Identification of Sources of Resistance for Peanut Aspergillus flavus Colonization and Aflatoxin Contamination

Université Gaston Berger, BP 211 Saint-Louis, Senegal Centre National de Recherches Agronomiques de Bambey, Institut Sénégalais de Recherches Agricoles, BP 211 Bambey, Senegal Centre d’Etudes Régional pour l’Amélioration de l’Adaptation à la Sécheresse, Route de Khombole, BP 3320 +iès, Senegal Conseil Ouest et centre Africain pour la Recherche et le Développement Agricoles, Avenue Bourguiba, BP 48 Dakar RP, Senegal


Introduction
Peanut (Arachis hypogaea L.) is an important staple crop in Senegal. e national peanut production was estimated at 1,050,042 tons during the rainy season of 2016 [1]. is crop is mainly produced in Fatick, Kaolack, Kaffrine, Louga, and ies regions, with more than 60% of the national peanut production [1]. Peanut seeds are widely used for food consumption and play a significant economic role for smallscale farmers and food industries in Senegal [2]. However, pre-and postharvest aflatoxin contamination in peanut is a serious threat for food safety and human health in Senegal [3]. It is one of the major constraints limiting sustainable and good quality seed production in the world [4]. Aflatoxin contamination is due to Aspergillus flavus (Link ex Fries, Teleomorph: Petromyces flavus) [5]. Damages caused by this facultative plant pathogen in maize, peanut, and sesame were reported in Senegal [6]. Considerable economic losses caused by this bacterium are mainly due to crop quality value and international trade restrictions on food stuffs charged in aflatoxin [7].
Aflatoxin is the name of a group of toxin known as G1, G2, B1, B2, M1, and M2 that produced the plant pathogen [8]. ese toxins occur naturally and have been found in a wide range of commodities, including peanuts used for animal and human consumption [9]. Aflatoxins are toxic, mutagenic, and carcinogenic compounds [10]. Depending on their levels, toxins can severely affect the liver and induce immune-suppressing effects [9].
To handle this issue, a wide range of preharvest aflatoxin contamination management methods were developed. Application of atoxinogenic isolates of A. flavus [11] and host genetic resistance were tested [12]. In Senegal, previous studies reported that varieties 55-437 and 73-3 were resistant to A. flavus [13]. Identification of new sources of resistance merits to be investigated for efficient peanut breeding program. First step of host genetic resistance is the seed colonization test. erefore, the present study was undertaken to identify promising peanut genotypes under laboratory conditions.

Plant Materials.
e plant material consisted of 67 genotypes including 58 chromosomal substitutions lines [14] and nine national released varieties. e chromosomal substitution lines belong to a cross between Fleur 11 and a synthetic amphidiploid parent (Table 1).

Isolation of Aspergillus flavus, Sporangial Suspension
Preparation, and Inoculation. Aflatoxinogenic strain provided from peanut seeds were purified by successive cultures on 5/2 agar medium. e aflatoxin concentration level was checked using the Reveal ® Q + Aflatoxin test kit (accesso peanut enterprise corporation, USA). e spore suspension of A. flavus was obtained by soaking colonized seeds in 50 ml of sterile distilled water. en, one drop of Tween 20 was added to the solution and thoroughly mixed for 10 minutes. Inoculation was carried out by introducing 100 μl of the supernatant of the spore suspension into each Petri dish.

Seed Colonization Test.
e seed colonization test was conducted following a modified Mehan and McDonald procedure. For each genotype, 50 seeds were sterilized and rinsed properly in sterile distilled water. en, the seeds were hydrated to about 20% moisture content. e 50 seeds of each genotype were placed in 5 Petri dishes containing 10 seeds, and each Petri dish was considered as a replication. e seeds were inoculated with a conidial suspension (60 µL containing approximately 1 × 10 8 mL −1 conidia of the aflatoxigenic strain of A. flavus). is preparation was kept at laboratory conditions (25 ± 0.12°C and 82 ± 0.42% relative humidity) for fifteen days.

Data Collection.
e seeds' colonization was observed during two weeks, and aflatoxin concentration was measured using the Reveal ® Q + Aflatoxin test kit (accesso peanut enterprise corporation, USA). e incidence was calculated using the following formula: number of seeds showing pathogen colonizaton total number of seeds × 100.
(1)  International Journal of Agronomy e severity scale of aflatoxin on seeds was estimated using a modified Tonapi et al. [15] scale. It was defined as follows: 0, noninfected seeds; 1, seeds whose surface covered by the fungus is less than 20%; 2, 20%-40% seed surface covered by the fungus; 3, 40%-60% seed surface covered by the fungus; 4, 60%-80% seed surface covered by the fungus; and 5, 80%-100% seed surface covered by the fungus. e severity calculation based on Tonapi et al. [15] formula was as follows: where p < 0.001 i is severity scale from 0 to 5 and N i is the number of seed corresponding to scale i of severity.

Data Analysis.
Data analysis was performed with the open-source statistical software R version 3.4.5 [16]. Descriptive statistics of recorded data were generated with pastecs package [17]. In order to find out variability of incidence and severity according to tested genotypes, data were subjected to Poisson regression analysis using glm (generalized linear model) function of package stats implemented in the R. Spearman's rank correlation test was performed to highlight relationship between incidence, severity, and aflatoxin concentration levels using correlation test function of package stats. Identification of different groups of genotypes based on incidence and severity was performed based on a principal component analysis and a hierarchical clustering with the functions PCA and HCPC of package FactoMineR [18], respectively. e Euclidean distance and Ward classification method were used to classify tested genotypes. e function fviz_pca_biplot [19] was used to plot the principal components analysis biplot in different clusters based on hierarchical classification.

Reaction of Peanut Genotypes to Aspergillus flavus.
Analysis of variance revealed highly significant (p < 0.001) variation of aflatoxin incidence and severity among the tested peanut genotypes (Table 2). e severity ranged between 0 and 44%, respectively, with a mean of 8.82 ± 0.45%. e recorded incidence ranged from 0 to 70% with an average value of 20.36 ± 0.80% (Table 3).
One genotype (12CS_104) showed aflatoxin concentration level less than 4 ppb. A total of 34 genotypes presented aflatoxin concentration level up to 2000 ppb ( Figure 1).
Out of the 67 genotypes, eight showed incidence less than 10% while 33 showed incidences between 10 and 20% and 16 with incidences ranged from 20 to 30% (Figure 2).

Correlation between Incidence, Severity, and Aflatoxin
Concentration Level. Spearman's rank correlation test revealed a strong relationship (r � 0.93, p < 0.001) between incidence and severity of peanut genotypes. Positive and significant correlations were detected between aflatoxin concentration levels and disease incidence (r � 0.28, p < 0.01) and aflatoxin concentration levels and disease severity (r � 0.35, p < 0.05) ( Table 4).

Classification of the Tested Genotypes according to Sensibility and Aflatoxin Concentration Level.
e factorial axes 1 and 2 explained 60.5 and 39.5% of overall variability, respectively ( Figure 3). Hierarchical classification performed on principal component analysis revealed three clusters of genotypes based on disease incidence and aflatoxin concentration levels (Figure 3). e clusters 1, 2, and 3 grouped 33, 20, and 14 genotypes, respectively. e incidence and aflatoxin concentration are significantly (p < 0.001) associated to cluster 1 (Table 4).
Mean values of these two variables in this cluster are less than the overall mean. erefore, cluster 1 is characterized by desirable genotypes which combine low incidence values and aflatoxin concentration levels. Cluster 2 is significantly (p < 0.001) related to the aflatoxin concentration level (Table 5). e mean value of aflatoxin concentration in cluster 2 (4075.5 ppb) is 190% which is higher than the overall mean (2143.8 ppb). us, this second cluster is characterized by genotypes with high level of aflatoxin. Incidence is linked to cluster 3 (Table 5). Mean value of this variable (35%) in cluster 3 is superior to overall mean (20.35%). us, the cluster 3 encompasses the most susceptible genotypes to A. flavus.
Based on the closest distance between each genotype and the respective cluster centres, 12CS_039, 12CS_010, and 12CS_050 were the first representative genotypes (paragon) of cluster 1, 2, and 3, respectively (Table 4). Based on the farthest distance from a genotype projected point in a cluster to the centres of the two others, clustering revealed that cluster 1, 2, and 3 were characterised by the genotypes 12CS_104, 78-936, and 12CS_021, respectively ( Figure 3, Table 5). Based on results, out of 67 genotypes, 33 promising genotypes (cluster 1) were noted (Figure 3).  International Journal of Agronomy

Discussion
In the present study, a wide phenotypic variation was observed among the tested genotypes for incidence, severity, and aflatoxin concentrations. is variation can be explained by the variability of seed coat structure of the tested genotypes. In fact, the seed coat can constitute a barrier to A. flavus seed invasion depending on its thickness and/or permeability [20], and Zhou and Liang [21] studies showed that genotypes seed coat with smaller hilum, more compact arrangement and thicker testa showed more resistance to A. flavus. In addition, implication of wax and cutin layers of seed coat was demonstrated to be related to genotypes resistance [22]. Another explanation of this wide variation in incidence, severity, and aflatoxin rate can be biochemical compounds' differential variability in the tested seeds. Lindsey and Turner [23] demonstrated that the presence of polyphenol compounds, specifically, tannins in seed can have inhibitor effect against A. flavus. Amaya et al. [24] and Liang et al. [25] showed the difference among seed coat biochemical compounds to determine sensibility to A. flavus. Liang [22] demonstrated that the presence of trypsin in seeds can also be related to resistance to A. flavus. Turner et al. [26] isolated and identified the 5,7-dimethoxyisoflavone as an inhibitor for A. flavus invasion in peanut seed. 12CS_104 was the most resistant genotype to aflatoxin contamination with an aflatoxin level lower than the European Union standards (4 ppb). However, except 12CS_104, all the genotypes have their aflatoxin concentration level higher than the Chinese (20 ppb) standards.   Indeed, the highest a atoxin concentration level was observed with genotype 78-936. e contrasting genotypes observed in this study can be used as positive and negative checks, respectively, for accurate eld experiment. Furthermore, these contrasted genotypes can be used to develop mapping population for genetic study such as inheritance of a atoxin and identi cation of quantitative trait loci (QTL). e varieties 55-437 and 73-30 showed incidence less than 15% as reported by the previous study realized 30 years ago by Zambettakis et al. [13], but their a atoxin concentration levels were largely up to the European Union standards.
e correlation test showed a positive relationship between A. avus colonization and a atoxin contamination.
is con rmed that the presence of A. avus induced aflatoxin production in seeds. Hierarchical classi cation highlighted three clusters according to incidence, severity, and a atoxin concentration levels. e relatively low values of incidence observed on the 33 genotypes belonged to cluster 1 should be con rmed under eld conditions. ese genotypes can be evaluated in di erent locations on infested elds.

Conclusion
is study uncovered that the lines 12CS_104 exhibited low values of incidence and severity. Furthermore, its a atoxin