Flowers are designed to attract animal visitors. Consequently, floral organs are exposed to the microbes carried by their visitors as well as the microbes transported in air and rain. This is particularly problematic because the nectar, stigmas and the transmitting tissue of styles are especially rich sources of nutrients, energy and water for the growth of microbes. Not surprisingly, many pathogens are known to invade plants via the nectaries and sexual organs of flowers and some pathogens are specifically adapted to be dispersed from plant to plant by floral visitors (Antonovics, 2005). So how do plants protect themselves (and their developing offspring) from the hoard of pathogens that could invade through floral organs? In an elegant study in this issue of New Phytologist, Huang et al. (pp. 997–1008) show that a volatile organic compound (VOC), (E )-β-caryophyllene, that is produced by the stigmas of Arabidopsis thaliana flowers, acts as a fumigant that limits the growth of bacterial pathogens such as Pseudomonas syringae pv. tomato (PST).
‘But, does ( E)-β-caryophyllene have anti-microbial properties or does it signal to the plant to up-regulate other anti-microbial defenses?’
PST is an important bacterial pathogen of tomato and plants in the cabbage family, such as Arabidopsis thaliana, that causes necrotic lesions on leaves but such damage is only occasionally seen on flowers. (E )-β-Caryophyllene is one of the most common compounds found in floral odors, but interestingly, it has never been shown to function in pollinator attraction (Huang et al.). Huang et al. used A. thaliana plants that were incapable of producing (E )-β-caryophyllene due to a transposon insertion into a key gene in the pathway leading to (E )-β-caryophyllene synthesis and inoculated the stigmas of these mutant plants and wild-type plants with PST. They found that the bacteria had greater proliferation and produced more symptoms on the mutant plants. Moreover, the seeds resulting from the mutants weighed less and were misshapen compared with wild-type plants but there were no differences in seed characteristics when the stigmas were not inoculated with PST. In short, (E )-β-caryophyllene seems to protect the plants from invading bacteria and that the pathogens, if unchecked at the stigma, adversely affect offspring quality. To strengthen these findings, they used transgenic plants that produced extra (E )-β-caryophyllene in the leaves. When the transgenic and wild-type plants were sprayed with PST, proliferation of the bacteria was greater on the wild-type plants and the wild-type plants had more necrotic lesions than the transgenic plants. These findings were consistent with the results of their stigma infection experiment: (E )-β-caryophyllene increases plant resistance to a bacterial pathogen. But, does (E )-β-caryophyllene have anti-microbial properties or does it signal to the plant to up-regulate other anti-microbial defenses? To answer these questions, Huang et al. cultured PST in liquid media with and without minute quantities of (E )-β-caryophyllene and they cultured PST on solid media and added minute quantities of (E )-β-caryophyllene to the air passing over these cultures. In both cases, bacterial growth was reduced in a dose dependent manner by (E )-β-caryophyllene. Finally, they performed a series of experiments using quantitative real-time PCR to show that other anti-herbivore and anti-pathogen defense related genes were not up-regulated by (E )-β-caryophyllene in A. thaliana. In short, the production of (E )-β-caryophyllene on the stigmas of A. thaliana plants functions as a direct defense against bacterial pathogens.
Floral traits are influenced by many selective pressures
Pollination biology and floral ecology historically have been concerned with the effects of floral traits, such as flower color, odor, shape, inflorescence architecture and the types and quantities of the floral rewards, on the primary functions of the flowers: the attraction of pollinators and the dispersal of pollen from the anthers and the deposition of pollen onto the stigma (e.g. Darwin, 1862). However, floral ecologists quickly learned that unwanted animal visitors, such as those that steal the nectar without affecting pollination and those that eat flowers, can have adverse effects on pollinator attraction and the number and quality of the offspring that are produced by plants through both the male (pollen) and female (seed production) roles. Moreover, these unwanted visitors are thought to have had a profound influence on the evolution of floral traits (e.g. Inouye, 1983; Strauss et al., 1996). For example, tubular flowers typically protect their nectar at the base of the tube with thick leathery calyces that cannot be easily punctured or chewed by nectar robbers. More recently, some floral volatiles have been shown to function as deterrents against flower feeding by herbivorous insects (e.g. Junker & Bluethgen, 2008; Kessler et al., 2008; Willmer et al., 2009). The study by Huang et al. suggests that the bouquet of a floral fragrance can also be influenced by pathogens. Clearly, floral traits evolve under a complex set of selective pressures: they must attract and reward pollinators for the proper dissemination of pollen to conspecifics and the proper deposition of pollen from conspecifics but they must also discourage unwanted floral visitors, including pathogenic microbes, or suffer the adverse fitness consequences of these visitors.
Do other floral traits function in pathogen resistance?
It is reasonable to expect that floral exposure to microbial pathogens would increase with time and the number and types of animal visitors. The longevity of unpollinated flowers ranges from a few hours in Datura spp., Cucurbita spp. and many other species to several weeks in some orchid species (Primack, 1985). Shykoff et al. (1996) noted that in natural populations of two species of dioecious Silene, the male plants had both longer floral lifetimes and a higher incidence of the pollinator vectored fungal disease, Microbotrium violaceum, than the female plants. They proposed that differences in the infection rate between the two sexes could be due to differences in floral longevity and later demonstrated that male and female plants differ in risk of infection per contact (Kaltz & Shykoff, 2001). These studies suggest that floral longevity may represent a tradeoff between expected fitness payoffs (additional pollen donation and pollen accumulation on the stigma) and the risk of pathogen infection.
In virtually all species, including the long-lived flowers of orchids, pollination, the entry of pollen tubes into the ovary or fertilization, is associated with a rapid senescence of the styles and other floral organs (e.g. Stead, 1992). In maize, the spores of several species of ear rot fungi are transported to the silks (styles) by air currents and by silk feeding corn root worm beetles. The fungal spores germinate and grow through the silks and infect the parent plant and its seeds. Maize silks rapidly collapse and senesce when pollen tubes enter into the ovary from the silks. During an unusually cool and wet summer that was ideal for the proliferation of ear rot fungi, Valdivia et al. (2006) showed under field conditions that a 12-h post-pollination delay in the senescence of silks resulted in a nearly three-fold increase (72% vs 25%) in the incidence of ear rot fungi in the resulting ears, suggesting that rapid style senescence may function in pathogen resistance.
Because nectar is composed primarily of simple sugars, it is a rich media for the growth of microbes and the nectars of many species are known to contain anti-microbial chemicals and proteins (Nicolson & Thornburg, 2007; Hillwig et al., 2010). Moreover, the nectary provides a relatively easy route for pathogen invasion into the vascular system of plants and several pathogens are known to enter plants via the nectaries of flowers (see Sasu et al., 2010a). In wild and cultivated squash (Cucurbita pepo) and other cultivated cucurbits, Erwinia tracheiphila, the causative agent of bacterial wilt disease, is transmitted by cucumber beetles (Fig. 1). The beetles acquire the pathogen when they feed on infected leaves and they transmit the pathogen in wild squash when they aggregate in the flowers of nearby plants to mate and their infected frass falls onto or near the nectary (Sasu et al., 2010a). Sasu et al. (2010b) showed that the nectar of wild squash is strongly antibiotic and inhibits the growth of E. tracheiphila in vitro as effectively as 5% ampicillin for 12 h and, in vivo, it dramatically reduces probability of infection. Interestingly, the flowers of both cultivated and wild C. pepo are only open for 5–6 h on one morning and they abscise 24–36 h after anthesis. Consequently, E. tracheiphila must traverse the nectary, enter the vascular system and move below the abscission zone within 24 h if infection is to occur. Consequently, the antibiotic nectar merely needs to retard the growth of E. tracheiphila in order to be effective.
Although floral ecologists are just beginning to look at the roles of floral traits in pathogen resistance, it is already apparent that a given floral trait is likely to play multiple roles. For example, pollinators associate the fragrance of a flower with nectar and pollen rewards but individual components in the bouquet may also serve to discourage floral herbivores and resist disease transmission. As noted by Huang et al., many anti-pathogen defenses, including defensive volatile compounds, evolved before flowering plants, and consequently, their roles in pollination (pollinator attraction and pollen dispersal) may be secondary and derived. In this regard, flavonols, many of which are known to have anti-microbial properties, accumulate on the stigmas of many species in response to pollination and damage to floral organs and have been shown to play a role in pollen germination and tube growth (Taylor & Hepler, 1997), but their role in floral pathogen resistance has not been explored. It is also worthwhile to note that sporophytic self-incompatibility in the Brassicaceae has strong similarities to anti-fungal defenses (Dickinson, 1994) while the Solanaceous-type of gametophytic self-incompatibility has strong homologies to RNase based bacterial and virus defenses (Hillwig et al., 2010). Pollination is risky to both the parent plant and its offspring … and floral ecologists are discovering that prudent plants possess traits that allow them to practice safe sex.