Abstract
Trichomes are large epidermal cells on the surface of leaves that are thought to deter herbivores, yet the presence of trichomes can also negatively impact plant growth and reproduction. Stomatal guard cells and trichomes have shared developmental origins, and experimental manipulation of trichome formation can lead to changes in stomatal density. The influence of trichome formation upon stomatal development in natural populations of plants is currently unknown. Here, we show that a natural population of Arabidopsis halleri that includes hairy (trichome-bearing) and glabrous (no trichomes) morphs has differences in stomatal density that are associated with this trichome dimorphism. We found that glabrous morphs had significantly greater stomatal density and stomatal index than hairy morphs. One interpretation is that this arises from a trade-off between the proportions of cells that have trichome and guard cell fates during leaf development. The differences in stomatal density between the two morphs might have impacts upon environmental adaptation, in addition to herbivory deterrence caused by trichome development.
Introduction
In Arabidopsis, trichomes are large epidermal cells that protrude from the surface of the leaves and petioles. Trichomes play important roles in both biotic defences and abiotic stress tolerance (Levin, 1973; Mauricio and Rausher, 1997; Handley et al., 2005; Dalin et al., 2008; Sletvold et al., 2010; Sletvold and Ågren, 2012; Sato and Kudoh, 2016). However, trichome development appears to impose a fitness cost on growth and reproduction (Mauricio, 1998; Sletvold et al., 2010; Kawagoe et al., 2011; Sletvold and Ågren, 2012; Sato and Kudoh, 2016). In addition to trichomes, stomatal guard cells represent another specialized cell type that is present on the leaf surface. Trichome initiation occurs prior to stomatal meristemoid development, and the patterning of trichomes and guard cells appears to be linked (Larkin et al., 1996; Glover, 2000; Bean et al., 2002; Bird and Gray, 2003). Therefore, there might be a trade-off between trichome and stomatal guard cell development during leaf formation (Glover et al., 1998).
We wished to determine whether trichome formation might be associated with changes in stomatal patterning in natural populations of plants. To achieve this, we investigated stomatal patterning in a naturally-occurring population of Arabidopsis halleri subsp. gemmifera that includes trichome-forming and glabrous morphs (Kawagoe et al., 2011; Sato and Kudoh, 2016). The glabrous morphs within this population harbour a large transposon-like insertion within the GLABRA1 (GL1) gene (Kawagoe et al., 2011). GL1 is also required for trichome formation in A. thaliana, with homozygous gl1 mutants being glabrous (Oppenheimer et al., 1991). Our experiments provide new insights into the relationship between stomatal and trichome patterning under natural conditions.
Methods
Study site and experimental model
This investigation used a well-characterized population of Arabidopsis halleri subsp. gemmifera that is located beside a small stream in central Honshu island, Japan (Fig. 1A) (35°06’ N, 134°55’ E) (Aikawa et al., 2010; Kudoh et al., 2018). Sampling occurred during September 2016 (photoperiod approximately 12 h, with dawn at 05:40 and dusk at 18:10). During this season, A. halleri bore larger rosette leaves that are well-suited for quantification of stomatal density (Fig. 1B).
Stomatal density measurement
Eight plants of each trichome morph (hairy or glabrous) were selected at the study site, with individuals chosen such that the replicate plants were distributed evenly across the site. Glabrous and hairy morphs were identified by visual inspection of the leaf surface. Stomatal density was measured by obtaining impressions from the adaxial surfaces of 3-5 rosette leaves of each plant. Data were obtained from 58 and 62 leaves of hairy and glabrous plants, respectively. We focused on the adaxial surface because this surface also harbours the majority of the trichomes. Between the times of 12:00 and 13:00, President Plus dental impression paste (Coltene) was applied to the adaxial side of each leaf to create a leaf surface impression (Fig. 1C). Solidified impression paste was removed from leaves and transported to the laboratory for further processing. First, each impression was assigned a randomly-generated number to ensure subsequent steps were performed blind. Each leaf impression was painted with transparent nail varnish (60 seconds super shine, Rimmel) that, after drying, was peeled away from the dental impression paste using transparent adhesive tape (Scotch Crystal). Next, the adhesive tape was used to attach the nail varnish impression to a 0.8 mm – 1 mm thick microscope slide. Leaf impressions were examined using an epifluorescence microscope in white light illumination mode. Images were captured from the centre of each leaf half, away from the midrib, using a Hamamatsu camera and Volocity software set to 20x zoom. Two images were captured from each impression, and the number of stomata and pavement cells was counted in an 800 µm x 800 µm square using the Fiji software to obtain cell density measures. Cell density measures were expressed as per mm2 (multiplication by 1.56). Stomatal index was calculated according to Equation 1. After all measurements, data were disaggregated according to a blinding/randomization scheme. The differences between hairy and glabrous plants were statistically tested by nested analysis of variance, whereby leaves were nested within the hairy and glabrous morphs. Tests were conducted using the R 3.6.0 software (R Core Team, 2019) and plots generated with the beeswarm R package (v0.2.3) and Inkscape v0.91. No adjustments were applied to images in Fig. 1.
Equation 1. Derivation of stomatal index, where SI is stomatal index, s is the number of stomata in the field of view, and p is the number of epidermal pavement cells in the field of view.
Results
We investigated stomatal patterning in naturally-occurring hairy and glabrous morphs of A. halleri (Sato and Kudoh, 2016). Approximately half of the A. halleri population at this study site is glabrous, whilst remaining plants have trichomes (Kawagoe et al., 2011). As trichome initiation occurs prior to stomatal meristemoid formation (Larkin et al., 1996; Glover, 2000), it is likely that trichome and stomatal patterning are linked (Bean et al., 2002), so we hypothesized that this might produce a difference in stomatal density between the two trichome morphs of A. halleri under natural conditions.
We found that the trichome formation dimorphism was accompanied by a difference in stomatal density (Fig. 2A; Supplemental Dataset S1). Glabrous morphs had significantly greater stomatal density compared with hairy-leaved morphs (glabrous: 31.4 ± 1.5 stomata mm-2; hairy: 23.7 ± 1.1 stomata mm-2; ± s.e.m) (Fig. 2A; Table S1A; Supplemental Dataset S1). Furthermore, the stomatal index was significantly greater in glabrous morphs (18.13 ± 0.41) compared with hairy morphs (16.11 ± 0.46) (Fig. 2B; Table S1B). The pavement cell density did not differ significantly between the morphs (Table S1C). Stomatal density ranged from 17 – 87 stomata mm-2 for hairy morphs and 27 – 119 stomata mm-2 for glabrous morphs (Fig. 2A). This stomatal density was lower than for Arabidopsis thaliana, which has reported stomatal densities of 180 – 350 stomata mm-2 depending on background accession and growth conditions (Gray et al., 2000; Zhang et al., 2008; Franks et al., 2015).
Discussion
Glabrous plants had significantly greater stomatal density and stomatal index compared with hairy plants (Fig. 2A; Fig. 2B). As the density of surrounding pavement cells did not vary between the morphs, these differences in stomatal density and index are due to the greater density of stomata in glabrous morphs compared with hairy morphs (Fig. 2B). Our field data are consistent with a laboratory-based study in which transgenic tobacco plants expressing an Antirrhinum myb-like transcription factor, which caused an excess of trichomes, also had significantly reduced stomatal density (Glover et al., 1998). Similarly, the trichome-bearing Col-0 accession of A. thaliana has lower stomatal density than the glabrous C24 accession (e.g. about 115 mm-2 for Col-0 and 180 mm-2 for C24) (Perazza et al., 1998; Lake and Woodward, 2008). This suggests that in natural populations of A. halleri, there could be a trade-off between trichome and stomatal development. Since the glabrous gl1 mutant of A. thaliana has a significantly greater density of stomatal units compared with the wild type (Berger et al., 1998) and the glabrous phenotype of A. halleri at this study site is associated with an insertion within GL1 (Kawagoe et al., 2011), it is possible that the GL1 haplotype influences the stomatal density within this population of A. halleri.
In some cases, there does not appear to be a tradeoff between stomatal and trichome density. For example, elevated CO2 decreases stomatal density (Woodward and Kelly, 1995), but might also reduces trichome density (Bidart-Bouzat et al., 2005). Therefore, in future, it could be informative to examine the relationship between stomatal and trichome density under a range of different experimental conditions that apply different types of selection pressure.
Interestingly, trichome production appears to impose a fitness cost. For example, glabrous A. halleri plants have 10% greater biomass than hairy plants when grown in the absence of herbivores (Sato and Kudoh, 2016). This cost of herbicide resistance arising from trichome formation also occurs in glabrous and hairy A. lyrata (Løe et al., 2007; Sletvold et al., 2010) and A. thaliana (Mauricio and Rausher, 1997; Mauricio, 1998) under experimental conditions excluding herbivores. Whilst this fitness advantage of glabrous over hairy leaves in the absence of herbivory might be due to trichome production (Mauricio and Rausher, 1997; Mauricio, 1998; Kawagoe and Kudoh, 2010; Sletvold et al., 2010; Kawagoe et al., 2011; Sletvold and Ågren, 2012), we suggest that glabrous morphs might also gain an advantage by having a greater density or number of stomata. It has been proposed that increasing the number of stomata could increase carbon assimilation (Lawson and Blatt, 2014). For example, Arabidopsis overexpressing STOMAGEN has greater stomatal density and a 30% increase in carbon assimilation compared with the wild type. However, these lines also have a higher transpiration rates and consequently lower water use efficiency (Tanaka et al., 2013).
Optimal stomatal density is important to achieve high photosynthetic rates. A low stomatal density restricts CO2 vertical diffusion through the leaf and reduces photosynthetic rates, whilst high-density stomatal clustering diminishes CO2 diffusion and causes low carbon assimilation (Lawson and Blatt, 2014). Both A. halleri morphs examined are likely to be within an optimal range of stomatal densities, having evolved and survived under natural conditions. However, the higher stomatal density in the glabrous morph might contribute to its faster growth in absence of herbivory (Sato and Kudoh, 2016). In future, it would be interesting to explore this by measuring the CO2 assimilation rate of these trichome morphs under laboratory and/or natural conditions. It would also be informative to determine whether the stomatal density difference between the two trichome morphs confers any advantages within microenvironments characterized by differences in water or light availability. The lower stomatal density of A. halleri compared with A. thaliana (Gray et al., 2000; Zhang et al., 2008; Franks et al., 2015) might reflect differences in growth conditions. An alternative explanation might relate to genome size, because there appears to be a negative correlation between genome size and stomatal density (Beaulieu et al., 2008), and the genome of A. halleri (250 Mb) is approximately double the size of the A. thaliana genome (125 Mb) (The Arabidopsis Genome Initiative, 2000; Briskine et al., 2017).
In summary, we found that glabrous morphs of A. halleri growing under natural conditions had higher stomatal density and stomatal index than a hairy morph. This might contribute to the reported fitness advantage of glabrous plants over hairy plants in absence of herbivores (Sato and Kudoh, 2017). This differing stomatal density phenotype might derive from the common upstream components in the pathways leading to trichome and guard cell development.
Conflict of Interests
The authors declare no competing financial interests.
Author contributions
NMLS, JS, MNH, SAT, GT, HK and AND performed experimentation and/or analysed data, and NMLS, HK and AND wrote the paper.
Data availability
All data generated during this study are included in the published article and Supplementary Information files.
Table S1. Nested ANOVA analysis of (a) stomatal density, (b) stomatal index and (c) pavement cell density. Df is degree of freedom; *, ** and *** indicates significant at p < 0.05, p < 0.01 and p < 0.001 respectively; NS, not significant at p > 0.05.
Dataset S1. Complete stomatal density data collected during experimentation.
Acknowledgements
We thank Dora Cano-Ramirez, Haruki Nishio and Tasuku Ito for experimental assistance. This research was funded by the UK Biotechnology and Biological Sciences Research Council (BBSRC; grant BB/J014400/1), The Royal Society (grant IE140501), and the Japan Society for Promotion of Science (JSPS; CREST no. JPMJCR15O1). This research was conducted using Joint Usage of the Center for Ecological Research, Kyoto University.