Overview of its biology and genetics

Despite the fact that O. cumana is a major threat to sunflower cultivation from the Mediterranean region to Central Asia, methods for a sustainable control of this pest are scant and most often ineffective. Thus, understanding the biology of the interaction between the parasite and its host is a necessary step towards the development of selective control methods. This aim should include notably deciphering the physiological and molecular mechanisms governing the parasite development and the establishment of the interaction with its host. Unfortunately, compared with other plant pathogens such as bacteria and fungi, our knowledge of broomrape biology is limited and comes from studies mainly made on Phelipanche ramosa, Phelipanche aegyptiaca, and Orobanche minor, three other noxious parasitic weeds. Thus, throughout this section the generic name “orobanche” will be used when referring to broomrape species and most of the biological events mentioned below could be applied to O. cumana.

Orobanche cumana belongs to the Orobanche section (synonym of the Osproleon section) from the Orobancheae class of the Orobanchaceae family. Orobanche cumana is a close relative of O. cernua, two different Orobanche species that have long been considered as a single one (Román et al., 2003). In terms of genetics and genomics, the diploid O. cumana species, exhibits a chromosome number (2n) of 38 and an estimated genome size of 1.42 Gb (Piednoël et al., 2012). To date, neither the reference transcriptome nor the complete genome sequence are available for this species. Only Expressed Sequence Tags from Phelipanche aegyptiaca are publicly available in the Parasitic Plant Genome Project website (PPGP, 2015).

Like any other holoparasite species, broomrapes are achlorophyllous root parasitic plants relying on a host for their nutrition and survival. Orobanche seeds are very small, 200-300 µm, and a large amount of seeds is produced annually, since only one plant can produce several tens of thousands of seeds. These seeds are easily disseminated by wind, water and other agents; for instance, contaminated crop seed lots of sunflower achenes (Castejon et al., 1991). Seed maturation and desiccation are phases of varying durations depending on the species, and environmental factors such as temperature. After release from the flower’s capsule, seeds are buried in the soil, where they remain viable in dormancy for decades (Prider et al., 2013). Among the remarkable morphological and physiological adaptations that characterize the orobanche species, the requirement for the seeds to perceive molecules produced by host roots to germinate is probably the most intriguing one. Indeed, the germination of broomrapes is a two-step process corresponding, first, to a conditioning period, thought to be required for the acquisition of their sensitivity to the germination stimulants (GS), followed by the chemical stimulation of the germination itself that ends with the radicle protrusion (Lechat et al., 2012). During the conditioning period, broomrape seeds are exposed to a moist environment and suitable temperatures ranging from 15 to 25°C depending on the species (Linke et al., 1989). If conditions are not favourable for the germination or if seeds are not stimulated by host roots, seeds enter into secondary dormancy (Matusova et al., 2004). In P. ramosa and O. cumana, it has been demonstrated that seeds need a minimum 4 days conditioning period at 21°C, during which a global DNA demethylation occurs, which is an epigenetic process required for the seed to become responsive to GS (Lechat et al., 2015). Once conditioned, broomrape seeds are still unable to germinate without stimulation by GS, chemicals compounds produced and exuded in the rhizosphere by surrounding host roots; this mechanism implies that the parasite is in the vicinity of the roots (3-5 mm) on which it will attach itself. The efficient GS for the germination of O. cumana seeds, which are present in sunflower root exudates, are the sesquiterpene lactones (Joel et al., 2011; Raupp & Spring, 2013), heliolactone (Ueno et al., 2014), and strigolactones (Yoneyama et al., 2011). Whatever the GS, it has been demonstrated that O. cumana seeds respond by an up-regulation of CYP707A1, an abscisic acid catabolic gene, that then triggers the germination process (Delavault et al., 2013).

During germination, the seed produces a radicle also called procaulome or germ tube of a few mm long that grows towards the host root probably thanks to a positive chemotropism. Seeds of P. ramosa and O. cumana exposed to root exudates of tobacco and sunflower, respectively, released indole-3-acetic acid even before development of the germ tube suggesting that auxin may be involved in the germination process (Slavov et al., 2004). When the procaulome of O. cumana reaches the host root, its apical cells differentiate into papillae producing a carbohydrate that allows parasite adhesion to the root (Joel & Losner-Goshen, 1994). This step of the broomrape life cycle constitutes the transition to parasitic phase. During parasite attachment, the germ tube produces a swollen structure at the apex, the appressorium also called pre-haustorium, which produces intrusive cells, some of which establish connections with the host vessel system. Penetration of the parasite into host root tissues toward sap-conducting vessels involves mechanical and enzymatic processes (Joel & Losner-Goshen, 1993). Intrusive cells progress between host cells that are pushed away under mechanical pressure. To facilitate its progression between host cells, the parasite also degrades the middle lamella of their cell wall. For instance, pectin degrading enzymes such as pectin methyl esterase (PME), polygalacturonases (PG), and rhamnogalacturonases are secreted by germinating seeds of P. aegyptiaca and P. ramosa (Ben-Hod et al., 1993; Véronesi et al., 2007). Using immunocytochemical methods, Losner-Goshen et al. (1998) detected PME in the cytoplasm and cell walls of O. cumana intrusive cells localized at the sunflower-parasite interface. The virulence of the parasite seems to be correlated with the production of these enzymes. Indeed, the intensity of these pectinolytic activities, PME and PG secreted by different races of O. cumana, was proportional to the virulence of each race (Véronési et al., 2005).

Once the appressorium penetrates the endodermis, parasite cells progress towards host vessels and then initiate a specialized endophytic organ called haustorium that proceeds to the establishment of vascular connections (Hibberd & Jeschke, 2001). The haustorium is also the transfer structure that enables the parasite to uptake water and nutrients (carbohydrates and nitrogen compounds) from host conductive system. Whereas host factors for the haustorium development (haustorium-inducing factors or HIF) have been characterized in several Orobanchaceae species, those factors have not yet been evidenced in Phelipanche and Orobanche species (Yang et al., 2015). Histological studies at the O. cumana-sunflower interface confirmed the existence of a direct connection between host and parasite xylem elements thanks to parasite cells differentiating into tracheid-like vessels (Labrousse, 2002). Moreover, some parasite cells were observed to be in close contact with phloem cells in host roots, suggesting that Orobanche are “phloem feeders”. In fact, when the phloem-mobile symplasmic tracer carboxyfluorescein is applied on source host-leaves, the molecule is translocated in a few minutes into the parasite demonstrating unequivocally the existence of a symplasmic continuum between host and parasitic phloem (Péron, 2010). Penetration of the intrusive cells into the xylem triggers their division leading to the formation of a tubercle, a callus-shaped bulge of parasite cells outside the root. Once vascular connections are established and the haustorium is functional, the parasite then acts as a supernumerary sink organ for the host plant by expressing genes involved in sucrose metabolism and the establishment of a strong sink strength (Draie et al., 2011; Péron et al., 2012). The attached parasite diverts hormones and photoassimilates from the host plant, mainly auxin and sucrose (Bar-Nun et al., 2008; Abbes et al., 2009). The tubercle, with a diameter of 0.5 to 2.5 cm, serves as a transient storage organ and develops numerous adventitious roots; the apical bud of this tubercle gives rise to a subterranean shoot and then to a flowering spike, after emergence from the soil, that rapidly produces a large amount of tiny seeds.

Thus, much remains to be studied of the biology of these intriguing interactions, especially that between O. cumana and its host, to identify vulnerabilities that would be specific targets for new control methods.

Race structure of O. cumana

The first mention to different races within the O. cumana species dates back to the beginning of the 20^th century. Efforts made by Soviet breeders at the Saratov experimental station resulted, by 1920, in the release of genetically resistant varieties of sunflower. Since these varieties effectively controlled the races of O. cumana from the Saratov and Voronezh regions but not those affecting sunflowers in the Rostov and Krasnodar regions, two races, namely A and B, were differentiated in Russia from then on. Both races had a different geographical distribution: race A was common in Saratov and Voronezh regions and race B was found in Rostov and Krasnodar regions (Antonova, 2014). Also in Moldova and Ukraine (former USSR) race B of O. cumana was identified in the 1930’s (Duca, 2014; Miladinovic et al., 2014). The control provided by sunflower hybrids with resistance to race B was overcome by race C of O. cumana in Moldova in the 1970’s (Duca, 2014). Thereafter, breeding works resulted in new varieties of sunflower with resistance to this race allowing an efficient control of the parasite in the former USSR until the 1990’s. Throughout the past decades a new failure in the genetic control of O. cumana has occurred in the Russian Federation as a consequence of crop intensification and short crop rotations, together with the use of genetic material from foreign breeding programmes which was therefore susceptible to the local races of the parasite (Antonova, 2014). Recently, new and highly virulent races of O. cumana from the southern regions of the Russian Federation have been identified as races E, F, G and H (Antonova et al., 2013). The presence of races E and F of the parasite has also been recently reported in the Republic of Moldova (Gisca et al., 2013; Duca, 2014), as well as races C and G in Khazakhstan (Antonova, 2014). In Ukraine, the increase in the area cropped to sunflower throughout the last decade, together with short rotations, has been associated with the frequent occurrence of race E followed by races F and G in the south and southeast of the country (Pototskyi, 2014).

Orobanche cumana was reported to be affecting sunflower in Romania in early 1940’s (Pacureanu, 2014). The first attempt to determine races of the parasite in that country was that of Vrânceanu et al. (1980), who used a set of sunflower genotypes, each of them carrying one single major gene of resistance. These genes (Or₁ to Or₅) were reported to control races A to E respectively in Romania, each of them also conferring resistance to the previously described races (what is termed as cumulative resistance) (Vrânceanu et al., 1980). As occurred in other countries, races A to E in Romania were effectively controlled through genetic resistance for some decades, until race F was identified in the mid-nineties (Pacureanu-Joita et al., 2008).

The first report of O. cumana in Serbia dates back to the 1950’s; the populations detected were probably race B, because they were controlled by Russian oil varieties and hybrids with resistance to this race (Miladinovic et al., 2014). Economically important incidences of broomrape on sunflower were again observed in that country in the 1990’s and identified as a race E (Mihaljčević, 1996). During the last 20 years, and even though the parasite being spread to new sunflower production areas in the country, no new races have occurred so far (Miladinovic et al., 2014). Infections of sunflowers by O. cumana were reported in Bulgaria in the first half of the 20^th century, when the parasite populations were characterized as races A and B. Some decades later, races C, D and E, were identified and thereafter effectively controlled by means of genetic resistance. Race E of O. cumana remained predominant in Bulgaria in the first years of the 21^st century until the sunflower oil production of the country was threatened again by race F (Shindrova, 2006). Nowadays, races E, F and possibly race G are widely spread in the sunflower growing areas of Bulgaria (Batchvarova, 2014).

With respect to other sunflower growing countries, in Hungary, populations of O. cumana were characterised as being low to moderately virulent (races A–D) (Zoltán, 2001). Currently, race E is the most frequent one in the country (Hargitay, 2014) although race F has also been identified (Molinero-Ruiz et al., 2014). Turkey is another country where infections of sunflowers by O. cumana endanger the crop production in every growing season. Race F is widely distributed in Trakya region (Kaya et al., 2004), which accounts for more than half of the total area of sunflowers in the country. Also, race G has been identified in several locations of this same region (Molinero-Ruiz et al., 2014). In the past few years, infections of sunflower hybrids by O. cumana have also become frequent in the Anatolia region, but the race/s of the parasite have not been determined (Kaya et al., 2012). The presence of O. cumana parasitizing sunflower in France was not reported until 2007 (Jestin, 2012). It is currently located mainly in the south and also in the west of the country (Jestin et al., 2014), although no information about race/s is available. Orobanche cumana race A was identified in the northeast of China in the 1990’s (Baichun et al., 1996). A widespread increase in new virulent parasite races has occurred since then, so that the wide distribution of races A to F there, as well as the identification of race G in Inner Mongolia, have recently been reported (Ma & Jan, 2014; Shi et al., 2015).

Spain is the country where the greatest efforts have been devoted to characterizing races of O. cumana during the past 40 years. Race B of the parasite was prevalent in Spain (Saavedra Del Río et al., 1994) until race C was detected in the 1970’s (González-Torres et al., 1982), when an unexpectedly high infection of O. cumana occurred in the oilseed sunflower variety Peredovik, which was obtained by Soviet breeders as being resistant to race B. Interestingly, populations of O. cumana race B from Spain did not fit the cumulative resistance of the differentials proposed in Romania (Melero-Vara et al., 1989) as would occur years later, when the inbred line L86, registered as resistant to race F (Fernández-Martínez et al., 2004), was extremely susceptible to the less virulent race E (Molinero-Ruiz et al., 2006). Also interesting about races of O. cumana in Spain is that in the 1980’s two different races of the parasite were identified, one in the south and another in the centre of the country, and characterized as races D and E, respectively (Melero-Vara et al., 1989). Pathogenic and molecular differentiation of O. cumana populations from Spain as related to geographical origin were again recently reported (Molinero-Ruiz et al., 2006; 2014; Pineda-Martos et al., 2013). The last outbreak of genetic resistance in the crop in Spain occurred twenty years ago after the identification of race F (Saavedra del Río et al., 1994; Alonso et al., 1996). During the early years of 2000, O. cumana race F became widely distributed in southern Spain (Molinero-Ruiz et al., 2006; 2009) and is also present in some sites of northern and central Spain (Fernández-Escobar et al., 2008; Molinero-Ruiz & Dominguez, 2014).

Important efforts of breeders have been devoted to the search for effective resistance against the increasingly virulent parasite populations and, as a result, resistant genotypes have been obtained and released (see ‘Sunflower breeding and genetics of resistance to O. cumana’ section). In each country, sunflower material identified as being resistant to O. cumana has been used to differentiate local races of increasing virulence and termed from A to E, F or G. Because no comparative studies have been conducted to test the correspondence of races between countries, an unbundled knowledge of pathogenic traits of the parasite is currently occurring. In spite of races A to G already having been identified in many countries (Table 1), very few works assess the similarity of those populations from different geographic origins and characterized as belonging to the same race (Molinero-Ruiz et al., 2014).

Table 1. Races of Orobanche cumana identified –in the past and at present– in several countries where sunflowers are grown and parasite infections are known to occur

When determining races of crop pathogens, a long list of terms is frequently used to designate a wide range of races, making the nomenclature system extremely complex. When this happens, mathematical codes are, by far, more advantageous than the use of consecutive numbers or letters given in chronological order of race discovery. To ease communication and the comparisons of results of race characterization, a universal adoption of the coded triplets system (Limpert & Müller, 1994) is frequent for many plant pathogens (Liebenberg & Pretorius, 2011; Gurung et al., 2013; Dreiseitl, 2014) including Plasmopara halstedii Farl. Berl. & de Toni, which causes sunflower downy mildew (Molinero-Ruiz et al., 2002). Due to: a) the diversity of races of O. cumana identified worldwide, and b) their characterization according to the reaction of genes of resistance in sunflower genotypes which are particular to each crop region, the use of the coded triplets system as a simple and global method to internationally determine the races of the parasite seems imperative at this moment. In this review we propose that this system be internationally adopted from now on by sunflower breeders and pathologists. The essential component of the coded triplets system is a group of sunflower lines, termed differentials, which are divided into sets of up to three lines each. In Table 2 we propose eight lines of sunflower that can be used as differentials for races of O. cumana, notwithstanding the fact that additional ones, showing a consistent resistant reaction to the current most virulent races of the parasite, could be included in the future. The system numerically assigns a value if the differential is susceptible. If the first differential of a set of three is susceptible, it imparts a value of 1; if the second differential is susceptible it imparts a value of 2; and if the third differential is susceptible it imparts a value of 4. The system is additive within each set. When applied to O. cumana, the name of each race of the parasite would be a 3-digit code, one digit from each of the three sets of sunflower lines. As an example, if the first and second lines are susceptible, they impart a value of 1+2 (3) or, if the three lines are susceptible, they will impart a value of 1+2+4 (7). If the set only groups two lines, the largest possible value will be 3. This is illustrated in Table 3, where the codes for historical races A to E are given, as well as those for parasite races of a virulence higher than that of E.

Table 2. Proposal for standardized characterization and nomenclature of Orobanche cumana populations using the coded triplets system, which is based on the use of eight sunflower lines -termed differentials-, grouped into three sets

Table 3. Proposal for characterization of populations of Orobanche cumana using the coded triplets system, and its correspondence with the traditional method based on the use of consecutive capital letters (A, B, C, etc.) given in chronological order of identification of races

Among genotypes identified as being resistant to O. cumana, particularly useful as differentials for race characterization are public inbred lines. Because its genetic background is known, public material can easily be exchanged between research groups. In addition to the sunflower genotypes proposed by Vrânceanu et al. (1980) as differentials of races A to E of O. cumana, other inbred lines were identified by scientists as having a clearly resistant or susceptible discriminating reaction against parasite populations in different countries and/or growing areas within countries. This is the case of line LC1093, resistant to race F in Romania (Pacureanu-Joita et al., 1998), and line P96 (Fernández-Martínez et al., 2004), which was identified as being resistant to race F from Spain and susceptible to race G from Turkey (Molinero-Ruiz et al., 2014). According to the results of the works by Vrânceanu et al. (1980), Pacureanu-Joita et al. (1998), Fernández-Martínez et al. (2004), Molinero-Ruiz et al. (2008), Antonova et al. (2013) and Molinero-Ruiz et al. (2014), several lines are clear differentials of races of O. cumana (Table 2). The use of this single group of differentials worldwide will facilitate a comparison of the results of the characterization of races of O. cumana in different countries. Moreover, the most virulent parasite races and their location will be known by scientists working on the achievement of sunflower material with resistance to O. cumana in any of the countries where the parasite occurs.

Finally, it is important to note that a diverse genetic composition of O. cumana populations was evidenced in Russia and also in Spain (Molinero-Ruiz et al., 2008; Antonova et al., 2013). Genetic heterogeneity within race F populations seems to be a natural and inherent trait, because no components with a virulence higher than F (potential new races) (Molinero-Ruiz et al., 2009) or molecularly distinguishable groups (Molinero-Ruiz et al., 2014), have so far been identified within them. At this moment, genetic studies of the parasite are crucial to bring the knowledge on the parasite to the same level as what known about the genetics of resistance to O. cumana in sunflower. Studies on the inheritance of avirulence genes in O. cumana have confirmed the gene-for-gene interaction in the O. cumana-sunflower parasitic system for races E/F and the dominant sunflower gene Or₅ (Rodríguez-Ojeda et al., 2013), but the inheritance of avirulence genes against other sources of resistance to the most virulent races is still unknown.

Sunflower breeding and genetics of resistance to O. cumana

Mechanisms of resistance against O. cumana

One of the most important aspects when selecting for broomrape resistance is to obtain information on the physiological basis of the different resistance sources, with the ultimate goals of conducting physiology-based breeding and pyramiding resistance genes underlying different resistance mechanisms (Pérez-Vich et al., 2013). The mechanisms of sunflower resistance to O. cumana have been studied for a long time. Early studies already mentioned the importance of the root system’s pH for broomrape resistance in sunflower. Morozov (1947) cited the findings of Rihter (1924), according to which low pH values in the root of the host plant promote susceptibility to this parasite. On the contrary, basic soils (pH Cl₂Ca 8.14) have been related to a lower number of broomrape stems per sunflower plant as compared to acid soils (pH Cl₂Ca 6.17) in Spain (Lozano-Cabello, 1999).

Defensive responses such as physical or chemical barriers that prevent parasite intrusion and development can be activated at three different stages of the parasite-host interaction: pre-attachment of the seedlings, pre-haustorial stage, or post-haustorial stage (Pérez-de-Luque et al., 2009). Results obtained by Morozov (1947) established that sunflower root cells contain substances which activate germination of broomrape seeds and in that way support the development of a sprout (stimulators). Long after that, Alonso (1998), Wegmann (1998), Matusova et al. (2004) and Höniges et al. (2008) also pointed out the importance of broomrape germination stimulants (see ‘Overview of its biology and genetics’ section for more details). Low exudation of germination stimulants by sunflower roots has been described as a pre-attachment resistance mechanism to race E in the sunflower line LR1, derived from H. debilis subsp. debilis (Labrousse et al., 2001). Another pre-attachment resistance mechanism is the exudation by sunflower roots of seed germination inhibitors and/or inhibitors of radicle exoenzymes (Höniges et al., 2008). Phytoalexins (compounds of a phenolic nature), in particular 7-hydroxylated simple coumarins, have been suggested as playing a defensive role by preventing broomrape germination and subsequent connection with sunflower roots in the sunflower variety Cortés (Eurosemillas, Córdoba, Spain) resistant to Spanish O. cumana populations collected between 1994 and 1996 (Serghini et al., 2001).

Pre-haustorial resistance mechanisms have also been reported in the sunflower-O. cumana interaction. These take place from the first contact of the parasite seedling with the host root up to the establishment of vascular connections through haustorium development (Pérez-de-Luque et al., 2009). Studies in resistant H. annuus (Dörr et al., 1994) and H. debilis (Labrousse et al., 2001) have identified the formation of an encapsulation layer that prevents intrusion of the broomrape radicle into the cortex. Physical barriers reinforcing host cortical cell walls through suberization and protein cross-linking (in cv. HE-39999, resistant to race F, Advanta Ibérica, Sevilla, Spain) (Echevarría-Zomeño et al., 2006), or callose depositions (in line LR1, resistant to race E) (Letousey et al., 2007) have been reported. In this last case, callose accumulation around the parasite penetration site was correlated with the induction of a callose synthase gene (HaGSL1) (Letousey et al., 2007). Accumulation of lignin and its precompounds in injured host cells has also been described in resistant sunflower genotypes (Panchenko & Antonova, 1975; Antonova, 1994; Jorrín et al., 1996). Lignification and suberization for cell wall fortification are processes linked to the activity of peroxidases (Echevarria-Zomeño et al., 2006), and the important role of these enzymes in relation to resistance to O. cumana was reported in early studies in sunflower by Suhorukov (1930) (cited by Morozov, 1947) and later by Antonova & ter Borg (1996). Chemical barriers have also been described in the above-mentioned sunflower genotype HA-39999 (resistant to race F), and again phenolics appear as toxic compounds in the form of accumulation and excretion of phytoalexins into the apoplast that create a toxic environment around the broomrape infection point (Echevarria-Zomeño et al., 2006).

Once the parasite develops vascular connections to the host, the parasitic phase begins and defence mechanisms are post-haustorial (Pérez-de Luque et al., 2009). These mechanisms are characterized by the death of broomrape tubercles during their initial developmental stages. A physical barrier consisting of sealing sunflower xylem vessels with gel or gum-like substances which block the flux of water and nutrients from the host to the parasite, has been observed in the resistant line LR1 against race E (Labrousse et al., 2001). As can be concluded for LR1 or HA-39999 genotypes, several resistance mechanisms occurring at different stages may operate within a single resistant genotype.

Genetic resistance to herbicides as a breeding alternative for the control of O. cumana

During the last decade another alternative for the control of O. cumana, based on the development of herbicide resistance in sunflower (Sala et al., 2012a) together with herbicide treatments, has been implemented. This option is generally used in combination with the available O. cumana resistance genes. The development of sunflower genotypes bearing an acetohydroxyacid synthase (AHAS, EC 4.1.3.18, also known as acetolactate synthase) mutant gene (Shaner et al., 1984; Ray, 1984), and therefore resistant to AHAS, has made it possible to successfully control broomrape regardless of the race composition of the populations of these weeds (Škorić & Pacureanu, 2010). At present, three different herbicidal technologies commercially available are based on the introduction of different AHAS alelles conferring resistance to particular AHAS inhibitors in the crop (Sala et al., 2012a).

The reduction in herbicide binding is caused by mutations at key sites in the genes coding for the catalytic subunit of AHAS. Several authors have reviewed known mutations of the AHAS genes that confer resistance to AHAS-inhibiting herbicides in weeds and crops (Preston & Mallory-Smith, 2001; Tranel & Wright, 2002; Tan et al., 2005). Based on molecular studies, Kolkman et al. (2004) identified and characterized three genes coding for the AHAS catalytic subunits in sunflower (Ahasl1, Ahasl2 and Ahasl3). Ahasl1 is a multilallelic locus and the only member of this small family in which all the induced and natural mutations for herbicide resistance have been described thus far in sunflower. Ahasl1-1, also known as Imr1 (Bruniard & Miller, 2001) or Ar_pur (Kolkman et al., 2004), harbours a C-to-T mutation in codon 205 (Arabidopsis thaliana nomenclature), which confers a moderate resistance to imidazolinone (IMI). Ahasl1-2 (also known as Ar_kan) shows a C-to-T mutation in codon 197 conferring high levels of sulfonylurea (SU)-resistance (Kolkman et al., 2004). Ahasl1-3 presents a G-to-A mutation in codon 122, which confers high levels of IMI-resistance (Sala et al., 2008a), and Ahasl1-4 harbours a G-to-T mutation in codon 574, which endows a wide range of resistances to four families of herbicides targetting AHAS (Sala & Bulos, 2011)

Imidazolinone and sulfonylurea herbicides have been seen to have a broad spectrum of weed control activity, flexibility in timing of application, low usage rates, and low mammalian toxicity (Brown, 1990; Tan et al., 2005). These herbicides inhibit the enzymatic activity of AHAS, the first enzyme in the pathway for the synthesis of the branched chain amino acids valine, leucine, and isoleucine (Singh, 1999). This same enzyme has been shown to be the site of action for the triazolopyrimidines (TZ, Subramanian & Gerwick, 1989), pyrimidyloxybenzoates (POB, Subramanian et al., 1990), and sulfonylaminocarbonyl-triazolinones (Santel et al., 1999). Because this technology depends on the availability of IMI- and SU-resistant hybrid cultivars, the development of IMI- or SU-resistant plants with altered AHAS genes and enzymes is imperative. Herbicide resistant (HR) plants have been discovered in sunflower, which permit the development and commercialization of at least four HR traits. Resistance in these traits is due to a form of the AHAS large subunit enzyme (AHASL) that is less sensitive to herbicide inhibition and is conferred by a single, partially dominant nuclear gene (Sala et al., 2012a).

The use of IMI-resistant sunflowers (also termed as Imisun) for the control of O. cumana was early advised by Alonso et al. (1998). The genetics behind this trait was well described (Bruniard & Miller, 2001). The first IMI-resistant trait in sunflower was developed by the BASF Company and in simultaneous use with IMI herbicides was called Clearfield technology (Tan et al., 2005). A set of different Clearfield herbicides were registered to be applied as post-emergence treatments in sunflower. These herbicides were formulated in several ways, using different imidazolinone molecules, and taking into account the specific characteristics of the weeds, soils and crop rotations in each sunflower production region. It is important to note that the moment of application is crucial for an effective control of O. cumana. Pulsar® (imazamox 40 g/L) was effective in split applications between the plant stages of 2 and 4 true leaves (Demurin & Perstenyeva, 2010; Masirevic et al., 2010). Due to the simultaneous control of both O. cumana and key weeds, the Clearfield technology for sunflower production reached about 50% market share in Trakya Region (European part of Turkey) and over 70% of the Turkish sunflower area in the last years (Kaya & Evci, 2009).

The second IMI-resistant trait in sunflower, known as CLPlus®, is controlled by the expression of the partially dominant nuclear allele Ahasl1-3, which was developed by seed mutagenesis and selection with imazapyr (Sala et al., 2008a). Based on a vast array of environmental conditions and on biochemical studies, it was determined that the CLPlus® trait affords greater resistance to IMI herbicides than the Imisun trait, also providing a better stability to cope with the unpredictable portion of the environmental variation (Sala et al., 2008a,b,c; 2012b). In fact, the CLPlus® trait displays the lowest level of inhibition of the AHAS enzyme extracts by imidazolinones, which results in a higher level of accumulation of biomass after imidazolinone application at the above-ground (Sala et al., 2012a) and root levels (Sala et al., 2012b).

SU-resistant and ExpressSun® sunflowers, both with resistance to SU herbicides, were developed separately from wild sunflower populations discovered in the USA (Al-Khatib et al., 1999) and by ethyl methanesulfonate mutagenesis over the line HA89 (Gabard & Huby, 2001), respectively. The resistance allele Ahasl1-2 from SU-resistant sunflowers was introgressed into cultivated sunflower by forward crossing and selection with the herbicide tribenuron, and gave rise to the trait known as Sures (Miller & Al-Khatib, 2004). Even though the inheritance of these traits has not been reported, it is well established that the target-site-resistance is the result of the mutation P197L at the Ahasl1 locus (Kolkman et al., 2004), and that damage differences between Sures-resistant breeding lines are the result of the presence of modifier genes (Miller & Zollinger, 2004). These mutations were independently used by some public institutions and by private companies to develop SU-resistant hybrid cultivars in many countries, increasing the range of available herbicides in sunflower (Jocić et al., 2011; Streit, 2012). The use of SU herbicides for the control of O. cumana in SU-resistant sunflowers was tested under field conditions. A split post-emergence application of metsulfuron methyl provided a good level of control and was not phytotoxic to sunflower (Gabard & Huby, 2001). When the response of one SU- and one IMI-resistant sunflower hybrids to seven doses of imazamox and tribenuron methyl was analysed in-field and in greenhouse experiments, imazamox proved to be more effective for the control of the parasite in IMI-resistant sunflower than tribenuron in SU-resistant sunflowers (Malidza et al., 2012).

Finally, a novel HR trait in sunflowers, still under development, is known as AIR. It is controlled by the Ahasl1-4 and presents a completely new pattern of cross-resistance for sunflower, since it shows a broad range level of resistance to different AHAS-inhibiting herbicides (IMI, SU, TZ and POB). Furthermore, this allele also presents a higher level of resistance to IMI and SU than lines carrying the Imisun and the Sures traits, respectively (Sala & Bulos, 2011). It is known that sunflower lines developed to cope with some AHAS-inhibiting herbicides are susceptible to foliar applications and, in many cases, to soil residues of other AHAS-inhibiting herbicides (Howatt & Endress, 2006). In these cases, the cross-resistance of Ahasl1-4 could allow sunflower hybrids carrying this allele to cope with the soil residues of other types of AHAS-inhibiting herbicides from the fallow or the previous crop.

Other candidate genes are now under study aiming at the development (Sala et al., 2008a; Sala & Bulos, 2011) and characterization (Sala et al., 2008b) of new HR traits, which could lead to the formulation of new herbicides, targeting different enzymes. Similarly, herbicide formulations and seed dressing technologies using different AHAS-inhibiting herbicides and which have proven to be useful for the control of Striga spp. in corn and sorghum (Tuinstra et al., 2009; Ransom et al., 2012), should be tested over the known HR sunflower technologies in order to provide new strategies for the control of O. cumana.

Other chemical options have a detrimental effect against O. cumana, such as inducers of seed germination leading to suicidal germination of the parasite in the absence of sunflower (Lachia et al., 2014), or inhibitors of the germination process (Okazawa & Benesh, 2011) but, as they are not coupled with genetic breeding of sunflower for traits of resistance to chemicals, they will not be covered in this review. Extensive information on chemical signals from hosts and their effect on parasite species can be found in other works (Smith et al., 1990; Gomez-Roldan et al., 2008; Umehara et al., 2008; Yoneyama et al., 2008). Finally, the availability of next-generation sequencing technologies, metabolomics and its applications to produce continuous and massive information about parasitic weeds (Piednöel et al., 2012; Westwood et al., 2012; Pineda-Martos et al., 2014) should be exploited. Previously known herbicide target genes obtained from model plants can be used for a rapid and effective characterization using different bioinformatics approaches on the creation of new herbicides specifically directed at O. cumana. The discovery of differences in metabolic pathways between this organism and sunflower will be vital for these developments. Also, the sequences generated will provide insights into the biology of parasitism and advance progress towards understanding parasite virulence.

In summary, the parasitic weed O. cumana poses a risk to sunflower oil production in countries of Southern and Eastern Europe every year, and causes, on average, 50% sunflower seed losses when susceptible hybrids are grown. Understanding the biology of the interaction between the parasite and its host is a necessary step toward the development of selective control methods. Physiological and molecular mechanisms governing the Orobanche development and the establishment of the interaction with its hosts, for example the chemical stimulation of parasite seed germination, or the role of degrading enzymes on the parasite progression between host cells, have been deeply studied and characterized. However, much remains to be done on deciphering the biological processes that are particular to O. cumana in sunflower. This knowledge would lead to the identification of vulnerabilities as specific targets for new control methods. Breeding for resistance is a recurrent, feasible and effective alternative for controlling O. cumana. Different races of the parasite have been identified in all the countries where it threatens sunflower oil production and, at the same time, important breeding works have been devoted to the search for effective genetic resistance against the increasingly virulent local parasite populations. Since information on pathogenic traits of the parasite is obtained locally, its validity is restricted to particular geographical areas within Europe. Comparisons of results giving the characterization of races at an international level, as well as the effectiveness of resistance sources in international sunflower breeding programmes, will be favoured by the universal adoption of the coded triplets system for nomenclature of O. cumana races. On the other hand, breeding sunflowers for resistance to AHAS-inhibiting herbicides has appeared during the past decade as another alternative for the control of O. cumana together with other weeds of the crop, and irrespective of parasite races. IMI- and SU- resistant plants were discovered in sunflower, which permitted the development and commercialization of four herbicide resistant traits. This alternative for the control of O. cumana is, in some instances, implemented together with genetic resistance to races of the parasite, and is likely to undergo an important development as long as novel HR traits are discovered.