Planta
Planta (1985)165:158-169
9 Springer-Verlag 1985
The developmental morphology and growth dynamics of the tobacco leaf R.S. Poethig* and I.M. Sussex Department of Biology, Yale University, New Haven, CT 06511, USA
Abstract. The developmental morphology and growth dynamics of the leaf of Nicotiana tabacum L. cv. Xanthi Nc. are described. Epidermal and internal cell patterns indicate that the leaf axis arises from approx. 100 cells in four layers of the shoot apex, while the lamina arises from several rows of cells in each of three layers of the leaf axis. Cell patterns at the apex and margin of the leaf do not support the classical view that these regions have a specialized meristematic function. Instead the development of the leaf appears to be largely dependent on intercalary growth. The pattern of growth within the lamina is surprisingly complex. In addition to a proximal-distal gradient in the duration of growth and cell division during development, localized transitory changes in the rate of these processes also occur. These observations are discussed in reference to previous discriptions of leaf development in tobacco. Key words: Cell division - Growth (leaves) - Leaf development - Nicotiana (leaf growth).
Introduction Research on the cellular basis of leaf morphogenesis has usually focused on two issues: the number of initial cells that give rise to the leaf, and the role of these cells in controlling leaf shape. A prominent, though by no means general, view in the classical literature (see review by Foster 1936) * Present address: University of Pennsylvania, Department of
Biology, Leidy Laboratories/G7, Philadelphia, PA 19104, USA S y m b o l s : R A= relative rate of growth in area; RE = relative rate
of growth in length; Rv = relative rate of growth in volume
is that leaves arise from a small group of initial cells situated at the tip of the leaf axis, and that the lamina arises from one or two rows of cells along the lateral margins of the leaf axis. These marginal and submarginal ceils constitute what is known as the marginal meristem, and are thought to control the shape of the lamina. However, recent studies do not support these conclusions. In the case of tobacco, for example, analyses of periclinal chimeras have shown that both the leaf axis and lamina are derived from at least three layers of cells (Burk et al. 1964; Dulieu 1968, 1970; Stewart and Burk 1970). Other types of genetic mosaics indicate that there may be 16 or more initial cells in each layer of the axis promordium (Dulieu 1968, 1970; Stewart and Dermen 1975), and several rows of cells in each layer of the lamina primordium (Dulieu 1968). Unfortunately there is no consensus about the behavior of these initial cells. Stewart and Dermen (1975) imply that leaf initials reside in an apical meristem during the extensinon of the axis, while Dulieu (1968) believes that each initial forms a unique longitudinal section of the axis, such that only initial cells in the very center of the leaf primordum contribute to the leaf apex (Dulieu 1968). There is more agreement about the behavior of cells at the leaf margin. Careful quantitative studies of the pattern of cell division in Xanthium italicum (Maksymowych and Erickson 1960), Trapaeolum majus (Fuchs 1975; 1976), Jasminum nudiflorum (Thomasson 1970), PauIownia tomentosa (Jeune 1972), Lycopus europaeus (Jeune 1983) and Nicotiana tabacum (Dubuc-Lebreux and Sattler 1980) demonstrate that the frequency of cell division is no higher at the leaf margin than in intercalary regions, and may in fact be significantly lower. Furthermore, the pattern of variegation in periclinal chimeras indicates that marginal cell lineages are highly variable in their behavior,
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf
and may contribute either very little or a great deal to the expansion of the lamina (Stewart and Dermen 1975; Bergann and Bergann 1984). These observations make it clear that traditional interpretations of the behavior and function of marginal cells need to be re-examined. In order to resolve some of the discrepancies between classical and more recent interpretations of leaf development, we decided to undertake a study of leaf development in Nicotiana tabacum L. cv. Xanthi Nc. This cultivar has the advantage of being relatively small in size, and of having genetic markers useful for clonal analysis (Poethig and Sussex 1985). Its leaves are ovate in shape and have a naked petiole, traits controlled by the br allele of the broad locus (van der Veen 1957). Information about the cellular parameters of leaf development in N. tabacum L. cv. Xanthi Nc. was obtained by clonal analysis and is presented in the following paper (Poethig and Sussex 1985). In this paper we introduce the experimental system used for this analysis, and describe the developmental morphology and growth dynamics of the leaf. Material and methods For this study we used a stock of Nicotiana tabacum L. cv. Xanthi Nc. heterozygous for the chlorophyll mutations al and a2 (=yg). Seeds of this stock were obtained from H. Dulieu (Station d'Amelioration des Plantes, I.N.R.A., Dijon, France) who has described the nature of these mutations in detail (Dulieu 1974, 1975; Dulieu and Dalebroux 1975; Singh and Dulieu 1976). The al + a2/+ stock was propagated vegetatively, starting with axillary shoots from nine different plants. Axillary shoots were released from growth arrest by removing the top half of the plant (topping), and were harvested two weeks later. After removing five or six bottom leaves, the axillary shoots were inserted in coarse vermiculite in 0.5-1 containers and were kept under a plastic-covered frame until they formed roots, 7-10 d later. Plants were grown in a controlled-environment room on tables watered automatically with a solution of Hyponex (1 g/l; Hyponex Co., Copley, O., USA). The room was operated at 27 ~ C, 60% relative humidity, and a 16:8 light:dark cycle. Illumination was provided by fluorescent (40 W Lifeline coolwhite; Sylvania Co., Seneca Falls, N.Y., USA) and incandescent (60 W) lamps, at an energy fluence rate of 150 W m -2.
Experimental system. Our observations concern the development of leaves on axillary shoots recently released from growth arrest. Because all axillary buds are arrested at approximately the same stage (six or seven leaves) and grow out rapidly after topping, this system was synchronous, reproducible and fast. Axillary buds were released from growth arrest by topping 15to 20-cm-tall plants, so as to leave five or six mature leaves. As axillary buds appeared, all but the topmost ones were removed. A fresh nutrient solution was added to the reservoir tanks immediately after topping, and was replenished with disfilled water as needed; no additional nutrients were supplied during the course of an experiment.
159
Growth analysis. The earliest phase of leaf growth was reconstructed by measuring the length of successive leaves (leaves 9, 10 and 11, counting from the base of the axillary shoot) in axillary shoots harvested at various times after topping. To construct a growth curve, the average lengths of these three leaves in each set of shoots were placed at 28-h intervals (the length of the plastochron), and sets of points from successive harvests were positioned relative to one another by finding the point at which they overlapped. The smallest leaf primordium (leaf 11 in the youngest set of shoots) was assumed to be no more than 28 h old because the next youngest leaf was not visible in these buds. This leaf primordium was arbitrarily assigned an age of 24 h in order to position the growth curve on a time scale. After leaves reached a length of 5 mm they could be measured non-destructively with a millimeter ruler. The later part of the growth curve in Fig. 14 represents the actual growth of a single leaf, in this case leaf 10. The growth of the lamina was studied by marking a leaf with India ink and then photographing it daily for 8 d. The areas of the regions delimited by the marks were then measured with a planimeter on enlarged photographic prints. The relative growth in area (RA) was calculated using the formula R A= ( l n A i - l n A o ) t -1 where A o is the original area of the region, A 1 is its area 1 d later, and t equals 1 d (Williams 1975, pp. 10~ 12). The relative rate of extension of the axis was calculated in the same way.
Histologicalprocedures. For the purpose of analyzing the distribution of mitotic figures, cell size, and the initial stages of vascular differentiation, leaf primordia were fixed in 3 : 1 (v/v) ethanol:acetic acid, stained according to the Feulgen procedure, dehydrated in ethanol and mounted in Permount (Fisher Scientific, Cincinnati, O., USA) (Jensen 1968). In the case of tobacco, this procedure generally yields considerably background staining and it is necessary to use a green filter to clearly visualize mitotic figures. For detailed analysis, specimens were fixed in 3% gultaraldehyde (sodium-phosphate buffer, pH 7), dehydrated in ethanol, and embedded in glycol methacrylate according to the procedure of Ruddell (1967) or the JB-4 procedure (Polysciences, Warrington, Pa., USA). Sections 1-2 Jam thick were cut with glass knives on a JB-4 microtome (Sorvall Instruments, Du Pont & Co., Wilmington, Del., USA), attached to slides, and stained with periodic acid-Schiff's-hematoxylin (Feder and O'Brien 1968). Sections and whole mounts were examined with a C. Zeiss (Oberkochen, FRG) brightfield microscope. Photomicrographs were taken with Technical Pan film 2415 (Eastman-Kodak, Rochester, N.Y., USA) which was developed for medium contrast (Kodak HC110, dilution F). The mitotic frequency within the plane of the lamina was determined using Feulgen-stained whole mounts. Specimens were examined at 375-fold magnification and the number of mitotic figures (metaphase through anaphase) in different interveinal regions was counted with the aid of an ocular grid. The number of cells in each region was then determined by multiplying the area of the region by the average number of cells per unit area, as determined from six to ten randomly chosen fields. These numbers were then used to calculate the mitotic index (percent cells in mitosis).
Scanning electron microscopy. Observations were made on flesh, uncoated specimens. Shoot apices were dissected just prior to examination and were kept on moist filter paper in small Petri dishes until use. Specimens were examined with an ETEC Autoscan (Perkin-Elmer, Norwalk, Ct., USA) operated at 2.5 kV.
160
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf
Results
Successive leaves on an axillary shoot differ in shape and size. The first six or seven leaves are initiated before growth arrest, and are small and variable in shape at maturity. Leaf size increases up to the ninth through eleventh nodes and then decreases with each successive node. Starting at about the eighth node the leaf blade becomes ovate in shape, that is, broader at the base than at the tip. Because space limitations made it necessary to use successive leaves to reconstruct the development of a single "typical" leaf, it was important to choose a series of leaves whose morphology and growth rate were reasonably similar. Attention was therefore restricted to leaves 9, 10 and 11 (counting from the base of the axillary shoot) because these leaves are essentially identical in shape, size, number of cells per unit area, and growth rate (Table 1).
The morphology of the leaf primordium The shoot apical meristem in tobacco is remarkably flat, and leaves arise within the plane of the apex (Figs. 1, 3). The first evidence at the surface of the shoot of the formation of a leaf is a lateral swelling called the leaf buttress (P1 in Fig. 1). The leaf buttress is about ten cells wide and is located about ten cells from the center of the apex. In longitudinal sections the buttress appears to encompass about four layers of the apical meristem (Fig. 3). The primordium emerges from the apex in the form of a low bulge, slightly flattened toward the shoot apex (P1 in Fig. 2). Immediately after initiation, a boundary forms on the adaxial side of the primordium. However the lateral margins of the primordium continue to expand around the apical
Table 1. Some morphological and developmental parameters of leaves 9, 10 and 11 of Xanthi Nc. tobacco, counting from the base of the axillary shoot. These data represent an average of 20 samples, except for the number of palisade cells/ram 2, which was calculated from five samples. The growth rate of the leaf represents the time it took the leaf to grow from ~ cm to one half its final length. The plastochron is the time between the emergence of successive leaves; for the sake of analysis, emergence was defined to occur at a length of 1 cm Leaf no.
9 10 11
Area (cm 2)
125+2 128___3 130_+3
Length/ width
1.9 1.9 1,9
Palisade cells/mm z
797+__17 795__+19 798_+21
Growth rate (h)
Plastochron (h)
124__+1 28_+1 1 2 6 _ + 1 30_+1 128_+1
dome until the leaf encompasses about one third of the circumference of the shoot. Epidermal cell lineages in young primordia (leaf length = 200 gin) do not emanate radially from the tip of the primordium, as would be expected if the primordium was derived from a small group of apical initials. Along the midline of the primordium, parallel cell files run from the tip to the base of the axis in apparent continuity with cell files in the internode below (P2 in Fig. 1); but towards the lateral margins of the axis the cell pattern is irregular, and cell lineages cannot be traced to the leaf tip (P2 in Fig. 1,
2). The initiation of the lamina The lamina, or leaf blade, arises at the boundary between the adaxial and abaxial faces of the leaf axis, and is first recognizable as a distinct ridge when the primordium is 500 gm long (Fig. 4). Prior to this stage, its point of origin is marked internally by a band of densely-stained cells running from the margins of the primordium through the central provascular strand (Fig. 5). The subsequent vacuolation of cells in the region surrounding the provascular strand leaves a distinctive group of cells at the margin (Fig. 6). This putative group of lamina initials is composed of between ten and twenty cells in the transverse plane of the axis - about eight cells in the epidermis, six in the sub-epidermal layer and one to three cells internal to these layers. Initially, the lamina extends in a smooth curve from the tip to the base of the leaf axis (Fig. 4). Immediately after initiation (leaf length = 700 gin), the lamina stops expanding in a narrow region located about 100 gm from the base of the axis (Fig. 4). This region gives rise to the petiole. In the genotype described here (br/br), the lamina on the petiole (also called the wing) undergoes little further expansion; in Br/Br stocks it becomes quite broad (van der Veen and Bink 1961). During the course of development the lamina undergoes a subtle change in shape. At first it is essentially elliptical. After the leaf emerges from the bud the basal half of the lamina begins to increase in width relative to the apical half so that the lamina becomes ovate. The dynamics of this growth pattern will be described in more detail later in this paper.
Cellular differentiation in the leaf primordium a. Epidermal hairs. Leaf primordia in tobacco develop a dense blanket of epidermal hairs soon after initiation. Hairs appear at the tip of the primor-
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf
161
Fig. l. A scanning electron micrograph (SEM) of a shoot apex of Xanthi Nc. tobacco with four leaf primordia (PI to P4). Primordium P1 is at a buttress stage, P2 and P3 are at stages prior to the initiation of the lamina, and P4 is at a stage after lamina initiation, eh = Epidermal hairs. • 260 ; bar = 100 lam Fig. 2. SEM of a shoot apex with two leaf primordia, x 240; b a r = 100 gna Fig. 3. Median longitudinal section of a shoot apex with leaf primordia at about the same stage as those in Fig. 2. x240; bar = 100 gm
dium and along its midline when the primordium is about 100 Ixm long (P2 in Figs. 2, 3). Hairs do not form along the lateral margins of the primordium until after the initiation of the lamina, but then arise with remarkable rapidity in this region
(Fig. 4). Subsequently they appear in the intercalary region of the lamina. Apart from this regional distribution, hairs do not display any obvious spacing pattern at the time of their initiation. It is not uncommon, for example, to find adjacent
162
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf
Fig. 4. Scanning electron micrograph of successive leaf primordia at the shoot apex of Xanthi Nc. tobacco. The smaller primordium is in the process of initiating a lamina; the lamina (la) and midrib (m) of the larger primordium are well defined. The indentation at the base of the lamina in the larger primordium is the prospective petiole, eh = Epidermal hairs, x 150; bar=100 gm
cells initiating hairs simultaneously. The precocious differentiation of epidermal hairs is important because hair-initiating cells are probably not meristematic. Hence the early apparance of hairs at the apex and margins of the leaf indicates that these regions do not contribute appreciably to the growth of the leaf. This conclusion is supported by the fact that marginal epidermal cells are considerably larger than adjacent epidermal and subepidermal cells (Figs. 7, 8). It is also interesting to note that marginal cells are elongated parallel to the margin rather than perpendicular to it. This feature contrasts with the generally accepted interpretation of marginal growth (Esau 1977, p. 336), which portrays the margin as contributing primarily to the transverse expansion of the lamina. The shape of marginal cells clearly indicates that these cells contribute more to the longitudinal extension of the leaf than to its transverse expansion.
b. Vascular differentiation. Lateral provascular strands first appear in the lamina when the leaf is 1 mm long, and by the time the leaf has reached a length of 3 mm, all six or seven major lateral veins have been initiated (Figs. 9-11). The first three or four veins arise simultaneously; later ones are initiated sequentially at the base of the lamina. Veins arise at an oblique angle to the midrib. When they first become visible, provascular strands appear to be three or four cells in width and are separated by four to ten interveinal cells (Fig. 9). Subsequently these lateral strands increase dramatically in width, often becoming wider than interveinal regions (Fig. 11). In a young leaf primordium the primary lateral veins make up the majority of tissue in the leaf blade. Later in development, interveinal regions expand considerably so that, in a mature leaf, primary lateral veins occupy only a small fraction of the lamina.
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf
163
Figs. 5, 6. Transverse sections of the leaf primordia of Xanthi Nc. tobacco just prior to (Fig. 5) and during (Fig. 6) the initiation of the lamina. Brackets indicate the regions from which the lamina will arise, pv=provascular tissue. x 275
c. Mesophyll tissue. The number of mesophyll layers in different regions of the lamina is four or five. During meristematic growth, cells in interveinal regions of the lamina are uniform in shape and size. As the leaf matures the lower three or four layers differentiate into spongy mesophyll, whereas the uppermost mesophyll layer develops a typical palisade morphology (Avery 1933). Cell division stops first in the epidermis, then in the lower three or four mesophyll layers, and finally in the upper mesophyll (palisade) layer. The expansion of the lamina
In tobacco, as in many other species (yon Papen 1935; Tenopyr 1918), the shape of a mature leaf is more closely correlated with variation in cell number than with variation in cell size and shape. As shown in Fig. 12, the number of palisade cells
Figs. 7, 8. The margin of a mature leaf of Xanthi Nc. tobacco. Note the large size of marginal epidermal cells (me) relative to intercalary epidermal (ie) and palisade mesophyll (pro) cells. Fig. 7. Scanning electron micrograph. Fig. 8. Paradermal section. eh = Epidermal hairs, x 96, bar= 100 pm
per unit area is fairly uniform throughout the lamina. Cell size does decrease toward the tip and base of the lamina (as indicated by the relatively high cell density in these regions), but this decrease is not great enough to completely account for the shape of the lamina (Fig. 12). Since neither epidermal cells nor palisade cells show any striking variation in shape within the lamina, it is clear that the variation in the width of the lamina is primarily associated with variation in cell number. Given this correlation it is important to determine how variation in cell number arises during lamina expansion, and to what extent patterns of cell division are correlated with the growth dynamics of the leaf. a. Mitotic frequency. The fiequency-distribution and orientation of cell division during the expansion of the lamina was characterized primarily by clonal analysis (Poethig 1984; Poethig and Sussex 1985), but to determine the validity of this approach we also examined the frequency of mitotic figures in cleared specimens. Because of the difficulty of recording mitotic figures in large specimens, observations were restricted to the early phase of leaf development. At the 6-ram stage, the primordium is still completely meristematic (Fig. 13A). Contrary to our expectations, the frequency of mitosis varies dramatically within the lamina at this stage. At two or three points along the length of the leaf, the mitotic frequency is twice what it is elsewhere. At present we do not know whether the number or position of these peaks is fixed, or whether they represent waves of cell division moving along the leaf. The frequency of mitosis does tend to be consistently high, however, in a region immediately behind the tip of the leaf. Cell division ceases at the tip of the leaf when the leaf is about 1.5 cm long, and then gradually stops in more basal regions (Fig. 13 C, D). During this period there is no evidence of periodic variations in mitotic frequency; however the sampling meth-
164
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf
Figs. 9-11. The differentiation of primary lateral veins (arrows) in the lamina of leaves of Xanthi Nc. tobacco. The leaves shown here are at different magnifications, and only the right half of the lamina is pictured. The actual length of the leaves is 1 mm (Fig. 9), 2 mm (Fig. 10), and 3 mm (Fig. 11). Specimens cleared and Feulgen-stained
ods used would have precluded observing minor variations within interveinal regions. Prior to the emergence of the leaf from the bud - which occurs when the leaf is about I cm long - the mitotic frequency in the basal two or three interveinal regions is low in comparison with more distal regions (Fig. 13A, B). When the leaf reaches a length of about 1.5 cm, the mitotic frequency in these regions increases, and eventually becomes greater than or equal to the mitotic frequency in more distal region of the lamina (Fig. 13 C, D). The pattern of cell division during later developmental stages was studied exclusively by clonal analysis (Poethig and Sussex 1985). Throughout this later period, cell division declines in frequency and then ceases at the distal end of the leaf, while occurring at a relatively rapid rate in the basalmost region of the lamina. Meristematic growth ceases when the leaf is one half its final size.
b. The dynamics of leaf growth. A semi-logarithmic plot of the growth in the length of the leaf axis is shown in Fig. 14A. The relative growth rate of the axis (RE) was calculated using values of In L
obtained from this plot (Fig. 14B). The relative growth rate is analogous to a compound interest rate and is a more informative measure of growth than the absolute growth rate because it is independent of size (see Green 1976, and Williams 1975, pp. 10-14, for an explanation of the relative growth rate). The relative growth rate is expressed in units of t i m e - l , which in our case is day-1. It is also worth noting that exponential growth is represented by a constant relative growth rate, and that an R E of 0.69 d - 1 represents a doubling time of 1 d. The R E of the axis is maximal during initiation and drops gradually over the next 2 d (Fig. 14B). When the axis is 1-2 mm long, RE increases and remains constant for a variable period of time. This period of exponential growth is usually quite brief, and ends about 4 d later when the leaf emerges from the bud at a length of about 1 cm. From this point on, R E gradually declines. Under our growing conditions leaves reached a mature length of 18 cm about three weeks after initiation. Throughout the growth of the leaf the petiole grows at a slower relative rate than the adjacent
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf
92 96
90
100
88
/
/
/ 40
30
20
10
Cell Number [x1031
Fig. 12. The number of palisade cells per unit area (0.1 mm 2) in different regions of the lamina (right), and the total number of cells in 0.i-ram-wide transverse section of the lamina (left) of Xanthi Nc. tobacco, calculated from the average cell density in that region. Note that the variation in cell number along the length of the leaf closely corresponds to the variation in the width of the leaf
165
region of the midrib (Fig. 15A, B). The midrib elongates uniformly early in development but immediately after emergence a basipetal gradient in R E develops. Growth stops first at the tip of the leaf (leaf length = 5 cm) and then in progressively more basal regions of the axis. The behavior of the lamina is more complex than that of the axis (Fig. 15 A). Prior to the emergence of the leaf, the lamina exhibits two peaks in its relative rate of expansion (RA), one near the tip of the leaf, the other near the base (Fig. 15A, panel a). During, or immediately after emergence, there is a general increase in the R A of the lamina, although to different extents in different regions (Fig. 15A, panels b, c). Soon thereafter, R A starts to decline in a basipetal fashion. The behavior of the base of the lamina is unique in two respects. Although this region has a relatively low R A prior to emergence (Fig. 15A, panel a), after emergence its R A becomes greater than that of any other region (Fig. 15 A, panel c). Furthermore, while apical regions of the lamina undergo a progressive decline in R A after emergence, the R A of the basal region declines only briefly and then increases once again later in development. Following this second increase, the R A of the basal lamina drops precipitously. The dramatic change in the R A of the basal
n
C B
2 A
2 4
2 4
Mitotic Index
2 4
Fig. 13A-D. Spatial variation in the mitotic index (% cells in mitosis) within the lamina of meristematic leaves of Xanthi Nc. tobacco. Leaf length: A 6mm;B12mm;C14mm;D20mm
166
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf
/ Jf
100
E
1C
E e| .a
1
0.1
/ 5 1
5
10
10 Days
15
20
15
5.8cm
20
Fig. 14. A Semi-logarithmic plot of the growth in the length of the axis of leaves of Xanthi Nc. tobacco. B Relative growth rate of the leaf axis per day. Data obtained from buds dissected at various times after topping (e) and from daily measurements of a single leaf (A)
A
a
b
o.s 1
o.~g--i-
c
o.5
d
i
o.~
i
9
!
o'5
015
g
1
0,5
1
Relative Growth Rate
B
lamina
ilirm,dr,b
1.0 .8
8.7cm
11.5cm
Fig. 16. The growth of a leaf of Xanthi Nc. tobacco from about one third to two thirds its final size. The lamina was marked with a matrix of ink dots when the leaf was 5.8 cm long
variation in the orientation of growth within the lamina is responsible for this change. This possibility was eliminated by marking leaves with a grid of equally spaced points and observing the deformation of this grid during development (Fig. 16). Although it is impossible to draw unequivocal conclusions about the orientation of growth from these data without undertaking detailed analyses like those of Richards and Kavanagh (1943), Erickson (1966) and Silk (1983), it is clear that the growth of the lamina is more or less isotropic. At least there is no evidence that growth is much more highly oriented in one region of the lamina than in another. We believe, therefore, that the change in the slope of the lamina in the br genotype is primarily the consequence of an increase in the R A of its basal region.
.6
Discussion a
b
c
d
e
f
a
b
c
d
e
f
Fig, 15. A Relative growth rate (d- 1) of the area (e) and length (o) of different transverse sections of a leaf on seven successive days. Panel a leaf length 1-1.5 cm; panel b 1.5-2.4 cm; panel c 2.4-3.7 cm; paneld 3.7-5.2 cm; panele 5.2-7.4 cm; panelf 7.4-9.6 cm; paneIg 9.6-11.3 cm. B The data in A, presented in a way that better illustrates the change in the relative growth rate of different regions of the leaf during development
region of the lamina is primarily responsible for transforming the lamina from an elliptical structure into an ovate one. This change in shape cannot be due to the prolonged growth of the base because cultivars with elliptical leaves have the same basipetal pattern of growth as Xanthi Nc. (Avery 1933; Poethig 1981). Nor it is likely that
Our observations confirm many of those made by other investigators (Avery 1933; Schwartz et al. 1952; Haber and Foard 1963; Hannam 1968; Williams 1975; Dubuc-Lebreux and Sattler 1980), but also disagree with some of these studies on a number of important points. Although some of these discrepancies may follow from the unique morphology of Xanthi Nc., most appear to have a more important basis. We will now review the major findings of this study, focusing in particular on those that conflict with existing interpretations of leaf development in tobacco.
The cell number of the primordium. Avery (1933) reported that the internal tissue of the leaf axis was derived from a single subapical initial and that
R.S. Poethig and I.M. Sussex: Morphologyand growth of tobacco leaf the internal tissue of the lamina arose from a single row of submarginal initials, but the histological structure and external morphology of the leaf primordium indicate that the primordium actually arises from a much larger group of cells. A leaf buttress has a surface area of about 50 cells and encompasses four layers of the apical meristem. This population arises within 28 h, the interval between the appearance of successive leaves. Taking the cell doubling time in the primordium to be 19 h (derived from the data of H a n n a m 1968), the number of cells in the epidermis of the primordium one plastochron before the appearance of the leaf buttress can be calculated using the formula N1 = No etln2/a, where N 1 is the number of cells in the epidermis of the buttress (50); N o is the number of cells one plastochron earlier, d is the cell doubling time (19 h), and t is the duration of the plastochron (28 h). Solving this formula for N o gives a value of 18. Therefore, on the assumption that a leaf primordium is determined no earlier than one plastochron before it becomes visible (Snow and Snow 1933), we conclude that the leaf axis in tobacco is derived from at least 18 cells in the epidermis of the apical meristem. Of course a leaf primordium does not arise solely from the epidermis of the shoot. Histological observations (Hannam 1968; this paper) and periclinal chimeras (Burk et al. 1964; Dulieu 1968; Poethig 1984)indicate that it encompasses at least two and perhaps three internal layers. Since it is reasonable to suppose that the primordium is derived from about the same number of cells (probably somewhat fewer) in these layers as in the epidermis, the tobacco leaf is estimated to arise from on the order of 50-100 initial cells. A similar argument can be made in the case of the lamina. When the lamina first becomes distinct it encompasses about 20 cells in a transverse section of the leaf axis. This population arises within less than a day. Thus, assuming that lamina initials have a cell doubling time of 13.8 h (Hannam 1968), and that the lamina is initiated in 24 h, we calculate that the lamina is derived from a minim u m of six cells in transverse section of the leaf axis. This is considered a minimum estimate because the time between the initiation and appearance of the lamina is probably less than 24 h.
The cell lineage of the leaf axis. Our observations also contradict the idea that the initials of a leaf primordium reside at the apex of the leaf axis (Avery 1933; Stewart and Dermen 1975). If this were true, one would expect cell files to emanate radially from the apex, as they do in the shoot
167
apical meristem. Instead, cell files run longitudinally in the axis and terminate along its lateral margins rather than at the apex. Thus the tobacco leaf a p p e a r s to be derived from "line source", much like a corn leaf, rather than from a " p o i n t source", such as an apical meristem.
The dynamics of leaf elongation. The relative rate of increase in the volume (Rv) of the tobacco leaf has been characterized by H a n n a m (1968), whose data were later re-examined by Williams (1975). These data, and the results presented in this paper, indicate that leaf growth can be divided into three phases: 1) a brief initiation phase characterized by a rapid but declining growth rate, 2 ) a stage of exponential growth, and 3) a maturation phase during which the relative rate of growth declines. Williams (1975) has described a similar pattern in Linum usitatissimum, Brassica oleracea and Lupinus angustifolius, and a review of published data indicates that it also exists in Xanthium pensylvanicure (Maksymowych 1959), Lactuca sativa (Bensink 1971), Cucumis sativus (Horie et al. 1979) and Trifolium repens (Denne 1966). The difference between the deceleratory and exponential growth pattern is relatively subtle and is easy to overlook. Nevertheless this distinction is real and must be explained. Williams (i 975) has argued that the dynamics of leaf growth are largely determined by physical interactions within the apical bud. According to this argument, growth is rapid early in leaf development because the leaf has not come in contact with other leaves, and that as it does, its growth rate declines and stabilizes at a lower exponential rate. As yet, however, there is no experimental support for this and any other possible explanation for the dynamics of leaf elongation. The dynamics of lamina expansion. Avery (1933) studied the dynamics of lamina expansion in tobacco by comparing the growth rate of different regions of the lamina with the growth rate of the lamina as a whole. This type of analysis permitted him to characterize the distribution growth within the lamina at a given point in time, but made it impossible to study changes in the growth rate of any particular region during the course of development. As our results demonstrate, such temporal variation in growth rate may be an important feature of leaf development. Indeed, one of the most striking aspects of leaf development in Xanthi Nc. is the dramatic increase in the R a of the basal region of the lamina that takes place after emergence. In contrast, the basal region of the lamina in Java - a variety with broad, elliptical leaves (genotype
168
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf
Br/Br) - only undergoes a transitory increase in R A after emergence (Poethig 1981). This observation lends support to the conclusion that the ovate character of the leaf of Xanthi Nc. is primarily a consequence of the increase in the R A of the leaf base, rather than a result of the prolonged growth or orientation of growth in this region. In Xanthi Nc., R A is maximal in the basalmost region of the lamina, whereas in the cultivar studied by Avery (1933) the growth rate was found to be maximal in a localized region near the center of the lamina (Richards and Kavanagh 1943). If Avery's data are correct then tobacco leaves should develop a bulge in this central region because it is surrounded by more slowly growing tissue. This rarely, if ever, occurs. Furthermore, an examination of the pattern of leaf growth in the cultivar Java, a variety of tobacco whose leaf shape is similar to that of the variety studied by Avery (1933), showed that R a is maximal near the base of the lamina, not at the center (Poethig 1981). These discrepancies cast doubt on Avery's interpretation of the pattern of growth of the tobacco leaf, and call for a more careful study of the dynamics of leaf expansion in tobacco than has hitherto been undertaken. The pattern of cell division during the expansion of the lamina. An accurate quantitative description of the pattern of cell division during leaf morphogenesis is essential for an understanding of the cellular basis of leaf shape. Among the most comprehensive of such studies are those of Fuchs (1966, 1975, 1976), Thomasson (1970) and Jeune (1972). By using cleared rather than sectioned specimens these investigators were able to obtain a remarkably detailed picture of the orientation and distribution of cell division within the plane of the lamina. They found that both of these parameters varied over time and within the plane of the lamina, and that this variation was correlated with the pattern of growth. This technique has also proved useful in the case of tobacco. Although our analysis of cleared specimens was not extensive, it nevertheless demonstrated that the frequency of cell division varied considerably within the lamina (during at least the early part of leaf expansion) in a manner consistent with the pattern of growth. The most obvious feature of this pattern is the basipetal gradient in the duration of meristematic growth. What is less well documented is the spatial variation in the frequency of cell division that occurs in meristematic tissue. This aspect of leaf development is described in more detail in the following paper (Poethig and Sussex 1985).
We are grateful to John Duesing (Ciba-Geigy Co., Greensboro, N.C., USA) for many helpful discussions, and to A. Pooley for his assistance with the scanning electron microscope. R.S.P. was supported by a National Science Foundation Predoctoral Fellowship and by National Institutes of Health Training Grant HD07180-01.
References Avery, G.S. (1933) Structure and development of the tobacco leaf. Am. J. Bot. 20, 565-592 Bensink, J. (1971.) On morphogenesis of lettuce leaves in relation to light and temperature. Meded. Landbouwhogeschool, Wageningen 71, No. 15 Bergann, F., Bergann, L. (1984) Zur Entwicklungsgeschichte des Angiospermenblattes. 4. Uber Periklinalchimfiren bei Sedum rubrotinctum R.T. Clausen, Biol. Zentralbl. 103, 147-171 Burk, L.G., Stewart, R.N., Dermen, H. (1964) Histogenesis and genetics of a plastid-controlled chlorophyll variegation in tobacco. Am. J. Bot. 51,713 724 Denne, M.P. (1966) Leaf development in Trifolium repens. Bot. Gaz. 127, 202-210 Dubuc-Lebreux, M.A., Sattler, R. (1980) D6veloppement des organes foliac6s chez Nicotiana tabacum et le probleme des m~rist~mes marginaux. Phytomorphology 30, 17-32 Dulieu, H. (1968) Emploi des chim~res chlorophylliennes pour l'~tude de l'ontog6nie foliaire. Bull. Sci. Bourgogne 25, 1-60 Dulieu, H. (1970) Les mutations somatiques induites et l'ontog~nie de la pousse feuill6e. Ann. Amelior. Plant. 20, 27-44 Dulieu, H. (1974) Somatic variations on a yellow mutant in Nicotiana tabacum L. (al +/al a2 +/a2). I. Non-reciprocal genetic events occurring in leaf cells. Mutat. Res. 25, 289-304 Dulieu, H. (1975) Somatic variations on a yellow mutant of Nicotiana tabacum L. (ai +/al a2 +/a2). II. Reciprocal genetic events occurring in leaf cells. Mutat. Res. 28, 69-77 Dulieu, H., Dalebroux, M.A. (1975) Spontaneous and induced reversion rates in a double heterozygous mutant of Nicotiana tabacum var. Xanthi nc. - dose-response relationship. Mutat. Res. 30, 63-70 Erickson, R.O. (1966) Relative elemental rates and anisotropy of growth in area: a computer programme. J. Exp. Bot. 17, 390M03 Esau, K. (1977) Anatomy of seed plants. John Wiley & Sons, New York Feder, N., O'Brien, T.P. (1968) Plant microtechnique: some principles and new methods. Am. J. Bot. 55, 123-142 Foster, A.S. (1936) Leaf differentiation in angiosperms. Bot. Rev. 2, 349-372 Fuchs, C. (1966) Observations sur l'extension en largeur du limbe foliaire du Lupinus albus L. C.R. Acad. Sci. Ser. D
263, 1212-1215 Fuchs, C. (1975) Ontogen~se foliaire et acquisition de la forme chez le Tropaeolum peregrinum L. I. Les premier stades de l'ontogen6se du lobe m6dian. Ann. Sic. Nat. Bot. Biol. Veg. 16, 321-390 Fuchs, C. (1976) Ontogen6se foliaire et acquisition de la forme chez le Tropaeolum peregrinum L. II. Le developpement du lobe apr+s la formation de lobules. Ann. Sci. Nat. Bot. Biol. Veg. 17, 121-158 Green, P.B. (1976) Growth and cell pattern formation on an axis : critique of concepts, terminology and modes of study. Bot. Gaz. 137, 187-202 Haber, A.H., Foard, D.E. (1963) Nonessentiality of concurrent cell divisions for degree of polarization of leaf growth. II. Evidence from untreated plants and chemically induced
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf changes in the degree of polarization. Am. J. Bot. 50, 937 944 Hannam, R.V. (1968) Leaf growth and development in the young tobacco plant. Aust. J. Biol. Sci. 21, 855-870 Horie, T., de Wit, C.T., Goudrian, J., Bensink, J. (1979) A formal template for the development of cucumber in its vegetative state. Proc. K. Ned. Akad. Wet. C 82, 433-479 Jensen, W.A. (1968) Botanical histochemistry, W.H. Freeman and Co., San Francisco, USA Jeune, B. (1972) Observations et expbrimentation sur les feuilles juveniles du Paulownia tomentosa H. Bn. Bull. Soc. Bot. France 119, 215-230 Jeune, B. (1983) Croissance des feuilles de Lycopus europaeus L. Beitr. Biol. Pflanz. 58, 253 266 Maksymowych, R. (1959) Quantitative anaIysis of leaf development in Xanthium pensylvanicum. Am. J. Bot. 46, 635 644 Maksymowych, R., Erickson, R.O. (1960) Development of the lamina in Xanthium italicum represented by the plastochron index. Am. J. Bot. 47, 451-459 Poethig, R.S. (1981) The cellular parameters of leaf development in Nicotiana tabacum L. Ph.D. thesis, Yale University Poethig, R.S. (1984) Cellular parameters of leaf morphogenesis in maize and tobacco. In: Contemporary problems in plant anatomy, pp. 235 259, White, R.A., Dickison, W.C., eds. Academic Press, New York Poethig, R.S., Sussex, I.M. (1985) The cellular parameters of leaf development in tobacco : a clonal analysis. Planta 165, 170-184 Richards, O.W., Kavanagh, A.J. (1943) The analysis of the relative growth gradients and changing form of growing organisms: illustrated by the tobacco leaf. Am. Nat. 77, 385-395 Ruddell, C.L. (1967) Embedding media for 1 2 micron sectioning. 2. Hydroxyethyl methacrylate combined with 2-butoxyethanol. Stain Technol. 42, 253-255 Schwartz, D., Renier, A., Cuzin, J. (1952) t~tude quantitative
169
de la croissance du tabac. II. Morphog6n6se foliaire. Ann. Inst. Exp. Tabac Bergerac 1, 55-95 Silk, W.K. (1983) Kinematic analysis of leaf expansion. In: The growth and functioning of leaves, pp. 89-108, Dale, J.E., Milthorpe, F . L , eds. Cambridge University Press, Cambridge, UK Singh, I.S., Dulieu, H. (1976) Effects g~n~tiques et variabilit~ g6n6tique pour plusiers charact6res chex un mutant monohybride de Nicotiana tabacum L. Ann. Amelior. Plant. 26, 579-590 Snow, M., Snow, R. (1933) Experiments on phyltotaxis II. The effect of displacing a primordium. Phil. Trans. R. Soc. London Ser. B 222, 353-400 Stewart, R.N., Burk, L.G. (1970) Independence of tissues derived from apical layers in ontogeny of the tobacco leaf and ovary. Am. J. Bot. 57, 1010-1016 Stewart, R.N., Dermen, H. (1975) Flexibility of ontogeny as shown by the contribution of shoot apical layers to leaves of periclinal chimeras. Am. J. Bot. 62, 935-947 Tenopyr, L.A. (1918) On the constancy of cell shape in leaves of varying shape. Bull. Torrey Bot. Club 45, 51-76 Thomasson, M. (1970) Quelques observations sur la rbpartition de zones de croissance de la feuille du Jasminum nudiflorum Lindl. Candollea 25, 297-340 van der Veen, J.H. (1957) Studies on the inheritance of leaf shape in Nicotiana tabacum L. Excelsior, Oranjeplein, The Hague van der Veen, J.H., Bink, J.P.M. (1961) Multiple effects of the leaf shape allele Pt in Nicotiana tabacum L. Genetica 32, 33-50 van Papen, R. (1935) Beitr~ige zur Kenntnis des Wachstums der Blattsbreite. Bot. Arch. 37, 159 206 Williams, R.F. (1975) The shoot apex and leaf growth, Cambridge University Press, Cambridge, UK Received 12 October; accepted 21 December 1984
Planta (1985)165:158-169
9 Springer-Verlag 1985
The developmental morphology and growth dynamics of the tobacco leaf R.S. Poethig* and I.M. Sussex Department of Biology, Yale University, New Haven, CT 06511, USA
Abstract. The developmental morphology and growth dynamics of the leaf of Nicotiana tabacum L. cv. Xanthi Nc. are described. Epidermal and internal cell patterns indicate that the leaf axis arises from approx. 100 cells in four layers of the shoot apex, while the lamina arises from several rows of cells in each of three layers of the leaf axis. Cell patterns at the apex and margin of the leaf do not support the classical view that these regions have a specialized meristematic function. Instead the development of the leaf appears to be largely dependent on intercalary growth. The pattern of growth within the lamina is surprisingly complex. In addition to a proximal-distal gradient in the duration of growth and cell division during development, localized transitory changes in the rate of these processes also occur. These observations are discussed in reference to previous discriptions of leaf development in tobacco. Key words: Cell division - Growth (leaves) - Leaf development - Nicotiana (leaf growth).
Introduction Research on the cellular basis of leaf morphogenesis has usually focused on two issues: the number of initial cells that give rise to the leaf, and the role of these cells in controlling leaf shape. A prominent, though by no means general, view in the classical literature (see review by Foster 1936) * Present address: University of Pennsylvania, Department of
Biology, Leidy Laboratories/G7, Philadelphia, PA 19104, USA S y m b o l s : R A= relative rate of growth in area; RE = relative rate
of growth in length; Rv = relative rate of growth in volume
is that leaves arise from a small group of initial cells situated at the tip of the leaf axis, and that the lamina arises from one or two rows of cells along the lateral margins of the leaf axis. These marginal and submarginal ceils constitute what is known as the marginal meristem, and are thought to control the shape of the lamina. However, recent studies do not support these conclusions. In the case of tobacco, for example, analyses of periclinal chimeras have shown that both the leaf axis and lamina are derived from at least three layers of cells (Burk et al. 1964; Dulieu 1968, 1970; Stewart and Burk 1970). Other types of genetic mosaics indicate that there may be 16 or more initial cells in each layer of the axis promordium (Dulieu 1968, 1970; Stewart and Dermen 1975), and several rows of cells in each layer of the lamina primordium (Dulieu 1968). Unfortunately there is no consensus about the behavior of these initial cells. Stewart and Dermen (1975) imply that leaf initials reside in an apical meristem during the extensinon of the axis, while Dulieu (1968) believes that each initial forms a unique longitudinal section of the axis, such that only initial cells in the very center of the leaf primordum contribute to the leaf apex (Dulieu 1968). There is more agreement about the behavior of cells at the leaf margin. Careful quantitative studies of the pattern of cell division in Xanthium italicum (Maksymowych and Erickson 1960), Trapaeolum majus (Fuchs 1975; 1976), Jasminum nudiflorum (Thomasson 1970), PauIownia tomentosa (Jeune 1972), Lycopus europaeus (Jeune 1983) and Nicotiana tabacum (Dubuc-Lebreux and Sattler 1980) demonstrate that the frequency of cell division is no higher at the leaf margin than in intercalary regions, and may in fact be significantly lower. Furthermore, the pattern of variegation in periclinal chimeras indicates that marginal cell lineages are highly variable in their behavior,
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf
and may contribute either very little or a great deal to the expansion of the lamina (Stewart and Dermen 1975; Bergann and Bergann 1984). These observations make it clear that traditional interpretations of the behavior and function of marginal cells need to be re-examined. In order to resolve some of the discrepancies between classical and more recent interpretations of leaf development, we decided to undertake a study of leaf development in Nicotiana tabacum L. cv. Xanthi Nc. This cultivar has the advantage of being relatively small in size, and of having genetic markers useful for clonal analysis (Poethig and Sussex 1985). Its leaves are ovate in shape and have a naked petiole, traits controlled by the br allele of the broad locus (van der Veen 1957). Information about the cellular parameters of leaf development in N. tabacum L. cv. Xanthi Nc. was obtained by clonal analysis and is presented in the following paper (Poethig and Sussex 1985). In this paper we introduce the experimental system used for this analysis, and describe the developmental morphology and growth dynamics of the leaf. Material and methods For this study we used a stock of Nicotiana tabacum L. cv. Xanthi Nc. heterozygous for the chlorophyll mutations al and a2 (=yg). Seeds of this stock were obtained from H. Dulieu (Station d'Amelioration des Plantes, I.N.R.A., Dijon, France) who has described the nature of these mutations in detail (Dulieu 1974, 1975; Dulieu and Dalebroux 1975; Singh and Dulieu 1976). The al + a2/+ stock was propagated vegetatively, starting with axillary shoots from nine different plants. Axillary shoots were released from growth arrest by removing the top half of the plant (topping), and were harvested two weeks later. After removing five or six bottom leaves, the axillary shoots were inserted in coarse vermiculite in 0.5-1 containers and were kept under a plastic-covered frame until they formed roots, 7-10 d later. Plants were grown in a controlled-environment room on tables watered automatically with a solution of Hyponex (1 g/l; Hyponex Co., Copley, O., USA). The room was operated at 27 ~ C, 60% relative humidity, and a 16:8 light:dark cycle. Illumination was provided by fluorescent (40 W Lifeline coolwhite; Sylvania Co., Seneca Falls, N.Y., USA) and incandescent (60 W) lamps, at an energy fluence rate of 150 W m -2.
Experimental system. Our observations concern the development of leaves on axillary shoots recently released from growth arrest. Because all axillary buds are arrested at approximately the same stage (six or seven leaves) and grow out rapidly after topping, this system was synchronous, reproducible and fast. Axillary buds were released from growth arrest by topping 15to 20-cm-tall plants, so as to leave five or six mature leaves. As axillary buds appeared, all but the topmost ones were removed. A fresh nutrient solution was added to the reservoir tanks immediately after topping, and was replenished with disfilled water as needed; no additional nutrients were supplied during the course of an experiment.
159
Growth analysis. The earliest phase of leaf growth was reconstructed by measuring the length of successive leaves (leaves 9, 10 and 11, counting from the base of the axillary shoot) in axillary shoots harvested at various times after topping. To construct a growth curve, the average lengths of these three leaves in each set of shoots were placed at 28-h intervals (the length of the plastochron), and sets of points from successive harvests were positioned relative to one another by finding the point at which they overlapped. The smallest leaf primordium (leaf 11 in the youngest set of shoots) was assumed to be no more than 28 h old because the next youngest leaf was not visible in these buds. This leaf primordium was arbitrarily assigned an age of 24 h in order to position the growth curve on a time scale. After leaves reached a length of 5 mm they could be measured non-destructively with a millimeter ruler. The later part of the growth curve in Fig. 14 represents the actual growth of a single leaf, in this case leaf 10. The growth of the lamina was studied by marking a leaf with India ink and then photographing it daily for 8 d. The areas of the regions delimited by the marks were then measured with a planimeter on enlarged photographic prints. The relative growth in area (RA) was calculated using the formula R A= ( l n A i - l n A o ) t -1 where A o is the original area of the region, A 1 is its area 1 d later, and t equals 1 d (Williams 1975, pp. 10~ 12). The relative rate of extension of the axis was calculated in the same way.
Histologicalprocedures. For the purpose of analyzing the distribution of mitotic figures, cell size, and the initial stages of vascular differentiation, leaf primordia were fixed in 3 : 1 (v/v) ethanol:acetic acid, stained according to the Feulgen procedure, dehydrated in ethanol and mounted in Permount (Fisher Scientific, Cincinnati, O., USA) (Jensen 1968). In the case of tobacco, this procedure generally yields considerably background staining and it is necessary to use a green filter to clearly visualize mitotic figures. For detailed analysis, specimens were fixed in 3% gultaraldehyde (sodium-phosphate buffer, pH 7), dehydrated in ethanol, and embedded in glycol methacrylate according to the procedure of Ruddell (1967) or the JB-4 procedure (Polysciences, Warrington, Pa., USA). Sections 1-2 Jam thick were cut with glass knives on a JB-4 microtome (Sorvall Instruments, Du Pont & Co., Wilmington, Del., USA), attached to slides, and stained with periodic acid-Schiff's-hematoxylin (Feder and O'Brien 1968). Sections and whole mounts were examined with a C. Zeiss (Oberkochen, FRG) brightfield microscope. Photomicrographs were taken with Technical Pan film 2415 (Eastman-Kodak, Rochester, N.Y., USA) which was developed for medium contrast (Kodak HC110, dilution F). The mitotic frequency within the plane of the lamina was determined using Feulgen-stained whole mounts. Specimens were examined at 375-fold magnification and the number of mitotic figures (metaphase through anaphase) in different interveinal regions was counted with the aid of an ocular grid. The number of cells in each region was then determined by multiplying the area of the region by the average number of cells per unit area, as determined from six to ten randomly chosen fields. These numbers were then used to calculate the mitotic index (percent cells in mitosis).
Scanning electron microscopy. Observations were made on flesh, uncoated specimens. Shoot apices were dissected just prior to examination and were kept on moist filter paper in small Petri dishes until use. Specimens were examined with an ETEC Autoscan (Perkin-Elmer, Norwalk, Ct., USA) operated at 2.5 kV.
160
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf
Results
Successive leaves on an axillary shoot differ in shape and size. The first six or seven leaves are initiated before growth arrest, and are small and variable in shape at maturity. Leaf size increases up to the ninth through eleventh nodes and then decreases with each successive node. Starting at about the eighth node the leaf blade becomes ovate in shape, that is, broader at the base than at the tip. Because space limitations made it necessary to use successive leaves to reconstruct the development of a single "typical" leaf, it was important to choose a series of leaves whose morphology and growth rate were reasonably similar. Attention was therefore restricted to leaves 9, 10 and 11 (counting from the base of the axillary shoot) because these leaves are essentially identical in shape, size, number of cells per unit area, and growth rate (Table 1).
The morphology of the leaf primordium The shoot apical meristem in tobacco is remarkably flat, and leaves arise within the plane of the apex (Figs. 1, 3). The first evidence at the surface of the shoot of the formation of a leaf is a lateral swelling called the leaf buttress (P1 in Fig. 1). The leaf buttress is about ten cells wide and is located about ten cells from the center of the apex. In longitudinal sections the buttress appears to encompass about four layers of the apical meristem (Fig. 3). The primordium emerges from the apex in the form of a low bulge, slightly flattened toward the shoot apex (P1 in Fig. 2). Immediately after initiation, a boundary forms on the adaxial side of the primordium. However the lateral margins of the primordium continue to expand around the apical
Table 1. Some morphological and developmental parameters of leaves 9, 10 and 11 of Xanthi Nc. tobacco, counting from the base of the axillary shoot. These data represent an average of 20 samples, except for the number of palisade cells/ram 2, which was calculated from five samples. The growth rate of the leaf represents the time it took the leaf to grow from ~ cm to one half its final length. The plastochron is the time between the emergence of successive leaves; for the sake of analysis, emergence was defined to occur at a length of 1 cm Leaf no.
9 10 11
Area (cm 2)
125+2 128___3 130_+3
Length/ width
1.9 1.9 1,9
Palisade cells/mm z
797+__17 795__+19 798_+21
Growth rate (h)
Plastochron (h)
124__+1 28_+1 1 2 6 _ + 1 30_+1 128_+1
dome until the leaf encompasses about one third of the circumference of the shoot. Epidermal cell lineages in young primordia (leaf length = 200 gin) do not emanate radially from the tip of the primordium, as would be expected if the primordium was derived from a small group of apical initials. Along the midline of the primordium, parallel cell files run from the tip to the base of the axis in apparent continuity with cell files in the internode below (P2 in Fig. 1); but towards the lateral margins of the axis the cell pattern is irregular, and cell lineages cannot be traced to the leaf tip (P2 in Fig. 1,
2). The initiation of the lamina The lamina, or leaf blade, arises at the boundary between the adaxial and abaxial faces of the leaf axis, and is first recognizable as a distinct ridge when the primordium is 500 gm long (Fig. 4). Prior to this stage, its point of origin is marked internally by a band of densely-stained cells running from the margins of the primordium through the central provascular strand (Fig. 5). The subsequent vacuolation of cells in the region surrounding the provascular strand leaves a distinctive group of cells at the margin (Fig. 6). This putative group of lamina initials is composed of between ten and twenty cells in the transverse plane of the axis - about eight cells in the epidermis, six in the sub-epidermal layer and one to three cells internal to these layers. Initially, the lamina extends in a smooth curve from the tip to the base of the leaf axis (Fig. 4). Immediately after initiation (leaf length = 700 gin), the lamina stops expanding in a narrow region located about 100 gm from the base of the axis (Fig. 4). This region gives rise to the petiole. In the genotype described here (br/br), the lamina on the petiole (also called the wing) undergoes little further expansion; in Br/Br stocks it becomes quite broad (van der Veen and Bink 1961). During the course of development the lamina undergoes a subtle change in shape. At first it is essentially elliptical. After the leaf emerges from the bud the basal half of the lamina begins to increase in width relative to the apical half so that the lamina becomes ovate. The dynamics of this growth pattern will be described in more detail later in this paper.
Cellular differentiation in the leaf primordium a. Epidermal hairs. Leaf primordia in tobacco develop a dense blanket of epidermal hairs soon after initiation. Hairs appear at the tip of the primor-
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf
161
Fig. l. A scanning electron micrograph (SEM) of a shoot apex of Xanthi Nc. tobacco with four leaf primordia (PI to P4). Primordium P1 is at a buttress stage, P2 and P3 are at stages prior to the initiation of the lamina, and P4 is at a stage after lamina initiation, eh = Epidermal hairs. • 260 ; bar = 100 lam Fig. 2. SEM of a shoot apex with two leaf primordia, x 240; b a r = 100 gna Fig. 3. Median longitudinal section of a shoot apex with leaf primordia at about the same stage as those in Fig. 2. x240; bar = 100 gm
dium and along its midline when the primordium is about 100 Ixm long (P2 in Figs. 2, 3). Hairs do not form along the lateral margins of the primordium until after the initiation of the lamina, but then arise with remarkable rapidity in this region
(Fig. 4). Subsequently they appear in the intercalary region of the lamina. Apart from this regional distribution, hairs do not display any obvious spacing pattern at the time of their initiation. It is not uncommon, for example, to find adjacent
162
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf
Fig. 4. Scanning electron micrograph of successive leaf primordia at the shoot apex of Xanthi Nc. tobacco. The smaller primordium is in the process of initiating a lamina; the lamina (la) and midrib (m) of the larger primordium are well defined. The indentation at the base of the lamina in the larger primordium is the prospective petiole, eh = Epidermal hairs, x 150; bar=100 gm
cells initiating hairs simultaneously. The precocious differentiation of epidermal hairs is important because hair-initiating cells are probably not meristematic. Hence the early apparance of hairs at the apex and margins of the leaf indicates that these regions do not contribute appreciably to the growth of the leaf. This conclusion is supported by the fact that marginal epidermal cells are considerably larger than adjacent epidermal and subepidermal cells (Figs. 7, 8). It is also interesting to note that marginal cells are elongated parallel to the margin rather than perpendicular to it. This feature contrasts with the generally accepted interpretation of marginal growth (Esau 1977, p. 336), which portrays the margin as contributing primarily to the transverse expansion of the lamina. The shape of marginal cells clearly indicates that these cells contribute more to the longitudinal extension of the leaf than to its transverse expansion.
b. Vascular differentiation. Lateral provascular strands first appear in the lamina when the leaf is 1 mm long, and by the time the leaf has reached a length of 3 mm, all six or seven major lateral veins have been initiated (Figs. 9-11). The first three or four veins arise simultaneously; later ones are initiated sequentially at the base of the lamina. Veins arise at an oblique angle to the midrib. When they first become visible, provascular strands appear to be three or four cells in width and are separated by four to ten interveinal cells (Fig. 9). Subsequently these lateral strands increase dramatically in width, often becoming wider than interveinal regions (Fig. 11). In a young leaf primordium the primary lateral veins make up the majority of tissue in the leaf blade. Later in development, interveinal regions expand considerably so that, in a mature leaf, primary lateral veins occupy only a small fraction of the lamina.
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf
163
Figs. 5, 6. Transverse sections of the leaf primordia of Xanthi Nc. tobacco just prior to (Fig. 5) and during (Fig. 6) the initiation of the lamina. Brackets indicate the regions from which the lamina will arise, pv=provascular tissue. x 275
c. Mesophyll tissue. The number of mesophyll layers in different regions of the lamina is four or five. During meristematic growth, cells in interveinal regions of the lamina are uniform in shape and size. As the leaf matures the lower three or four layers differentiate into spongy mesophyll, whereas the uppermost mesophyll layer develops a typical palisade morphology (Avery 1933). Cell division stops first in the epidermis, then in the lower three or four mesophyll layers, and finally in the upper mesophyll (palisade) layer. The expansion of the lamina
In tobacco, as in many other species (yon Papen 1935; Tenopyr 1918), the shape of a mature leaf is more closely correlated with variation in cell number than with variation in cell size and shape. As shown in Fig. 12, the number of palisade cells
Figs. 7, 8. The margin of a mature leaf of Xanthi Nc. tobacco. Note the large size of marginal epidermal cells (me) relative to intercalary epidermal (ie) and palisade mesophyll (pro) cells. Fig. 7. Scanning electron micrograph. Fig. 8. Paradermal section. eh = Epidermal hairs, x 96, bar= 100 pm
per unit area is fairly uniform throughout the lamina. Cell size does decrease toward the tip and base of the lamina (as indicated by the relatively high cell density in these regions), but this decrease is not great enough to completely account for the shape of the lamina (Fig. 12). Since neither epidermal cells nor palisade cells show any striking variation in shape within the lamina, it is clear that the variation in the width of the lamina is primarily associated with variation in cell number. Given this correlation it is important to determine how variation in cell number arises during lamina expansion, and to what extent patterns of cell division are correlated with the growth dynamics of the leaf. a. Mitotic frequency. The fiequency-distribution and orientation of cell division during the expansion of the lamina was characterized primarily by clonal analysis (Poethig 1984; Poethig and Sussex 1985), but to determine the validity of this approach we also examined the frequency of mitotic figures in cleared specimens. Because of the difficulty of recording mitotic figures in large specimens, observations were restricted to the early phase of leaf development. At the 6-ram stage, the primordium is still completely meristematic (Fig. 13A). Contrary to our expectations, the frequency of mitosis varies dramatically within the lamina at this stage. At two or three points along the length of the leaf, the mitotic frequency is twice what it is elsewhere. At present we do not know whether the number or position of these peaks is fixed, or whether they represent waves of cell division moving along the leaf. The frequency of mitosis does tend to be consistently high, however, in a region immediately behind the tip of the leaf. Cell division ceases at the tip of the leaf when the leaf is about 1.5 cm long, and then gradually stops in more basal regions (Fig. 13 C, D). During this period there is no evidence of periodic variations in mitotic frequency; however the sampling meth-
164
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf
Figs. 9-11. The differentiation of primary lateral veins (arrows) in the lamina of leaves of Xanthi Nc. tobacco. The leaves shown here are at different magnifications, and only the right half of the lamina is pictured. The actual length of the leaves is 1 mm (Fig. 9), 2 mm (Fig. 10), and 3 mm (Fig. 11). Specimens cleared and Feulgen-stained
ods used would have precluded observing minor variations within interveinal regions. Prior to the emergence of the leaf from the bud - which occurs when the leaf is about I cm long - the mitotic frequency in the basal two or three interveinal regions is low in comparison with more distal regions (Fig. 13A, B). When the leaf reaches a length of about 1.5 cm, the mitotic frequency in these regions increases, and eventually becomes greater than or equal to the mitotic frequency in more distal region of the lamina (Fig. 13 C, D). The pattern of cell division during later developmental stages was studied exclusively by clonal analysis (Poethig and Sussex 1985). Throughout this later period, cell division declines in frequency and then ceases at the distal end of the leaf, while occurring at a relatively rapid rate in the basalmost region of the lamina. Meristematic growth ceases when the leaf is one half its final size.
b. The dynamics of leaf growth. A semi-logarithmic plot of the growth in the length of the leaf axis is shown in Fig. 14A. The relative growth rate of the axis (RE) was calculated using values of In L
obtained from this plot (Fig. 14B). The relative growth rate is analogous to a compound interest rate and is a more informative measure of growth than the absolute growth rate because it is independent of size (see Green 1976, and Williams 1975, pp. 10-14, for an explanation of the relative growth rate). The relative growth rate is expressed in units of t i m e - l , which in our case is day-1. It is also worth noting that exponential growth is represented by a constant relative growth rate, and that an R E of 0.69 d - 1 represents a doubling time of 1 d. The R E of the axis is maximal during initiation and drops gradually over the next 2 d (Fig. 14B). When the axis is 1-2 mm long, RE increases and remains constant for a variable period of time. This period of exponential growth is usually quite brief, and ends about 4 d later when the leaf emerges from the bud at a length of about 1 cm. From this point on, R E gradually declines. Under our growing conditions leaves reached a mature length of 18 cm about three weeks after initiation. Throughout the growth of the leaf the petiole grows at a slower relative rate than the adjacent
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf
92 96
90
100
88
/
/
/ 40
30
20
10
Cell Number [x1031
Fig. 12. The number of palisade cells per unit area (0.1 mm 2) in different regions of the lamina (right), and the total number of cells in 0.i-ram-wide transverse section of the lamina (left) of Xanthi Nc. tobacco, calculated from the average cell density in that region. Note that the variation in cell number along the length of the leaf closely corresponds to the variation in the width of the leaf
165
region of the midrib (Fig. 15A, B). The midrib elongates uniformly early in development but immediately after emergence a basipetal gradient in R E develops. Growth stops first at the tip of the leaf (leaf length = 5 cm) and then in progressively more basal regions of the axis. The behavior of the lamina is more complex than that of the axis (Fig. 15 A). Prior to the emergence of the leaf, the lamina exhibits two peaks in its relative rate of expansion (RA), one near the tip of the leaf, the other near the base (Fig. 15A, panel a). During, or immediately after emergence, there is a general increase in the R A of the lamina, although to different extents in different regions (Fig. 15A, panels b, c). Soon thereafter, R A starts to decline in a basipetal fashion. The behavior of the base of the lamina is unique in two respects. Although this region has a relatively low R A prior to emergence (Fig. 15A, panel a), after emergence its R A becomes greater than that of any other region (Fig. 15 A, panel c). Furthermore, while apical regions of the lamina undergo a progressive decline in R A after emergence, the R A of the basal region declines only briefly and then increases once again later in development. Following this second increase, the R A of the basal lamina drops precipitously. The dramatic change in the R A of the basal
n
C B
2 A
2 4
2 4
Mitotic Index
2 4
Fig. 13A-D. Spatial variation in the mitotic index (% cells in mitosis) within the lamina of meristematic leaves of Xanthi Nc. tobacco. Leaf length: A 6mm;B12mm;C14mm;D20mm
166
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf
/ Jf
100
E
1C
E e| .a
1
0.1
/ 5 1
5
10
10 Days
15
20
15
5.8cm
20
Fig. 14. A Semi-logarithmic plot of the growth in the length of the axis of leaves of Xanthi Nc. tobacco. B Relative growth rate of the leaf axis per day. Data obtained from buds dissected at various times after topping (e) and from daily measurements of a single leaf (A)
A
a
b
o.s 1
o.~g--i-
c
o.5
d
i
o.~
i
9
!
o'5
015
g
1
0,5
1
Relative Growth Rate
B
lamina
ilirm,dr,b
1.0 .8
8.7cm
11.5cm
Fig. 16. The growth of a leaf of Xanthi Nc. tobacco from about one third to two thirds its final size. The lamina was marked with a matrix of ink dots when the leaf was 5.8 cm long
variation in the orientation of growth within the lamina is responsible for this change. This possibility was eliminated by marking leaves with a grid of equally spaced points and observing the deformation of this grid during development (Fig. 16). Although it is impossible to draw unequivocal conclusions about the orientation of growth from these data without undertaking detailed analyses like those of Richards and Kavanagh (1943), Erickson (1966) and Silk (1983), it is clear that the growth of the lamina is more or less isotropic. At least there is no evidence that growth is much more highly oriented in one region of the lamina than in another. We believe, therefore, that the change in the slope of the lamina in the br genotype is primarily the consequence of an increase in the R A of its basal region.
.6
Discussion a
b
c
d
e
f
a
b
c
d
e
f
Fig, 15. A Relative growth rate (d- 1) of the area (e) and length (o) of different transverse sections of a leaf on seven successive days. Panel a leaf length 1-1.5 cm; panel b 1.5-2.4 cm; panel c 2.4-3.7 cm; paneld 3.7-5.2 cm; panele 5.2-7.4 cm; panelf 7.4-9.6 cm; paneIg 9.6-11.3 cm. B The data in A, presented in a way that better illustrates the change in the relative growth rate of different regions of the leaf during development
region of the lamina is primarily responsible for transforming the lamina from an elliptical structure into an ovate one. This change in shape cannot be due to the prolonged growth of the base because cultivars with elliptical leaves have the same basipetal pattern of growth as Xanthi Nc. (Avery 1933; Poethig 1981). Nor it is likely that
Our observations confirm many of those made by other investigators (Avery 1933; Schwartz et al. 1952; Haber and Foard 1963; Hannam 1968; Williams 1975; Dubuc-Lebreux and Sattler 1980), but also disagree with some of these studies on a number of important points. Although some of these discrepancies may follow from the unique morphology of Xanthi Nc., most appear to have a more important basis. We will now review the major findings of this study, focusing in particular on those that conflict with existing interpretations of leaf development in tobacco.
The cell number of the primordium. Avery (1933) reported that the internal tissue of the leaf axis was derived from a single subapical initial and that
R.S. Poethig and I.M. Sussex: Morphologyand growth of tobacco leaf the internal tissue of the lamina arose from a single row of submarginal initials, but the histological structure and external morphology of the leaf primordium indicate that the primordium actually arises from a much larger group of cells. A leaf buttress has a surface area of about 50 cells and encompasses four layers of the apical meristem. This population arises within 28 h, the interval between the appearance of successive leaves. Taking the cell doubling time in the primordium to be 19 h (derived from the data of H a n n a m 1968), the number of cells in the epidermis of the primordium one plastochron before the appearance of the leaf buttress can be calculated using the formula N1 = No etln2/a, where N 1 is the number of cells in the epidermis of the buttress (50); N o is the number of cells one plastochron earlier, d is the cell doubling time (19 h), and t is the duration of the plastochron (28 h). Solving this formula for N o gives a value of 18. Therefore, on the assumption that a leaf primordium is determined no earlier than one plastochron before it becomes visible (Snow and Snow 1933), we conclude that the leaf axis in tobacco is derived from at least 18 cells in the epidermis of the apical meristem. Of course a leaf primordium does not arise solely from the epidermis of the shoot. Histological observations (Hannam 1968; this paper) and periclinal chimeras (Burk et al. 1964; Dulieu 1968; Poethig 1984)indicate that it encompasses at least two and perhaps three internal layers. Since it is reasonable to suppose that the primordium is derived from about the same number of cells (probably somewhat fewer) in these layers as in the epidermis, the tobacco leaf is estimated to arise from on the order of 50-100 initial cells. A similar argument can be made in the case of the lamina. When the lamina first becomes distinct it encompasses about 20 cells in a transverse section of the leaf axis. This population arises within less than a day. Thus, assuming that lamina initials have a cell doubling time of 13.8 h (Hannam 1968), and that the lamina is initiated in 24 h, we calculate that the lamina is derived from a minim u m of six cells in transverse section of the leaf axis. This is considered a minimum estimate because the time between the initiation and appearance of the lamina is probably less than 24 h.
The cell lineage of the leaf axis. Our observations also contradict the idea that the initials of a leaf primordium reside at the apex of the leaf axis (Avery 1933; Stewart and Dermen 1975). If this were true, one would expect cell files to emanate radially from the apex, as they do in the shoot
167
apical meristem. Instead, cell files run longitudinally in the axis and terminate along its lateral margins rather than at the apex. Thus the tobacco leaf a p p e a r s to be derived from "line source", much like a corn leaf, rather than from a " p o i n t source", such as an apical meristem.
The dynamics of leaf elongation. The relative rate of increase in the volume (Rv) of the tobacco leaf has been characterized by H a n n a m (1968), whose data were later re-examined by Williams (1975). These data, and the results presented in this paper, indicate that leaf growth can be divided into three phases: 1) a brief initiation phase characterized by a rapid but declining growth rate, 2 ) a stage of exponential growth, and 3) a maturation phase during which the relative rate of growth declines. Williams (1975) has described a similar pattern in Linum usitatissimum, Brassica oleracea and Lupinus angustifolius, and a review of published data indicates that it also exists in Xanthium pensylvanicure (Maksymowych 1959), Lactuca sativa (Bensink 1971), Cucumis sativus (Horie et al. 1979) and Trifolium repens (Denne 1966). The difference between the deceleratory and exponential growth pattern is relatively subtle and is easy to overlook. Nevertheless this distinction is real and must be explained. Williams (i 975) has argued that the dynamics of leaf growth are largely determined by physical interactions within the apical bud. According to this argument, growth is rapid early in leaf development because the leaf has not come in contact with other leaves, and that as it does, its growth rate declines and stabilizes at a lower exponential rate. As yet, however, there is no experimental support for this and any other possible explanation for the dynamics of leaf elongation. The dynamics of lamina expansion. Avery (1933) studied the dynamics of lamina expansion in tobacco by comparing the growth rate of different regions of the lamina with the growth rate of the lamina as a whole. This type of analysis permitted him to characterize the distribution growth within the lamina at a given point in time, but made it impossible to study changes in the growth rate of any particular region during the course of development. As our results demonstrate, such temporal variation in growth rate may be an important feature of leaf development. Indeed, one of the most striking aspects of leaf development in Xanthi Nc. is the dramatic increase in the R a of the basal region of the lamina that takes place after emergence. In contrast, the basal region of the lamina in Java - a variety with broad, elliptical leaves (genotype
168
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf
Br/Br) - only undergoes a transitory increase in R A after emergence (Poethig 1981). This observation lends support to the conclusion that the ovate character of the leaf of Xanthi Nc. is primarily a consequence of the increase in the R A of the leaf base, rather than a result of the prolonged growth or orientation of growth in this region. In Xanthi Nc., R A is maximal in the basalmost region of the lamina, whereas in the cultivar studied by Avery (1933) the growth rate was found to be maximal in a localized region near the center of the lamina (Richards and Kavanagh 1943). If Avery's data are correct then tobacco leaves should develop a bulge in this central region because it is surrounded by more slowly growing tissue. This rarely, if ever, occurs. Furthermore, an examination of the pattern of leaf growth in the cultivar Java, a variety of tobacco whose leaf shape is similar to that of the variety studied by Avery (1933), showed that R a is maximal near the base of the lamina, not at the center (Poethig 1981). These discrepancies cast doubt on Avery's interpretation of the pattern of growth of the tobacco leaf, and call for a more careful study of the dynamics of leaf expansion in tobacco than has hitherto been undertaken. The pattern of cell division during the expansion of the lamina. An accurate quantitative description of the pattern of cell division during leaf morphogenesis is essential for an understanding of the cellular basis of leaf shape. Among the most comprehensive of such studies are those of Fuchs (1966, 1975, 1976), Thomasson (1970) and Jeune (1972). By using cleared rather than sectioned specimens these investigators were able to obtain a remarkably detailed picture of the orientation and distribution of cell division within the plane of the lamina. They found that both of these parameters varied over time and within the plane of the lamina, and that this variation was correlated with the pattern of growth. This technique has also proved useful in the case of tobacco. Although our analysis of cleared specimens was not extensive, it nevertheless demonstrated that the frequency of cell division varied considerably within the lamina (during at least the early part of leaf expansion) in a manner consistent with the pattern of growth. The most obvious feature of this pattern is the basipetal gradient in the duration of meristematic growth. What is less well documented is the spatial variation in the frequency of cell division that occurs in meristematic tissue. This aspect of leaf development is described in more detail in the following paper (Poethig and Sussex 1985).
We are grateful to John Duesing (Ciba-Geigy Co., Greensboro, N.C., USA) for many helpful discussions, and to A. Pooley for his assistance with the scanning electron microscope. R.S.P. was supported by a National Science Foundation Predoctoral Fellowship and by National Institutes of Health Training Grant HD07180-01.
References Avery, G.S. (1933) Structure and development of the tobacco leaf. Am. J. Bot. 20, 565-592 Bensink, J. (1971.) On morphogenesis of lettuce leaves in relation to light and temperature. Meded. Landbouwhogeschool, Wageningen 71, No. 15 Bergann, F., Bergann, L. (1984) Zur Entwicklungsgeschichte des Angiospermenblattes. 4. Uber Periklinalchimfiren bei Sedum rubrotinctum R.T. Clausen, Biol. Zentralbl. 103, 147-171 Burk, L.G., Stewart, R.N., Dermen, H. (1964) Histogenesis and genetics of a plastid-controlled chlorophyll variegation in tobacco. Am. J. Bot. 51,713 724 Denne, M.P. (1966) Leaf development in Trifolium repens. Bot. Gaz. 127, 202-210 Dubuc-Lebreux, M.A., Sattler, R. (1980) D6veloppement des organes foliac6s chez Nicotiana tabacum et le probleme des m~rist~mes marginaux. Phytomorphology 30, 17-32 Dulieu, H. (1968) Emploi des chim~res chlorophylliennes pour l'~tude de l'ontog6nie foliaire. Bull. Sci. Bourgogne 25, 1-60 Dulieu, H. (1970) Les mutations somatiques induites et l'ontog~nie de la pousse feuill6e. Ann. Amelior. Plant. 20, 27-44 Dulieu, H. (1974) Somatic variations on a yellow mutant in Nicotiana tabacum L. (al +/al a2 +/a2). I. Non-reciprocal genetic events occurring in leaf cells. Mutat. Res. 25, 289-304 Dulieu, H. (1975) Somatic variations on a yellow mutant of Nicotiana tabacum L. (ai +/al a2 +/a2). II. Reciprocal genetic events occurring in leaf cells. Mutat. Res. 28, 69-77 Dulieu, H., Dalebroux, M.A. (1975) Spontaneous and induced reversion rates in a double heterozygous mutant of Nicotiana tabacum var. Xanthi nc. - dose-response relationship. Mutat. Res. 30, 63-70 Erickson, R.O. (1966) Relative elemental rates and anisotropy of growth in area: a computer programme. J. Exp. Bot. 17, 390M03 Esau, K. (1977) Anatomy of seed plants. John Wiley & Sons, New York Feder, N., O'Brien, T.P. (1968) Plant microtechnique: some principles and new methods. Am. J. Bot. 55, 123-142 Foster, A.S. (1936) Leaf differentiation in angiosperms. Bot. Rev. 2, 349-372 Fuchs, C. (1966) Observations sur l'extension en largeur du limbe foliaire du Lupinus albus L. C.R. Acad. Sci. Ser. D
263, 1212-1215 Fuchs, C. (1975) Ontogen~se foliaire et acquisition de la forme chez le Tropaeolum peregrinum L. I. Les premier stades de l'ontogen6se du lobe m6dian. Ann. Sic. Nat. Bot. Biol. Veg. 16, 321-390 Fuchs, C. (1976) Ontogen6se foliaire et acquisition de la forme chez le Tropaeolum peregrinum L. II. Le developpement du lobe apr+s la formation de lobules. Ann. Sci. Nat. Bot. Biol. Veg. 17, 121-158 Green, P.B. (1976) Growth and cell pattern formation on an axis : critique of concepts, terminology and modes of study. Bot. Gaz. 137, 187-202 Haber, A.H., Foard, D.E. (1963) Nonessentiality of concurrent cell divisions for degree of polarization of leaf growth. II. Evidence from untreated plants and chemically induced
R.S. Poethig and I.M. Sussex: Morphology and growth of tobacco leaf changes in the degree of polarization. Am. J. Bot. 50, 937 944 Hannam, R.V. (1968) Leaf growth and development in the young tobacco plant. Aust. J. Biol. Sci. 21, 855-870 Horie, T., de Wit, C.T., Goudrian, J., Bensink, J. (1979) A formal template for the development of cucumber in its vegetative state. Proc. K. Ned. Akad. Wet. C 82, 433-479 Jensen, W.A. (1968) Botanical histochemistry, W.H. Freeman and Co., San Francisco, USA Jeune, B. (1972) Observations et expbrimentation sur les feuilles juveniles du Paulownia tomentosa H. Bn. Bull. Soc. Bot. France 119, 215-230 Jeune, B. (1983) Croissance des feuilles de Lycopus europaeus L. Beitr. Biol. Pflanz. 58, 253 266 Maksymowych, R. (1959) Quantitative anaIysis of leaf development in Xanthium pensylvanicum. Am. J. Bot. 46, 635 644 Maksymowych, R., Erickson, R.O. (1960) Development of the lamina in Xanthium italicum represented by the plastochron index. Am. J. Bot. 47, 451-459 Poethig, R.S. (1981) The cellular parameters of leaf development in Nicotiana tabacum L. Ph.D. thesis, Yale University Poethig, R.S. (1984) Cellular parameters of leaf morphogenesis in maize and tobacco. In: Contemporary problems in plant anatomy, pp. 235 259, White, R.A., Dickison, W.C., eds. Academic Press, New York Poethig, R.S., Sussex, I.M. (1985) The cellular parameters of leaf development in tobacco : a clonal analysis. Planta 165, 170-184 Richards, O.W., Kavanagh, A.J. (1943) The analysis of the relative growth gradients and changing form of growing organisms: illustrated by the tobacco leaf. Am. Nat. 77, 385-395 Ruddell, C.L. (1967) Embedding media for 1 2 micron sectioning. 2. Hydroxyethyl methacrylate combined with 2-butoxyethanol. Stain Technol. 42, 253-255 Schwartz, D., Renier, A., Cuzin, J. (1952) t~tude quantitative
169
de la croissance du tabac. II. Morphog6n6se foliaire. Ann. Inst. Exp. Tabac Bergerac 1, 55-95 Silk, W.K. (1983) Kinematic analysis of leaf expansion. In: The growth and functioning of leaves, pp. 89-108, Dale, J.E., Milthorpe, F . L , eds. Cambridge University Press, Cambridge, UK Singh, I.S., Dulieu, H. (1976) Effects g~n~tiques et variabilit~ g6n6tique pour plusiers charact6res chex un mutant monohybride de Nicotiana tabacum L. Ann. Amelior. Plant. 26, 579-590 Snow, M., Snow, R. (1933) Experiments on phyltotaxis II. The effect of displacing a primordium. Phil. Trans. R. Soc. London Ser. B 222, 353-400 Stewart, R.N., Burk, L.G. (1970) Independence of tissues derived from apical layers in ontogeny of the tobacco leaf and ovary. Am. J. Bot. 57, 1010-1016 Stewart, R.N., Dermen, H. (1975) Flexibility of ontogeny as shown by the contribution of shoot apical layers to leaves of periclinal chimeras. Am. J. Bot. 62, 935-947 Tenopyr, L.A. (1918) On the constancy of cell shape in leaves of varying shape. Bull. Torrey Bot. Club 45, 51-76 Thomasson, M. (1970) Quelques observations sur la rbpartition de zones de croissance de la feuille du Jasminum nudiflorum Lindl. Candollea 25, 297-340 van der Veen, J.H. (1957) Studies on the inheritance of leaf shape in Nicotiana tabacum L. Excelsior, Oranjeplein, The Hague van der Veen, J.H., Bink, J.P.M. (1961) Multiple effects of the leaf shape allele Pt in Nicotiana tabacum L. Genetica 32, 33-50 van Papen, R. (1935) Beitr~ige zur Kenntnis des Wachstums der Blattsbreite. Bot. Arch. 37, 159 206 Williams, R.F. (1975) The shoot apex and leaf growth, Cambridge University Press, Cambridge, UK Received 12 October; accepted 21 December 1984
