The Genetic Control Of Stomatal Development Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Stomata are pores upon the surface of leaves and stems of all terrestrial plants, formed by a pair of specialised parenchyma cells, known as guard cells, which flank the stoma and control pore size via varying turgor. Stomata play a vital role in gaseous exchange, by mediating the concentration of water vapour and carbon dioxide within the tissues (1). Carbon dioxide is able to diffuse through the stomata into the mesophyll layer, where it may be converted into organic compounds such as carbohydrate, while water vapour diffuses out of the plant, as a necessary side effect of stomatal opening. Stomatal are predominantly situated upon the abaxial (lower) surface of the leaf in order to minimise water loss, with a few exceptions which develop stomata upon the adaxial (upper) leaf surface only - notably plants with floating leaves, such as the water lily (A). Transpiration results in the mass flow of minerals and essential nutrients through the xylem but must be controlled in order to prevent excessive water loss - as transpiration is proportional to CO2 intake, this ratio is a permanent compromise for the plant (B).

Plants are able to vary their gas exchange levels according to environment, in two ways. Firstly, plants may adjust guard cell turgor to directly reduce the stomatal aperture, decreasing both water loss and gas exchange; and secondly, during development, by altering stomatal numbers in order to increase or decrease the stomatal index - the ratio of stomata number to epidermal cell number (2).

There is evidence to show that the physical conditions of mature leaves affect stomatal index in developing leaves possibly by modification of existing development pathways, controlling both stomatal frequency and patterning (2). At high levels of carbon dioxide, a reduction in stomatal index may reduce water loss while maintaining carbon dioxide uptake. In many species (such as Arabidopsis thaliana), stomatal index decreases in developing leaves when mature leaves are exposed to increased CO2 concentration (D). For example, in a study of 100 plants grown at twice the level of atmospheric carbon dioxide (700 p.p.m. compared to 350 p.p.m.), 60% of a selection of plant species exhibited an average stomatal frequency reduction of 9% in CO2-shielded developing leaves (2). However, in some species (for example, maize) this may be due to an increase in epidermal cell size, rather than an increase in stomatal cell number as stomatal to epidermal cell ratio appears to be maintained (C). In other species, no such effect has been detected at all. This evidence suggests that it is highly likely that there is a signalling pathway between mature leaves and developing leaves in many types of plant, and thus adaptable genetic control of stomatal development to environmental conditions. However, there appears to be no absolute predictor of stomatal index change with varying environmental conditions - with much variation between subspecies, and factors such as habitat temperature often insignificant. (Woodward and Kelly, 1995)

The genetic control of stomatal development may be a contributing factor to the great success of land plants. Indeed, the first fossil "true" stomata have been dated to around 50 to 60 million years ago, with developing precursors dated to 410 million years ago, and their structure has changed little since - it has been suggested that rather than modifying stomatal design, the genetic control may have been reined in, with greater control over development according to environmental conditions (4). Others suggest that this original structure may have remained unchanged for millions of years as it is simply very successful and needs little alteration - "a winning design." ( Franks, PJ and Beerling, DJ (2009)

10 µM

Fig. 1. Coalified stomata excavated from Downton Castle Sandstone (Shropshire), belonging to the sterile axis of an unidentified plant species. The stomata are likely to have originated from the Pridoli epoch, around 416 to 418.7 m.y.a., in the Silurian period. Copied from Edwards et al, 1998.

Stomata would have probably been evolutionarily advantageous for a plant - it is likely that they developed from epidermal pores, and would have provided the plant with more carbon dioxide, thus increasing photosynthesis, and assisting with desiccation resistance (3). There have been conflicting viewpoints concerning the original purpose of stomata, however - for example, a suggestion has been proposed that 'stomata first appeared on plants as structures that facilitated sporophyte drying out before spore discharge' (Duckett et al), raising the possibility that pseudostomatal pores could have originally evolved to allow water out of the plant's tissue, not to allow carbon dioxide in (4). However, the lower plant upon which this theory is based, Sphagnum, is unusual in that it does not have airspaces leading from the stomata, or potassium ion regulation of opening (4). There is no doubt that stomata would have provided a significant evolutionary advantage, however. As well as the advantages previously described, plants with stomata are better able to maintain low leaf temperature, are less likely to experience a fatal xylem embolism during low water availability, and experience faster nutrient transport (3). The fossil record generally has a lack of evidence for a reasonably precise evolutionary pathway, according to several researchers, notably Edwards et al. (1998). It is therefore difficult to determine how and why stomata evolved.

Stomatal have almost certainly driven evolutionary selection (Hetherington, Nature, 2003) particularly during the Phanerozoic era (400 m.y.a) when several new phyla bearing stomata evolved, such as the angiosperms. This is probably the most widespread and highly successful plant group today. Angiosperms may also have a greater level of control over their stomata when compared with, for example, conifers and ferns (5). In a 2009 paper by Brodribb et al. , it has been shown that angiosperms exhibit a very fast stomatal closure or opening response to both elevated and decreased carbon dioxide levels, while other major vascular lineages (such as conifers) exhibit minimal stomatal closure in response to elevated carbon dioxide (5). This would suggest some sort of signalling pathway present in angiosperms but not in other plant groups, and may have contributed to their evolutionary success. Brodribb et al. suggest that the evolution of genetic signalling pathways may have given flowering plants a slight advantage with regards to water use optimization, and that this may have been a contributing factor to their success (5). These pathways are still not fully understood.

In this project, the genetic control of stomata development will be examined in depth, and the signalling pathway between plants and their stomata will be critically analysed and described. The project will concentrate on notable aspects of genetic control - including the patterning of stomata, the differentiation of stomatal cells from the epidermis, and the loss-of-function experiments that have been carried out upon plants, as well as further comparisons between other development and signalling pathways.

Early Genetic Control of Stomatal Development

As stomata develop, they follow a specialised, dedicated cell lineage. The division pathway of specific cells in order to make stomata is so important that almost every gene involved in stomatal development controls division in one way or another (6). This cell lineage pathway originates in developing epidermis, and begins with the asymmetric division of a developing epidermal cell, via a postprotodermal meristemoid mother cell, into a meristemoid, a small precursor and indicator of stomatal development(6). The meristemoid now either divides several times asymmetrically - possibly affecting stomatal cell spacing, and amplifying the number of cells in the epidermis in such a way that almost all epidermal cells may be produced via the stomatal cell lineage (Sack and Nadeau, 2000) - or it may convert into a guard mother cell (GMC) which divides once symmetrically forming a pair of guard cells, which then undergo morphogenesis to develop into the functioning stomata (Fig.2). Mutations of regulatory genes, such as gene knockout or knockdown, or overexpression, can be a useful tool for visualising their function. For example, TOO MANY MOUTHS (TMM) gene, encoding a leucine-rich-repeat receptor like protein, has a tmm mutant phenotype which results in excess leaf stomatal development, while the tmm-1 mutant phenotype results in early conversion of the meristemoid to a GMC. This would suggest that TMM is responsible for repressing stomatal division, and also for recognising and reacting to spacing control signals.(6) (Bhave and Sack 2008)

Fig. 2. The simplified stomatal cell lineage. (1) A protodermal cell develops into a meristemoid mother cell, via an asymmetric division, (2) further developing into a smaller meristemoid, a precursor of stomatal development, and a larger stomatal lineage ground cell, an SLGC. (3) The meristemoid either asymmetrically undergoes amplifying divisions to become a pavement cell, (4) or undergoes conversion into a guard mother cell, which (5) divides once symmetrically to make a pair of guard cells. (6) The pavement cell has differentiated from the SLGC and may now flank the guard cells. Own diagram adapted from Bergmann and Sack, 2006 (6)

The first point of interest when considering the genetic control of the stomatal lineage would be how protodermal cells are "triggered" to start dividing into stomatal precursor cells originally. In order to consider this, we must examine three particular genes, known as SPEECHLESS (SPCH), MUTE and FAMA, all of which are involved in stomatal development, forming a "three-step transcriptional cascade" (10). These genes encode a trio of basic helix-loop-helix proteins which share strong sequence similarity, with highly similar bHLH domains, and carboxy-terminal. We will firstly examine SPCH, which controls the transition from meristemoid mother cell, to meristemoid. In order to determine the function of SPCH, Torii et al. (2007) carried out a loss-of-function mutation of the gene using a T-DNA insert. This spch mutant plant exhibited an almost complete lack of stomata. Next, SPCH was over-expressed using a constitutive CaMV35S promoter. This mutant plant exhibited many small epidermal cells that strongly expressed a GUS protein, linked with the TMM gene, showing that these cells have entered the stomatal lineage (Torii and Sloan, 2007). When SPCH is expressed ectopically, it can even trigger differentiation in adult pavement cells (10). It has thus been suggested that SPCH is responsible for initiating the first asymmetric division of the stomatal lineage. In order to test this, Pillitteri and Torii (2007) examined a weaker mutation of the SPCH gene, a spch-2 allele, which has a reduced stomatal index and fewer entry divisions than the wild-type. Also, fewer SLGCs (Fig.2) had formed in relation to the number of stomata. In total, SPCH probably has a role in entering cells into the stomatal lineage, as well as promoting amplifying divisions via SLGC formation, and it is necessary and sufficient for stomatal lineage entry(10)(Bergmann, 2006)(Torii and Sloan, 2007). There are still many questions to be answered, however. Pillitteri and Torii (2007) have speculated how SPCH "chooses" a set of cells to differentiate into meristemoids, when all express SPCH in equal amounts (10).

Fig.3. Petal epidermis, (1) with no stomata present in the wild type, and (2) after strong MUTE expression using the CaMV35S strong promoter. Petals do not normally express stomata; suggesting MUTE is able to induce stomatal expression in an organ. Scale bar = 20µm. Photograph from Pillitteri et al. (2008) (11)

The second of these highly similar bHLH protein-encoding genes is MUTE, which has been shown to have a strong role in the conversion of a meristemoid to a guard mother cell (GMC) (10). MUTE is detected by GUS protein linked expression just before the transition to a GMC - using the natural MUTE promoter to express GUS (Sloan, 2007). GUS could not be detected in young meristemoids, or in guard cells. Next, loss- and gain-of-function mutants were used in order to determine MUTE function - an excellent tool for determining both spatial and temporal action of genes (11). The loss-of-function mutation, mute, results in a weak and light-coloured plant with no stomata at all. The plant is probably weak as it cannot regulate gas exchange or transpiration properly, showing that stomata are an essential feature of plants (Torii and Sloan, 2007). Meristemoids with mute also failed to exhibit a specific marker, E994, which signifies the transition to a guard cell in normal plants. By again using the CaMV35S strong promoter to over express MUTE throughout an A. thaliana test plant, this resulted in every epidermal cell developing into stomata (Fig.3). MUTE even resulted in stomatal development in organs that are normally stomata-free, including petals. Of the three bHLH proteins, MUTE overexpression is the only one causing whole stomata development, rather than excessive divisions, for example(11). Without MUTE expression, meristemoids continue to divide until physically restricted. Sloan et al. (2007) suggested that MUTE plays a role in terminating asymmetry in meristemoids and converting them to GMCs, while Pillitteri et al. (2008) later suggested that MUTE was a "necessary and sufficient" prerequisite for stomatal differentiation throughout an entire plant(Sloan, 2007)(11). Pillitteri et al. (2007) have also speculated how MUTE is activated after a number of cell divisions, which seems to vary between one and three (10).

Fig.4. (A) Wild type A. thaliana with functioning FAMA in the cotyledon. (B) Mutant with fama-1 loss-of-function gene - rows of epidermal "tumors" are visible. Scale bar =10µm. Photograph from Bergmann et al. (2006) (21)

The third bHLH protein-encoding gene is FAMA which plays a role in GMC to guard cell transition (21) and is, along with MUTE, dependent upon SPCH being functional (Macalister). Bergmann et al. used loss- and gain-of-function mutants, as described previously, in order to determine this gene's function. Using a T-DNA insertion to create a fama-1 mutant, Bergmann et al. described a plant with no "true" stomata, but with rows of thin epidermal "tumors" where stomata should normally develop (Fig.4)(21). A second, 3' end T-DNA insertion, called fama-2, resulted in healthy plants which occasionally exhibited mutant stomata with 3 or 4 guard cells (Fig. 4)(21). FAMA was fused to a GFP reporter, in order to view its expression pattern - it was found to be expressed in immature guard cells and GMCs(21). When GFP was fused with several stomatal markers (such as TMM for example), it was also found that these markers continued expressing strongly throughout the stomatal lineage in fama "tumor" cells, but in wild type FAMA plants, expression was weakened by the final guard cell stage (21). Next, FAMA overexpression could be induced using estrogen (Zuo 2000) which can result in gene expression up to eight times higher than that achieved with the CaMV35S promoter, as well as excellent control of expression, and good inducibility. This resulted in excessive unpaired guard cell production, with many kidney-shaped cells developing, even from cells that do not enter the stomatal lineage - for example, root cells- and in non-epidermal cells, FAMA overexpression could cause partial transformation to guard cell phenotype (21). The expression of single guard cells was interpreted by Bergmann et al. (2006) to mean that FAMA causes guard cell differentiation, and is a "master regulator" of whether GMCs divide or differentiate.

In further studies, it could be suggested that using oestrogen as an overexpression inducer would be a slightly improved experimental design unless physically impossible. So far, only FAMA seems to have been induced using this method, while SPCH and MUTE have been over expressed with a CaMV35S promoter, while oestrogen induction is more controllable with greater gene expression (Zuo, 2000). There may also be more in-depth studies of the genetic origins of the bHLH protein-encoding genes. Bergmann et al. (2009) have recently determined orthologs of FAMA, MUTE and SPCH in maize and rice, which have significant common functions. Pillitteri et al. have suggested that a genetic duplication event may have resulted in the development of the paralogous bHLH genes. They also hypothesize that "the variation of stomatal patterning among land plants could in part be a result of the diversification of the stomatal bHLHs" over time (10). This may result in a greater understanding of the function and origin of these genes.

Spacing control is a greatly important part of the genetic control of stomatal development, and there are several theories as to how the spacing of stomata occurs. Stomata in Arabidopsis are virtually always surrounded by pavement cells only (99.4%) (Geisler 1998). This probably is a natural adaptation - stomata require an ion flow between guard cells and the surrounding epidermis, and may not be able to reach proper turgor if they are in contact with other stomata. Spacing could occur in a number of different ways - for example, lateral inhibition by guard cells of the surrounding epidermis. Larkin et al. (1997) have suggested that the placement and orientation of meristemoid cells controls stomatal patterning to some extent, possibly suppressing their development. More recently, the SDD1 (stomatal density and distribution) gene has been implicated in stomatal spacing control (9). The sdd1-1 mutant allows many more protodermal cells to enter the stomatal lineage, increases the proportion of stomata on the abaxial epidermis, as well as altering meristemoid orientation so that stomatal "clusters" form (Berger, 2000). However, overexpression of SDD1 using CaMV35S promotion does not result in reduced stomatal density below normal levels. Berger et al. have suggested that the stomata-free space that normally surrounds each stomata is absent in sdd1-1 mutants, and thus implies SDD1 may play a major role in controlling stomatal patterning. The protease that this gene encodes, SDD1, probably processes a signal precursor which controls stomatal density and patterning (12). Pillitteri et al. (2007), among others, have conducted further research which shows this unknown signal is "interpreted by the transmembrane receptor TOO MANY MOUTHS (TMM) and three ERECTA family receptor-like kinases ERECTA, ERECTA-LIKE 1 (ERL1) and ERL2" (Pillitteri termination 2007) .

What, then, is the identity of this unknown signal? Torii, Bergmann et al. have suggested a candidate gene - EPIDERMAL PATTERNING FACTOR 1, or EPF1, which is expressed in signalling cells only (12). It encodes a small protein. Overexpression of EPF1 results in a proportional decrease in stomatal density but no other specific phenotypes other than secondary poor growth. The loss-of-function mutant, epf1-1, results in stomatal clustering. In total, this would suggest that EPF1 is "a mobile signal that negatively regulates stomatal formation and enforces the orientation of spacing divisions" (13). This research has great further application - Franssen et al. (Franssen)**** have suggested that "peptide and non-peptide hormone-activated signalling cascades are linked in plants as they are in animals".

We have previously discussed the TOO MANY MOUTHS (TMM) gene phenotype, and the effect of over-expression or knock-down of this gene. How, though, does TMM interact with other genes in order to control spacing? Torii, Bergmann et al. have proposed a model involving proteins encoded by TMM and ERECTA family genes, which are respectively a leucine-rich-repeat receptor-like protein and leucine-rich-repeat receptor kinases, where EPF1 acts as a ligand for these receptors (12). This model replaced a previous theory which suggested that SDD1 might encode a receptor which receives a signalling molecule (such as that encoded by EPF1) (13). The proof for this new model came from the discovery that tmm and er:erl1:erl2 mutations are epistatic (inhibitory) to the overexpression phenotype of EPF1- the stomatal density decrease in the overexpression phenotype is thus dependent upon functional TMM, ER, ERL1 and ERL2 but not upon SDD1 (12). However, Nadeau (2009) has suggested that, as EPF1 is only present from meristemoid stages onwards, another signal must co-ordinate the entry of protodermal cells into the stomatal lineages (13). Nadeau has suggested that EPF1paralogs in A.thaliana may fulfil this role, a suggestion that could perhaps be tested by determining whether spch (or any other early acting gene) mutations are epistatic to the overexpression phenotype of any of these paralogs.

Fig.5. A graphical depiction of (A) a wild type A.thaliana plant, (B) an er;erl1;erl2 triple mutant, resulting in excessive stomatal numbers and spacing defects. This mutant exhibits a phenotype similar to the erecta single mutant, but is more pronounced, with additional patterning defects not exhibited by the latter. Own diagram developed from descriptions by Serna (2009) (20), and Bergmann and Sack (2006) (6).

The ERECTA gene family therefore is likely to receive a signal encoded by EPF1 but it is also likely that each member of the family plays a different role in the control of stomatal development. Using mutations of each family member, a loss-of-function erecta mutant was found to have overabundant stomata, and compacted inflorescence, while erecta like1 and erecta like2 mutations had no visible phenotype but enhanced erecta, with triple mutant er;er1;erl2 exhibiting a very strong erecta mutant phenotype and also spacing defects (shpak 2004)(6). Phylogenetic analysis of the two ERECTA-LIKE genes suggested that they are paralogs of ERECTA (Shpak 2004). They all appear to function slightly differently in the stomatal lineage however. For example, while ER promotes meristemoid differentiation, ERL1 inhibits it(6). It has been proposed that the ERECTA family leucine-rich-repeat receptor-like-kinases (LRR-RLK) form heterodimers or homodimers, in order to become a signalling complex(6). This complex is unlikely to simply include TMM and the ERECTA family however, as TMM has varying functions according to organ, and Bergmann and Sack (2006) have thus suggested that unknown genes may complete this model complex(6).

It has been determined that a signalling cascade falls downstream from the receptors - a "mitogen-activated protein kinase (MAPK) cascade" consisting of five genes - the map kinase kinase kinase, YODA (YDA), two map kinase kinases called MKK4 and MKK5, and two map kinases, MPK3 and MPK6(13). By identifying mutations that could be lethal in seedlings, Bergmann et al. (Somerville 2004) were able to identify yda mutants with excessive stomata, which were also clustered. With regards to experimental design, the examination of mutations that are lethal early in development can often be a useful tool in order to discover mutations in early-acting genes, which may not be identified if plants are only examined once fully grown. By using dynamic observation of yda mutants - i.e., examining a gene's action over a period of time - Bergmann et al. were able to visualise at which point YDA would normally be implemented. They found that the yda plants are sterile with abnormal flowers. Also, yda mutants did not enter the stomatal lineage early, but after 72 hours post-germination, more cells than normal began to differentiate into stomata, without reference to the wild-type "one-cell spacing rule"(Somerville 2004).

also lukowitz 2004

minus 297