Analysis of Clonal Sectors of Altered Epidermis on EMS Treated Arabidopsis Plants

M. David Marks (1), David G. Oppenheimer (2), and Edward Garon (1)

1) Department of Genetics and Cell Biology and
Plant Molecular and Genetics Institute
University of Minnesota
St. Paul, MN 55108
Phone # (612) 625-6737

2) Present Address
Department of Biological Sciences
University of Alabama
Tuscaloosa, AL 35487


Wild-type Arabidopsis seeds were treated with EMS and then grown to maturity. Plants were analyzed for altered epidermal sectors. Sectors of three general types were seen. Both trichomes and pavement cells of some sectors were altered. Some sectors contained normal pavement cells and trichomes with previously characterized mutant phenotypes. Finally, one sector contained altered trichome subsidiary cells.


Wild-type Arabidopsis seeds were treated with EMS and the resulting seedlings were examined for sectors of altered epidermis. This control experiment is a prelude to generating sectors homozygous for embryo lethal gene mutations (emb/emb; (Meinke and Sussex, 1979) ) in heterozygous plants (emb/+). This approach has been used to uncover seedling functions for the FUS1 gene (Miséra et al., 1994) . Mutations in FUS1 result in accumulation of anthocyanin during late embryogenesis and seedling lethality. To generate fus1/fus1 somatic sectors in adult plants, seeds of fus1/+ plants were treated EMS. Approximately 1% of the resulting seedlings displayed subepidermal purple sectors which had a phenotype similar to the subepidermis of dying fus1 mutant seedlings. These results support the validity of the clonal sector approach to uncovering somatic functions for EMB genes. However, in contrast to the fus1 analysis in which a particular type of phenotype might be predicted, the seedling phenotypes of most emb mutations cannot be predicted. Therefore, it is important to know the types of sectors that might be encountered on wild-type control plants. In this report, we document the occurrence of mutant epidermal sectors on plants arising from EMS treated wild-type seeds.

Material and methods

Arabidopsis seeds (WS ecotype) were suspended in a solution of either 0.4% or 0.6% EMS for 8 to 10 hours. Seeds were rinsed several times with water, sown on flats of vermiculite saturated with modified Hoaglands solution, and grown under previously described conditions (Feldmann and Marks, 1987) . After one week, the number of surviving seedlings was assessed and during weeks two and three, plants containing epidermal sectors were identified by analysis under a Nikon SMZ-U stereomicroscope. Stereoscopic images were generated using ring fluorescent lighting and a Nikon AFX-IIA controller. Scanning Electron Microscope (SEM) images were made with a Phillips 500 SEM microscope containing a cryostage. Samples were pre-frozen in liquid nitrogen and scanned at 1.5 kV under various magnifications.


The epidermal surface of Arabidopsis leaves contains four general types of differentiated epidermal cells. As shown in Figure 1, these are pavement cells, guard cells, trichomes, and trichome subsidiary cells. Wild-type trichomes contain a stalk and usually between two and four branches. The surface of the trichomes contains numerous bumps or papillae. Distinct subsidiary cells surround the base of trichomes. The pavement cells are the most abundant cell type and acquire a jig-saw puzzle piece shape on mature leaves. The guard cells, which comprise the stomata, are the smallest cell type.

Figure 1. SEM of wild-type leaf epidermal cells. A) The solid arrow denotes a mature pavement cell and the open arrow highlights a guard cell. The bar in the lower left represents 10 µ. B) The solid arrows points to a mature trichome and the open arrow denotes a trichome subsidiary cell. An immature trichome with a smooth outer surface also is present. The bar in the lower left represents 13 µ.

In three separate experiments, 5000 wild type seeds were treated with either 0.4% EMS (twice) or 0.6% EMS (once) for 10 hours and 8 hours, respectively (see Table 1). Between 25% and 34% of the treated seeds germinated and survived. The number of germinating seedlings that did not survive beyond the two leaf stage was not determined. The seedlings were analyzed for the presence of visible chlorophyll and epidermal surface alterations (Table 1). Depending on the treatment, between approximately 6% and 14% of the surviving plants exhibited sectors deficient in chlorophyll synthesis (see figures 1A and B). The chlorophyll sectors were very similar in size and position to those reported previously .

Table 1

Treatment Survivors(a) % Chlorophyll Sectors(b)% Epidermal Sectors
0.4% EMS, 10 hrs X112608.21.2
0.4% EMS, 10 hrs X2199213.82.7
0.6% EMS, 8 hrs17165.61.3

Table 1 as text

(a) number of seedlings out of 5000 seeds for each treatment that survived to the two leaf stage.
(b) percent of surviving seedlings with either chlorophyll deficient or alter epidermal sectors.

Between 1.2% and 2.7% of surviving plants had altered sectors of leaf epidermis. In some cases, the leaf shape was not changed by the sector; in other cases, leaf shape was drastically altered. The shape of the sectors on normal shaped leaves did not differ significantly from that of the chlorophyll sectors (see figures 2 and 3). This suggests that the fate map for the epidermis is likely to be similar to that previously reported for the subepidermis .
A common type of epidermal alteration is shown in figures 2D and 3A and B. Eleven plants (0.2%) contained sectors of "bubbled" epidermis. The cells were rounded and had a glassy appearance. As seen in figure 3A and B, the commitment to form trichomes apparently was not altered. These sectors contained trichome-like cells that have a phenotype similar to those on distorted trichome mutants (dis; (Haughn and Somerville, 1988; Marks et al., 1991; Marks and Esch, 1992) ). However, they can be distinguished from true dis trichomes by the lack of surface papillae (compare figures 3B and 3D). Stomata are not apparent; however, the small rounded cells could be disfigured guard cells.

Figure 2. Analysis of whole leaf sectors. A) and B) Leaves containing subepidermal chlorophyll deficient sectors. C) Normal shaped leaf with a mutant trichome sector. D) Abnormal leaf with a bubbled cell sector. The mutant cells are abnormally shaped and cause the leaf to expand asymmetrically. E) Wild-type leaves.

Two other examples of general defects in epidermal cell expansion are shown in figure 3. Figures 3C and 3D show a sector containing dis-like trichomes. However, unlike dis1 and dis2 mutants, the shape of the pavement cells in the sectors are altered (figure 3D). The epidermal cells are elongated and do not exhibit the typical jigsaw puzzle shape of ordinary epidermal cells.
A unique sector (found on only one seedling) running through the center of a leaf is shown in figures 3 E through 3H. Two types of trichomes are seen in this sector. Slightly distorted unbranched trichomes are present toward either edge of the sector. The surface of the bottom half of these trichomes exhibit papillae, but the top half of the trichomes are smooth. This type of mutant trichome has not been reported previously. Toward the middle of the sector, aborted trichomes are found. These trichomes have an appearance similar to those found on gl2 mutants (Rerie et al., 1994) . The epidermal cells between the mutant trichomes also are altered. Pavement cells in the sector are shaped abnormally and guard cells are very rare. Given the gradient of trichome phenotypes in this sector, it is possible that the underlying mutation is acting in a non-cell autonomous fashion.

Figure 3. SEM analysis of mutant leaf sectors affected in both trichome formation and general epidermal cell expansion. A) and B) Low and high magnifications of a bubbled cell sector. Rudimentary trichomes also are visible in the sector. Guard cells are not seen. C) A mutant sector containing distorted-like trichomes and abnormally shaped epidermal cells is present toward the bottom half of the figure. D) Higher magnification of a distorted-like trichome present in the mutant sector shown in C). E) Normal shaped leaf with sector containing abnormal trichomes and epidermal cells. F) High magnification of sector border. The mutant sector contains under-expanded pavement cells and lacks guard cells. G) Toward the edge of the sector shown in E) the trichomes are unbranched. The lower half of the surface of these trichomes is papillate in appearance whereas the top half is smooth. H) Only rudimentary trichomes are present in the middle of the sector shown in E). The bar in the lower left of each figure represents 100 µ.

Eighteen seedlings had sectors with mutant trichomes and normal intervening pavement and guard cells (figure 4). Figure 4A shows a leaf with a sector of unbranched trichomes. These unbranched trichomes are similar to those found on stichel (sti) mutants (Hülskamp et al., 1994) . The mutant trichomes in figures 4C and 4D resemble trichomes found on gl2 mutants and those in figure 4 E through 4H resemble those found on the dis1 and dis2 mutants (Haughn and Somerville, 1988; Marks et al., 1991; Marks and Esch, 1992)

Figure 4. SEM analysis of mutant sectors only affected in trichome formation. A) and B) Sector containing unbranched trichomes. C) and D) Sector containing rudimentary trichomes. E) and F) Sector containing distorted-like trichomes. G) and H) Sector containing trichomes with an abnormally expanded stalk. A), C), E), and G) are lower magnifications of the same images shown in B), D), F), and H), respectively. The bar in the lower left of each figure represents 100 µ.

One leaf sector contained trichomes with a novel phenotype (figure 5). The trichomes were more slender than wild-type and exhibited an altered branching pattern. The morphology of the subsidiary cells also was altered. These cells expanded perpendicular to the epidermal surface resulting in elevated trichomes.

Figure 5. SEM analysis of a mutant sector affected in trichome subsidiary cell expansion and trichome branch formation. A) Field of trichomes on a base of abnormally expanded subsidiary cells. B) Higher magnification of an abnormally branched trichome on a base of longitudinally expanded subsidiary cells. The bar in the lower left of each figure represents 100 µ.


Treatment of wild-type seeds of Arabidopsis resulted in plants with altered epidermal sectors. These sectors fell into two general classes. In the first class, all cells in a sector were altered (see figure 3), the most common phenotype being bubble shaped cells (See figures 2 D, 3 A and B). Other less extreme phenotypes also were seen. For example, figures 2C and 2D show a sector containing dis trichomes and intervening epidermal cells with a slightly altered morphology. Another type of alteration resulted in sectors with epidermal phenotypes similar to those of known trichome mutants (see figure 4). In these cases, the intervening pavement cells were normal. A unique class was represented by a single mutant sector with a normal epidermis, but containing trichomes with a new phenotype (figure 5). These trichomes exhibited an altered branching pattern and were perched atop a tall base of subsidiary cells.

The molecular nature of the lesions that result in mutant sectors is unknown. EMS is believed to induce alterations such as single base pair changes and chromosome breakage. To account for the sectors on diploid wild-type plants, four possibilities are proposed. First, a mutation could result in the formation of a dominant antimorphic allele that inhibits the function of a wild-type allele. An example of such a mutation is found in maize. The C1-I allele can inhibit the wild-type C1 allele's ability to induce the expression of genes required for anthocyanin pigment production (Goff et al., 1991) . Second, a sector could be induced by a dominant gain of function mutation that results in either ectopic expression or overexpression of a gene. Knotted-1 provides an example of a mutant phenotype resulting from ectopic expression (Smith et al., 1992) . Third, a mutation could be recessive. These recessive mutations could be uncovered or perhaps generated by some type of somatic recombination, gene conversion, or paramutation-like mechanism. Finally, the mutations could be due to the incomplete dominance of a wild-type allele.

These results indicate that it should be possible to use clonal analysis to uncover somatic functions for embryo lethal genes. The likelihood that such an approach will work is bolstered by the phenotype of the plants with bubble-like sectors. It is highly unlikely that embryos homozygous for such mutations would be viable. Indeed, the shape of the cells in the early embryos of the raspberry1 and raspberry2 mutants is similar to that found in these sectors (Yadegari et al., 1994) . However, the results in this report also indicate that caution should be exercised in using a clonal analysis to uncover vegetative functions of genes required for embryogenesis. The EMS treatments used in this study were similar to those previously used to generate somatic sectors in Arabidopsis. For example, (Hülskamp, et al., 1994) used 0.4% or 0.6% EMS for eight hours and they reported a rate of 15 trichome sectors per 5,000 plants (0.3%) heterozygous for a particular trichome mutation. This is approximately the same rate (0.2%) at which the bubble-like sectors were generated in wild-type plants in this study.


We thank Pam VanderWiel, Scott Sattler, Susan Pollock, and Dan Szymanski for helpful comments. This work was supported by National Science Foundation grant DCB9118306 and USDA grant 930372.


Feldmann, K. A. and Marks, M. D. (1987). Agrobacterium-mediated transformation of germinating seeds of Arabidopsis thaliana: A non-tissue culture approach. Mol. Gen. Genet. 208, 1-9.

Goff, S. A., Cone, K. C. and Fromm, M. E. (1991). Identification of functional domains in the maize transcriptional activator C1: comparison of wild-type and dominant inhibitor proteins. Genes Devel 5, 298-309.

Haughn, G. W. and Somerville, C. R. (1988). Genetic control of morphogenesis in Arabidopsis. Dev. Genet. 9, 73-89.

Hülskamp, M., Miséra, S. and Jürgens, G. (1994). Genetic dissection of trichome cell development in Arabidopsis. Cell 76, 555-566.

Irish, V. F. (1992). The fate map of the Arabidopsis embryonic shoot apical meristem. Development 115, 745-753.

Marks, M. D., Esch, J., Herman, P., Sivakumaran, S. and Oppenheimer, D. (1991). A model for cell-type determination and differentiation in plants. In Molecular Biology of Plant Development. G. Jenkins and W. Schuch, ed. (Cambridge, CB2 3EJ: The Company of Biologists Limited). 77-87.

Marks, M. D. and Esch, J. J. (1992). Trichome formation in Arabidopsis as a genetic model for studying cell expansion. Current Topics in Plant Biochemistry and Physiology 11, 131-142.

Meinke, D. and Sussex, I. (1979). Embryo-lethal mutants of Arabidopsis thaliana. Dev. Biol. 72, 50-61.

Miséra, S., Müller, A., Weiland-Heidecker, U. and Jürgens, G. (1994). The FUSCA genes of Arabidopsis: negative regulators of light responses. Mol. Gen. Genet. 244, 242-252.

Rerie, W. G., Feldmann, K. A. and Marks, M. D. (1994). The GLABRA2 gene encodesa homeo domain protein required for normal trichome development in Arabidopsis. Genes Devel 8, 1388-1399.

Smith, L. G., Greene, B., Veit, B. and Hake, S. (1992). A dominant mutation in the maize homeobox gene, Knotted-1, causes its ectopic expression in leaf cells with altered fates. Development 116, 21-30.

Yadegari, R., de Paiva, G., Laux, T., Koltunow, A., Apuya, N., Zimmerman, J., Fischer, R., Harada, J. and Goldberg, R. (1994). Cell differentiation and morphogenesis are uncoupled in Arabidopsis raspberry embryos. Plant Cell 6, 1713-1729.