Dancing Together. Social Controls in Parasitic Plant Development

WHAT ARE PARASITIC ANGIOSPERMS?

It has been estimated that 1% of all flowering plants are parasitic. In most cases, this parasitism is not obvious. For common eyebright, Indian paintbrush (Castilleja coccinea), gerardia (Agalimis purpurea), and even witchweed (S. asiatica), the only distinguishing feature is the subterranean attachment to host plants. As shown in Figure1, the cellular anatomy of theS. asiatica root meristem differs little from that of other small seeds such as Arabidopsis (Dolan et al., 1993). Longitudinal sections of the 1-d-old S. asiaticaseedling reveal an even more simplified architecture containing a relatively small number of cylindrical layers (Fig. 1A). The most remarkable feature is the apparent absence of a root cap, the meristem surface being covered by the developing epidermal cells (Fig. 1B). The cellular density of the inner stele also appears much reduced because cell vacuoles emerge very early in the central zone. A direct assignment of each individual developing file cannot be made with certainty. In addition, a quiescent center is not readily apparent without a thorough mitotic analysis.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

The cellular structure of the S. asiatica seedling. A, Longitudinal section of a 12-h-old seedling, stained with toluidine blue, showing the radicle and root meristem. The shoot meristem does not emerge until contact with the host is established; the shoot meristem and cotyledons are still enclosed within the dark-staining seed coat. B, Enlargement of the root meristem in A. The simple cellular structure is apparent in the small number of files, each with a comparable level of vacuolarization. C, Longitudinal section of the 20-h-old haustorium. The first 12 to 14 cells along each file swell radially. The arrows indicate the haustorial hairs emerging from the epidermal layer. Bars = 50 μm.

WHAT SIGNALS HOST COMMITMENT?

The breakage of seed dormancy in S. asiaticais dependent on small, diffusing, host recognition molecules, or xenognosins, originating from the roots of its monocotyledonous hosts. The agronomic impact of this parasite on grain production in sub-Saharan Africa, as well as its spread to the southeastern United States, has resulted in extensive efforts to identify these molecular signals. Several xenognostic germination signals have been discovered (Boone and Lynn, 1995) that in some cases are able to define precise chemical potential gradients around the host roots (Fate and Lynn, 1996). The breakage of dormancy is the most critical step in this commitment, starting a viability clock that runs only a few days without host attachment.

The second level of host integration, the development of the host-parasite attachment organ (Fig. 1C), was naturally assumed to be dependent on similar diffusible xenognosins (Riopel, 1979). Haustorial development is not initiated when isolated seedlings are grown in culture, and screens of commercially available plant exudates suggested that small molecules were both necessary and sufficient for induction of haustorial development. The identified xenognosins were unlike any plant hormone known to mediate developmental commitments (Lynn et al., 1981), and surprisingly were not found to be present in host root exudates. Active compounds did appear after host wounding, leading to the proposal that the parasite released oxidative enzymes, similar to cutinase exudation by Fusarium solaniduring haustorial penetration (Shaykh et al., 1977), to liberate xenognosins from the host cell surface (Chang and Lynn, 1986). This hypothesis was tested directly by screening for enzymatic activity. Young S. asiatica seedlings were found, however, to be particularly deficient in cell wall-localized oxidative enzyme activity at all points during xenognosis (Kim et al., 1998).

WITHOUT EXUDED ENZYMES, HOW ARE THE XENOGNOSINS RELEASED FROM HOST CELLS?

The answer initially came from studies with redox-sensitive probes, and later fluorescein diacetate, suggesting that theS. asiatica root meristem constitutively produces hydrogen peroxide (H₂O₂). Exogenous applications of the enzyme catalase removes extracellular H₂O₂, and blocks the haustorial inducing activity of both isolated host cell wall preparations and simple phenolic inducers (Kim et al., 1998). Therefore, extracellular H₂O₂ is necessary and can be limiting in haustorial induction.

The epidermal cells of the S. asiatica meristem constitutively produce H₂O₂. As shown in Figure2, oxidation of CeCl₃ occurs specifically within the apoplasmic space of the cells along the surface of the meristem (Keyes et al., 2000; W.J. Keyes, unpublished data). Conceptually analogous with radar detection, a chemical potential gradient of H₂O₂ would be maintained at the root meristem and released from parasite epidermal cells that are low in peroxidases and cell wall phenols to exploit host epidermal cells rich in both (Fig. 3A). Upon contact with host cell surfaces, the H₂O₂ concentration serves as a peroxidase cosubstrate, oxidatively releasing simple benzoquinone xenognosins from host cell walls and creating a comparable benzoquinone potential gradient radiating from the host cells (Fig. 3B). This mechanism may be general; benzoquinone xenognosins are known to induce haustorial development in many parasites (Lynn and Chang, 1990;Matvienko et al., 2001). In Figure 3B, benzoquinone release is shown in a second step downstream of the probing process, but these processes occur simultaneously, together providing the continuous cross feeding of benzoquinone over the extended time period necessary to ensure commitment to haustoriogenesis (Smith et al., 1990).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

Localization of H₂O₂ accumulation in the seedling tissue. Twelve-hour-old S. asiaticaseedlings were incubated in freshly prepared 5 mmCeCl₃ and 0.1 mm KCl for 2 h and fixed in 1.25% (v/v) glutaraldehyde/1.25% (v/v) paraformaldehyde in 50 mm sodium cacodylate (CAB) buffer for 1 h. After two 10-min washes in 50 mmCAB buffer, the seedlings were dehydrated in a graded ethanol series and exposed twice to propylene oxide. The fixed seedlings were embedded in epoxy plastic, thin sectioned (70–80 nm), and observed with a JEM-1210 transmission electron microscope (JEOL, Tokyo) at 80 kV. Electron-dense deposits of cerium perhydroxides accumulated at interstitial locations between epidermal cells specifically in the meristematic zone. The apoplastic junction between two cells along the meristem surface are shown with cerium perhydroxides deposits (arrows). Bar = 200 nm.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

Proposed mechanism for the induction of haustorial development. A, The growing S. asiatica seedling produces H₂O₂ at the root meristem, giving rise to a localized chemical potential gradient (red) diffusing into the solution. H₂O₂ is necessary and generally the limiting substrate in peroxidase oxidation of cell wall-localized phenolics, such as sinapic acid, into benzoquinones (Kim et al., 1998). B, The accumulating quinones establish a similar chemical gradient (blue), diffusing back to the parasite seedling and signaling haustoriogenesis. Radial swelling and hair formation occur, leading ultimately to the formation of the functional attachment organ within 18 to 20 h.

WILL THE CONSTITUTIVE RELEASE OF H₂O₂IMPACT OTHER CELLULAR PROCESSES?

Production of reactive oxygen species in plants is well documented and a key component in plant defensive responses (Lamb and Dixon, 1997). H₂O₂ is a major product of the oxidative burst, and its presence may be both directly toxic to invading organisms and/or contribute to structural reinforcement of the plant cell wall via cross-linking reactions catalyzed by wall peroxidases. Here, curiously, the invading organism exudes H₂O₂, certainly not the best camouflage for a parasite, particularly considering that reactive oxygen species generally activate systemic resistance in plants. However, only low levels of H₂O₂ accumulate alongside host cells. In normal defense responses, H₂O₂ localized to the apoplast is consumed by wall peroxidases in both cross-linking reactions with pectic-associated phenols and in oxidative cleavage of these phenols (Chang and Lynn, 1990; Kim et al., 1998). Under these conditions, it might be expected that little H₂O₂ would access the host cell cytoplasm to activate counter resistance. What about the parasite cells? Exogenously added fluorescein diacetate requires cytoplasmic esterases to liberate the free phenol for oxidation, and allowed us to stain the cytoplasm of the parasite meristem cells (Keyes et al., 2000). This staining pattern is not altered by exogenous catalase treatment. Therefore, at least part of the H₂O₂ is produced cytoplasmically in the parasite meristem, making it necessary to consider how these cells cope with constitutive H₂O₂ production. It will be equally important to consider how the host cells experience the gradient of the benzoquinones.

HOW ARE THE BENZOQUINONES PERCEIVED?

At the host/parasite interface, which consists of a surface contact between the epidermal cells of both organisms, the host cells cross feed simple benzoquinones from their wall to the parasite cells (Fig. 3B). Receptors for the benzoquinone xenognosins have not yet been identified, but the available data suggest that they must function in a series of one-electron benzoquinone oxidoreductions (Smith et al., 1996). Terminal commitment to the host attachment organ occurs only after several hours of xenognosin exposure, consistent with a time-dependent accumulation of the necessary molecular components. Without sufficient exposure time, the meristem reverts to vegetative growth (Smith et al., 1990, 1996). It could be argued that this exposure time requirement prior to terminal commitment improves the probability of attaching to a viable host site, reporting on a minimal threshold of benzoquinone precursor at the host cell surface.

As is evident in the longitudinal sections of Figure 1C, the inner cells surrounding the forming vascular bundles swell radially, gradually forming an expanded surface area sufficient for host attachment. Somewhat later, the epidermal cells develop haustorial hairs that function as host attachment anchors. Consistent with the two distinct cellular expansion events, at least two unique expansin transcripts, saExp1 and saExp2, are induced by the benzoquinone. Concomitant with that expression, the seedling expansin, saExp3, is down-regulated (O'Malley and Lynn, 2000). As haustorial expansin message accumulation plateaus, the meristem cells terminally commit to forming the new organ. At intermediate xenognosin exposure times, haustorial expansin transcript levels fall after benzoquinone removal. The time necessary for re-induction correlates with the magnitude of expansin message depletion. Benzoquinone perception appears to function as a “molecular capacitor,” allowing accumulation to a molecular charge threshold necessary for commitment to haustorial development. Following this analogy, the capacitor is both leaky and capable of being recharged, ensuring developmental plasticity.

IS EXPRESSION OF THE HAUSTORIAL EXPANSINS SUFFICIENT FOR ORGAN DEVELOPMENT?

Because the attachment organ can be viewed simply as the expansion of preexisting cells, induction of saExp1 andsaExp2 may be sufficient to stage further haustorial development. Micromanipulative addition of expansin induces out of sequence development of leaf primordia in tomato (Lycopersicon esculentum; Fleming et al., 1997); however, these structures rarely developed past the initial stages, and other factors appear to be involved. Auxin transport into the meristem regulates expansin expression and induces complete leaf development (Reinhardt et al., 2000), suggesting that hormonal signaling regulates both initiation and radial position of organogenesis in the shoot meristem. With this background, the haustorial expansin genes are expected to be necessary for haustorial development and provide valuable molecular markers for the commitment, but not to be sufficient for complete haustoriogenesis.

ARE PLANT HORMONES NECESSARY FOR HAUSTORIAL DEVELOPMENT?

The answer to this question is not clear, but quite relevant. Exogenous cytokinins will induce haustorial development inS. asiatica, whereas auxins strongly inhibit both cytokinin and benzoquinone induction (Keyes et al., 2000). Haustoria originate as a recommitment of the root meristem pattern inS. asiatica, in this case a meristem poised for rapid transition to the parasitic mode, so it is not surprising that plant hormones effect development. These hormones function in accord with their site of synthesis, auxin in shoots and cytokinins in roots (Davies, 1995), at the very least maintaining the poles of the plant established in the embryo. However, exogenous addition of the cytokinins does not establish physiological relevance to haustoriogenesis, nor should it imply that the benzoquinones and cytokinins necessarily share common or even overlapping induction pathways.

Attachment organs produced by cytokinin induction are morphologically abnormal. This abnormality could arise from exposure to nonnatural cytokinin structures, or from differences in temporal or spatial application. For example, pleiotropic effects of the hormone on spatially isolated cell types within the root meristem could be expected to induce abnormal swelling events. In contrast, no such abnormalities appear when the seedlings are incubated with the benzoquinones. The benzoquinone xenognosins may be perceived only at specific cell-cell contact surfaces and not across many cell types. Contacts and cross feeding between the host and parasite epidermal cells may establish local redox gradients, and only later, stage the radial swelling of the internal cells and hair formation along the epidermis through different mechanisms.

SO HOW COULD REDOX ACTIVATION MEDIATE SOCIAL CONTROLS?

All cells maintain a redox stationary state. The plant cell in particular carries nuclear, mitochondrial, and plastid genomes, and the intracellular communication between them depends on redox signaling mechanisms (Mackenzie and McIntosh, 1999). Recent evidence connects H₂O₂ in signaling cell swelling in both stomatal control (Pei et al., 2000) and root gravitropism (Joo et al., 2001). The mammalian circadian rhythm, another system where events are clocked over a period of time, appears to be entrained by the redox state of the NAD cofactors (Rutter et al., 2001). Therefore, numerous possible mechanisms exist at the contact surface where cross-feeding H₂O₂ and the benzoquinone xenognosins could regulate both recognition and induction processes. As additional genetic markers become available for different stages of the developmental commitment (Matvienko et al., 2001; Ouedraogo et al., 2001), it should become increasingly possible to define the signals and molecular responses that allow epidermal cells of a parasite to dance with its host.

The text on parasitic plants edited by Press and Graves (1995) points out very clearly that cellular development is only one aspect of plant physiology that has been impacted by an understanding of these organisms. Parasitic members constitute a significant percentage of all angiosperms and many elements of their emergence, evolution, and ecology are noteworthy. Organelle genome evolution and biochemical integration within the plant cell is altered dramatically when assimilate comes from another plant. Overall source-sink relationships are altered in both host and parasite with attachment. Organisms that drain water, minerals, and even defensive chemistries from another plant establish unique dependencies and advantages. Given the severe economic impact these organisms still have on cereal production in Africa, the Mediterranean Basin, and India, as well as timber production in North America, we should expect continued investigations of these unique members of the plant kingdom to have both a basic and practical impact throughout the 21st century.