CLONAL ANALYSIS OF LEAF DEVELOPMENT IN

Transcription

1 American Journal of Botany 85(3): CLONAL ANALYSIS OF LEAF DEVELOPMENT IN COTTON 1 LIAM DOLAN 2 AND R. SCOTT POETHIG Plant Science Institute, Department of Biology, University of Pennsylvania, Philadelphia, Pennsylvania Clonal analysis has been used to describe the cellular parameters of leaf development in American Pima cotton (Gossypium barbadense). Sectors (clones) induced before leaf initiation indicate that the leaf primordium arises from 100 cells on the flank of the shoot meristem. An analysis of sector frequency during the period of leaf expansion suggests that the rate of cell division is fairly uniform throughout the length of the leaf, but is lower at the margin of the lamina than in intercalary regions. The shapes of marginal sectors indicate that the orientation of cell division (as defined by the orientation of the new cell wall) in this region is more often parallel to the margin than perpendicular to it, although the degree of polarization varies along the length of the margin. There is a slight gradient in the duration of cell division along the length of of the lamina late in development, with cell division ceasing progressively from the lamina tip to the base over two cell cycles. The parameters of cell division in cotton are therefore similar to those described for tobacco with the notable exception of the behavior of cells at the leaf margin. Key words: American pima cotton; clonal analysis; Gossypium barbadense; leaf development; Malvaceae. Leaves develop from populations of cells on the flanks of shoot meristems. Cells in the different meristematic layers undergo rounds of coordinated cell division to form a primordium. Cells on the flanks of this primordium in turn undergo rounds of cell division and expansion to form the lamina. Much of the understanding of the cellular processes of leaf morphogenesis has come from quantitative histological studies on developing leaves. Although these studies provide useful information regarding the direction, duration, and relative rates of cell division in young leaves from a range of species, they give no indication of the eventual contribution of these cells to the mature structure. Misinterpretation of these studies has led to spurious conclusions concerning the roles played by particular cells during leaf morphogenesis. Clonal analysis yields information about the cellular parameters of leaf development similar to that obtained by traditional histological methods; it also provides information about cell lineage (for reviews see Poethig, 1984, 1987; Dawe and Freeling, 1990; Irish, 1991). Clonal analysis depends upon the ability of ionizing radiation to induce chromosomal rearrangements that permit the phenotypic expression of recessive cell autonomous mutations. Cell lineages derived from cells that have undergone this event are permanently marked. Irradiations carried out prior to the initiation of a structure provide information about the fate and number of founder cells that give rise to a structure, whereas irradiations carried out later in development can be used to study the relative rate and duration of cell division within a structure (Poethig, 1984; Poethig and Sussex, 1985b; Poethig and Szymkowiak, 1995). 1 Manuscript received 21 January 1997; revision accepted 7 July The authors thank Drs. E. Turcotte, R. Percy, D. Stelly, and W. Raschke for seed stocks and invaluable advice. We would also like to thank Mary Lou Oellert and Donald Verlenden and others at the greenhouse and gardens of the University of Pennsylvania, and Sue Bunnewell for her help in preparing the final figures. 2 Current address: Department of Cell Biology, John Innes Centre, Norwich, NR4 7UH, UK. 315 Tobacco is the only dicot species in which a detailed clonal analysis of leaf development has been carried out (Poethig and Sussex, 1985b). Several of the results of this study conflicted with previous descriptions of leaf development in tobacco (Avery, 1933) and other dicots (reviewed in Cusset, 1986). Among these was the observation that the margin of the lamina played a much less significant role in leaf expansion than was previously thought. In order to test the generality of this conclusion, we decided to undertake a clonal analysis of a morphologically different leaf type. American Pima cotton (Gossypium barbadense L.) was chosen for the analysis because it has lobed, palmate leaves, rather than entire, pinnate leaves as in tobacco, and because of the availability of genetic markers suitable for clonal analysis. Foliage color in Pima cotton (2n 4x 52) is determined by the relative doses of wild-type and mutant copies of the cell-autonomous V 1 and V 7 loci (Turcotte and Feaster, 1973). Plants containing two copies of the mutant v 7 gene and two copies of the wild-type V 1 gene are virescent (yellow green). More than two copies of the wild-type V 1 allele or mutant v 7 allele results in the production of darkgreen or pale-green cells, respectively. Somatic recombination between homeologous alleles (i.e., V 1 and v 7 ) results in the production of single-colored sectors or twin spots depending upon the genotype of the daughter cells (Barrow, Chaudhari, and Dunford, 1973; Barrow and Dunford, 1974). The number, size and location of these sectors are easily observed on the virescent background of V 1 V 1 v 7 v 7 cells. We took advantage of this genetic system to determine the number and fate of founder cells of the cotton leaf and to ascertain the contribution of submarginal cells to the development of the lamina. A preliminary report of some of these results has been published (Dolan and Poethig, 1991). MATERIALS AND METHODS Seed stocks A V 1 V 1 v 7 v 7 stock of the Pima S-5 cultivar of G. barbadense (American Pima cotton) was used in this study. This stock is yellow-green in appearance and produces green and white sectors upon

2 316 AMERICAN JOURNAL OF BOTANY [Vol. 85 irradiation. Seed was provided by Dr. R. Percy from the USDA-ARS Maricopa Agriculture Centre, Arizona, USA. Seeds were sown in the greenhouse in a commercial potting medium and seedlings were subsequently transferred to soil after 2 3 wk. The third and fourth leaves after the cotyledons were used for the determination of cell lineage of the leaf primordium, and leaf 3 alone was used to determine the pattern of cell division in the developing lamina and at the leaf margins. Experimental procedure Irradiations of meristems before leaf initiation were carried out on seeds that had been allowed to germinate for h in aerated water or on wet tissue in petri dishes at 30 C. Dissection of the imbibed seed and examination with a stereomicroscope indicated that only the first two leaves had been initiated at this time. Leaves 3 and 4 differ only slightly in size and shape and were used for the estimation of the number of cells that form the leaf primordium. Irradiations of leaf primordia, i.e., after leaf initiation, were carried out on seedlings that had been planted directly in soil, and were performed after the cotyledons had expanded. Because cotton exhibits pronounced heteroblasty, all analyses were carried solely on leaf 3. For later irradiations, the size of irradiated leaves was determined directly with a stereomicroscope. It was not necessary to remove any leaves to make these measurements, therefore obviating any artifacts that might result from such defoliation. Primordia were gamma irradiated with a 137 Cs source (200 or 300 R at a rate of 75 R / min). Fully expanded leaves were photographed on a glass plate elevated 15 cm above a white background, using lateral illumination. Scoring procedure Green and white sectors located in the lamina, petiole, and/or stem were scored in leaves irradiated before initiation for the determination of the apparent cell number (ACN). These leaves were subsequently photographed. For the examination of smaller sectors induced after leaf initiation, leaves were either photographed or photocopied with a transparent grid placed over their surface. Subsequently, the leaf was submerged in a tray of water and overlaid with a glass plate inscribed with a grid (25-mm 2 squares) corresponding to that used in the photocopy. The number of sectors in each square was determined using a stereomicroscope and recorded on the photocopy of the leaf. Sectors were most easily visualized using a combination of incandescent illumination from above and weak dark-field illumination from below. The mutation frequency in a local region of the lamina (number of clones divided by the number of cells present in that region at the time of irradiation) was estimated as outlined by Poethig and Sussex (1985b). For this purpose, the number of cells present in the region at the time of irradiation was estimated by dividing the number of cells in this region by the average number of cells per clone located in the same region. The number of progenitor cells in the horizontal dimension of the leaf primordium was determined by dividing half the circumference of the petiole by the width of the sector in that petiole. Mutation frequency was determined along a line parallel to, and 10 mm from, the midvein of the leaf. Marginal cell division The analysis of marginal cell division was carried out on leaf 3. The frequency of marginal and intercalary sectors was determined in leaves that were irradiated at lengths of 3.4, 4.1, 6.1 and 7.1 mm. Black and white photographic prints were made of sectored leaves and the number of sectors located along an arbitrary section of the leaf margin was determined and compared to the number of sectors traversing a parallel line of equal length drawn 5 mm from the margin of the leaf. The area of sectors at the leaf margin was compared to the area of sectors from an intercalary portion of the leaf using data obtained from photographic enlargements of the leaf. The area of these sectors was measured using a stereomicroscope equipped with an ocular grid micrometer. Samples from several leaves were used to obtain sufficient numbers of sectors. In a leaf irradiated later in development (0.8 mm in length), the average sector area was compared to intercalary sectors situated between the nearest main vein and the margin. Data from leaves irradiated even later in development ( mm in length) were taken from a single leaf at each developmental stage. In each case, the average size of a random sample of sectors traversing a line drawn 5 mm from the margin was used as the intercalary sample. Student s t test was used to determine the significance of the difference between the area of marginal and nonmarginal sectors. The dimension of sectors at the leaf margin was represented by their length:width ratio. The length of a sector was defined by the extension of a sector along an axis perpendicular to the leaf margin, while the width was defined as the greatest dimension perpendicular to this long axis. Mesophyll cell measurements Mesophyll cells were observed by viewing fresh, water- infiltrated leaf segments with an epifluorescence microscope (blue excitation filter) and measured using an ocular grid micrometer. RESULTS Approximately 100 cells form the primordium of the cotton leaf The number of founder cells that gives rise to the leaf can be estimated from sectors induced before leaf initiation. Because such events are relatively rare, each sector is assumed to represent the descendents of a single cell. The contribution of a single founder cell to the population in the leaf can therefore be determined from the size of the sector produced by this marked cell. Specifically, the inverse of the fraction of the leaf occupied by the sector gives an estimate of the number of founder cells present in the leaf primordium at the time of irradiation (i.e., its apparent cell number, ACN) (Coe and Neuffer, 1986). Different regions of the lamina expand to different extents; consequently, the dimensions of the leaf primordium are most readily estimated from sectors located in the petiole (see Materials and Methods). Sectors induced in the mesophyll one to two plastochrons before leaf initiation indicate that leaves 3 and 4 arise from 11 ( ; N 14) cells in the horizontal dimension of the shoot meristem (Fig. 1). Because sectors induced in the G1 phase result in sectors that are twice as large as sectors induced after DNA replication in the G2 phase (McDaniel and Poethig, 1988), estimates of ACN based on individual sectors should range twofold. In this experiment, single sectors produced ACN estimates ranging from seven cells to 18 cells in the horizontal direction. The minimum number of founder cells in the vertical dimension of the leaf primordium can be estimated from the positions that sectors occupy within the leaf (Fig. 1). Sectors induced prior to leaf initiation suggest that the cotton leaf primordium encompasses at least three cells in the vertical dimension. This conclusion is based on observations that there are three types of sectors within the subepidermal layer of the leaf: (1) sectors restricted to the distal portion of the lamina, (2) sectors restricted to the central portion of the lamina that do not extend to the lamina margin or the petiole, and (3) sectors that run through the leaf from the internode above to the internode below. If the leaf primordium were derived from a single horizontal row of cells, then sectors in the subepidermal layer of the leaf should extend continuously around the leaf. The existence of sectors isolated in the middle of the leaf (types 1 and 2, above) demonstrates that the primordium must be composed of at least three cells in the

3 March 1998] DOLAN AND POETHIG COTTON LEAF CELL LINEAGE 317 Fig. 1. The size and distribution of sectors induced one to two plastochrons before leaf initiation. (A) Schematic illustration of the different types and distribution of sectors observed in this study. (B), (C) Leaves with sectors. vertical direction. Periclinal chimeras demonstrate that at least three cell layers contribute to the leaf primordium in cotton (Dolan and Poethig, unpublished data). Since the primordium contains at least 11 cells in its horizontal dimension and three cells in the vertical dimension, it can be concluded that 100 cells form the leaf primordium (three layers with 33 cells per layer). Nevertheless the large variability in sector size, which predicts a horizontal ACN of 7-13 cells, indicates that the number of cells that form the leaf primordium could, in principle, range from 63 to 162 cells. Cells that form the primordium have variable fates The fates of the founder cells that contribute to the leaf were determined from an analysis of clones induced one to two plastochrons before the leaf primordium was visible as a bump at the surface of the shoot. Sectors induced at this stage in development run along any of the three main veins, enter the lamina, and may abut the leaf margin as illustrated by Fig. 1. Some of these sectors are also present in the petiole. Sectors in any one region of the leaf are generally similar in size and shape in different leaves, but the shape of sectors varies in different parts of the leaf blade. Sectors in the basal region of the lamina, for example, tend to be broader than sectors in more apical regions, suggesting that the overall orientation of cell division is different in these regions of the leaf. Some of the sectors that abut the leaf margin cross from adaxial side to the abaxial side or vice versa, suggesting that these domains of the leaf are not clonally independent, as in tobacco. The sinus between two lobes is formed by a single lineage that arises at the base of the leaf where the main and lateral veins intersect. Sectors rarely cross from one side of the main vein to the other, except near the leaf tip. This suggests either that the vein does not differentiate at the tip of the leaf until relatively late in leaf development, or that fewer cells are sequestered into the vein at the leaf tip than in basal regions of the leaf (see Poethig and Sussex, 1985b). A diagrammatic representation of the contribution of founder cells to the growth of the cotton leaf is presented in Fig. 2. Variation in sector frequency and size in the developing lamina In order to interpret variation in the size and frequency of sectors throughout the lamina, it was first important to determine whether this variation was due to variation in cell size. The average area of palisade mesophyll cells in several locations throughout the leaf (Fig. 3) suggests that there are no major differences in cell size in intercalary regions of the lamina, although cells are slightly smaller at the tip of the leaf. In contrast, cells at the margin of the leaf are approximately half the size of cells in the intercalary regions (Table 1). Thus, Fig. 2. Schematic representation of the lineages in the leaf based on sectors induced before initiation. Fig. 3. The distribution of palisade cell areas in the lamina of leaf 3 (mature). Cell areas are represented in m 2.

4 318 AMERICAN JOURNAL OF BOTANY [Vol. 85 TABLE 1. The area ( m 2 1 SE) of palisade cells in intercalary and marginal regions of the leaf. Location Marginal Intercalary M/I variation in cell size may account for some of the difference between the relative size of sectors at the leaf margin and in adjacent intercalary regions of the lamina, but probably does not contribute significantly to variation in these parameters along the length of the leaf. The mutation frequency (number of sectors per total number of cells present at the time of irradiation) takes into account the number of cells present at the time of irradiation and is therefore a more accurate indicator of rate of cell division than the frequency of sectors in a unit area of a mature leaf (Poethig and Sussex, 1985a, b). The mutation frequency was estimated along the length of the leaf by taking into account the predicted number of cells present in various regions of the leaf at the time of irradiation, as outlined in the Materials and Methods. The data from a leaf irradiated at 7.1 mm (Fig. 4) indicate that the rates of cell division are relatively uniform throughout the length of the lamina, although there may be some local variation (Fig. 4B, C). Because the rate of cell division is similar throughout much of the length of the leaf, the proximal-distal gradient in sector size (Fig. 4D) must be due largely to differences in the duration of cell division. The fourfold difference in clone size between the tip and the base of the cotton leaf indicates that cell division ceases at the base of the leaf approximately two cell cycles after it ceases at the tip. The pattern of cell division at the leaf margin To investigate the role played by the marginal subepidermal cells in lamina development, the frequencies, size, shape, and orientation of sectors at the margin and in intercalary regions of the lamina were compared (Fig. 5). 1) Sector frequencies. The frequency of marginal sectors in leaves irradiated at various stages of development is presented in Table 2. Sector frequency at the leaf margin and in the interior portion of the leaf was determined in leaves irradiated after the lamina had been initiated, when primordia were 3.4, 4.1, 6.1, and 7.1 mm in length. Over 200 sectors scored in the intercalary and marginal region of these four leaves yielded average sector frequencies of 0.14 sectors/mm and 0.13 sectors/mm, respectively. Although these data suggest that the rates of cell division are similar at the margin and in the intercalary region during the development of the lamina, it is more appropriate to represent sector frequency in terms Fig. 4. Characterization of sectors induced in a leaf primordium 7.1 mm long. (A) The fully grown leaf used for this analysis; the sectors are too small to be seen at this magnification. (B) The distribution of sector frequency along the length of the leaf. (C) The distribution of mutation frequency along the length of the leaf. (D) The log of sector area plotted as a function of leaf length.

5 March 1998] DOLAN AND POETHIG COTTON LEAF CELL LINEAGE 319 TABLE 2. The frequency of sectors in intercalary and marginal regions of the lamina. Primordium size a (mm) a At time of irradiation. Intercalary frequency (sectors/mm) Marginal frequency (sectors/mm) Fig. 5. Marginal sectors. (A D) sectors induced when primordium was between 3.4 and 7.1mm in length. Leaves are oriented with distal regions (leaf tip) upward and proximal regions (leaf base) at the bottom of the figure. (E) Composite schematic showing the shapes and locations of marginal sectors induced after the initiation of the lamina. of numbers of sectors per cells rather than number of sectors per unit area. We did not determine cell number in irradiated leaves, but our analysis of cell size in unirradiated controls demonstrated that cells at the leaf margin are about half the size of cells in intercalary parts of the leaf (Table 1). Thus, a unit area at the leaf margin contains twice as many cells as the same size region in the intercalary part of the leaf, implying that sector frequency at the leaf margin is actually lower (on a per cell basis) than in intercalary regions. 2) Sector area. A comparison was made of sector areas in marginal and intercalary regions of leaves irradiated at different stages of development (a total of 326 sectors were examined). Since mutation frequency and, therefore, the rate of cell division are uniform throughout the leaf, sector areas reflect the duration of cell division. In all but the final two irradiations, there was no significant difference between the area of marginal and intercalary sectors (Table 3). In these later irradiations (carried out at lengths of 6.1 and 7.1 mm) marginal sectors are approximately half the area of intercalary sectors. The twofold difference in cell size between the margin and intercalary regions of the leaf (Table 1) accounts for the difference in sector area between marginal and intercalary regions of leaves irradiated late in development. Cell size has less of an effect on marginal sector area in leaves irradiated earlier in development because the sectors in these leaves are relatively large and encompass both large and small cells. Small cells therefore represent a much greater fraction of the cells in marginal sectors induced later in development and would be expected to reduce their area significantly. Therefore cell division ceases at approximately the same time at leaf margins as in the surrounding intercalary regions of the lamina. 3) Shape of marginal sectors. Fig. 5 shows the shape of sectors induced in different portions of the lamina early in development. Sector length is defined as the extent of the sector along an axis perpendicular to the leaf margin, while width is defined as their extent perpendicular to this length axis. Sectors exhibit a range of shapes. Sectors near the tips of lobes and at the leaf base occasionally exhibit length/width values of less than one, indicating that they are elongated parallel to the leaf margin. Otherwise, sectors are generally elongated perpendicular to the margin. An analysis of sector shape at the margin of the third leaf showed that the sectors are, on TABLE 3. The size ( 1 SD) of sectors in intercalary and marginal regions of the lamina. Primordium size a Mean sector area (mm2 ) Intercalary Marginal df P Early b Early b 0.8 mm 3.4 mm 4.1 mm 6.1 mm 7.1 mm c 0.01 c a At the time of irradiation. b Primordium size was not determined for these leaves, but based on sector size they were probably between 0.8 mm and 3.4 mm in length. c Statistically significant.

6 320 AMERICAN JOURNAL OF BOTANY [Vol. 85 average, 1.5 times longer than wide, with a mean length of mm and a mean width of mm. DISCUSSION A clonal analysis of leaf development in cotton was carried out to determine the generality of previously described cellular parameters of leaf development in tobacco (Poethig and Sussex, 1985b). The results may be summarized as follows: (1) the cotton leaf is derived from a population of 100 cells. (2) Marginal mesophyll cells contribute more to the growth of the lamina in cotton than in tobacco because of a difference in the orientation of cell division in this region. However, in cotton, as in tobacco, the rate and duration of cell division at the leaf margin are either the same or slightly less than in intercalary regions of the leaf. (3) Cell division ceases basipetally in the lamina over the course of about two cell cycles. Thus, in general, the cell lineage of the cotton leaf is similar to that of tobacco, although there are some obvious differences that may have some significance for the morphogenesis of these two different leaf types. The number and fate of founder cells Approximately 100 cells are sequestered from the flank of the shoot meristem during the initiation of the cotton leaf primordium. A similar number of leaf founder cells has been estimated for both tobacco and maize using similar methods (Poethig, 1984; Poethig and Sussex, 1985b; Poethig and Szymkowiak, 1995). In the transverse direction of the cotton leaf primordium there are 11 cells, compared to 13 cells in tobacco. As demonstrated by the variability in sector size and position, the fate of these founder cells is quite variable. Furthermore, if the leaf were derived from a set of apical initials, many of the sectors induced early in leaf development would run from the leaf tip to its base. Because few of the sectors induced early in development run to the tip of the leaf, it is clear that the the leaf is not derived from a group of apical initials. Sectors induced early in leaf development run from the midveins through the lamina to various regions of the leaf margin. As in tobacco, sectors rarely cross over these major veins (see Fig. 1A, B, C), except at the tip of the leaf. In tobacco, this was interpreted to result from the fact that the vasculature occupies a majority of the tissue in the lamina early in leaf development (Poethig and Sussex, 1985a). The morphology of the cotton primordium (Hammond, 1941; L. Dolan, unpublished data) suggests that this may also be the case in this species. Development of the lamina The distribution of sector frequencies and mutation frequencies suggests that the rate of cell division is relatively uniform throughout the length of the developing lamina, unlike tobacco where there are significant differences between the tip and the base. Previous studies involving the analysis of mitotic frequencies in whole mounts of young leaves of a variety of species (Thomasson, 1970; Fuchs 1975; Jeune, 1978, 1984) have revealed that the rate of cell division is fairly uniform in the proximal-distal dimension of the lamina during the early phase of leaf development, when the primordium is entirely meristematic. Later in development a proximal-distal gradient in the rate of cell division develops due to the cessation of cell division in apical regions. Conventional histological methods and clonal analyses reveal that cell division ceases progressively from tip to base over six cell cycles in tobacco (Poethig and Sussex, 1985a). The three to fourfold difference in sector size between the tip and base of the cotton leaf indicates that cell division also ceases basipetally in this species, but over a period of only two cell cycles. Cell division at the margin Sectors induced in marginal mesophyll cells (those mesophyll cells that lie below the epidermis at the leaf edge) of the tobacco leaf are narrow and elongated parallel to the margin. This indicates that these cells contribute little to lamina growth and that the intercalary cell division accounts for most of the cell division in the developing lamina. However, in cotton the orientation of the marginal sectors relative to the leaf margin varies depending on the location of the sector in the leaf. Marginal sectors near the leaf tip or in the basal regions of the leaf can be elongated parallel to the leaf margin. Nevertheless, these sectors are much less polarized than in tobacco, indicating that marginal cells in these regions of the cotton leaf contribute considerably to the development of the lamina. The difference in the behavior of marginal cells in cotton and tobacco may account at least in part for the development of the different leaf shapes in these species, although exactly how these marginal cell divisions might impinge upon leaf shape is unclear. In other regions of the leaf, sectors are oriented approximately perpendicular to the leaf margin, and the earlier in development they are induced the farther into the lamina they extend. This suggests that the orientation of marginal cell divisions is such that more cells are dividing parallel rather than perpendicular to the leaf edge late in lamina development. The size of these sectors indicates that marginal cells divide for approximately the same length of time as intercalary cells and at approximately the same or at a slightly slower rate. Therefore, marginal and intercalary cell lineages contribute equally to the expansion of the lamina. The results presented in this study show that although the marginal cells contribute to the cell population of the developing leaf, the cotton leaf does not have a marginal meristem in the sense described by Avery (1933). If the lamina were derived exclusively from a group of marginal cells (a marginal meristem), most sectors would be located at or near the margin and sectors induced early in development would run a great distance into the lamina. Such sectors are never observed. Analysis of mitotic indices and the direction of cell division in sectioned specimens of Xanthium, Lupinus and tobacco (Maksymowych and Erickson, 1960; Fuchs, 1968; Dubuc-Lebreaux and Sattler, 1980) provided no evidence for the existence of a marginal meristem. A quantitative analysis of mitotic figures in whole-mount specimens of developing leaves at different developmental stages in Jasminum (Thomasson, 1970), Tropaeolum (Fuchs 1975), and several other species (Jeune, 1978, 1984) also failed to show the existence of a marginal meristem. Each of these studies revealed that cell division takes place throughout the developing lamina, and that the frequency of mitosis at the leaf margin was either identical to or lower than the frequency of mitosis in the intercalary part of the

7 March 1998] DOLAN AND POETHIG COTTON LEAF CELL LINEAGE 321 lamina (for a review of histological data relating to this issue see, Cusset, 1986). In conclusion, the cellular parameters of leaf development in cotton appear to be generally similar to those of tobacco. This raises the issue of the relationship between these cellular parameters and leaf morphology. Although several investigators have demonstrated a relationship between the rate of cell division and regional variation in the expansion of the lamina (e.g., Poethig and Sussex, 1985a), there is no evidence that these are causally related. Nor can an explanation for leaf morphology be found in cell lineage patterns, since the fate of any particular primordial cell is quite variable. The goal of future studies must be to define the levels at which leaf morphogenesis is regulated and the factors involved in this process. LITERATURE CITED AVERY, G. S Structure and development of the tobacco leaf. American Journal of Botany 20: BARROW, J. R., AND M. P. DUNFORD Somatic crossing over as a cause of chromosome multivalents in cotton. Journal of Heredity 64: 3 7., H. CHAUDHARI, AND M. P. DUNFORD Twin spots on leaves of homozygous cotton. Journal of Heredity 6: 3 7. COE, E. H., JR., AND M. G. NEUFFER Embryo cells and their destiny in the corn plant. In S. Subtelny and I. M. Sussex [eds.], The clonal basis of development, Academic Press, New York, NY. CUSSET, G La morphogenèse du limbe des Dicotylédones. Canadian Journal of Botany 64: DAWE, R. K., AND M. FREELING Cell lineage and its consequences in higher plants. Plant Journal 1: 3 8. DOLAN, L., AND R. S. POETHIG Genetic analysis of leaf development in cotton (Gossypium barbadense L.). Development (supplement) 1: DUBUC-LEBREUX, M. A., AND R. A. SATTLER Dévelopment des organes foliacés chez Nicotiana tabacum et la probléme des meristémes marginaux. Phytomorphology 30: FUCHS, M. C Localisation des divisions dans le meristéme marginaul des feuilles des Lupinus albus L., Tropaeolum peregrinum L., Limonium sinuatum L. Miller et Nemophila maculata Benth. Comptes Hebdomadaires des Seances de l Academie des Sciences 267: Ontogenése foliare et acquisition de la forme chez le Tropaeolum peregrinum L. I. Les premiers stades de l ontogenése du lobe médian. Annales des Sciences Naturalles Botanique et Biologie Vegetale 16: HAMMOND, D. 1941a. The expression of genes for leaf shape in Gossypium hirsutum L. and Gossypium arboreum L. I. The expression of genes for leaf shape in Gossypium hirsutum. American Journal of Botany 28: IRISH, V. F Cell lineage in plant development. Current Opinion in Genetics and Development 1: JEUNE, B Sur le déterminisme de la forme de feuilles de dicotylédones. Adansonia 18: Position et orientation des mitoses dans la zone organogéne de jeunes feuilles de Fraxinus excelsior, Glechoma hederacea et Lycopus europaeus. Canadian Journal of Botany 62: MAKSYMOWYCH, R., AND R. O. ERICKSON Development of the lamina in Xanthium italicum represented by the plastochron index. American Journal of Botany 47: MCDANIEL, C. N., AND R. S. POETHIG Cell lineage patterns in the shoot apical meristem of the germinating maize embryo. Planta 175: POETHIG, R. S Cellular parameters of leaf morphogenesis in maize and tobacco. In R. A. White and W. C. Dickinson [eds.], Contemporary problems in plant anatomy, , Academic Press, New York, NY Clonal analysis of cell lineage patterns in plant development. American Journal of Botany 74: , AND I. M. SUSSEX. 1985a. The developmental anatomy and growth dynamics of the tobacco leaf. Planta 165: , AND. 1985b. The cellular parameters of leaf development in tobacco: a clonal analysis. Planta 165: , AND E. J. SZYMKOWIAK Clonal analysis of leaf development in maize Maydica 40: THOMASSON, M Quelques observations sur la répartition de zones de croissance de la feuille du Jasminum nudiflorum Lindl. Candolea 25: TURCOTTE, E. L., AND C. V. FEASTER The interaction of two genes for yellow foliage in cotton. Journal of Heredity 64: