A SYBR Green qPCR assay for specific detection of Colletotrichum ocimi , which causes black spot of basil

Summary. Colletotrichum ocimi causes black spot of basil ( Ocimum basilicum ) and is a serious threat to basil cultivation as it compromises leaf production. The pathogen also infects seeds, which could become primary sources of inoculum for spread of black spot. A SYBR Green real-time PCR assay was developed to detect Colletotrichum ocimi in basil leaves and seeds, based on the partial β-tubulin ( tub2 ) gene sequence. Two primer sets were designed and tested. The selected primer pairs produced amplicons of 130 bp. The real-time PCR assay was validated for analytical specificity, sensitivity, selectivity, repeatability and reproducibility. The assay was specific for C. ocimi with respect to ten Colletotrichum spp. and to another 12 pathogens of basil plants. Sensitivity was 1 pg µL -1 of genomic fungal DNA and amplification analyses were not influ-enced by basil genomic DNA. The assay detected and quantified C. ocimi in artificially inoculated basil leaves. This is the first specific primer set for C. ocimi , which allows rapid detection and quantification of the pathogen is a useful tool for diagnostics in plants. Detection in seeds would also be possible, but will require an optimized extraction method. The qPCR detection of C. ocimi in planta can contribute to adoption of effective preventive disease management strategies.


INTRODUCTION
Colletotrichum includes important species pathogenic to a range of plant hosts (Dean et al., 2012). Colletotrichum ocimi causes black spot of basil (Ocimum basilicum), an economically important crop cultivated in many countries, and, particularly, in the Mediterranean area (Damm et al., 2014). Black spot is the common name for this anthracnose disease, which can affect basil leaves and seeds (Gullino et al., 1995), and is a serious threat to seed-pro-duction companies and basil producers (Cacciola et al., 2020). Gullino et al. (1995) reported isolation of a Colletotrichum sp. from basil, which was initially identified as C. gloeosporioides. (Damm et al. (2014) later allocated these isolates to the C. destructivum species complex (SC), describing it as C. ocimi. Several other Colletotrichum species have been reported on basil: C. capsici in India (Alam et al., 1981), C. destructivum in Italy (Cacciola et al., 2020), C. siamense in Malaysia (Ismail et al., 2021) and Colletotrichum sp. in Florida, United States of America (Alfieri et al., 1984).
Before the 1990's, species of Colletotrichum were identified using a combination of morphological and cultural characters. The most investigated features were shape and size of conidia and appressoria, presence or absence of setae, sclerotia, acervuli, sexual morphs, and cultural characteristics (Cannon et al., 2000). However, these characters alone were considered inadequate for species identification, due to their variability depending on environmental and culture conditions ). Molecular tools have been considered more reliable and were subsequently used along with morphological observations. Particularly, multi-locus phylogenetics is used for Colletotrichum species identification . Up to 13 different loci are available for delineation of species within a SC (Liu et al., 2016;Talhinhas and Baroncelli, 2021). Bhunjun et al. (2021) considered the ITS region as useful for reaching a SC-level identifications, with glyceraldehyde-3-phosphate dehydrogenase (gapdh) and β-tubulin (tub2) as the most informative loci for species-level identification. However, the polyphasic approach suggested for Colletotrichum species identification is laborious and time-consuming Du et al., 2021).
The quantitative real-time PCR technique has been used as a rapid and sensitive diagnostic method for the identification of Colletotrichum spp. (Mirmajlessi et al., 2015). Schena et al. (2017) developed a duplex qPCR TaqMan method to detect and quantify C. godetiae and C. acutatum sensu stricto in olive tissues. An assay for the simultaneous detection of C. acutatum and C. gloeosporioides from infected strawberry leaves by real-time PCR was developed by Rahman et al. (2019). Du et al. (2021) elaborated a real-time PCR method to detect, quantify and monitor C. siamense infecting rubber trees, and Kamber et al. (2021) developed a specific quantitative real-time TaqMan PCR assay that allowed rapid and reliable detection and quantification of C. lupini in infected seeds and plant material. These specific molecular techniques provide effective and rapid tools for Colletotrichum species identification and quantification in symptomatic and asymptomatic plant tissues. These methods could be implemented by phytosanitary servic-es and companies to assess the health status of seeds and propagation material (Kumar and Gupta, 2020). However, no methods have yet been developed to identify C. ocimi or other species of the C. destructivum SC.
Considering the important losses that black spot disease can produce on basil seedlings and plants, the present study aimed to develop a SYBR Green qPCR assay to facilitate the detection and quantification of C. ocimi in basil leaves and seeds.

Fungal isolates and DNA extraction
Isolates of C. ocimi (Table 1) were taken from stock cultures maintained at -80°C in the culture collection of the AGROINNOVA Centre of Competence (University of Torino), and grown on potato dextrose agar (PDA, VWR Chemicals) amended with 25 µg µL -1 of streptomycin sulphate (PDA-S, Sigma-Aldrich). Genomic DNA was extracted from 0.1 g of mycelium, using the E.Z.N.A. Fungal DNA Mini Kit (Omega Bio-Tek), following the manufacturer's instructions, and was stored at -20°C. The DNA concentration of each sample was measured using NanoDrop 2000 (Thermo Fisher), and was then adjusted to 1-50 ng µL -1 . Damm et al. (2014) reported that the partial tub2 gene sequence of C. ocimi was less than 97% similar to sequences of this locus of other Colletotrichum spp., so this region was selected for primer design. Sequences of the tub2 region retrieved from the Gen-Bank database were: Colletotrichum ocimi (CBS 298 C. lini CBS 172.51, JQ005849, and C. utrechtense CBS 130243, KM105481). Selection of species within the C. destructivum SC was based on phylogenetic distance within the complex, with closely related species selected. All sequences were aligned with the software Mega v. 7 (Kumar et al., 2016), and a region with high polymorphism with respect to the non-target species was selected for primer design. Two primer pairs were designed and validated in silico using the reference strains listed above, using NCBI Primer-Blast (Ye et al., 2012): TubOc_23fw -TubOc_190rev amplifying 168 bp and TubOc_68fw -TubOc_197rev amplifying 130 bp (Table 2).

End-point PCR amplification, analytical specificity and sensitivity
Preliminary tests were carried out to define the optimal final concentrations of MgCl 2 (0.5, 1 and 1.5 mM), primers (0.5 and 1 µM), DMSO (0.5, 0.7 and 1.2 µL) and template (20, 50 and 100 ng µL -1 ). The end-point PCR reactions were carried out using a Taq DNA polymerase kit (Qiagen), in a total volumes of 25 µL each. The optimized mixture composition was as follow: 2.5 µL of Qiagen Buffer 10×, 0.5µL MgCl 2 (25mM), 0.5 µL dNTPs (10mM), 0.5 µL of each primer (10µM), 0.2 µL of Qiagen Taq DNA polymerase (5U) and 1 µL of DNA as template (20-50 ng µL -1 ) The thermal cycler conditions were: 94°C for 3 min, followed by 30 cycles each of 94°C for 45 s, 64°C for 45 s, and 72 °C for 1 min, and a final extension cycle at 72 °C for 5 min. The PCR products were examined by electrophoresis on 1% agarose gels (VWR Life Science AMRESCO R biochemicals), stained with GelRed TM in Tris-acetate buffer. The primers were tested to evaluate their specificity using strains of species within the C. destructivum SC (Damm et al., 2014) and of species reported as pathogens on basil plants and seeds (Garibaldi et al., 1997;Gilardi et al, 2018). All strains used are listed in Table 1. Analytical sensitivity tests were carried out by conducting PCR reactions using C. ocimi DNA of strain CVG190 (Guarnaccia et al., 2019), which was 10-fold serially diluted (10 ng µL -1 to 100 fg µL -1 ). The Limit of Detection (LOD) of these tests was assessed.

SYBR green real-time PCR (qPCR) assay development
The primers which provided the best results in conventional PCR analyses were selected and used for qPCR with SYBR Green. The real-time PCR assays were carried out using a StepOne Plus™ Real-Time PCR System thermal cycler (Applied Biosystems). Preliminary tests were conducted to define the optimal final concentrations of the primers (3, 1 or 0.3 µM) and the optimal annealing temperature (60 or 64°C). Reactions were carried out with the optimised mixture composition in a total volume of 10 µL, using 5 µL of 10× Power SYBR Green Mastermix, 0.3 µL of each primer (100 µM) and 1 µL of DNA as template (20-50 ng µL -1 ). The optimal amplification conditions were: 95°C for 3 min, followed by 40 cycles each at 95°C for 15 s and at 64°C for 35 s. The melting curves were acquired after each run by ramping the temperature from 60°C to 95°C. Each reaction was performed in triplicate, and the results were displayed using the StepOne software.

SYBR green real-time PCR (qPCR) assay validation
The protocol was verified by evaluation of analytical specificity, sensitivity, selectivity, repeatability and reproducibility. For specificity, qPCR was performed in triplicate on all the strains listed in Table 1, using strains of C. ocimi as positive controls. Analytical sensitivity of the qPCR assay was evaluated by a standard curve obtained using C. ocimi DNA of strain CVG190 10-fold serially diluted, ranging from 10 ng µL -1 to 100 fg µL -1 . The LOD of the method was assessed, along with the correlation coefficient (R 2 ) between the cycle threshold (Ct) and the initial concentration of genomic DNA and the mean relative efficiency. Another test was conducted on C. ocimi genomic DNA mixed with host plant DNA, to simulate interference from host plant DNA and to establish selectivity. Plant DNA was extracted after grinding in liquid nitrogen, using an E.Z.N.A. Plant DNA kit according to the manufacturer's instructions. Colletotrichum ocimi DNA of strain CVG190 was ten-fold serially diluted (from 10 ng µL -1 to 100 fg µL -1 ) with basil genomic DNA (10 ng µL -1 ) extracted from healthy leaves. Each reaction was performed in triplicate and the results were displayed through the StepOne software. The standard curve was used as internal control to quantify C. ocimi DNA in different samples. Three independent assays were conducted to determine the repeatability of the method. The reproducibility was assessed by two different operators on different days.

Detection of Colletotrichum ocimi in artificially inoculated basil leaves
Nine artificially inoculated leaf samples of O. basilicum were collected from basil plants of the 'Genovese' type cultivar 'Italiano classico' (Royal Seeds). Conidial suspensions (10 6 conidia mL -1 ) were sprayed onto pathogen-free 2-month-old plants, which were considered pathogen-free because no visual symptoms were found on leaves examined under a stereo-microscope (Leica EZ4). Control plants were sprayed with sterile water. The plants were then covered with transparent plastic film to maintain high relative humidity, and were transferred to a growth chamber maintained at 25°C with a 12 h light 12 h dark regime. The plastic film was removed at 3 d postinoculation (Guarnaccia et al., 2019). DNA was extracted from 0.1 g fresh weight of symptomatic or non-inoculated leaves. DNA was measured using NanoDrop 2000 (Thermo Fisher Scientific) and diluted to reach a concentration of 10 ng µL -1 . The obtained DNA was analysed with the SYBR green Real-time PCR (qPCR) assay described above.

Detection of Colletotrichum ocimi in artificially inoculated and non-inoculated commercial basil seeds
Conidial suspensions in water could not be used to inoculate seeds, due to gum production by basil seeds after their immersion. Glycerol has been reported as plasticiser of this gum (Amini et al., 2015). Preliminary tests were therefore conducted by soaking seeds in glycerol solutions to find the best concentration to avoid gum production (data not shown). To inoculate basil seeds, a conidial suspension of C. ocimi was prepared in 70% glycerol (VWR Chemicals), with a final concentration of 10 6 conidia mL -1 . Seed samples were exposed to thermal shock for 30 s in liquid nitrogen and then soaked for 1 h in the suspension. Control seed samples were soaked in 70% glycerol. During soaking, samples were gently shaken at 95 rpm. Each seed sample was then recovered on plastic plates with sterile absorbent paper and incubated at 25°C for 72 h in the dark. Two inoculation trials were conducted, each using six batches of seeds (1 g of seeds for each batch) of the 'Genovese' type cultivar 'Edwina' (Enza Zaden), with three replicates per batch for a total of 18 samples. After the first trial, DNA of each inoculated sample was extracted directly after the inoculation. After the second trial, each inoculated sample was first washed with 5 mL of sterile distilled water to remove glycerol residuals, and DNA extraction was then carried out. Ten commercial batches of seven basil cultivars of the 'Genovese' type were each sampled in triplicate (30 samples) of commercial basil seeds. (Table 3). The seeds had no visible symptoms under a stereo-microscope (Leica EZ4). The cultivars were selected among those known to be susceptible to black spot (Guarnaccia et al., 2019;Cacciola et al., 2020). After grinding in liquid nitrogen, DNA of 0.1 g of each seed sample was extracted using an E.Z.N.A. Plant DNA kit, according to the manufacturer's instructions. DNA concentrations were measured using NanoDrop 2000. The obtained DNA was ana-lysed with the SYBR green Real-time PCR (qPCR) assay described above.

Species-specific primer design, analytical specificity and sensitivity
The two designed primer sets were tested to assess their analytical specificity and sensitivity in end-point PCR. The primer pair TubOc_23fw -TubOc_190rev amplified DNA of C. ocimi strains, but also gave nonspecific amplification with DNA of C. bryoniicola (CVG257), C. americae-borealis (CBS 136232), C. lini (CBS 172.51), C. destructivum (CBS 136228) and Sclerotinia sp. The primer pair TubOc_68fw -TubOc_197rev gave best results, by amplifying only DNA of C. ocimi strains (Figure 1, A and B). In analytical sensitivity tests, the LOD of the end-point PCR was 1 ng µL -1 for the primer set TubOc_23fw -TubOc_190rev, and 0.1 ng µL -1 for the primer set TubOc68fw -TubOc_197rev (Figure 2). The analytical specificity and sensitivity assays showed that the primer pair TubOc68fw -TubOc_197rev was appropriate for subsequent assays (Supplementary Figure 1).

SYBR green real-time PCR (qPCR) assay validation
The SYBR green real-time PCR with the selected primer pair was able to amplify the 130 bp fragment of the tub2 partial gene of the C. ocimi strains, but not of the other species listed in Table 1. The specificity of the primers was also confirmed by presence of a single dissociation peak in the melting curve at 81.84 ± 0.22°C (Figure 3). The DNA of C. ocimi, serially diluted from 10 ng µL -1 to 100 fg µL -1 in sterile distilled water, was used to build a standard curve to evaluate the analytical sensitivity of the detection method (Figure 4, Supplementary Table 1). The LOD of the test was at 1 pg µL -1 (Ct = 36.45 ± 0.44). The correlation coefficient (R 2 ) between the cycle threshold (Ct) and the initial concentration of genomic DNA was >0.99. The mean value of the regression slope was -3.22, and the mean relative efficiency was 104%, which showed good qPCR efficiency (Adams, 2006). The presence of plant DNA together with fungal DNA had no influence on the selectivity of the primers in the SYBR green assay (Supplementary Table 2). A similar PCR efficiency and a reliable correlation between the Ct values and the amount of DNA of C. ocimi was found from the amplification of C. ocimi DNA of strain CVG190 diluted in basil genomic DNA, compared to amplification of the samples diluted in water. The assay was performed on different days by two different operators and its reproducibility was confirmed.  Table 1. N = negative control. B) End-point polymerase chain reaction (PCR) specificity of the primer set TubOc_23fw TubOc_190rev. The numbers of the tested strains correspond to those reported in Table 1. N = negative control.

Detection of Colletotrichum ocimi in artificially inoculated basil leaves
The SYBR green assay was used to detect and quantify C. ocimi in symptomatic basil leaves. The assay detected presence of C. ocimi in eight of the nine samples tested. In seven samples, the pathogen was detected at 0.247 to 0.001 ng µL -1 . One sample was below the LOD of the test (Ct = 38.26 ± 0.72). DNA samples obtained from non-inoculated basil leaves were negative for the presence of C. ocimi.

Detection of Colletotrichum ocimi in artificially inoculated and commercial basil seeds
The SYBR green assay was used to detect and quantify C. ocimi in artificially inoculated and in commercial basil seeds. For the artificially inoculated seeds, the first assay was conducted with DNA of basil seeds directly extracted after inoculation. The Ct values determined from four samples (one replicate each) were below the LOD of the assays (Ct = 37.13;38.20;39.03 or 38.15). For all other replicates, the assay was negative for the presence of C. ocimi. A second assay was therefore conducted on DNA of seeds extracted after inoculated seed washings. This second assay was detected presence of C. ocimi in four out of 18 tested samples. In two samples, the pathogen was found in a range of 0.004 and 0.001 ng µL -1 . The other two samples were below the LOD (Ct = 37.73 ± 0.09 and Ct = 38.93 ± 0.26). DNA samples obtained from non-inoculated basil seeds were confirmed as negative for the presence of C. ocimi. The assay was negative for the presence of C. ocimi in the commercial basil seeds sampled. DISCUSSION Molecular techniques are reliable and effective for the specific detection and quantification of plant pathogens (Mirmajlessi et al., 2015). The present study describes the development of end-point PCR and SYBR Green real-time qPCR assays for the specific detection of C. ocimi in basil leaves and seeds. Both techniques were used to assess their analytical specificity and sensitivity. Selectivity, repeatability and reproducibility of the SYBR Green real-time qPCR assay were also determined. The amplification reactions were performed by increasing the annealing temperature to 64°C. With the obtained increased stringency, one primer set (TubOc_23fw -TubOc_190rev) still gave non-specific amplification. However, the amplification reaction with the other primer set (TubOc68fw -TubOc_197rev) allowed the specific detection of C. ocimi, avoiding the problem of cross-amplification. The LOD of the test was 1 pg µL -1 of C. ocimi DNA. This detection threshold was within the range of different assays that target other Colletotrichum spp. such as C. lupini (10 pg µL -1 ), C. godetiae and C. acutatum s.s. (10 pg µL -1 ) and C. theobromicola (1.4 pg µL -1 ) (Schena et al., 2017;Kamber et al., 2021;Kaur et al., 2021). Both analytical sensitivity and selectivity tests showed good qPCR efficiency within the established range of acceptability for developing a new detection method, from 90 to 110% (Adams, 2006).
This new assay can be applied to DNA directly extracted from fresh plant material and was not affected by co-extracted plant DNA. The assay allowed detection and quantification of C. ocimi in symptomatic artificially inoculated basil leaves. The pathogen was also detected from 0.1 g fresh weight of symptomatic leaves, in a range of 0.247 and 0.001 ng µL -1 . Future studies are planned to test the method also on naturally symptomatic leaves.
For seeds, the assay conducted with DNA directly extracted from seeds inoculated with conidial suspension of C. ocimi in glycerol permitted the detection of the pathogen on one replicate of four samples. The obtained values were positive, but below the LOD of the method, and since they were observed only in one replicate, the repeatability parameter of the assay could be affected (Cardwell et al., 2018). A second assay was therefore conducted washing inoculated seeds before DNA extraction to reduce glycerol residuals. This second trial gave detection the pathogen on four inoculated samples out of 18, and quantification was possible on two of these samples. The second extraction method was used for quantification of C. ocimi on a subset of commercial seeds selected for testing, where the pathogen was not detected since no amplification curves and no peaks in the melting curves were observed during the qPCR analyses.
Globalization of markets and the centralization of companies in some regions have contributed to long distance exchanges of plant materials and seeds. Due to this international trade, different pathogens can be transmitted via uncertified material, causing disease outbreaks in new areas (Gullino et al., 2014;Munkvold and Gullino, 2020). Specific pathogen detection methods are therefore required. The assay developed in the present study is the first molecular diagnostic tool developed to detect C. ocimi, and is a useful tool for the rapid identification and quantification of the pathogen on symptomatic basil leaves. For seeds, the method had a low detection success rate. Further research is therefore required to improve DNA extraction. This could be the most important step, due to the production of gum and presence of possible inhibitors within the basil seed coats or endosperm.
Traditional approaches to identify Colletotrichum spp. based on morphological and molecular data require isolation, DNA extraction and multi-locus sequencing, are time-consuming, and are applicable only to a restricted range of samples due to the high cost of the analyses (Bhunjun et al., 2021). Although a precise cost analysis should be carried out, SYBR Green realtime PCR with melting curve analysis is reported to be cost-effective, easy to use, and have optimal efficiency for small amplicons (Capote et al., 2012). For this reason, this technique was selected to develop the protocol described in the present study. The technique produced reliable and accurate diagnoses, and can reduce the necessary analysis time to a few days instead of weeks.
Colletotrichum ocimi can be transmitted via basil seeds, as reported in previous studies (Guarnaccia et al., 2019;Cacciola et al., 2020). Thus, detection of the pathogen on inoculated seeds, demonstrated in the present study, even at a low rate, confirms that this diagnostic tool could be useful for further investigations of commercial seeds, with an improved DNA extraction method. Furthermore, the end-point PCR assay could be used as a screening method to detect the presence or absence of C. ocimi, and to estimate its relative quantity through agarose-gel visualization.
No specific primers have been previously designed to detect C. ocimi. The developed and validated assay described here was shown to be specific for its target organism. Further studies are planned, aiming to improve detection of the pathogen on seeds or seedlings and to develop effective and preventive disease management strategies for basil production. ACKNOWLEDGEMENT This study was supported by PSR project "MONI-TORA". LITERATURE CITED