears to be no reason, however, why cells in which localised growth occurs should not be regarded as special cases of uniform growth, and, indeed, it was in the attempt to explain wall morphology in cells in which growth is localised (root hairs, sporangiophores, etc.) that the theory was first developed.

The multinet theory does not, however, explain the presence of well defined and well oriented longitudinal bands of microfibrils, which occur on the outside of many cells; for example, parenchyma cells of oat coleoptiles (Wardrop and Cronshaw, 1958), collenchyma cells (Beer and Setterfield, 1958), cortical parenchyma cells of bean stems (Probine, 1963), etc. These bands have often been classified as secondary thickening, but they appear to be present during elongation in even very young cells. What is very puzzling, however, is that they occur on the outer wall of the cell, remote from the cytoplasm.

Roelofsen (1958) has attempted to explain these longitudinal bands by an extension of the multinet theory and by invoking a purely physical process. It does not, however, adequately explain all features relating to these bands.

Although the presence of these longitudinal bands in some cells is not satisfactorily explained, there is little doubt that the multinet theory accounts for most features of the microfibrillar morphology of cell walls with ‘transverse’ synthesis. It has nothing to say, however, about the mechanism by which the microfibrils come to be oriented transversely at the inner surface of the wall in the first instance, or anything about fundamental mechanism of wall extension. These topics will be treated in a later section.

(ii) ‘Crossed-Fibril’ Structure

In the almost spherical cell, Valonia (an alga), the first formed wall has no preferred direction of microfibril orientation—the microfibrils are completely random in the plane of the wall.

page 123

Fig. 5: Disordered microfibrils on the outer surface of a cell — typical ‘multinet’ appearance. (X 21,600).

Fig. 6: Electronmicrograph of microfibrils on the inner wall of Cladophora rupesfris (X 24,000). — also typical of Valonia. The cell axis lies along the long edge of the paper. (reproduced from Frei & Preston, 1961).

page 124

Transitional lamellae develop, however, which show successively less scatter of microfibrils until, finally, a type of structure is developed which is completely different from ‘multinet’ structure referred to above. The cellulose microfibrils in Valonia ventricosa, for example, are present in three orientations, in separate lamellae. The two ‘major’ directions (A and B) lie on an average at rather less than at right angles to each other; the third orientation (x), which is much less frequent, forms a bisector of this angle. The repeat from one lamellae to the next can be something like ABABAxBABAxBABAxBABABAB. Within each lamella the alignment of the microfibrils is very perfect (Fig. 6).

The Valonia cell remains roughly spherical in shape as it grows so that the microfibrils undergo no major oreorientation as do the transverse fibrils in elongating cells which give rise to ‘Multinet Structure’. Valonia does show another type of wall change during growth, however. Steward and Muhlethaler (1953) have reported that growing sporelings, about sixteen hours old, show a complete primary wall. With further cell growth this wall is stretched and finally torn, so that fragments of this first-formed wall may be seen as patches on the lamellated wall which follows it. The outermost layers of the lamellated wall also tear, but in this case the tears are along lines parallel to the microfibrils. The spaces which open up are not filled with new microfibrils. This is called ‘tearing growth’. It is not much in evidence in walls of higher plants—probably because they are much less crystalline and the microfibrils can slip over and past one another more easily in the gel-like amorphous matrix.

This type of structure is not confined to spherical cells, however. A rather similar structure is also formed by the filamentous green algae Cladophora and Chaetomorpha. In Chaetomorpha growth is entirely intercalary and elongation is confined to the part of cell nearer the base of the filament. The outermost lamellae may be seen to be torn and roll back as flakes forming a collar, which arises from this rupture of outer lamellae in the region of growth. In Cladophora growth is confined to part of a cell nearer the apex of the filament but the lamellae do not roll away—possibly because the amorphous matrix contains pectic compounds in the outer region and this may well make the wall more extensible. Frei and Preston (1961) have found, however, that lamellae from different depths in the wall differ markedly in structure. These differences are of the kind to be expected if the wall has been passively extended during growth, i.e. microfibrils which were laid down a little off axis become more nearly longitudinal and are very straight; those with almost transverse orientation become less nearly parallel to each other.

The microfibrils in these ‘crossed-fibril’ walls therefore, undergo translation and rotation as a consequence of cell extension, in general harmony with the ‘multinet growth’ hypothesis.

page 125

Fig. 7: Diagram of fibril orientation.

Organisation of the Secondary Wall

In presenting this brief picture of secondary wall organisation the main methods of investigation will be indicated. The conifer tracheid will be singled out for special attention because it is very important technologically.

1. When conifer tracheids are viewed in transverse section in a polarising microscope it is apparent that the secondary wall consists of three layers distinguished by their different optical properties. They are usually indicated by the labels S₁, S₂ and S₃. The S₁ layer, which is nearest the primary wall, and the S₃ layer, which is nearest the lumen, are usually rather thin. The central, or S₂ layer is variable in thickness, being thin in early (spring) wood and thick in late (summer) wood.

The polarising microscope observations may be interpreted as meaning that the orientation of the cellulose microfibrils in the S₁ and S₂ layers trace out a flat helix around the wall. In the central layer (S₂) the microfibrils trace out a steep helix around the wall (Bailey and Kerr, 1935). This interpretation was confirmed by Wardrop and Preston (1947) using a more sophisticated optical analysis. A diagram showing wall structure in Pinus radiata tracheids is shown in Fig. 7 (this anticipates a different interpretation of the S₁ layer, which was discovered using the electron microscope).

It was discovered by Preston (1934, 1948) that there was a systematic variation of fibrillar helix angle with cell length. In long page 126 tracheids the microfibrils in the S₂ layer had a more nearly axia orientation than they had in short tracheids. Preston and Wardrop (1949) have shown that this sort of relation holds for the S₃ layer also. Expressed analytically:

L = A + cot theta

where theta = helix angle

L = tracheid length

A and B are constants

A curve for a sample of Pinus radiata is shown in Fig. 8. The dependence of anisotropy of cell structure on cell length has some significance from the point of view of anistropy of wood properties and will be referred to again later.

Fig. 8: Variation of microfibrillar helix angle theta with cell length in Pinus radiata (theta is defined in Fig. 7).

Fig. 9: X-ray diffraction diagrams of Pinus radiata late wood. Ninth annual ring (left) and third annual ring (right).

2. The technique ofX-ray diffraction has also been used to investigate the wall structure of conifer tracheids (Preston and Wardrop, 1949). Two typical diagrams of Pinus radiata (late wood) reproduced in Figs. 9a and b. They are from the third and ninth annual rings, respectively, of a thirteen-year-old tree. The beam was directed through the wall in the radial direction. The helix angle can be estimated from the distribution of X-ray intensity around the 002 arcs, and for the examples given are 11° and 32°, respectively. Because the middle layer of late wood is normally thicker than the other two layers, its structure dominates the X-ray diagrams.

It is not obvious from the X-ray method that there is a three-layered wall structure. The method does, however, provide additional information:

3. The electron microscope provides a third and newer method of investigating cell wall structure. With the advent of this method, microfibrils could be seen for the first time and the pattern of wall organisation determined in much greater detail. The structure of conifer tracheids determined by other methods was largely confirmed. The S₁ layer, however, which lies closest to the primary wall, was found to have a more complex structure than had been thought previously. It has a crossed fibrillar arrangement rather than a single helix of microfibrils (Wardrop, 1957). The sign of the helix changes in alternate layers. The angle between the microfibrils in adjoining layers is about 100° and the microfibrils in each layer make an angle of about 50° with the cell axis. The fibrils are not equally shared between the two directions—one direction dominates the other. This accounts for the apparently misleading evidence of the polarising microscope.

The S₂ and S₃ layers are also laminated, but do not possess a crossed fibrillar arrangement. The signs of the helices are:

Harada et al. (1958) consider that in addition to the S₁, S₂ and S₃ layers there are a number of lamellae of intermediate orientation between the layers S₁ and S₂ and between S₂ and S₃, but there seems little doubt that they are few in number (Harada et al., 1958). Wardrop (private communication) pictures the wall as being organised as shown in Fig. 10.

It has been considered by many people that there is another wall layer inside the S₃ layer which is so different in properties as to deserve the name,Tertiary cell wall. Very little clue as to its nature has been found from staining reactions, but it is resistant to a wide range of reagents. It is covered with ‘warts’.

The warty layer has been fully described by Wardrop et al. (1959). They consider that the first indication of the formation of a wart is a localised withdrawal of the plasmalemma from the cell wall and the formation of localised thickenings on the wall. In most cases the indentation of the plasmalemma could be caused by localised wall thickening—but cause and effect here are difficult to separate. Inside each wart. at the tip, there is a small spherical body. This is assumed to be the denatured remains of an organelle which is enclosed by the tonoplast and plasmalemma as the sytoplasm dries on the surface of the lumen. Since each wart seems to have a small spherical body associated with it, it is interesting to speculate on the association between this organelle and the cell wall thickening.

Fig. 10: Arrangement of microfibrils in the various wall lamellae of Pinus radiata showing primary wall and transition lamellae (modified from Wardrop—private communication). Compare Fig. 7.

The warty layer can be cross-linked with formalin and when the cellulose is removed with sulphuric acid the layer can be recovered in beautiful sheets. When stripped in this way the warts are often seen in a systematic array as though arranged in a series of helices—this may give some clue to the mechanism of microfibrillar arrangement.

Wardrop (private communication) does not, however, regard it as a ‘Tertiary’ wall layer, and considers it, as was implied above, to be the remains of the plasmalemma and tonoplast.

Returning to the consideration of the secondary wall proper, it is reasonable to assume that the almost axial arrangement of microfibrils in the S₂ layer in tracheids of Pinus radiata would result in an anistropy of mechanical properties. This will not be the only factor modifying properties—e.g. anatomy of the specimen and the nature of intercellular adhesion must also be taken into account. Nevertheless, cell wall organisation is important. This was shown by Wardrop (1951) who measured the variation in breaking load in tension of tangential longitudinal sections of Pinus radiata. He showed that there was a close parallelism between tracheid length (and hence of microfibrillar angle) and breaking load in tension. The longer the tracheids, or the more nearly axial the arrangement of fibrils, the higher the breaking load.

In the case of single fibres, the dependence of mechanical properties on structure is, perhaps, even more striking than in the case of wood because one is dealing with a single cell, and not a group of cells aggregated to form a tissue. Preston (undated) has reported for sisal the figures given in Table 2.

It is generally considered that extension growth of cell walls ceases at, or about, the time that the secondary wall begins to be laid down. Wardrop (private communication) suggests that the tip of differentiating fibres may still be in the extension growth stage after the deposition of the secondary wall has begun. Electron microscope and autoradiographic evidence suggests that the change in microfibril orientation during enlargement of the primary wall is consistent with Roelofsen's multinet hypothesis. Other investigations have shown that the secondary wall is formed by the deposition of successive lamellae of cellulose microfibrils. It does not appear, however, that apposition occurs uniformly over the cell wall. Wardrop has examined developing fibres using the interference microscope and found that the wall is thicker near the centre of the cell than at the ends.

He therefore imagines that secondary wall lamellae are initiated near the centre of the cells and grow in the direction of microfibril orientation towards the tips. This means that the tip may still be in the extension stage after deposition of the secondary wall has already begun at the centre of the cell.

Extension Growth of the Primary Wall

(e) ‘Plastic’ Extension

We now come to consider the evidence for the plastic-extension type of mechanism. The initial suggestion that plastic stretching of the wall by turgor pressure was the primary mechanism of extension growth, was made by Heyn (1955). He measured the elastic behaviour of coleoptiles severed at their bases and placed in humid chambers. Cutting and application of hormone took place an hour and a half after decapitation. Subsequent bending under a weight page 134 showed increased plasticity of coleoptiles treated with growth substances as compared with untreated controls. Permanent curvature resulted from bending treated plants only. Heyn also placed severed coleoptiles in water, thus allowing the walls to be stretched by turgor pressure only, and found that plants provided with growth substances underwent considerably more elongation in comparison with untreated plants. Contraction on plasmolysis was about the same for treated and untreated showing that the greater extension of treated plants was permanent. This permanent elongation of treated plants took place in water at 1°C. at which temperature increase of cell wall substance could be assumed to be interrupted. This was taken to indicate that elongation is dependent on the plasticity of the wall as the primary factor.

More recently, Professor Preston and the author have re-examined this relationship between wall properties and growth (Probine and Preston, 1961, 1962, and Probine, 1963) and the findings of this work are summarised below. The internodal cells of Nitella were used since they are most convenient objects for the study of wall structure and growth. They are so large that the physical properties of the wall can be measured with comparative ease. Further, the wall grows uniformly over the whole wall surface as the cell

Fig. 11 : Load/extension curves of strips of Nitella cell wall cut either along the cell axis (longitudinal strip) or at right angles to it (transverse strip).

page 135 elongates. Averaged over the whole thickness of the wall, the microfibrils have a perferred orientation in the transverse direction (multinet structure). The strength of the wall can be readily determined for strips of wall cut with their length at any required angle to the cell axis. A typical example of load extension curves for (a) a longitudinal strip and (b) a transverse strip, each stretched in the direction of its length, is given in Fig. 11. It can be seen that a load of, say, 6 units, causes a longitudinal strip to elongate almost 5 times as much as a transverse strip. It follows, therefore, that the relative resistance to stretching is less in the longitudinal direction than in the transverse. This is clearly associated with the fact that there are fewer microfibrils reinforcing the longitudinal direction. It follows, therefore, that under internal turgor pressure the cell will expand preferentially in the longitudinal direction which incidentally is the direction of maximum growth rate.

It is not clear, however, that this has anything to do with growth, as the extensions concerned are static extensions. When the phenomenon of plastic flow is looked at, however, the situation is even more significant. When a strip of wall cut parallel to the transverse axis is loaded it stretches slightly, but no matter how long the load is left applied no further change in length occurs. When a strip cut parallel to the longitudinal axis is loaded, however, it continues to get longer and longer — it exhibits the phenomenon of plastic flow.

In order to attempt an assessment of any connection between plastic flow and wall growth, the rate of flow was measured on wall strips cut from a series of cells, the growth rates of which were known at the time of cutting. A strip cut from a cell which had grown 16% in length in 24 hours increased its length by plastic flow to the extent of about 13% during the 100 minutes that a load of 0.8 units was applied. On the other hand, a strip cut from a cell had grown in length only 1% in 24 hours, increased its length by plastic flow to the extent of only 1% under the same load. See Fig. 12 for complete results.

These curves should not be taken as literal statements of the behaviour of the material in situ, in the cell, since the load is applied uniaxially in the test specimens, but multiaxially in the cell. Nevertheless, they do show a difference in physical properties between a growing and a non-growing wall, which is consistent with the idea that plastic stretching of the wall by turgor pressure might be involved in the growth process. Further, the direction in which wall plasticity is greatest, is the direction in which growth rate of the cell is greatest. Since wall plasticity in different directions (and, if the plastic flow hypothesis is correct, cell growth) is determined by the arrangement of the reinforcing microfibrils in the wall, it could well be that it is the pattern in which microfibrils are laid down which determines the shape which a cell will ultimately page 136 acquire (for cells which grow uniformly over their length). It would appear that the direction of microfibrils is closely controlled by the protoplasm as part of genetically determined differentiation, and therefore, through this mechanism, cell shape may be determined.

The idea that growth is indeed controlled by wall properties is reinforced by the fact that spiral growth in Nitella can be accounted for by wall structure (Probine, 1963). Further, it has been found possible to modify wall structure and thereby to produce changes in cell shape (Probine, 1963 (b)). There is not space, however, to discuss this here.

(f) The Pectin Hypothesis

If plastic stretching of the wall is indeed involved in growth, it is to the plastic properties of the amorphous matrix we must look for a further understanding of cell wall growth. The cellulose microfibrils will only vary the plasticity of the wall in different directions. The view is generally held in botanical literature that the walls of growing cells contain a high proportion of pectin (polygalacturonic acid partially esterified with methanol). Because of its ability to form gels it was thought it was the pectic matrix which was responsible for the ‘plasticity’ of growing walls.

It has been shown by Ordin, Cleland and Bonner (1957) that esterification of pectin by methyl-derived carbon is an auxin controlled reaction. This suggests that methylation of carboxyl groups of adjacent pectin molecules, under influence of auxin, may be involved in the splitting of anhydride or calcium bridges which contribute to the mechanical properties of the wall. The bond splitting may require methyl esterification as a primary part of the reaction or may require it to stabilise the split once made. Bennet-Clark (1955) has summarised the effect on elastic properties of the sort of mechanism discussed above as follows: ‘It will probably be agreed that the plastic and elastic extensibility of a poly-galacturonic acid, or in general, of an oxidised hemicellulose, will be markedly controlled by the condition of the carboxyl groups. If these long chain molecules are associated with multivalent cations, minimum extensibility will be found owing to electrovalent binding together of adjacent molecules. If the carboxyls are free, hydrogen bonding will provide considerable tensile strength, but much less than that formed in presence of cations and, finally, when or if, they are converted to methyl esters, there will be a minimal tensile strength as hydrogen bonding will be replaced by Van der Waals forces and so extensibility will be maximal.’

A difficulty in the way of the pectin hypothesis is that Jansen et al. (1960) reported that the hot-water-soluble fraction is almost fully esterified, and in any case this is only 20% of the total cell wall uronic acid. It appears, therefore, the number of possible double salt cross-links, in which auxin does promote methylation, is very small indeed, and it is hard to see how these could be critical in cell page 137

Fig. 12 : Plasticity of longitudinal strips of Nitella cell wall plotted as a function of the applied stress. The growth rate of the cell from which the strip was cut is indicated in the legend.

wall growth. Again, Bishop et al. (1958) analysed oat coleoptiles and concluded that they contained less than 1% of polyuronides or pectin (Table 1). It is generally considered that their figure was too low, but it did have the effect of awakening investigators to the possibility that hemicelluloses rather than the polyuronides might be involved.

One may conclude this section by pointing out that the issue of cell wall growth is far from resolved. At this stage it is only possible to conclude that, on balance, it seems likely that the primary mechanism of extension is cell wall plasticity, but the changes in the wall which induce the changes in plasticity, which lead to extension growth, must surely be under metabolic control. The nature of this control is, however, not at all clear.

Conclusions

The arrangement of microfibrils in the plant cell wall is highly specific and complex — particularly in the secondary wall. It seems clear that the direction of microfibril deposition must be controlled by the protoplasm, but the mechanism of this control is not yet understood. There are some clues, however. Probine and Preston (1958) have shown that, in Nitella, there is a highly significant correlation between the direction of protoplasmic streaming and the direction in which microfibrils are laid down in the wall. This is not to say that the microfibrils are actually oriented by streaming; merely that the direction of streaming is an expression of a polar structure in the cytoplasm — in particular, a structural polarity of the periphal cytoplasm adjacent to the cell wall (plasmalemma?).

In this regard, it may be significant that in the first-formed wall in Valonia, microfibrils are laid down at random and it is only after the polarity of the cell is established that an oriented system of microfibrils is established. Again, Frei and Preston (1960) have shown that when Chaetomorpha cells are plasmolysed, microfibrils are laid down at random. When the cell recovers from plasmolysis the normal crossed-fibrillar structure is resumed — after the cytoplasm has regained its original configuration.

That microfibril orientation is related to protoplasmic streaming has often been discounted on the grounds that the direction of streaming and the direction of microfibrils do not coincide. There seems to be no difficulty here, however. There is abundant evidence that microfibrils can be laid down by the cytoplasm in at least three directions relative to some reference direction; for example. Valonia. Frei and Preston (loc. cit.) have pointed out that the wall patterns of Valonia and four species of Cladophora and Chaetomorpha, conform to a general pattern, so that it is formally possible to pass from the wall structure of one plant to another merely by changing the orientation of a rectangle (representing the principal axis of the cell) lying over a common grid (Fig. 13).

It may well be that the three directions in Valonia constitute a basic pattern of microfibril synthesis. In a given type of cell, it may be that microfibrils are laid down in only one of these directions giving, say, transverse synthesis which leads to multinet structure in many primary walls. In another type of cell, it may be that two of the possible directions are selected giving the ‘crossed-fibril’ type of structure. In a few cells, e.g. Valonia, all three are selected. The pattern may not only vary from species to species, however, but may also vary in the same cell stages of differentiation (as in the S₁, S₂ and S₃ layers of the conifer tracheids).

This generalisation, however, leaves the details of the mechanism far from understood. Perhaps the next advance will come from an understanding of the surface properties of the plasmalemma.

Regarding the mechanism of extension growth, much more remains to be done. A great deal of evidence points to wall plasticity being fundamental to extension, but the way in which this comes under metabolic control is far from clear. For example: how is plasticity controlled in those cells in which growth is localised?

If wall plasticity is indeed fundamental to extension growth, then the arrangement of microfibrils in the wall (which determines its plasticity in different directions) will set the pattern of wall extension, and, hence, determine the ultimate shape which the cell will acquire. This, clearly, has morphogenetic implications. There is no doubt that, in general, there is a correspondence between cell wall structure and cell shape, and this has been recognised in the form of an often quoted generalisation, viz., microfibrils tend to be laid down at right page 139

Fig. 13: Diagram representing the relationships between the structure of the wall in Chaetomorpha, Cladophora and Valonia. The two major orientations of microfibrils correspond to the two sets of double lines. The third orientation is represented by single lines. The longer edge of each rectangle is taken parallel to the axis of the corresponding cell. a, Chaetomorpha melagonium. b, Cladophora rupestris. c, Chaetomorpha princeps. d, Cladophora prolifera. e, Valonia ventricosa. (after Frei & Preston, 1961).

angles to the direction of growth (which is the inverse way of acknowledging the correspondence we have noted).

Microfibrillar morphology is, of course, not the only factor determining cell shape. For example, Valonia and Cladophora have a similar wall structure but one is nearly spherical and the other is nearly cylindrical. The difference is not one of wall structure, but of localisation of growth. Valonia grows uniformly over its surface and, as one would expect from its structure, grows into a sphere. In Cladophora growth is localised at the apex of the cell so that it grows into a cylinder. Again, in a tissue, while cell shape may be more or less dependent on wall structure, it will also depend to some extent on interactions with neighbouring cells.

Although there is evidence against ‘active’ growth of the wall being the primary cause of extension, there is nevertheless a parallelism between extension growth and synthesis, which cannot be ignored. For example, while wall extension and synthesis are not directly related in Nitella (Green, 1958), they do follow one another sufficiently closely to make one feel that they cannot be entirely unrelated. The mechanism of this interrelation is, however, obscure. The number of cases in which the plant seems to be able to compensate for increased mechanical stress, for example, suggests that a ‘feedback’ mechanism may be involved.

It was pointed out in the discussion of Table 1 that the chemical constitution of primary and secondary walls differed greatly. The page 140 detailed chemical changes which take place in the wall and in the cell cell during each stage of differentiation are also important to an understanding of wall formation and growth. There has not been space to review this here and the reader is referred to a paper by Jensen (1961) who has combined histochemical analysis of thin sections with cytochemical localisation procedures.

A number of workers have studied cytoplasmic organisation in differentiating cells in the hope of throwing additional light on the problem of cell wall differentiation, but the results so far have been disappointing.

Perhaps the best way to close is to say: Much has been done; little is yet really understood; much more remains to be done. The result of this future work should be rewarding.