Collection of samples, isolation and morphologic characterization of strains

Tomato plants with symptoms of vascular wilt were gathered from four locations: two from the town of Maneadero and two from the town of San Quintín, in the municipal area of Ensenada, Baja California. In total, 60 tomato plants were collected, with different levels of vascular wilt symptoms. Five samples of the Cherry variety were taken from San Quintín Este, from an open-air plantation. Ten plants established in bags were taken from a greenhouse in San Quintín Norte, five of which were Cherry Zebra and five, Strawberry tomatoes. From Maneadero Sur, three greenhouses with plants grown in the soil were sampled; two Cherry tomato plants were taken from the first one, along with three Heirloom tomato plants. From the second one, five Hierloom samples were taken, and in the third one, five G5 samples, three of the RG-871 variety and one TC. Finally, in Maneadero Norte, 20 samples were taken of Cherry tomato plants grown in the open field.

The samples were labelled and placed in polyethylene bags and moved to the Phytopathology Laboratory of the Scientific Research and Higher Education Center of Ensenada (Centro de Investigación Científica y de Educación Superior de Ensenada, CICESE) for processing. The plants were washed with tap water, dried with paper towels and, on the stem, approximately 10 cm above the root crown, three cross sectional cuts were made and pieces of symptomatic tissue were taken. They were submerged in alcohol at 75% and quickly flamed. Later, using a heat-sterilized blade, 0.5 cm pieces were cut and placed in a potato-dextrose-agar (PDA) culture medium, supplementing with chloramphenicol (15 μg mL^-1), and the Petri dishes were incubated in the dark at 25 ± 2 °C. To obtain the pure cultures from the fungal growth found, seven days after harvesting, hyphae tips were transferred into a PDA medium and their putative identification was carried out using morphological observation. From those primary plantations, monosporic strains were obtained, which were kept in glycerol at -20 °C until their use.

The fungal isolates were grown in PDA and agar-carnation (CLA) (Fisher et al., 1982), two of the standard media used to identify Fusarium species (^{Leslie and Summerell, 2008}). The Petri dishes were stored at 27 °C in complete darkness, following recommendations for their identification. After 15 days, the morphology and pigmentation of the culture in PDA were analyzed, along with the production of sporodochia, sclerotia, reproductive structures and macroconidia in a CLA medium using an Olympus SZX12 stereoscopic microscope and an Axiovert 200 Zeiss^® inverted microscope.

Molecular characterization of races of F. oxysporum f. sp lycopersici

For the extraction of DNA, the strains were grown in a potato dextrose liquid, shaking at 27 °C for four days. The mycelium was recovered by filtration, frozen and lyophilized in 2 ml microcentrifuge tubes, and pulverized using glass rods with pointed tips. The genomic DNA was obtained using the CTAB method at 3% and finally, it was adjusted to 50 μg mL^-1.

To identify the races of Fol, two pairs of previously reported oligonucleotides (^{Hirano and Arie, 2006}) were mixed to carry out a multiplex PCR. The sp13-f (5’-GTCA GTCCATTGGCTCTCTC-3´) and sp13-r (5’-TCCTTGACACCATCACAGAG-3’) pair amplify a 445 pb fragment of the gene of one exopolygalacturonase (pgx4) in races 1 and 3, but not in race 2; the pair of sp23-f (5’-CCTCTTGTCTTTGTCTCACGA-3’) and sp23-r (5’-GCAACAGGTCGTGGGGAAAA-3’) was designed in the site of deletion of two nucleotides (nt 259-260) in the gene of an endopolygalacturonase (pg1) and it amplifies a fragment of 518 pb in races 2 and 3, but not in race 1. In the isolates of F. oxysporum f. sp. radicis-lycopersici, these oligonucleotides did not amplify any fragments (^{Hirano and Arie, 2006}). After testing several conditions of amplification by combining the oligonucleotides, the reaction mixture used consisted of 1X buffer (Axygen), 0.2 μg mL^-1de BSA, 0.2 mM of dNTPs, 2.5 mM of MgCl₂, 0.4 µM of each of the four oligonucleotides, 0.1 U μg^-1 of Taq polymerase (5 U, Axygen) and 50 ng of DNA in a final volume of 25 µL. Amplification consisted of a 94 °C cycle for 1 min, followed by 45 cycles at 94 °C for 1 min, 61 °C for 1 min, 72 °C for 1 min and a final extension at 72 °C for 10 min, using a BioRad T100 thermocycler. The products were evaluated by electrophoresis in agarose gel at 1% and photographed using a BioRad Quantity One photodocumenter. The positive control was DNA taken from race 1 Fol and race 2 Fol, kindly provided by Dr. Raymundo García of the Food Research and Development Center (Centro de Investigación en Alimentos y Desarrollo, CIAD), Culiacán.

After analyzing the PCR products, and to confirm that the amplified fragment matched the expected sequence, five isolates were chosen at random from the determined races. The amplified products were cut out of the gel and purified using the DNA Clean and Concentrator kit (Zymo Research), following the instructions by the manufacturer and the fragments generated were sent for sequencing to Eton Bioscience Inc. (San Diego, California). Once the sequences were obtained, they were compared using the Basic Local Alignment Search Tool (BLAST) algorithm from the National Center for Biotechnology Information (NCBI). Sequences from the fragments of pgx4 and pg1 were recovered and they were all aligned using the program BioEdit v7.2.3 (copyright (c) 1997-2013, Tom Hall).

To confirm that the isolates obtained and analyzed with the multiplex PCR were from F. oxysporum, some of them were selected and a fragment of approximately 700 pb from the TEF1-α region was amplified by PCR, using oligonucleotides EF-1 and EF-2 (^{O’Donnell et al., 1998}), used earlier to distinguish between species of Fusarium (^{Lievens et al., 2009}). The reaction of the PCR was similar to the one used to amplify the polygalacturonase genes, but the concentration was reduced to 1.5 mM of MgCl₂. The amplification consisted of an initial denaturalization of 94 °C for 3 min, followed by 40 cycles of 1 min at 94 °C, 1 min at 61.5 °C, 1 min at 72 °C and a final extension of 10 min at 72 °C. The fragments were evaluated by electrophoresis in agarose gel, purified and sequenced as described earlier. For the comparative analysis of the sequences, we accessed the Fusarium-ID (http://isolate.fusariumdb.org) and NCBI databanks, using the BLAST algorithm. Some sequences that showed the maximum similarity with those obtained in this work were recovered from the genebank, and they were all aligned using the program BioEdit v7.2.3. Finally, the phylogram was obtained using the program MEGAX (^{Kumar et al., 2018}).

Identification of races of Fol with the use of differential varieties

Four tomato genotypes were used to confirm the identity of the races determined by multiplex PCR: Bonny Best (without resistance), Manapal (resistant to Fol race 1), Walter (resistant to Fol race 2) and I₃R₃ (resistant to Fol race 3) (^{Ascencio-Álvarez et al., 2008}).

Seeds of each variety were planted in polyethylene trays with 72 pits in a 2:1 mixture of Peat moss-perlite and after 30 days, they were transplanted into 1 L polyethylene pots. Twenty-one days after transplanting, six plants of each variety were inoculated with one out of five of the isolates selected (the same ones used in the sequencing analysis), adding 10 mL of a suspension of 1X10^-7 conidia mL^-1 two centimeters from the stem (^{Baysal et al., 2009}). Fol strain races 1 and 2 were used as positive controls and the negative control consisted of 10 mL of water. The plants were kept under greenhouse conditions at a variable temperature of 36±16 °C, distributed in a completely random design. Damages were evaluated on a daily basis until the observation of symptoms in the control strains (approximately 15 days after inoculation), based on the arbitrary damage scale proposed previously (^{Vakalounakis and Fragkiadakis, 1999}), where: 0 = plant without symptoms, 1 = plant with slight wilting or similar to a lack of water, 2 = plant similar to scale 1 + yellow or dry leaves in 50% of the foliage, 3 = plant similar to scale 1 + yellow or dry leaves in 50% or more of the foliage and 4 = plant completely wilted. Each one of these scales represents the percentages of damage of 0%, 25%, 50%, 75% and 100%, respectively. The plants that displayed a percentage of vascular wilting of 20% or less were considered resistant to the strains inoculated.

Evaluation of the genetic variability of Fusarium spp. by randomly amplified DNA polymorphisms

To determine the inter and intraspecific variability between the isolates, we used the RAPD, or Random Amplified Polymorphic DNA technique (^{Assigbetse et al., 1994}; ^{Jiménez et al., 2001}; ^{Luna-Paez et al., 2004}). The size and number of fragments generated in strains of Fol race 1 and Fol race 2 were determined by using the universal oligonucleotides OPA-01 (5’-CAGGCCCTTC-3’), OPA-03 (5’-AGTCAGCCAC-3’), OPA-05 (5´-A GGGGTCTTG-3´), OPA-11 (5´-CAATCGCCGT -3´), OPA-15 (5’-TTCCGAACCC-3’), OPA16 (5’-AGCCAGCGAA-3’) and OPA 17 (5’-GACCGCTTGT-3’) (Operon Technologies Inc). The reaction mixture consisted of 0.75 mM of Taq master Mix (Mercury, Cat# 790005), 25 ng of DNA and 0.1 mM of one of the oligonucleotides in a final volume of 25 μL. The conditions for amplification were 94 °C for 2 min, followed by 35 cycles at 94 °C for one min, 36 °C for 1 min and 72 °C for 1.5 min, with a final extension of 10 min at 72 °C. The amplified fragments were analyzed in 1.5% agarose gels and their size was determined using the program Quantity one (1D- Analysis Software), with a 100bp molecular weight marker as a reference. Based on the results, the oligonucleotides selected were those which displayed different striping patterns to the amplicons obtained for each race, and they were used for the typing of the strains obtained in this study.

Morphological characterization of fungal isolates

In PDA medium, the isolates displayed four types of colony morphology, named A, B, C and D in this investigation. The isolates in group A (Voip-1 to Voip-5) presented white cottonlike mycelia, irregularly shaped, with a violet center in the top and bottom (Figure 1A). The isolates from group B (Voip2 to Voip4, Voip7, Voip10, Voip12, Voip13, Voip23, Voip25, Voip26 and Voip28), presented cottonlike mycelia, a filamentous edge with a faint violet color and the presence of yellow or colorless sporodochia (Figure 1B and 1E). The isolates from group C (RG-2, RG-4, RE-2, RE-4, SO-4, SO-5, SO-8, SO-9, SO-11, SO-13, Voip6, Voip8, Voip9, Voip11, Voip14 to Voip22, Voip24 and Voip27) presented abundant cottonlike mycelia with a filamentous shape and a circular edge with a faint violet color at the top and white to violet at the bottom (Figure 1C). Finally, the isolates in group D (RG-3, RE-3, SO-2, SO-7 and SO-15) presented cottonlike mycelia, a filamentous shape and edges, with white to cream color at the top and cream colored at the bottom, and the presence of yellow or colorless sporodochia and structures that seemed similar to sclerotia (Figure 1D and 1F).

In the CLA culture medium, with the exception of isolate SO-7, they all presented typical F. oxysporum structures (^{Leslie and Summerell, 2008}), including macroconidia (Figures 1G and 1K), microconidia (Figure 1G), monophyllids (Figura 1M), conidiogenic structures (Figure 1L-1O) and false heads (Figures 1L-1O).

Molecular identification of races of Fol

The fragments generated by multiplex PCR using the pairs of oligonucleotides sp13 and sp23, indicated the presence of Fol race 1 and Fol race 3 in Baja California. According to this analysis, 36 isolates out of a total of 45 corresponded to Fol race 1, found in the four areas sampled; seven to Fol race 3 and of the two remaining (SO-2 and SO-11) no fragments were amplified with any of the pairs of oligonucleotides used (Figure 2).

Figure 1 Characteristics of Fusarium oxysporum f. sp lycopersici isolates obtained from tomato. Cultures of representative groups A) Voip5, B) Voip23, C) Voip14, D) SO-15. E) Sporodochia, F) sclerotia observed in some cultures. Micro graphs of macroconidia (G-K) and false heads (L-P) observed in the isolates of the four Fusarium morphological groups and in F. solani. G and L) Voip5, H and M) Voip23, I and N) Voip14, J and O I) SO-15 E K and P) SO-7. 

In order to compare results, the amplified fragments were sequenced with the pair of oligonucleotides sp13 of pgx4 of isolates Voip4, Voip6, Voip8, Voip9 and Voip14, identified as Fol race 1 and isolates RG-3, RE-2, SO-13 and SO-15, identified as Fol race 3. When comparing the sequences obtained from pgx4 with the NCBI database, all isolates displayed a similarity of 99% with sequence AB256797.1 (^{Hirano and Arie, 2006}), which confirmed that they belonged to Fol race 1. Only in isolate SO-13 (race 3) were nucleotide polymorphisms found in positions 64 and 65 with sequence AB256797.1, with a G replacing A. On the other hand, the comparison of sequences of the fragments amplified with the pair of oligonucleotides sp23 of pg1 of the strains identified as race 3, RG-3, RE-2, SO-13 and SO-15 have a T instead of C in the nucleotide position 342 and do not present nucleotides A and T in sites 54 and 66, which would identify them as belonging to race 3 and they showed a similarity of 99% with sequence AB256794.1, which belongs to an isolate of F. oxysporum f. sp. tulipae and with AB256778.1 of F. oxysporum f. sp. melonis. RE-3, it displayed a G in site 342, differentiating from the rest of the isolates.

Amplification of the elongation factor (TEF1α)

Based on the analysis of the multiplex PCR, the TEF1α of the following isolates were amplified: Voip4, Voip6, Voip8, Voip9, Voip14, SO-7 (Fol race 1), RG-3, RE-2, SO-13, SO-15 (Fol race 3), SO-2 and SO-11 (not identified by the multiplex PCR). In the majority, fragments of around 700pb were amplified, except for SO-11 (around 650 pb) and SO-2, where no amplicons were obtained, despite having checked the integrity of the DNA and having attempted the PCR several times. When analyzing the amplified sequences of TEF1-α, we verified that all isolates, excluding SO-7, belong to the species F. oxysporum. Isolate SO-7 was grouped with the sequences of F. solani (KT357542.1 and KT357549.1) (Figure 3).

Figure 2 Fragments generated by the pairs of oligonucleotides sp13 (which amplifies a fragment of 445 pb of the gene pgx4 in races 1 and 3) and sp23 (which amplifies a fragment of 518 pb in gene pg1 in races 2 and 3) in Fusarium oxysporum isolates. A fragment of approximately 445 pb was obtained in isolates RG-2, RE-4, SO-4, SO-7, SO-8 and SO-9. Meanwhile, isolates RG-3, RG-4, RE-2, RE-3, SO-5, SO 13 and SO15 generated two fragments of ap proximately 445 and 518 pb. Isolates SO-2 and SO-11 showed no fragments. 

Identification of races using different tomato cultivars

The pathogenicity tests of the Fol isolates in tomato plants showed the susceptibility of the varieties according to the variety and the isolate of the fungus. In the Bonny Best cultivar, all isolates caused wilting. Voip9 caused the most damage (41%), followed by Voip14, RE-2, SO-13 and Fol R2 (38%); Voip-8, RG-E, SO-11 and Fol R1 (32%) and Voip4 and RE-3 (29%). The least damaging (25%) were isolates Voip6, SO-15 and SO-7. Plants not inoculated presented no damage (Figure 4a).

In Manapal, isolate Voip9 caused the most damage (54%), followed by SO-15 (41%); Voip14, RG-13, RE-2 and SO-13 and Fol R2 (33%); Fol R1 (29%); RE-3 and SO-7 (25%), Voip4, Voip6, Voip8 and SO-11 and the control plants (9%) (Figure 4b).

In Walter, all isolates produced over 10% of vascular wilting, except Fol R2, which caused 9%; Voip8, Voip9 and Fol R1 (45%); Voip14, RE-3, SO-7 (41%); RG-3 y SO-15, (37%); RE-2 and SO-11 (33%); SO-13, Voip4 and Voip6, less than 20% (Figure 4c). In I3R3, isolate Fol R1 produced the least wilting (58%), followed by Voip8, Voip14 and SO-13 (53%); RG-3 (50%); RE-2 (45%), Voip9 (41%), Vop6 and RE-2 (38%); SO-15 (32%), SO-11 (25%), Voip4 (19%), SO-7 (8%) and RE-3 (3%) and the control (8%) (Figure 4d). Table 2 shows a summary of the results obtained in the tests for the identification of races according to comparisons between the presence of fragments generated by the sets of race-specific oligonucleotides, the sequences obtained and the in planta tests.

Genetic variability of isolates of Fol (RAPD)

Oligonucleotides OPA-01, OPA-15, OPA-16 and OPA-17 produced no fragments in Fol races1 and 2, and were therefore discarded for the analysis of variability between races used as a type. The approximate number and sizes of the fragments observed with OPA-03 were eight in Fol race 1, sized 1430, 519, 691, 752, 988, 1380, 1413 and 1926 pb, and four in Fol race 2, sized 519, 717, 1000 y 1430 pb. With the OPA-05 primer, three fragments of approximately 790, 954 and 1430 pb in Fol race 1 were observed, along with four in Fol race 2 sized 636, 790, 954 and 1616 pb. Finally, with primer OPA-11, four fragments were obtained in Fol race 1 with approximately 664, 776, 1273 and 1660 pb, and in Fol race 2, five fragments sized 430, 607, 725, 1284 and 1637 pb (Figure 5A).

In the evaluation of the isolates obtained in this investigation and using oligonucleotide OPA-05, no specific patterns were found (data not shown), therefore its use was discarded. With OPA-11, seven stripe patterns were obtained, the first of which displayed four fragments sized 664, 776, 1273 and 1660 pb, and it was found in 29 of the isolates obtained, identified as Fol race 1 in the previous tests (Voip1-Voip16, Voip18-Voip26, Voip28-Voip30 and SO-11). The second pattern, found in strain Voip15 consisted of two fragments, sized 730 and 1266 bp. The third, with four fragments sized 508, 1343, 1730 and 2134, was found in isolates RG-3 and SO-4. The fourth, sized 745 bp, was found in RG-4. The fifth was sized 500, 775, 871, 1211, 1343, 1645 and 2175 bp in size and found in RE-2, RE-3 and RE-4. The sixth had five fragments sized 742, 1160, 1312, 1632 and 2202, and the seventh, with three fragments, 886, 1166 and 1525 in SO-9. No specific striping pattern was noticed using OPA-11 in the strains that were identified as race 3 using the amplification of polygalacturonase genes (RG-3, RE-2, SO-13 and SO-15) (Figure 5B).

Figure 3 Phylogram using the Neighbor-Joining method, based on sequences of the TEF1-α region with 500 replicas. There was a total of 591 positions in the final set of data. The evolutionary analyses were carried out in MEGA X (^{Kumar et al., 2018}). The tree is drawn to scale, with the length of the branches in the same units as the evolutionary dis tances used to infer the phylogenetic tree. Evolutionary distances were calculated using the Maximum Composite Likelihood method. 

Figure 4 Percentage of wilting caused by isolates Voip4, Voip6, Voip8, Voip9, Voip14, RG-3, RE-2, RE-3, SO-13, SO-15, SO-7 and SO-11, in four different tomato cultivars a) Bonny Best, susceptible to the three races, b) Manapal, re sistant to race 1, c) Walter, resistant to races 1 and 2, and d) I3R3, resistant to race 3.