• Angiosperm stigmas exhibit high levels of peroxidase activity when receptive to pollen. To explore possible function(s) of this peroxidase activity we investigated amounts of reactive oxygen species (ROS), particularly hydrogen peroxide, in stigmas and pollen. Because nitric oxide (NO) was recently implicated in pollen tube growth, we also investigated amounts of NO in pollen and stigmas.
• Reactive oxygen species accumulation was assessed with confocal microscopy and light microscopy using ROS probes DCFH2-DA and TMB, respectively. NO was assayed using the NO probe DAF-2DA and confocal microscopy.
• Stigmas from various different angiosperms were found to accumulate ROS, predominantly H2O2, constitutively. In Senecio squalidus and Arabidopsis thaliana high amounts of ROS/H2O2 were localized to stigmatic papillae. ROS/H2O2 amounts appeared reduced in stigmatic papillae to which pollen grains had adhered. S. squalidus and A. thaliana pollen produced relatively high amounts of NO compared with stigmas; treating stigmas with NO resulted in reduced amounts of stigmatic ROS/H2O2.
• Constitutive accumulation of ROS/H2O2 appears to be a feature of angiosperm stigmas. This novel finding is discussed in terms of a possible role for stigmatic ROS/H2O2 and pollen-derived NO in pollen–stigma interactions and defence.
The interaction between pollen and stigma is one of the most important stages in the life cycle of a flowering plant, because its outcome determines whether fertilization will occur and thus whether seed will be set. This critical cellular dialogue between the haploid pollen (grain and tube) and the diploid cells of the stigma (and style) is one of the most precisely adapted of all activities of the plant – morphologically, physiologically and biochemically (Heslop-Harrison, 1978) – and has become a paradigm for the study of cell recognition and cell signalling in plants.
For fertilization to be achieved, pollen must establish molecular congruity/compatibility with the stigma and then, following production of a pollen tube, with the transmitting tissue of the style and ovary as the pollen tube grows through the pistil to deliver its two sperm cells to an ovule. Thus there must be a continuous exchange of signals, both physical and chemical, between pollen and pistil from the moment a pollen grain arrives on the stigma to the moment the pollen tube enters the ovule. Identifying these signals and the responses they induce has been the subject of intense research for the past three decades and a picture is emerging of a diverse array of signals that influence pollen germination and pollen tube growth and guidance within the pistil (Franklin-Tong, 2002; Johnson & Preuss, 2002; Feijo et al., 2004; Dresselhaus, 2006). Recently the animal neurotransmitter, gamma-aminobutyric acid (GABA), was identified as a potential chemoattractant for pollen tubes in Arabidopsis (Palanivelu et al., 2003), whilst in Lilium longiflorum, nitric oxide (NO) has been implicated in pollen tube guidance as a putative negative regulator of pollen tube growth able to induce tip reorientation (Prado et al., 2004).
When a pollen grain lands on a stigma, specific recognition events must take place to establish that: (a) the object that has alighted is a pollen grain and not a fungal spore or bacterium; (b) it is a pollen grain of the correct species, or a closely related species (interspecific hybridization is fairly common in angiosperms); and (c) in most hermaphrodite flowering plants, it is not a self-pollen grain (Heslop-Harrison, 1978; Franklin-Tong, 2002; Hiscock, 2004). Whilst the last of these three recognition events (self-incompatibility) has been studied extensively (reviewed in Hiscock & McInnis, 2003), relatively little is known about molecular signals and interactions mediating the first two recognition events.
The stigma surface is only receptive to pollen for a relatively short period, so the timing of pollination is critical. Pollination either side of this period of optimal female receptivity results in reduced seed set, or no seed set (Herrero, 2003). It has long been known that receptive stigmas ‘ripe’ for pollination are characterized by high levels of peroxidase activity (Dupuis & Dumas, 1990; McInnis et al., 2006) and tests most widely used to determine pistil receptivity measure stigma peroxidase activity (Dafni & Motte Maues, 1998). Nevertheless, the function of these ubiquitous enzymes in stigmas is not known (McInnis et al., 2006).
Recently we identified and characterized a stigma-specific peroxidase (SSP) from the ragwort Senecio squalidus (McInnis et al., 2005). As part of ongoing work to determine the function of SSP and stigmatic peroxidases generally, we showed that Senecio stigmas accumulate high amounts of ROS, particularly H2O2, in their epidermal cells (papillae) where SSP is localized (McInnis et al., 2006). The presence of such high amounts of ROS/H2O2 in the papillae, which receive and discriminate pollen, suggested that ROS/H2O2 (and, by potential association, SSP) may be important for stigma function. ROS/H2O2 have a variety of roles in cell metabolism but also act as signalling molecules mediating a range of cellular processes from development to defence, often in association with NO (Hancock et al., 2006), so it was not unreasonable to speculate that ROS/H2O2 might be involved in pollen–stigma interactions. Here we expand upon these preliminary findings in Senecio and show for the first time that stigmas from a range of different species, including Arabidopsis thaliana, accumulate high amounts of ROS/H2O2. In species with papillate stigmas ROS/H2O2 accumulation was confined almost exclusively to the papillae. We further show that pollen of S. squalidus and A. thaliana produce NO and that exogenous NO can reduce ROS/H2O2 amounts in stigmas. These observations are discussed in the context of possible functions for ROS/H2O2, peroxidases and NO in pollen–stigma interactions and defence.
Materials and Methods
Senecio squalidus L. plants were grown in a glasshouse under a 16 h day : 8 h night photoperiod according to Hiscock (2000). Arabidopsis thaliana (Col-O ecotype) were grown under essentially the same conditions but in a growth cabinet (16 h day : 8 h night, 60–100 µE m−2 s−1, 20°C). Flowers of Magnolia sp., Laurus nobilis, Drimys winteri, Lilium longiflorum, Juncus sp., Carex sp., Berberis darwinii, Ilex aquifolium, Pittosporum crassifolium, Echium canariensis, Menyanthes trifoliate, Aesculus hippocastanum, Ipomoea sp., Antirrhinum hispanicum, Apium sp., Sorbus aria, and Senecio squalidus were obtained from the University of Bristol Botanic Garden. Flowers of Anigozanthos sp., Freesia sp., and Ranunculus sp. were obtained from a local florist.
Reactive oxygen species measurement was performed using the fluorescent ROS indicator dye DCFH2-DA (2′,7′-dichlorodihydrofluorescein diacetate; Calbiochem, Darmstadt, Germany). Pistils of S. squalidus and A. thaliana were excised from 1 d postanthesis flowers and immediately immersed in 5 ml of 50 µm DCFH2-DA in MES-KCl buffer (5 µm KCl, 10 mm MES, 50 µm CaCl2, pH 6.15) for 10 min followed by a wash step in fresh buffer for 15 min. Other pistils were treated with 1 m sodium pyruvate (Sigma-Aldrich, Poole, UK) in MES-KCl buffer for 30 min, followed by a wash step, or treated with 500 µm SNP (sodium nitroprusside dehydrate, Sigma-Aldrich) in MES-KCl buffer before finally treating with DCFH2-DA as above. Negative controls were treated with buffer only.
NO measurement was performed using the fluorescent NO indicator dye DAF2-DA (diaminofluorescein diacetate; Calbiochem) largely according to Desikan et al. (2002). Pistils were excised and immediately immersed in MES/KCl buffer for 10 min, transferred to 10 µm DAF2-DA for 10 min, followed by a wash step (with MES/KCl buffer) for 15 min. Negative controls were treated with buffer only.
Imaging with confocal microscopy was performed on a Nikon PCM2000 (excitation 488 nm, emission 515–560 nm) and data analysed with SCION IMAGE software (Scion, Frederick, MD, USA). Relative fluorescence intensity values were calculated as average intensities from several pistils analysed in different experiments.
Reactive oxygen species/H2O2 measurement was performed using the H2O2 indicator dye TMB (3,5,3′,5′-tetramethylbenzidine-HCl; Sigma-Aldrich) (Barceló, 1998). Whole pistils or stigmas were excised from 20 different angiosperm species selected from genera covering a broad range of angiosperm phylogenetic diversity, as follows – basal monocolpates: Magnolia sp., Laurus nobilis, Drimys winteri; monocots: Lilium longiflorum; Juncus sp., Carex sp., Anigozanthos sp., Freesia sp.; basal eudicots: Ranunculus sp., Berberis darwinii; eudicots: Ilex aquifolium, Pittosporum crassifolium, Echium canariensis, Menyanthes trifoliate, Aesculus hippocastanum, Ipomoea sp., Antirrhinum hispanicum, Apium sp., Sorbus aria, Senecio squalidus. Depending on the size of the gynoecia, whole pistils or pistil segments consisting of stigma and a portion of style were excised from newly open flowers and immersed in TMB solution (0.1 mg ml−1 in Tris-acetate, pH 5.0) until a blue colour was observed (Barceló, 1998; Barcelóet al., 2002). Stained stigmas were observed using a Nikon dissecting microscope and images gathered with a Nikon Coolpix digital camera.
ROS accumulate constitutively in stigmatic papillae of Senecio squalidus and Arabidopsis thaliana and consist principally of H2O2
Confocal microscopy following treatment of pistils of S. squalidus with DCFH2-DA revealed relatively intense fluorescence in stigmas, indicating an abundance of ROS (Fig. 1a,b). Fluorescence was mostly localized to the specialized epidermal cells (papillae) of the stigma surface (Fig. 1a). Treatment of pistils with sodium pyruvate, a H2O2 scavenger, before addition of DCFH2-DA resulted in diminished fluorescence in papillae, suggesting that H2O2 is the predominant ROS in stigmas (Fig. 2a). ROS/H2O2 production was then investigated in stigmas of A. thaliana using confocal microscopy (Fig. 1c,d). As in S. squalidus, stigmatic papillae of A. thaliana fluoresced strongly in the presence of DCFH2-DA, indicating constitutive accumulation of ROS (Fig. 1c). Pretreating A. thaliana pistils with sodium pyruvate also quenched DCFH2-DA fluorescence, again suggesting that H2O2 is the most abundant stigmatic ROS (Fig. 2b). Treatment of pistils of S. squalidus and A. thaliana with FDA resulted in equal fluorescence in all cells of the pistil (data not shown), indicating that differential fluorescence intensities observed with DCFH2-DA were not a consequence of differential cell permeability of the dye.
Stigmas from a range of different angiosperms accumulate ROS/H2O2 constitutively
To determine whether stigmas of other angiosperms accumulate ROS/H2O2 constitutively and whether ROS/H2O2 accumulation is a general property of the angiosperm stigma, the presence of stigmatic ROS/H2O2 was investigated in 20 species spread among the major lineages of angiosperms: basal monocolpates, monocots, basal eudicots and eudicots using the ROS/H2O2-specific stain 3,5,3′,5′-tetramethylbenzidine (TMB). This assay is based on the H2O2-dependent oxidation of TMB by cellular peroxidases (Barceló, 1998). TMB was used instead of DCFH2-DA so that species with larger stigmas could be screened rapidly by conventional light microscopy – TMB turns blue in the presence of H2O2 (Barceló, 1998; Barcelóet al., 2002). Stigmas from 14 of the 20 species tested gave a strong positive reaction within as little as 5 min after treatment with TMB (Fig. 3). Stigmas from the remaining six species gave a positive reaction within 30 min, indicating the presence of ROS/H2O2 in stigmas from all 20 species tested. In species with papillate stigmas, close inspection revealed that staining of the stigma was associated exclusively with the papillae (Fig. 3). Interestingly, in those species with nonpapillate stigmas, staining was localized to specific regions of the stigma; for instance, in Ilex aquifolium, staining was associated with the stigma periphery (Fig. 3g), whereas in Pittosporum crassifolium, staining was concentrated at the centre of the stigma (Fig. 3h). Stigmas were also tested for peroxidase activity using the guaiacol peroxidase assay and all gave a positive reaction for the presence of peroxidases, mostly colocalizing with the presence of ROS/H2O2 (data not shown).
Pollen produces NO
The NO-specific fluorescent probe DAF2-DA was used in conjunction with confocal microscopy to investigate NO accumulation in stigmas of S. squalidus and A. thaliana. Low intensities of DAF2-DA fluorescence were observed in stigmas of both species, reflecting the presence of low amounts of NO (Figs 2, 4a). However, relatively higher intensities of fluorescence were observed in pollen of both S. squalidus and A. thaliana (Fig. 4b,c). Interestingly, in both S. squalidus and A. thaliana we observed consistently high intensities of fluorescence in pollen tubes compared with ungerminated or germinating pollen that had not yet produced a pollen tube.
Amounts of stigmatic ROS/H2O2 are reduced in the presence of NO
Observation of pollinated stigmas of S. squalidus revealed variable amounts of DCFH2-DA fluorescence associated with pollen, but fluorescence intensities in pollen grains and pollen tubes were generally much lower than for stigmatic papillae. Interestingly, where germinating pollen grains were in contact with the stigma, intensities of DCFH2-DA fluorescence in stigmatic papillae below the pollen grain appeared lower than in stigmatic regions not in contact with pollen grains (Fig. 5a,b) suggesting that in these regions, amounts of ROS/H2O2 were reduced. To investigate the possibility that pollen NO may be affecting amounts of stigmatic ROS/H2O2, stigmas of S. squalidus were exposed to sodium nitroprusside (SNP), which generates NO in the presence of light, before treatment with DCFH2-DA and observation with confocal microscopy. Stigmas treated with SNP showed a significant (∼80%) reduction in DCFH2-DA fluorescence (Figs 2, 5c,d), indicating a significant reduction in stigmatic ROS/H2O2 by NO. Repeating this treatment for A. thaliana stigmas gave a similar, though less dramatic (∼60%), reduction in DCFH2-DA fluorescence in the presence of SNP (Fig. 2a).
Here we show for the first time that stigmas from a range of different angiosperms accumulate high amounts of ROS, predominantly H2O2, when they are most receptive for pollination. In species with papillate stigmas (Heslop-Harrison & Shivanna, 1977), ROS/H2O2 were detected principally in the papillae – cells essential for pollen-stigma recognition and pollen development (Wolters-Arts et al., 1998). Confocal microscopy showed conclusively that, in S. squalidus and A. thaliana, ROS/H2O2 accumulation is confined to stigmatic papillae. Whilst DCFH2-DA is a general fluorescent probe for ROS (and also reacts with NO), we conclude that the majority of DCFH2-DA fluorescence in stigmatic papillae reflects the presence of H2O2 because: (a) DCFH2-DA fluorescence was reduced after application of the H2O2 scavenger sodium pyruvate (McInnis et al., 2006); (b) the TMB reaction, which is dependent on the presence of H2O2 (Barceló, 1998), gave a strong signal in stigmatic papillae (Fig. 3); and (c) amounts of NO in stigmas were very low as assessed by DAF-2DA (Fig. 2). Caution must always be exercised when interpreting the effects of sodium pyruvate, because as well as scavenging H2O2 it can also affect a number of metabolic processes, most notably the citric acid cycle (Sundqvist et al., 1989) and alternative oxidase pathway (Millar et al., 1993; Ribas-Carbo et al., 1995), which could lead to an indirect reduction in H2O2 through a general elevation of metabolism. Nevertheless, when the sodium pyruvate data are combined with the TMB data, we consider it reasonable to assume that H2O2 is the principal ROS in stigmas (McInnis et al., 2006). The possibility that differential fluorescence in stigmas with DCFH2-DA (and DAF-2DA) was a consequence of differential cell permeability was ruled out because treating pistils with fluorescein diacetate (FDA) resulted in equally bright fluorescence in all cells of the pistil.
The presence of such high constitutive amounts of ROS/H2O2 in stigmatic papillae was unexpected and suggests a functional role. The relationship between high amounts of stigmatic ROS/H2O2 and high levels of peroxidase activity in receptive stigmas is intriguing and may indicate a regulated turnover of H2O2 in this tissue (McInnis et al., 2006). Stigmatic peroxidases such as SSP (McInnis et al., 2005, 2006) might therefore be expected to help maintain controlled amounts of H2O2 through regulated degradation, or, alternatively, they may contribute to the generation of H2O2, as has been shown for certain plant peroxidases (Blee et al., 2001; Bolwell et al., 2002). A further possibility is that stigma peroxidases utilize H2O2 in peroxidation reactions such as the cross-linking of pectins and extensins in the cell wall (Barceló, 1998). These reactions could be important for pollen adhesion and/or pollen tube growth and development on and within the stigma (McInnis et al., 2006).
It is now well known that H2O2 and other ROS play important roles in cell signalling in both animals and plants (Hancock et al., 2001), and in plants, ROS/H2O2 signalling is important for several cellular processes, including stomatal closure, adventitious root development, root hair growth and gravitropism, as well as in responses to pathogen attack (Laloi et al., 2004; Torres & Dangl, 2005) and in fungus–plant mutualistic interactions (Tanaka et al., 2006). Stigmatic ROS/H2O2 could therefore be involved in signalling networks that promote pollen germination and/or pollen tube growth on the stigma. Recent studies on tobacco pollen growing in vitro have shown that H2O2 can promote pollen tube growth (N. Smirnoff and V. Zarsky, pers. comm.). This suggests a potential analogy with root hairs (which, like pollen tubes, elongate by tip growth) where ROS/H2O2 produced by an NADPH oxidase acts as a positive regulator of cell growth (Foreman et al., 2003). In most cases where ROS/H2O2 has been implicated in regulatory cell signalling networks, the enzymatic source of ROS/H2O2 is one or more NADPH oxidase (NOX) isoforms (Torres & Dangl, 2005), even though other enzymes, such as peroxidases, amine oxidases and xanthine oxidases, can also generate ROS/H2O2 (Neill et al., 2002). It is therefore possible that different stimuli activate specific ROS/H2O2-generating enzymes. An important next step is therefore to establish the enzymatic source(s) of stigmatic ROS/H2O2.
The second possible function of ROS/H2O2 in stigmas is in defence against attack by pathogens. ROS abundance is known to escalate in plant cells during the oxidative burst associated with the hypersensitive response (HR, Lamb & Dixon, 1997). These ROS play important signalling roles during initiation of resistance to invading pathogens (Lamb & Dixon, 1997; Desikan et al., 1998; Torres & Dangl, 2005). The role of ROS signalling in regulation of programmed cell death (PCD) associated with the HR is complex (Torres & Dangl, 2005). Despite evidence for induction of PCD by H2O2 during the HR (Levine et al., 1994), recent evidence shows that in A. thaliana NOX-derived ROS acts as a negative regulator of PCD and is antagonistic to the PCD-activator salicylic acid (Torres et al., 2005).
Despite producing secretions conducive to ‘appropriate’ pollen development, the stigmas of flowering plants are stubbornly resistant to attack by pathogens. It is possible that the high amounts of stigmatic ROS/H2O2 contribute to this resistance in a toxic manner similar to that proposed for ROS in the protection of nectar from pathogens (Carter & Thornburg, 2000; Carter & Thornburg, 2004). Like stigmatic secretions, nectar is a potential source of nutrients for microbes, but nectar accumulates very high amounts of ROS and never experiences microbe attack (Carter & Thornburg, 2004). So perhaps by analogy with nectar, stigmas are protected from microbial infection by the high toxic amounts of ROS/H2O2.
If high amounts of stigmatic ROS/H2O2 are toxic to potential pathogens, they are clearly not toxic to pollen, nor do they induce PCD in the stigma. Under this scenario, pollen must respond differently to stigmatic ROS/H2O2 than potential pathogens – a signalling dialogue that may be part of the system that identifies a pollen grain as a pollen grain and not a fungal spore. In this context, the relationship between pollen NO and stigmatic ROS/H2O2 may be important. Ours is not the first observation of NO production in pollen tubes (Prado et al., 2004), but it is the first report of NO in ungerminated pollen. When we treated stigmas of S. squalidus and A. thaliana with the NO donor SNP, fluorescence intensities in the presence of DCFH2-DA were reduced by as much as 80%, indicating that NO can reduce ROS/H2O2 abundance in stigmas, perhaps by interfering with their production or by scavenging them. It is therefore intriguing to speculate that the reduced DCFH2-DA fluorescence observed in stigmatic papillae directly beneath developing pollen grains may be the result of pollen-derived NO. The potential for a causal relationship between the presence of NO in pollen and the down-regulation of ROS/H2O2 in stigmatic papillae is exciting and demands further investigation. We have yet to observe any differences in the response by S. squalidus stigmas to compatible vs self-incompatible pollen, or pollen from other species of Senecio, so it is possible that pollen NO acts as a signal telling the stigma that a pollen grain (as opposed to a fungal spore) has alighted. Given that NO has been shown to scavenge ROS and counteract the effect of ROS in animal systems (Stamler, 1994), we speculate that NO from pollen may be reducing ROS/H2O2 abundance in stigmatic papillae. In support of such a hypothesis, there are compelling reports that NO can counteract the effects of ROS in plants (reviewed in Beligni & Lamattina, 2001), for instance in certain host–pathogen interactions (Beligni & Lamattina, 1999) and during the establishment of legume-Rhizobium symbioses (Herouart et al., 2002). The real possibility of NO/ROS ‘cross-talk’ during the initial stages of pollen-stigma recognition is extremely exciting, because, if proven to be correct, it will mirror the situation in animal systems, where it is becoming evident that ROS and NO interact in mammalian reproductive processes such as sperm capacitation and sperm–egg interactions (Ford, 2004).
We thank Steve Neill and Jo Bright for helpful discussions. This work was funded by a pump priming grant from the Faculty of Science, University of Bristol, which supported SMM for 6 months. RD is supported by a Fellowship from The Leverhulme Trust. Work in JTH's laboratory is supported by the BBSRC, and work in SJH's laboratory is supported by the NERC.