By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Due to essential maintenance the subscribe/renew pages will be unavailable on Wednesday 26 October between 02:00- 08:00 BST/ 09:00 – 15:00 SGT/ 21:00- 03:00 EDT. Apologies for the inconvenience.
Author for correspondence: G. EjetaTel: +1 765 494 4320Fax: +1 765 496 2926Email: firstname.lastname@example.org
• Breeding for resistance to Striga in maize (Zea mays), with paucity of donor source and known mechanisms of resistance, has been challenging.
• Here, post-attachment development of S. hermonthica was monitored on two maize inbreds selected for field resistance and susceptibility reactions to Striga at the International Institute of Tropical Agriculture. Haustorial invasion of the parasite into roots of these inbreds was examined histologically.
• Morphological differences were observed between roots of the susceptible and the resistant inbreds. The resistant maize had fewer Striga attachments, delayed parasitic development and higher mortality of attached parasites compared with the susceptible inbred. Striga on the susceptible inbred usually penetrated the xylem and showed substantial internal haustorial development. Haustorial ingress on the resistant inbred was often stopped at the endodermis. Parasites able to reach resistant host xylem vessels showed diminished haustorial development relative to those invading susceptible roots.
• These results suggest that the resistant inbred expresses a developmental barrier and incompatible response against Striga parasitism.
The parasitic weed Striga (Striga spp.) threatens cereal grain production in tropical and subtropical regions of Africa and Asia. Striga infests 40% of the cereal-producing areas of sub-Saharan Africa (Lagoke et al., 1991). In West Africa, Striga is believed to infest over 50 million ha (Lagoke et al., 1991), and the weed continues to expand its range. Yield losses as a result of Striga infestation in Africa depend on the crop cultivar, weather, and degree of infestation. Mboob et al. (1986) estimated an annual yield loss of US$7 billion in the Savannah regions of Africa alone. Infestation is often more serious under low soil fertility and increased moisture stress conditions. Plasticity in adaptation of Striga to diverse environments and the complexity of its host–parasite relationship have made its control more difficult. The slow pace of development and deployment of Striga-resistant cultivars is mostly attributable to paucity of sources of resistance, the complex genetics of resistance, and scant knowledge about specific mechanisms associated with expression of resistance in maize to the parasite. Sources of Striga resistance in maize have been scarce, perhaps because early evolution and adaptation of the maize crop took place in the absence of the parasite. Striga seeds have very specific requirements for after-ripening, conditioning (warm stratification that enables Striga seeds to become responsive to germination stimulants), and stimulation by chemical compounds exuded by host and nonhost plants before they can germinate to successfully parasitize the host plant. Subsequent haustorial development, attachment, and penetration, as well as further growth and development of the parasite also require signals or resource commitment from the host plant (Joel et al., 2007).
Greater understanding of the specific mechanisms associated with resistance would facilitate the development of improved selection methods and strategies to improve maize for Striga resistance. Field testing of crop resistance to Striga provides little information concerning the basis of resistance in the host plant (Joel et al., 2007). Characterization of defenses against the parasite and identification of developmental stages of the parasite that are vulnerable to such defense are important (Hood et al., 1998). Information of this kind has been gained from histological comparison of the interaction between Striga haustoria in resistant and susceptible plant roots (Hood et al., 1998; Gurney et al., 2003, 2006). All parasitic angiosperms infect host plants by first forming a haustorium (Visser & Dörr, 1987), an organ that structurally and physiologically joins host and parasite (Kuijt & Toth 1976). The haustorium forms a vascular bridge that provides the parasite with direct access to host-plant nutrients (Saunders, 1933; Okonkwo & Nwoke, 1978; Ba, 1987; Riopel & Baird, 1987).
Striga hermonthica (Del.) Benth. seeds used for this study were collected in Abuja, Nigeria, in 2003 and shipped to the Purdue University Parasitic Plant Containment Facility. The resistant (Z. Diplo.BC4-19-4-1-#-3-1-B-1-B-B) and susceptible (5057) maize inbreds used for this study were selected based on their contrasting reactions to Striga from 25 inbred lines screened under artificial field infestation and in the glasshouse through regional trials in and outside west and central Africa by the International Institute of Tropical Agriculture's (IITA) maize improvement program. Detailed descriptions of the genetic background and development of the resistant maize inbreds have been published (Kling et al., 2000; Menkir, 2006; Menkir et al., 2006). Z. Diplo.BC4-19-4-1-#-3-1-B-1-B-B (referred to by Menkir (2006) and hereafter denoted as ZD05) is a progeny of backcrosses between IITA-developed tropical maize germplasm and Z. diploperennis, a wild progenitor of Zea mays L. It was registered as a Striga-resistant line, TZSTRI108 (Menkir et al., 2006).
Striga seed conditioning
Striga seeds were sterilized by immersion in 25 ml of 75% (v/v) ethanol for 2 min, rinsed three times with sterile double-distilled (dd) water, then washed in a sonicator for 2 min with 25 ml Metricide 28 (Metrex, Inc., active ingredient: glutaraldehyde, 2.5%), a disinfectant, followed by three sterile water rinses for 2 min in a laminar flow hood. After being rinsed three times in sterile dd water, the Striga seeds were soaked overnight in a 0.015% (w/v) solution of benomyl, and then transferred to glass fiber paper moistened with the benomyl solution in a Petri dish. The dishes were sealed with paraffin tape and incubated for 11 d in the dark at 29°C. Preconditioned (warm stratified) Striga seeds were treated overnight with 1.2 × 10−6m solution of artificial germination stimulant GR-24; this ensures that maize roots are infested with uniformly germinated preconditioned Striga seeds.
Maize seed germination
Maize seeds were sterilized with 25 ml 1.05% sodium hypochlorite solution for 30 min and rinsed with distilled water several times, after which they were soaked overnight in 5% (w/v) Captan slurry (active ingredient: N-[trichloromethyl] thio-4-cyclohexene-1,1-dicarboimiide, 39%), a nonsystemic fungicide. Sterilized maize kernels were pre-germinated on filter paper kept wet in sterilized Petri dishes at 30°C for 48 h.
Determination of growth and infection
A bioassay called Sand Packed Titer Plate Assay (SPTPA) described by I. O. Amusan (manuscript under review) was used to study the reaction and post-attachment events of pre-germinated Striga on maize roots. In brief, SPTPA is an in vitro system for maize or sorghum and Striga coculture. SPTPA allows nondestructive sampling and progressive observation of resistance reactions during the post-attachment phase of parasitic development, and enabled us to identify specific mechanisms of resistance deployed by crop cultivars. Eleven days after transplanting maize plants into SPTPA, its roots were infested with 2 mg (approx. 200 seeds) of pre-germinated Striga seed by carefully aligning the weed seeds along the host root with a camel hair brush. Three replicates of SPTPA were established for each treatment in a completely randomized design.
Stages of development of attached Striga on host roots were recorded at 3, 9, 15 and 22 d after infection (dai) using an HP 4600 see-through scanner. Developmental stages of the parasite on host root were defined as follows: S1, attached Striga with seed coat intact; S2, emergence of first leaf primordia; S3, Striga shoots having two to four scale leaf pairs; S4, Striga shoots having five or more scale leaf pairs; D, attached Striga dead on host root.
Striga hermonthica development and penetration within the host root were monitored by dissecting representative haustoria of the parasite attached to the host roots at 3, 9, 15 and 22 dai. Immediately after cutting, the Striga and section of the maize root to which it was attached were put in a fixative solution consisting of 2.5% (w/v) paraformaldehyde and 2% (v/v) glutaraldehyde in 0.1 m cacodylate buffer at pH 6.8. In this primary fixative solution, excess maize root was trimmed to within 0.5 mm of the attachment point. After 2 h in the primary fixative, each sample was washed three times with 0.1 m cacodylate buffer then placed in a secondary fixative of 2% (w/v) OsO4 in 0.1 m cacodylate buffer for 1 h followed by three washes of 0.1 m cacodylate buffer. Tissue dehydration over several days was achieved by soaking in increasing concentrations of ethanol (10, 30, 50, 70, 90 and 100%) followed by two soaks in propylene oxide. Infiltration was in increasing concentrations of SPURR (10, 25, 45, 60, 75, and 90 to 100%). Tissues were embedded in flatbed molds after orienting the Striga attachments so that sectioning would result in cross-sections of maize root and longitudinal sections of endophytic Striga tissue. Polymerization was achieved by baking at 60°C for 48 h. Two micron sections were cut on a microtome, attached to adhesive-coated microscopic slides and stained with toluidine blue. The micrographs shown are sections through the center of the Striga–maize contact captured digitally at ×10 or ×20 magnification.
Marked differences in the growth and development of S. hermonthica were observed on the roots of two inbred lines, 5057 (susceptible) and ZD05 (resistant), of maize (Fig. 1). At 3 dai, all the Striga seedlings that attached to the roots of the susceptible inbred still had an intact seed coat. At 9 dai, 10, 38, 27 and 25% of attached S. hermonthica on the roots of 5057 attained S1, S2, S3 and S4 stages of development, respectively. By 15 dai, 85% of the parasites that attached to the root of 5057 had reached S4. By 22 dai, 96% of S. hermonthica that attached to the root of the susceptible maize attained S4, and < 5% of attached parasites were dead. At 3 dai, S. hermonthica had reached the same developmental stage on the resistant maize as on 5057. In contrast to development of S. hermonthica on 5057 at 9 dai (which had 10% of attached parasite on its root at S1), 65% of S. hermonthica that attached to the root of the resistant maize remained at S1, while 12 and 24% reached S2 and S3, respectively. By 15 dai, 50% of attached parasites were dead on the roots of ZD05, while 19, 6 and 25% of attached Striga seedlings attained S1, S3 and S4, respectively. At 22 dai, 88% of attached parasites on the resistant inbred ZD05 were dead, and only 12% reached S4.
On maize line 5057, the parasite had rapid shoot development and formed secondary haustoria which connected to other sites on the maize root system (Fig. 2a). Although 12% of attached Striga seedlings on ZD05 were at S4 by 22 dai, their shoots were stunted with poorly emerged scale leaf pairs and no secondary parasitic roots (Fig. 2b). The two maize lines differed in root morphology. The resistant inbred had fewer, thin branched roots in the upper profile compared with the susceptible maize (Fig. 2a,b). These differences in root characteristics were apparent before infection. Since we did not compare roots of uninfected maize plants at ages corresponding to 3–22 dai, we cannot say whether Striga infection affected these differences between the inbreds.
Histological evidence (Figs 3–5) revealed critical differences in the invasion of host roots of 5057 and ZD05 by S. hermonthica haustoria. On inbred 5057 (Fig. 3) by 3 dai, the attached parasite (Fig. 3a) had its seed coat intact with numerous primary root hairs and a viscid substance at the point of contact with the host root. Sectioning of this attached parasite (Fig. 3b) showed that the haustorium had successfully penetrated the entire host root cortex. The apical tissue of the endophyte (part of parasite inside the host root tissue) was in contact with the endodermis of the host plant. The haustorial tissue consisted mainly of parenchyma-like cells. By 9 dai, the attached parasite had developed three scale leaf pairs (S3) and a distinct swelling at the attachment site (Fig. 3c). A cross-section through this swelling and the host root (Fig. 3d) revealed complete parasitic invasion of the host stele, and highly differentiated haustorial tissues. The haustorium consisted of three well defined tissues, including the vascular core (VC) encircled by hyaline body (HB) and the endophyte (E). The endophyte penetrated the lumen of the host xylem, formed parasite–host xylem-to-xylem connection, and its lateral outgrowth invaded the stele. The parasite vascular core consists of reticulated xylem elements (xylem treachery). These tracheids were characterized by spiral thickenings that transmit from the endophyte to the vascular core. The hyaline body was densely stained and had abundant organelles.
By 15 dai, S. hermonthica had five scale leaf pairs (S4). The internal structures and tissue development of the haustorium were similar to that observed at 9 dai; however, the haustoria were much bigger with a well differentiated vascular core (see Supplementary material, Fig. S1). At 22 dai (Fig. 3e), the parasite had 10 scale leaf pairs and had formed secondary haustoria (indicated with white arrowheads in Fig. 2a) which invaded the host root. The secondary haustorium is formed by lateral or terminal contact of secondary parasite roots with host roots (Okonkwo, 1966a). Longitudinal section of the primary haustoria (Fig. 3f) exhibited the same structural components seen in the micrograph of the haustoria growing on the susceptible maize at 9 dai. The entire haustorium was much bigger by 22 dai. In addition, histological evidence revealed that the endophyte of the secondary haustoria penetrated the host root endodermis and made contact with the xylem cell wall (Fig. S2). However, secondary haustoria tissue differentiation had not begun at this period because the parasite had not invaded the lumen of the host's xylem.
Shoot and haustorial development of S. hermonthica on the root of ZD05 are presented in Fig. 4. The interaction between S. hermonthica and ZD05 at 3 dai (Fig. 4a,b) had features similar to those observed on the parasite attached to susceptible maize 5057 at the same 3 dai. However, several distinctive responses to the invasion of the resistant host root by Striga haustoria were observed at subsequent sampling days. At 9 dai (Fig. 4c), the attached parasite had only one scale leaf pair (S2). Although the parasite bridged the lumen of the host xylem, there was no evidence of tissue differentiation in the haustoria (Fig. 4d), in contrast to haustoria on susceptible maize at the same age (Fig. 3d).
In rare cases, by 22 dai the parasite on the root of resistant maize produced four pairs of poorly emerged scale leaves on stunted and purpling shoot tissue (Fig. 4e). Longitudinal sections of these parasites (Fig. 4f) revealed features similar to those earlier described on Striga-resistant sorghum (Maiti et al., 1984). Eventually, haustoria penetrated through the host endodermis, but on reaching the host xylem, apparent occlusions were formed in the cavity of the xylem vessels (Fig. 4f). There were a few lines of tracheids in the haustoria, which was probably insufficient to allow proper development of the parasite. There was no clearly defined hyaline body in the haustorial tissue.
Other manifestations of resistance to Striga in ZD05 are presented in Fig. S3 and Fig. 5. In some instances, at 9 dai, Striga seedlings attached to ZD05 remained with their seed coat unbroken (Fig. S3a). The haustorium barely penetrated the root cortex of the resistant maize (Fig. S3b). In other instances (Fig. S3c,d), haustorial ingression of the host root stopped at the endodermis. At 15 dai, most parasite seedlings on resistant maize had no shoot development, S1 (Fig. 5a). The haustorium penetrated the host cortex, but could not penetrate the endodermis. A deposition of an unknown substance (X) was observed at the parasite's haustorium and host endodermis interface (Fig. 5b). At 22 dai, two additional responses to invasion of resistant maize roots by haustoria were observed. In the first, the visible Striga tissue became dark brown (Fig. 5c). Sections through the host–parasite connection (Fig. 5d) revealed that the Striga haustorium penetrated the host endodermal tissue, but did not reach the lumen of host's xylem elements. In the second response, the parasite had poorly emerged scale leaves with stunted shoot growth (Fig. 5e). Examination of these sections (Fig. 5f) revealed no host–parasite xylem-to-xylem bridge. The haustoria were inside the stele, in contact with the outside wall of the host's xylem vessels, but with no clear haustorial incursion into the lumen of the host's xylem. This is suggestive of haustorial collapse, or a severed host–parasite xylem-to-xylem connection that must have been formed earlier. In addition, haustoria lacked a definite hyaline body and vascular core.
Resistance to parasitic weeds can be expressed before or after host–parasite vascular bridge formation (Rispail et al., 2007). In addition to a reduced number of attachments, we found multiple post-attachment barriers to Striga parasitism in ZD05, including physiological/biochemical incompatibility to parasite growth and development after formation of vascular continuity between Striga haustoria and the resistant maize root, and/or a mechanical barrier to vascular connection. The genetics behind these barriers are not yet determined. Any one of these resistance mechanisms could potentially be controlled by multiple genes, and be durable from a genetic standpoint. An advantage of stacking different mechanisms of resistance into a single genotype is to enhance stability of resistance and durability of genes. Resistance to Striga can be greatly affected by factors such as drought, soil type and fertility levels. Striga populations may also differ in their virulence; it is possible that having multiple resistance mechanisms would provide some insurance under varying parasite populations. If these different resistance mechanisms are associated with different genes, ZD05 could provide a durable resistance to Striga parasitism. Similar multilevel disruptions to Striga parasitism have been reported for sorghum (Arnaud et al., 1999).
Fewer Striga attached to the resistant inbred roots relative to those of the susceptible, even when equal numbers of germinated parasite seed were placed directly on each root system. There is delayed parasitic post-attachment development, and higher mortality of attached parasites on the resistant maize inbred (ZD05) compared with the susceptible inbred (5057). The observations made in this laboratory study reflect, and to some degree explain, those made in the field. Fewer shoots emerge in field plots of ZD05 than in plots of 5057 (Menkir, 2006), though occasionally Striga do emerge and may set seed on the resistant host. Although the underlying mechanism for reduced attachment frequency needs further exploration, the post-attachment resistance expressed in ZD05 is a powerful mechanism and unlike any hitherto reported in maize.
Successful parasitism on susceptible maize was evident by rapid parasite shoot development as well as the formation of secondary haustoria that connected to other sites on the maize root system. Primary haustoria are formed by contact between a host root and the radicle of germinated parasitic seed (Okonkwo & Nwoke, 1978). Secondary haustoria are formed as a result of contact between host root and parasitic secondary roots (Kuijt, 1977). Nwoke & Okonkwo (1978) reported that roots of all root parasites formed haustoria along points of contact with host root, with the exception of a few members of the Orobanchaceae family. Higher root volume on the susceptible maize probably enhances formation of secondary haustoria. These secondary haustoria constitute additional water and mineral nutrient absorptive organs for the parasite (Nwoke & Okonkwo, 1978; Gurney et al., 2003). Striga development on the resistant ZD05 was retarded and often led to an early death of parasites attaching to its roots, as opposed to rapid and pervasive shoot development of parasites on the roots of susceptible maize.
Quantification of root length density and other root characteristics is not the focus of this study; however, it is worth mentioning that apparent differences in root architecture between the two maize inbred lines may account for an additional mechanism of avoidance in the resistant inbred line. Cherif-Ari et al. (1990) showed that low root length density might be one mechanism adopted by certain sorghum varieties to avoid Striga parasitism, resulting from fewer host root/parasitic seed interactions. Striga reportedly stimulates root growth of infected maize (Kim et al., 1999). Although differences in root morphology were apparent before infection, it is unclear whether increased Striga infection in 5057 exaggerated these differences, because root measurements were not made or compared with uninfected controls. Root morphology might enable the resistant inbred to avoid Striga parasitism in the field, but would not have affected results from this experiment because equal numbers of parasites were applied directly to the roots of both genotypes.
Penetration of host root cortex by haustoria has been suggested to be largely a mechanical process aided by the voluminous air space in the host root cortex (Joel & Losner-Goshen, 1994; Dörr, 1997). Others have hypothesized enzymatic degradation of host tissue by the endophyte (Saunders, 1933; Kuijt, 1977; Okonkwo & Nwoke, 1978). Although Striga haustoria successfully penetrated the root cortex of both resistant and susceptible maize inbreds in our study, the root cortex of the resistant inbred often offered a strong barrier against penetration by haustoria. Differentiation of haustoria is an essential and necessary factor for growth and development of invading Striga (Gurney et al., 2003). Hence, a continuous connection between the vascular tissue of the host and parasite is necessary to ensure constant supply of mineral solutes, water and photosynthate from the host to Striga.
It has been generally reported that host–parasite xylem-to-xylem connections are made in Striga (Dörr, 1997), while haustorial invasion of a host's sieve tube or phloem tissue have not been widely reported. With the overwhelming evidence provided in the literature for dependency of Striga on the host plant for carbohydrate (Press, 1995), it is likely that xylem feeding may not be the only source of carbon for successful parasitism of Striga on some of its hosts. The few storage reserves in the extremely small seed of Striga are depleted within a few days after germination (Chang & Lynn, 1987). A Striga seedling is holoparasitic in its subterranean stage, unable to assimilate carbon, entirely dependent on the host for carbohydrate (Graves, 1995). Okonkwo (1966b) reported that both subterranean and emerged Striga seedlings obtained water, minerals, and photosynthate from the host. Graves et al. (1989, 1990) estimated host-derived carbon to be between 35 to 76% in S. hermonthica parasitic on sorghum and millet even after emergence. And 85% of the carbon transferred from a range of hosts was used in respiration by S. hermonthica (Press & Graves, 1989). Graves et al. (1989) emphasized that competition between sorghum and S. hermonthica for organic solutes is important and may account for 20% reduction in host biomass. Anatomical observations in our study revealed parasitic invasion of other tissue in the host stele besides the xylem vessels of susceptible maize. This is evident by the observed spread of Striga tissue (lateral outgrowth) within the stele of susceptible maize. This observation corroborates the findings of Saunders (1933), who reported formation of lateral outgrowths of Striga haustoria into the cortex of sorghum root for absorption of organic solutes from the host. So although there may be no direct phloem-to-phloem contact, Striga likely taps the resources of its host's phloem through these intrusions.
Nonhost resistance to Striga could result from lack of nutritional components necessary for sustaining development of attached Striga, or constitutive or induced general resistance that prevents parasitism (Hood et al., 1998). Poor haustorial differentiation and development on ZD05, in contrast to those parasitizing 5057 in spite of host–parasite xylem-to-xylem connections, suggests a fundamental physiological or biochemical incompatibility between parasite and the resistant host. Gurney et al. (2003) reported similar incompatibility between S. hermonthica and a wild relative of maize, Tripsacum dactyloides. An induced biochemical defense reaction may lead to inhibition of nutrient uptake by the parasite or a physiological process that increases water and nutrient retention in the host root tissue (Arnaud et al., 1999). These physiological/biochemical defense reactions might include the formation of occlusions in the cavity of the host xylem in response to haustorial penetration. It is unlikely that the endophyte is capable of acquiring nutrients or other compounds necessary for growth from the host cortex. As indicated by multiple emerged scale leaf pairs on the Striga attached to ZD05 (Fig. 5e), an earlier invasion of host xylem cavity by the parasite must have occurred. The poor expansion of those scale leaves, however, may indicate an induced physiological/biochemical defense or enzymatic response from ZD05, in which the parasite–host vascular connection was severed.
The endodermis is generally considered a substantial barrier to vascular penetration by root parasitic weeds. This has been reported to be the site of resistance expression in varieties of several crop species, including rice to S. hermonthica (Gurney et al., 2006), sorghum to S. asiatica (Maiti et al., 1984) and sunflower (Helianthus annuus) and vetch (Vicia sativa) to Orobanche cumana (Dörr et al., 1994; Labrousse et al., 2001). In our study, we have not made an observation of degree of lignification of endodermal cell wall in ZD05. It is possible that the endodermis possesses tough mechanical tissue, sclerenchyma, serving as a barrier to haustorial penetration. Some researchers (Wegmann et al., 1991; Westwood et al., 1998; Hood et al., 1998; Goldwasser et al., 1999) have reported the role of a chemical barrier to parasitism, a secretion at the host–parasite interface. Host-induced compounds, such as phenols, phytoalexins and pathogenesis-related proteins, are among compounds produced by resistant plants that discourage parasitism. Histological evidence in our findings revealed accumulation or deposition of an unidentified substance at the haustoria–host interface in ZD05. Further analysis of the chemical composition of this substance may lead to isolation of a specific compound controlling this resistance to Striga parasitism.
In our laboratory, coculture method ZD05 displays a multilevel resistance to Striga in that fewer parasites attach to its roots and those that do are less likely to develop and have a higher mortality than those on susceptible maize. This corroborates field trials in which this inbred has fewer emerged Striga per plot than other entries (Menkir, 2006). Further genetic studies are needed to determine the mode of inheritance as well as loci involved in the expression of this trait. The superior resistance of ZD05 was, however, stable over years and locations in which the field trials were conducted (Menkir et al., 2006) as well as under artificial infestation using a mixture of Striga seeds collected from two locations in Nigeria (Menkir, 2006). Because the resistance occurs post-germination, maize expressing this resistance could help deplete the amount of Striga seeds in the soil. ZD05 provides a good source germplasm for breeding maize cultivars with broad resistance to Striga in West Africa where it has been extensively tested (Menkir et al., 2006).
The authors would like to express their gratitude to Dr Debra Sherman and Chia-Ping Huang of the Purdue University Life Science Microscopy Facility for their technical assistance in the tissue preparations and microscopy. This work has been conducted with financial support from the United States Agency for International Development (USAID) through a linkage grant between the International Institute for Tropical Agriculture (IITA) and Purdue University.