Induction of epidermal cell fate in Arabidopsis shoots

Abstract

Land plants have evolved a cuticle-bearing epidermis to protect themselves from environmental stress and pathogen attack. Despite its important role, little is known about the molecular mechanisms regulating shoot epidermal cell identity. In a recent study, we found that the Arabidopsis thaliana ATML1 gene is possibly a master regulator of shoot epidermal cell fate. We revealed that ATML1 has the ability to confer shoot epidermis-related traits to non-epidermal cells of the seedlings. These data are consistent with the previous loss-of-function mutant analyses, which implied a positive role of ATML1 in epidermal cell differentiation. Importantly, ectopic epidermal cells induced in ATML1-overexpressing lines provide a novel tool to assess the intrinsic properties of epidermal cells and to study epistatic interactions among genes involved in epidermal/mesophyll differentiation. Using this system, we obtained data revealing that ATML1 negatively influenced mesophyll cell fate. In addition, we provided a working model of how division planes in epidermal cells are determined.

The specification of plant shoot epidermal cells during development remains unknown. It has been suggested that ATML1, an HD-ZIP class IV transcription factor, is a key regulator of epidermal cell differentiation.¹ ATML1 is specifically expressed in the shoot epidermis, and loss-of-function mutations in ATML1 and its closest homolog PDF2 severely affect epidermal cell differentiation of leaves.^1-4 We recently reported ATML1 gain-of-function experiments for the first time, and revealed that ATML1 is able to confer epidermal cell fate in non-epidermal tissues of the seedlings.⁵ In this addendum, we summarize and further discuss our major findings and provide a perspective for future research in this field.

ATML1 Activates the Expression of Epidermis-Specific Genes

Our overexpression studies revealed that ATML1 activated expression of several epidermis-specific genes, which contain ATML1-binding sites (L1 box) in their promoters, suggesting that ATML1 acts on the promoters of these genes.^5,6 However, not all the L1 box-containing genes tested were upregulated in ATML1-overexpressing lines.⁵ For example, PDF1 was not significantly upregulated in ATML1-overexpressing plants, whereas loss of ATML1 and PDF2 activity resulted in decreased PDF1 expression.^1,5 In addition, endogenous ATML1 mRNA was slightly increased in ATML1-overexpressing plants, although the ATML1 promoter was ectopically activated in the inner tissues of these plants.⁵ These results may suggest a negative feedback regulation of transcription levels or stability of endogenous ATML1 and PDF1 mRNA to offset excess accumulation of these gene products. Alternatively, these observations may indicate that other co-factors are required to activate these genes at a higher level.

In general, positive feedback regulation potentially amplifies subtle changes in gene expression, which could make a system unstable and may lead to gene misregulation. Therefore, it is possible that ATML1-mediated positive feedback is tightly controlled by other factors to avoid misexpression of genes.

ATML1 is a Positive Regulator of Shoot Epidermal Cell Fate

The most striking phenotype of the ATML1 overexpressor is formation of ectopic epidermal cell features such as stomata and trichomes in the inner tissues of the seedlings (Fig. 1B-D).⁵ We concluded that ATML1 has the ability to induce pluripotent protodermal cells that can produce several cell types found in the shoot epidermis (Fig. 1E). Importantly, post-embryonic induction of ATML1 was sufficient to induce protodermal cells in the inner tissues of the cotyledons, suggesting that the inner tissues of cotyledons are still competent to respond to ATML1 at the time of germination and are able to change their cell fate.

Figure 1. Overexpression of ATML1 affects epidermal/mesophyll cell differentiation in leaves. (A) Transverse section of a wild-type leaf. Red arrowheads, stomata. Red arrow, trichome. (B-D) Transverse sections of mature leaves from 14-d-old ATML1-overexpressing plants. (B) Green mesophyll cells decreased in the inner tissues of the ATML1-overexpressing leaf. (C) Magnified view of the inset in B. A pair of chloroplast-containing guard cells was observed. (D) A large cell (black arrow), expressing the trichome cell marker GL2-GUS (shown in blue), was found in the inner tissues. This GL2-GUS positive cell was surrounded by small accessory cell-like cells. The section was stained with Safranin O. (E) Our results suggest that ATML1 promotes epidermal differentiation and negatively regulates mesophyll cell differentiation in the outermost cell layers of the leaves. Scale bars, 100 µm in A and B and 20 µm in C and D.

We could not conclude whether ectopic epidermal cells were induced at later stages of leaf development in response to ATML1 expression because the RPS5A promoter used in this study was active only in young proliferative cells.⁷ Further misexpression studies using constantly active promoters are necessary to reveal the stage-dependent effects of ATML1 on epidermal or mesophyll cell differentiation.

Epidermal Differentiation May be Associated with Loss of Mesophyll Cell Characters

Most epidermal cells lack chloroplasts, whereas inner mesophyll cells develop chloroplasts. Our results suggest that ATML1 negatively influences mesophyll cell differentiation. The proportion of green mesophyll cells decreased with ATML1 overexpression.⁵ This result is consistent with a previous study in which mesophyll cells developed in place of epidermal cells on the surface of the atml1;pdf2 leaves.¹ Our marker gene analyses suggest that green surface cells in the atml1;pdf2 leaves exhibited both epidermal and mesophyll cell identities.⁵ Collectively, these observations suggest that ATML1 is a negative regulator of mesophyll cell differentiation in epidermis and that ectopic mesophyll cell fate is induced on the surface of the organ in the absence of ATML1 and PDF2 activity (Fig. 1E).

From an evolutionary point of view, it is tempting to speculate that epidermis has evolved from chloroplast-bearing parenchymal cells (i.e., mesophyll-like cells) through repression of chloroplast development and acquisition of cuticle biosynthetic machinery. Submerged aquatic angiosperms do not need to conserve water and tend to have thinner cuticle layers than those of terrestrial plants.^8,9 In fact, some aquatic angiosperm species possess chloroplasts in the epidermis, which may indicate a partial loss of epidermal identity in these species.^10-12 Future studies including large-scale identification of target genes for ATML1 and analysis of ATML1 orthologs in other species, including non-vascular plants and aquatic angiosperms, will help to understand the roles of ATML1 in the evolution of epidermis.

Cell Division Planes

Shoot epidermal cells mostly exhibit anticlinal divisions in the wild-type. However, in the ATML1 overexpressing plants, division planes that separate 2 guard cells were not necessarily perpendicular to the surface of the leaves, indicating that epidermal identity itself was not sufficient for anticlinal cell division.⁵ In fact, ectopic stomatal pores in ATML1 overexpressors tended to face toward air spaces in the inner tissues, suggesting that guard mother cells (GMCs) divide perpendicular to the inner surface.⁵ These observations raised 2 fascinating possibilities. The first is that “surface proximity” as well as “epidermal identity” is required for anticlinal cell division. This idea is also consistent with our observation that ectopic stomata in the L2 layer were formed mostly by anticlinal cell divisions.⁵ The second possibility is that ectopic epidermal cells have a tendency to choose a division plane so that daughter cells are on the same plane as their neighboring epidermal cells to maintain a single cell layer. Under this assumption, a surface-positioned GMC would be formed by “anticlinal” cell division of a surface-positioned meristemoid (MM), as 2 rounds of “periclinal” or “oblique” cell divisions are expected to displace the GMC from the surface, considering that 2 rounds of asymmetric cell division of MM are usually necessary to generate a GMC.¹³ In this scenario, the resulting surface GMC should again divide anticlinally to maintain a single layer of epidermal cells. Therefore, apparent “anticlinal” cell division of the ectopic surface GMCs may reflect the intrinsic nature of the epidermal cells to maintain a single cell layer. This idea is consistent with the localization of ACR4 receptor kinase in the basal and lateral sides of epidermis, thus implying the importance of intercellular communication within the same layer for the coordination of cell behaviors.^14,15 Future experiments including detailed 3-dimensional and time-course observations are required to test these possibilities.

Future Perspectives

Our results show that misexpression of ATML1 causes ectopic differentiation of epidermal cells in inappropriate positions. Therefore, ATML1 expression should be tightly controlled and restricted to the outermost cell layer throughout development. It will be a challenge for future research to identify the positional information that ensures epidermis-specific activation/autoregulation of ATML1.

Glossary

Abbreviations:

ATML1

ARABIDOPSIS THALIANA MERISTEM LAYER 1 or ARABIDOPSIS THALIANA MERISTEM L1 LAYER

ACR4

ARABIDOPSIS THALIANA HOMOLOG OF CRINKLY4 or ARABIDOPSIS CRINKLY4

GL2-GUS

a transgenic line expressing the reporter gene uidA under control of the GLABRA2 promoter

PDF1

PROTODERMAL FACTOR1

PDF2

PROTODERMAL FACTOR2

RPS5A

RIBOSOMAL PROTEIN S5A

Takada S, Takada N, Yoshida A. ATML1 promotes epidermal cell differentiation in Arabidopsis shoots. Development. 2013;140:1919–23. doi: 10.1242/dev.094417.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.