To date, very few species have been investigated for nectar proteins, and the few studies that have appeared in the literature demonstrate that floral nectar contains a large, heterogeneous assemblage of defence proteins (Carter and Thornburg, 2000, 2004; Naqvi et al., 2005; Carter et al., 2007; Kram et al., 2008; Hillwig et al., 2010, 2011). The floral nectar of an ornamental tobacco (Nicotiana langsdorffii × Nicotiana sanderae) contains five nectar proteins (nectarins) that function in a novel biochemical pathway – the nectar redox cycle (NRC), and this pathway produces high concentrations of hydrogen peroxide that protect against micro-organisms.

Although the NRC appears to be present in some other unrelated species (Carter and Thornburg, 2000), it was not found to occur in Petunia, even though this belongs to the same family (Solanaceae) as tobacco (Hillwig et al., 2010). The floral nectar of Petunia hybrida contains several RNases, a peroxidase and an endochitinase that have antimicrobial activity (Hillwig et al., 2010, 2011).

A lipase (JNP1) was reported from the floral nectar of Jacaranda mimosifolia, although its antimicrobial activity has not yet been demonstrated (Kram et al., 2008). Extrafloral nectar also has its own enzymatic defence against invasion by micro-organisms in that the predominant enzymes present are chitinase, β-1,3-glucanase and peroxidase (Gonzales-Teuber et al., 2009, 2010). Defence proteins, such as glucosidases, chitinases, hydrolases and thaumatine-like proteins, have also been detected in pollination drops (Wagner et al., 2007). These secretions are produced by gymnosperm ovules and have a similar chemical composition to angiosperm nectar (Nepi et al., 2009).

Four of the 24 proteins recognized by 2-D electrophoresis were identified by MS and most closely resemble β-d-xylosidases from Arabidopsis thaliana, with some homology to a β-d-xylosidase from Medicago varia (Table 1). Results of the MS/MS analysis are summarized in Table 1, where the spot numbers match those reported in Fig. 1C and D. Accession number in the UniProtKB database, protein name, species, peptide sequence and Mascot Score/Mascot threshold score are also included. As the complete genome sequence of C. pepo has not yet been determined, the Mascot peptide sequence search was carried out setting a large taxonomy range to Viridiplantae. As a consequence, the majority of the peptide sequences found match the amino acid sequence of β-d-xylosidase in A. thaliana, one of the completely sequenced plant organisms. It is interesting to note that these peptides probably represent the most conserved of the amino acid sequences of β-d-xylosidase from A. thaliana and C. pepo. From the peptide sequence analysis, it is clear that the four spots can be assigned to β-d-xylosidase, as some common peptides have been found between them. The four spots had a molecular mass of approx. 70 kDa and a pI ranging from 5·6 to 6·3 (Fig. 1C). Having identical molecular weights, but slightly different isoelectric points (Fig. 1C, D), they are thought to be isoforms of the same protein, probably resulting from specific post-translational modifications. On the basis of spot quantification analyses, the four hypothetical isoforms were present in relatively similar quantities in both male and female flowers (Fig. 1E), with the exception of polypeptide 4, which appeared to be slightly more abundant in the female nectar.

The presence of xylosidases was supported by the enzymatic assay. This revealed xylosidase activity of 0·23 ± 0·04 and 0·29 ± 0·06 µm min⁻¹ mL⁻¹ in female and male nectar, respectively. The occurrence of such activity under conditions similar to those that occur in nature (temperature = 30 °C) was confirmed by the increase in xylose concentration and a corresponding decrease in the concentration of the xylo-oligosaccharides – and especially xylobiose – in the mixture nectar + oligosaccharides after incubation at 24 and 48 h (Fig. 2).

Neither xylose nor the xylo-oligosaccharides tested was detected in male and female nectar, indicating that lack of xylosidase activity is simply due to the absence of the appropriate substrate.

Endoxylanases and xylosidases are key enzymes in the degradation of xylans, the major hemicelluloses found in the secondary walls of most higher plants. Xylans have a relatively complex structure based on a β-1,4-linked d-xylose backbone, substituted to varying degrees (Minic et al., 2004). Endo-β-1,4-xylanases hydrolyse the insoluble xylan backbone into shorter, soluble xylo-oligosaccharides, while β-d-xylosidases hydrolyse xylo-oligosaccharides and xylobiose from their non-reducing ends to liberate d-xylose (Minic et al., 2004, and references therein). Side chain-cleaving enzymes, such as α-l-arabinofuranosidase, are also important, and they are recognized as a limiting step in achieving efficient hydrolysis of the polysaccharide polymer (Tuncer and Ball, 2003).

Plants use these enzymes for dynamic regulation of cell wall morphology, structure and composition during their development (Minic et al., 2004). These same classes of enzymes represent important components in the offensive arsenal of phyto-pathogens, both fungi and bacteria, and they are used to degrade cell wall polymers when invading plant tissue (Beliën et al., 2006).

The interaction between plants and pathogens induces a diverse array of responses from both sides. Plant defence responses, including cell wall strengthening, production of antimicrobial compounds, ethylene biosynthesis and rapid, localized cell death (Aro et al., 2005; Beliën et al., 2006, and references therein), are frequently triggered by pathogen- or plant-derived molecules that have been termed ‘elicitors’ (Bucheli et al., 1990; Esquerré-Tugayé et al., 2000). Plant responses can be stimulated by the direct interaction of a specific pathogen peptide with the plant cell and do not involve intermediate compounds. Specific fungal xylanases are reported to be able to stimulate plant responses directly (Sharon et al., 1993; Noda et al., 2010).

Xylo-oligosaccharides are recognized as important signals for defence responses in plants, and are most probably involved in the elicitation of phytoalexins, ethylene synthesis, PR (pathogenesis-related) proteins (Ryan and Farmer, 1991) and xylanase inhibitor proteins (Beliën et al., 2006) following plant tissue invasion by fungi. At the same time, these wall-derived molecules increase the pathogenic potential of micro-organisms. It was demonstrated that the production of plant cell wall-degrading enzymes in micro-organisms (cellulase, hemicellulase, pectinase and ligninases) can be induced by the presence of wall polymers, or molecules derived from these polymers. For example, the presence of xylobiose and various other oligosaccharides in cultures of the fungus Trichoderma reesei is known to induce cellulase and xylanase expression (Aro et al., 2005).

In the present study, the authors propose a functional role for β-xylosidase in C. pepo nectar that takes into account all the above observations. The invasion of the nectar by micro-organisms is followed by damage to the nectary cell walls due to the action of cellulases and xylanases produced by the pathogens. This action may induce the release of several oligosaccharides from the cell walls. After the invasion of the nectar by micro-organisms, it is likely that the relative abundance of the plant cell wall-derived oligosaccharides displaying differing degrees of polymerization (DP) is important to the plant. For example, their presence at very low concentrations may be useful in ‘alerting’ the plant's defence mechanism against micro-organisms (Shibuya and Minami, 2001), but increased levels may be detected promptly by the latter, thus resulting in their increased ability to damage the plant cell wall via xylanase activity. Therefore, the degradation of surplus xylo-oligosaccharides may help keep pathogens present in nectar at a reduced state of activity. It is important to point out that the DP of xylo-oligosaccharides capable of inducing the xylanase activity of micro-organisms varies widely: xylobiose (DP 2) is reported to be the elicitor of xylanases in the fungi Aspergillus and Trichoderma (Aro et al., 2005), while the same enzyme activity is stimulated by xylo-oligosaccharides with a DP of 6–30 in the bacterium Prevotella bryantii (Miyazaki et al., 2005). Thus, it is very likely that the nectar defence system, which must be effective against a wide range of micro-organisms, is equipped with a complete set of enzymes involved in xylan degradation (endo-β-1,4-xylanases, β-d-xylosidases and α-l-arabinofuranosidases), although only two of them were identified in the present study (Table 1). Another strategy involving the direct inhibition of microbial xylanase activity has been reported for Nicotiana (Harper et al., 2010). Here, the nectarin NEC4 functions as a defence agent that inhibits a xyloglucan-specific endoglucanase produced by fungi during pathogenesis. Thus, it appears that protection of nectar from invasion by micro-organisms may be direct (by inhibiting pathogen enzyme activity involved in cell wall degradation, as in Nicotiana) or indirect (by regulating the concentration of oligosaccharides released by cell walls and that elicit the cell wall-degrading activity of the micro-organism).

An alternative explanation for the function of xylosidases found in the floral nectar of C. pepo takes into account the possibility that these enzymes are derived from nectary cell walls, from where they are mobilized by nectar flow. Although the more common method for nectar exudation is via the modified stomata of the nectary epidermis, nectar exudation through the epidermal cell wall cannot be ruled out (Nepi et al., 1996; Nepi, 2007). Xylosidases and endo-xylanases may be involved in the reorganization of cell walls during the development of nectary parenchyma and the changes in cell shape and volume that accompany this process (Nepi et al., 1996). Reorganization of cell walls may also increase wall permeability to nectar. These enzymes, once released into the nectar, help defend the latter from fungal or bacterial invasion, as described above.

It is interesting to note that a xylosidase was also found in the pollination drop of Juniperus communis (Wagner et al., 2007). Both pollination drop and nectar are secretions that have a very similar chemical composition, even though their functions are entirely different (Nepi et al., 2009). Whereas the former is very common amongst gymnosperms, and is the landing site for pollen, the latter is widely distributed amongst angiosperms and is the most common reward for pollinators. Since both are sugary solutions and are more or less exposed to the environment, they are equally subject to contamination and invasion by micro-organisms. The presence of xylosidases in both these secretions may account for a certain degree of conservatism in the defence proteins of these two groups of plants.