At least two functions for BdMUTE during the development of stomatal complexes in Brachypodium distachyon

Abstract

Fundamental to multicellular organisms is the capacity to produce different cell types with specialized attributes. Understanding the way cells acquire their fates is a major and exciting challenge in developmental biology. Grasses develop a unique stomatal structure consisting of two dumbbell-shaped guard cells (GCs) flanked by two lateral subsidiary cells (SCs) (Stebbins & Shah, 1960; Rudall et al., 2017; Hepworth et al., 2018; Nunes et al., 2020). Due to its simplicity and accessibility, because they are part of the epidermal tissue, and with the current possibilities to isolate molecular markers of its different cell types and to genetically manipulate its development, this four-celled complex is an excellent model system to delve into the genetic control of cell fate determination and differentiation.

In grasses, among them Brachypodium (Brachypodium distachyon), stomatal complexes develop in stomatal-forming cell files, from the base to the tip of the leaves, and through an invariant pattern of cell divisions (Stebbins & Shah, 1960; Serna, 2011; Hepworth et al., 2018; Nunes et al., 2020; Fig. 1a). The first cell division produces the guard mother cell (GMC). Before GMC division takes place, cells from ontogenetically unrelated files on either side of newly formed GMC acquire subsidiary mother cell (SMC) identity. SMCs divide by producing small cells flanking the GMC that differentiate as SCs. The GMC then divides, like SMCs with its cell division plane being parallel to the main axis of leaf growth, and it yields the paired dumbbell-shaped GCs. This pattern of cell divisions is different from that which takes place during stomatal development in Arabidopsis (Arabidopsis thaliana; Serna & Fenoll, 2000; Bergmann & Sack, 2007; Fig. 1b). In Arabidopsis, where epidermal cells are not organized in rows, stomata formation commences from the leaf tip and proceeds basipetally (Peterson et al., 2010; Vatén & Bergmann, 2012). The first cell division that initiates the stomatal lineage produces the first meristemoid (M). These Ms usually undergo additional self-renewing asymmetric divisions, in an inward spiral, until they assume GMC identity. Then GMCs divide to give rise to a pair of kidney-shaped GCs. Cells in direct contact with the GCs and ontogenically related to them, can differentiate into pavement cells, or they can produce secondary Ms.

Although with different functions in Arabidopsis and grasses, and different regulatory relationships even within grasses (McKown et al., 2023), one of the key factors that regulates stomatal development is the basic helix–loop–helix transcriptional factor MUTE. In grasses, the primary role of these transcriptional factors, which laterally move from the GMC, where their genes are expressed, to the neighbouring epidermal cell files, is promoting SC recruitment (Raissig et al., 2017; Wang et al., 2019). Consistently, the bdmute mutant of Brachypodium lacks SCs (Raissig et al., 2017). This mutant has also impaired GMC fate and GC morphogenesis with most GMCs developing functional stomata with two incompletely differentiated GCs, and some of them failing to specify the orientation of the GMC division plane and/or undergoing excessive cell divisions (Raissig et al., 2017). In domesticated grasses like maize (Zea mays) or rice (Oryza sativa), mute mutants (bzu2/zmmute and osmute) also fail to recruit SCs and, in contrast to bdmute, completely abort stomatal development (Wang et al., 2019; Wu et al., 2019). In contrast to grass MUTE proteins, the MUTE protein of Arabidopsis does not move between cells. AtMUTE expression and the localization of the protein encoded by this gene are restricted to Ms and GMCs (MacAlister et al., 2007; Pillitteri et al., 2007; Wang et al., 2019). AtMUTE plays a crucial role for transition from M to GMC (MacAlister et al., 2007; Pillitteri et al., 2007) as well as from GMC to GCs (Han et al., 2018). According to the functions assigned to AtMUTE, atmute exhibits Ms arrested after having undergone an excess of self-renewing cell divisions (MacAlister et al., 2007; Pillitteri et al., 2007), and its overexpressing converts all their epidermal cells into stomata (MacAlister et al., 2007; Pillitteri et al., 2007). Therefore, grass MUTE acquired at least a new role driving the recruitment of SCs, which is associated with their mobility.

In grasses, despite having impaired GMC fate, and because SC recruitment occurs before GMC division, the possible role of grass MUTE regulating GMC fate and/or GC morphogenesis was still unclear. Here, I discuss recent evidence on how BdMUTE regulates these cell fates during the development of stomatal complexes in Brachypodium and underline the importance of BdMUTE mobility for SC recruitment (Spiegelhalder et al., 2024).

The primary role of BdMUTE, which moves laterally from the GMC, where BdMUTE is expressed, to the epidermal cells of neighbouring files, is to recruit SCs (Raissig et al., 2017). Similarly, both OsMUTE and ZmMUTE induce the recruitment of the SCs, and the protein encoded by OsMUTE, and more probably by ZmMUTE, moves laterally into the neighbouring cells (Wang et al., 2019; Wu et al., 2019). In Arabidopsis, AtMUTE, with its expression and the localization of the protein encoded by this gene restricted to Ms and GMCs (MacAlister et al., 2007; Pillitteri et al., 2007; Wang et al., 2019), has no effect in the epidermal cells that surround the stomata.

But what is the nature of BdMUTE function during SC recruitment? Is BdMUTE mobility indeed required for BdMUTE-mediated SC recruitment? Or alternatively, can BdMUTE recruit SCs independently of its mobility? The analysis of several lines expressing different levels of a functional BdMUTEp:3×GFP-BdMUTE construct, but with reduced mobility due to the 3×GFP tag, in the bdmute background sheds light on these questions. While the strongest expressing 3×GFP-BdMUTE line almost fully rescued SC development, those expressing lower levels of 3×GFP-BdMUTE showed a gradual decrease in their potential to rescue the development of SCs (Spiegelhalder et al., 2024). It is then likely that in the weakly expressing 3×GFP-BdMUTE lines, 3×GFP-BdMUTE levels in many epidermal cells adjacent to the GMCs of neighbouring files are below those capable of inducing SC recruitment. Supporting this view, the strongest expressing 3×GFP-BdMUTE line exhibited a temporal delay in the SC recruitment relative to the control YFP-BdMUTE line (Spiegelhalder et al., 2024). Taken together, this strongly suggests that BdMUTE mobility is indeed required for BdMUTE-mediated SC recruitment, which, in addition, takes place in a dosage-dependent manner. Unexpectedly, BdFAMA, whose expression and protein localization overlap in GMCs and GCs of both wild-type (WT) plants and bdmute, when expressed under its own promoter, partially compensates the lack of SCs in bdmute with many SCs arising from diagonal divisions (McKown et al., 2023). But, considering that SC recruitment depends on BdMUTE mobility (Spiegelhalder et al., 2024), how could BdFAMA, which is unable to move between cells, induce SC recruitment? One possibility is that BdFAMA could activate some unknown factor capable of moving from the GMC to the neighbouring cells, where it would induce the SMC fate. However, considering that in bdmute; BdFAMAp:YFP-BdFAMA many SMCs were unable to properly orientate their asymmetric divisions (McKown et al., 2023), it is likely that the movement of the transcription factor could be necessary to correctly polarize the SMC. Taken together, these results support a model, which attributes to BdFAMA a compensatory role for the induction of SMC fate in the absence of BdMUTE, and to BdMUTE mobility a later role in polarizing asymmetric SMC division (McKown et al., 2023). This compensatory ability of BdFAMA does not depend on its activation through BdMUTE (McKown et al., 2023).

Given that BdMUTE-mediated SC recruitment requires the mobility of BdMUTE, the absence of SC in Arabidopsis could be due to the permanence of AtMUTE in the cells where it is produced. However, when the YFP-BdMUTE genetic construct, under the control of the GMC specific AtMUTE promoter (MacAlister et al., 2007; Pillitteri et al., 2007), is expressed in Arabidopsis, YFP illuminates stomatal precursors and also adjacent epidermal cells (Raissig et al., 2017; Wang et al., 2019), but this construct does not have the ability to induce the recruitment of SCs (Raissig et al., 2017; Wang et al., 2019). The same thing happens when the AtMUTEp:YFP-ZmMUTE construct is expressed in Arabidopsis (Wang et al., 2019). This suggests that the inability of AtMUTE to recruit SCs does not depend on its lack of mobility but on the nature of stomatal development in this plant species. Like BdFAMA in Brachypodium, AtFAMA can rescue atmute (McKown et al., 2023), highlighting the overlapping functions of these transcriptional factors (McKown et al., 2023).

In Arabidopsis, as stated previously, AtMUTE guides the transition from M to GMC (MacAlister et al., 2007; Pillitteri et al., 2007) and from GMC to paired GCs (Han et al., 2018). It is assumed that Brachypodium, like the rest of the grasses, does not have Ms because protodermal cells destined to produce directly generate the GMC through an asymmetric division (Raissig et al., 2016; Nunes et al., 2020; Serna, 2020). But could BdMUTE regulate the transition from GMC to paired GCs? The phenotype of bdmute, with c. 25% of GMC aborting GCs formation (Raissig et al., 2017), suggests that, perhaps redundantly with other factors, this could be the case. However, given that the bdmute mutant is completely devoid of SCs (Raissig et al., 2017), and that the recruitment of the SCs takes place before the GMC division (Stebbins & Shah, 1960; Serna, 2011; Hepworth et al., 2018; Nunes et al., 2020), the observed defects of the GMC division could simply be a consequence of the absence of SCs.

Interestingly, the weakest expressing 3×GFP-BdMUTE line in the bdmute background rescued stomatal production almost completely (c. 2% aborted GMC fate) but failed to recruit SCs (Spiegelhalder et al., 2024). This is telling us that the defects of the GMC division in the bdmute mutant are not simply a consequence of the lack of SCs, but that BdMUTE indeed regulates stomatal production independently of BdMUTE mobility and SC recruitment. According to this, although the line with the highest expression of 3×GFP-BdMUTE almost fully rescued the bdmute phenotype exhibiting c. 98% of mature four-celled complexes like those of WT plants, a considerable number of GMCs, at the time of their division, had rescued no SCs due to SC recruitment being delayed (Spiegelhalder et al., 2024).

Furthermore, in Brachypodium the abortion of functional paired GCs is strongly correlated with the orientation of the GMC division plane (Spiegelhalder et al., 2024). Consistent with their mutant phenotype, OsMUTE and ZmMUTE also seem to function by correctly orienting the division plane of their GMCs (Wang et al., 2019; Wu et al., 2019). However, in contrast to the bdmute mutant, both bzu2/zmmute and osmute mutants completely abort stomatal development (Wang et al., 2019; Wu et al., 2019). Interestingly, the lack of even one functional copy of BdFAMA in the bdmute mutant aborts stomatal development phenocopying bzu2/zmmute and osmute (McKown et al., 2023). This indicates that BdFAMA regulates the transition from GMC to GC in the absence of BdMUTE, allowing the development of stomata in the bdmute mutant. However, considering that BdMUTE represses early BdFAMA activity, BdMUTE-mediated targeting of the GMC cleavage plane does not require BdFAMA activity in WT plants (McKown et al., 2023). The almost null expression of FAMA in both bzu2/zmmute and osmute suggests that in maize and rice, in contrast to Brachypodium, FAMA expression depends on MUTE activity (Wang et al., 2019; Wu et al., 2019). Although direct evidence is lacking, this MUTE-dependent activation of FAMA could explain the complete abortion of stomatal development in the bzu2/zmmute and osmute mutants.

Like in Arabidopsis (Pillitteri et al., 2007; Lee et al., 2014; Han et al., 2018), in Brachypodium, an adequate progression of stomatal development requires reciprocal regulation between BdMUTE and BdFAMA, with BdMUTE promoting BdFAMA expression and BdFAMA repressing BdMUTE (McKown et al., 2023). In bzu2/zmmute, the expression of ZmFAMA is reduced significantly (Wang et al., 2019), and the same occurs in osmute, which shows very low levels of OsFAMA (Wu et al., 2019). This indicates that MUTE positively regulates FAMA expression in domesticated grasses. Although there is still no evidence of the possible regulation of MUTE by FAMA in these species, it is possible that this regulation is also necessary for a proper stomatal development.

The bdmute mutant has impaired GC morphogenesis, exhibiting shorter GCs and with a less pronounced dumbbell shape than those of WT plants (Raissig et al., 2017; Spiegelhalder et al., 2024). The absence of SCs in this mutant makes us question whether the development of functional but undifferentiated GCs is due to the lack of BdMUTE or, conversely, the absence of SCs. To answer this question, Spiegelhalder et al. (2024) analysed an expressing 3×GFP-BdMUTE line in the bdmute background, which develops a similar number of complexes recruiting zero, one or two SCs. The number of SCs, and therefore the degree of BdMUTE mobility, which in turn depends on the level of BdMUTE expression in the GMC, correlates with the degree of GC differentiation so that GCs in complexes with two SCs showed more signs of maturity than those of complexes with only one SC. The GCs from complexes without SCs showed the least signs of differentiation.

But what is behind these phenotypes? The fact that bdfama, whose GCs are immature, develops SCs (McKown et al., 2023), suggests that the presence of SCs is not enough to trigger GC morphogenesis. Similarly, the absence of SCs in c. 8% of the lsc-1 mutant stomatal complexes, which does not noticeably affect the differentiation of its GCs, points in the same direction (Cui et al., 2023). Therefore, it is likely that BdMUTE, independently of SC recruitment, guides, in a dosage-dependent manner, GC morphogenesis. Supporting this view, BdMUTE is also expressed in GCs (Raissig et al., 2017). Interestingly, the shape of the GCs in those complexes that develop only one SC was different, with the GC flanked by a SC showing greater signs of morphogenesis compared with the nonflanked one (Spiegelhalder et al., 2024). This may suggest that, in addition to a possible autonomous role of BdMUTE in regulating GC morphogenesis, the presence of lateral SCs, and thus the mobility of BdMUTE towards SMCs, also influences this process.

Finally, the 3×GFP-BdMUTE construct rescued, in a dosage-dependent manner, not only stomatal morphology but also stomatal function in the bdmute background (Spiegelhalder et al., 2024). As established previously (see, for example, Zhang et al., 2022; Durney et al., 2023; Liu et al., 2024), this confirms that the unique stomatal complexes of grasses trigger a more effective stomatal movement and, consequently, gas exchange.

The results described here highlight the two most important findings related to the functions exerted by BdMUTE during Brachypodium stomatal development. They are (Spiegelhalder et al., 2024): the need for MUTE mobility for MUTE-mediated SC recruitment; and the BdMUTE regulation of GMC division orientation independently of BdMUTE mobility. These findings show that BdMUTE drives cell fate transitions in a dosage-dependent manner (Spiegelhalder et al., 2024), adding complexity to the mechanism that regulates stomatal development, and questioning whether this also acts as a regulatory mechanism for stomatal development in all plant species. Future challenges include revealing whether BdMUTE regulates GC morphogenesis and understanding the mechanism underlying the lateral but not radial movement of BdMUTE. The phenotypic differences between bdmute and domestic grass mute are telling us that the regulatory mechanisms of stomatal development Brachypodium differ from those of domestic grasses. A systematic understanding of these differences will contribute to genetic improvements in grass crops.

None declared.

The New Phytologist Foundation remains neutral with regard to jurisdictional claims in maps and in any institutional affiliations.