Temporal expression of the sor 1 gene and inhibitory effects of Sorghum bicolor L . Moench on three weed species

The temporal expression of gene sor1 and the inhibitory effect of Sorghum bicolor L. Moench against weeds were studied by semiquantitative polymerase chain reaction and intercropping management, respectively. To quantify sor1 expression, seeds were sown in pots and RNA was collected from the roots at 5, 10, 15, 20 and 30 days after emergence (DAE). In the inhibition assay, cotton and three weeds were evaluated during single cropping and during intercropping with S. bicolor. The assay was completely randomized, with eight replications. We found early expression of sor1 in most S. bicolor accessions by 5 DAE, and a gradual reduction thereafter. Only one of the accessions showed sor1 expression up to 30 DAE. In the inhibition assay, the most significant effects were related to the dry matter production (shoots and roots) of the weeds Cenchrus echinatus and Cynodon dactylon. The intercropping of cotton and S. bicolor had no apparent deleterious effects.


Introduction
Weeds are a serious problem in all kinds of crops, because they affect the development cycle and yield, especially in herbaceous species before flowering.Although various control methods are available, the most effective is the use of synthetic herbicides, which unfortunately have high costs in terms of management, as well as posing risks to human health and the environment (Bolonhezi et al. 2005;Cheema et al. 2003).The use of such herbicides in a single-cropping system can also lead to further selection of herbicide-resistant weeds (Neves 2005).
Despite the numerous benefits that classical breeding has brought to crop yields, it is still limited in its capacity to generate weed-resistant cultivars.In contrast, the current techniques, which are based on the use of biotechnology-derived herbicide-resistant crops, have contributed to minimizing the use of herbicides, thereby providing greater crop protection and productivity (Parker et al. 2005;Correia & Durigan 2007;Imura et al. 2011).However, the management of such crops can still require the use of chemical herbicides.
Weed control through the use of natural defensive agents obtained from other plants is an agroecological alternative often adopted by farmers of small plots of land (less than 10 ha).In general, a solution containing extracts from several herbicide species is sprayed weekly between rows of short-cycle crops, such as beans, corn, cotton and peanuts (Melo et al. 2013).In the literature, several studies have demonstrated the potential of allelopathic herbicides, at low concentration, to control monocot and dicot weeds (Bertin et al. 2003;Weston & Duke 2003;Albuquerque et al. 2010).This practice is beneficial to the environment and offers the potential for biorational weed control.It can also minimize the cost of management, given that many of the herbicide species are widely distributed on croplands.
According to Souza Filho et al. (2006), allelochemicals produced by plants affect local vegetation and the succession of others plants, as well as playing a role in the induction of dormancy and seed preservation.Such chemicals therefore constitute an important mediator of population dynamics because they determine the pattern and density of vegetation, in natural and crop systems alike.Numerous plants have been reported to be allelopathic, including lichens, alfalfa, cucumber, rye, barley, wheat, rice, soybeans, and sorghum (Singh et al. 2003;Belz 2007;Albuquerque et al. 2010).
According Meazza et al. (2002) and Hejl & Koster (2004), the role of sorgoleone as a natural herbicide is due to its effect on electron transport in the chloroplasts, whereby it inhibits the production of p-hydroxyphenylpyruvate dioxygenase (HPPD) in a manner similar to that of synthetic herbicides.Inhibition of HPPD disrupts the biosynthesis of carotenoids and results in the bleaching of leaves, due to a loss of chlorophyll.In the field, sorgoleone has an inhibitory effect on the growth of several species, including beans, wheat, and soy, although its greatest effect has been shown to be on the growth of weeds (Roth et al. 2000;Souza Filho et al. 2006).Sorgoleone is biosynthesized, during early seedling establishment, by root hairs, which possess the entire metabolic machinery (Dayan 2006;Baerson et al. 2008aBaerson et al. , 2008b)).According to Marchi et al. (2008), sorgoleone is rapidly degraded in soil and is mainly produced in younger plants, its production peaking by 10 days after emergence (DAE).However, other authors have reported that sorgoleone has a long half-life in the soil, with long-term effects on many cellular targets (Dayan et al. 2007;Baerson et al. 2008aBaerson et al. , 2008b;;Barbosa et al. 2010).
Molecular studies have described the metabolic pathway of sorgoleone, and a number of genes involved in the event cascade have been identified and characterized (Buchanan et al. 2000;Dayan et al. 2003;Pan et al. 2007;Baerson et al. 2008aBaerson et al. , 2008b)).According to Dayan et al. (2003), the gene sor1, which encodes a membrane desaturase, is the main precursor of sorgoleone synthesis.
Although the cellular localization and the steps of the biosynthetic pathway of sorgoleone have been determined (Dayan et al. 2003;Pan et al. 2007;Baerson et al. 2008aBaerson et al. , 2008b)), there is a lack of studies investigating its differential expression among genotypes and its bioactivity in vivo.Therefore, the evaluation of accessions that show variability for sorgoleone synthesis could contribute to the selection of the top lines in breeding programs for food production, as well as for weed control, in farming systems.
In the present study, we evaluated the temporal expression of sor1 in Sorghum accessions.We also quantified its inhibitory effect in an intercropping system with three herbaceous weeds.

Experimental procedure and RNA extraction
Seeds of five Sorghum bicolor accessions, one each of the cultivars IPA 467-4-2, IPA 7301011, IPA 4202, CNB 9040, and Sudan, were sown in the greenhouse of the Department of Agronomy of the Federal Rural University of Pernambuco, in 5-L pots containing commercial substrate (Plantmax; Eucatex, São Paulo, Brazil).At 5, 10, 15, 20, 30 and 50 DAE, rootlets were collected for RNA extraction (Plant RNA Mini-Spin Invisorb kit; Invitec, Berlin, Germany) from fresh tissue samples (100 mg), in accordance with the manufacturer's recommendations.The concentration and purity of the RNA were estimated with a spectrophotometer (BioPhotometer Plus; Eppendorf, Hamburg, Germany).

In vivo effect of Sorghum bicolor density on the growth of weeds and cotton
We performed the in vivo inhibition assay in a greenhouse, using the Sorghum bicolor accession that showed the highest sor1 expression in the semiquantitative RT-PCR.We first evaluated the effect of S. bicolor on the weeds Cyperus rotundus, Cynodon dactylon, and Cenchrus echinatus.Plastic pots (17 cm in diameter) were filled with organic substrate (Plantmax; Eucatex), plus N:P 2 O 5 :K 2 O fertilizer (20:10:20).On the basis of the work carried out by Trezzi & Vidal (2004), we planted a total of nine plants in each pot, at the following weed-S.bicolor ratios: 9:0, 5:4 and 0:9.To investigate the possible allelopathic effect of S. bicolor on cotton, we added an arrangement of S. bicolor and cotton (Gossypium hirsutum L., cv.CNPA 8H), in pots filled with the same substrate plus N:P2O5:K2O-fertilizer (60:60:60), at sorghum-cotton plant ratios of 9:0, 7:2, and 0:2.The experimental design adopted was totally randomized, with eight replications.Pots were watered daily.Each individual plant height (cm) and dry biomass (g) were measured at 56 days.Statistical analysis were performed with the GENES program (Cruz 2006).

Results and discussion
The semiquantitative expressions of sor1 and β-actin in Sorghum bicolor accessions are shown in Fig. 1.The expression profile of sor1 differed among accessions, as well as among roots of different ages.In general, the expression of sor1 was highest during the first 20 DAE, varying among accessions and trending downward.We observed sor1 expression as early as 5 DAE, except in the IPA 4202 cultivar (Fig. 1), from which the gene was absent at that stage.
The best sor1 expression was seen in the Sudan and IPA 7301011 cultivars, the latter distinguished by expression that was more uniform and prolonged, being detectable at up to 30 DAE (Fig. 1).This result is interesting because the longer duration of the sor1 activity might translate to more effective control of the surrounding weeds, especially in denser cropping.Another aspect is that, in a Sorghum bicolor improvement program, hybridization between the Sudan and IPA 7301011 cultivars could result in high-sorgoleone progenies, prolonging the effects and consequently improving protection against various weeds.Because sor1 expression was observed at 30 DAE in the Sudan and IPA 4202 cultivars, we also investigated sor1 expression in both at 50 DAE; however, no activity was found.On the basis of those results, the Sudan cultivar was selected for the in vivo assay involving S. bicolor density with weeds and cotton.We found that S. bicolor had no major effect on the plant height of the cotton or the weeds (Fig. 2).However, it did have an effect on the dry masses of shoots and roots in Cenchrus echinatus and Cynodon dactylon (Fig. 3).
In the cotton and Sorghum bicolor treatment, we found no phytotoxic effect on dry mass.Instead, we observed increases of 122% and 141% in the mass of cotton canopy and roots, respectively.Cheema et al. (2003) also reported a beneficial effect of intercropping S. bicolor with wheat.The authors carried out a study involving the control of weeds and wheat competition by foliar application of S. bicolor water extract (12 L ha −1 at 30 and 40 days after sowing) and observed a 33-53% reduction in weed biomass, together with a 7-14% increase in wheat yield.These results seem to be associated with the density of S. bicolor in management with other cultures.Hallak et al. (1999) found anatomical changes in beans (Phaseolus vulgaris L.) grown in the presence of root exudates of S. bicolor at 0.10 and 0.15 mM.The authors observed a reduction in the number of cell layers, thickening of the cellulose portions of the collenchyma, and deformation of vessel elements.In a similar study, Souza et al. (1999) found sorgoleone to be phytotoxic to wheat, beans, soy, and pigweed, even resulting in death, at 0.10 mM.
Although Sorghum bicolor is considered to be a highly allelopathic species, its toxic effect is not the same for other crops, because it depends on many factors, such as genotype concentration, plant density, fertility, and soil moisture (Rice, 1984;Kruse et al. 2000;Yang et al. 2004).However, the efficacy of control against some weeds might not occur immediately, as does that of commercial herbicides, but could build over time, as with most natural products (Belz 2007).Netzley et al. (1988) reported that the concentration of sorgoleone exuded into the soil from S. bicolor roots can reach 10-100 μM.According to Czarnota et al. (2001), the post-emergence application of sorgoleone inhibits the   growth of various weeds, especially broad-leaf weeds, at the same concentration as the synthetic herbicide atrazine (0.6 kg a.i.ha −1 ).Weston & Czarnota (2001) also showed S. bicolor toxicity in pre-emergent small-seeded weeds.
The results of the present study confirm the findings of Wu et al. (1999), who showed that there is genetic variability in sorgoleone production among Sorghum accessions at the intraspecific and interspecific levels.At the molecular level, Yang et al. (2004) demonstrated such variability from the amino acid sequence of the SOR1 protein among species of tomato, potato, tobacco, sesame, beans, rice, and sorghum.These findings open a range of opportunities for genetic improvement of the germplasm containing this gene, in order to select crops with high allelopathic ability.Despite the great benefit of crops with allelopathic properties for the agroecological management of weeds, scientists emphasize the need to understand the genetic control of allelopathy (Baerson et al. 2008b;Albuquerque et al. 2010).
On the basis of our results, not only those related to the sor1 expression in the roots of the different Sorghum bicolor accessions but also those related to the inhibitory effect of the Sudan cultivar on the biomass production of two major weeds, we suggest a pyramid breeding scheme involving the Sudan and IPA 7301011 cultivars with other high-yield top lines, in order to generate progenies for further use in single cropping or intercropping agroecological management.It should be borne in mind that, prior to intercropping management, the absence of an allelopathic effect on the other main crop should be verified, in order to ensure that S. bicolor will be phytotoxic only to the weeds.