MATERIALS AND METHODS

Native maize germplasm. For this research, 55 native maize populations, previously collected and characterized (^{Briones-Reyes et al. (2015}), were evaluated. The populations were collected at 19 sites in five states located between 1498 and 2520 masl: Montaña de Guerrero region (13), southeast State of Mexico (11), Mixteca Alta of Oaxaca (13), Puebla (6) and Tlaxcala (12) (Table 1). The maize collections were kept in a cold room at -4 °C since the date of collection until they were used.

Source of inoculum and seed inoculation. The source of inoculum were teliospores collected in 2014 at a maize lot (hybrid Cardenal, Asgrow^®) in the municipality of Tenango del Valle, Mexico. The teliospores were removed from sori, passed through a 117 micron sieve (Mont Inox^®), stored in plastic containers with envelopes containing CaCl₂ (J. T. Baker and Macron Fine Chemicals^® 94%), and kept at 20 ± 2 °C until they were used. The inoculum was morphologically characterized using the methodology of ^{Vánky (2012}), and molecularly characterized based on the methodology proposed by ^{Márquez-Licona et al. (2018}). The morphological characterization of the inoculum was conducted by observing pathogen’s teliospores placed on slides. For this procedure, 100 teliospores were measured and characterized with a 100 X compound microscope (CX31RBSFA, Olympus^®). The molecular characterization was performed using monobasidial cultures and amplifying the ITS region (Internal Transcribed Spacer), as well as part of the LSU rDNA region (Large Subunit Ribosomal DNA), using ITS1F/NL4 primers (^{Gardes and Bruns 1993}; ^{O’Donnell et al., 1998}).

The extraction of total DNA of the monobasidial cultures was made with the Ultra Clean^® Microbial DNA Isolation Kit commercial kit (MoBio Laboratories Inc). The DNA integrity was verified in 1% agarose gel and the amount of DNA was determined by spectrophotometry at 260 nm absorbance (Nanodrop^® ND-1 V 3.2.1). For amplification, a mixture was prepared using 25 μL, buffer My Taq 1X, 10 pm of each primer, 1U MyTaq™ DNA Polymerase (Bioline Germany) and 100 ng of DNA. The amplification was carried out in a Mastercycler Pro thermocycler (Eppendorf^®) at 94 °C initial denaturation temperature for 1 min, 55 °C for 1 min and 72 °C for 1 min, plus one cycle at 72 °C for 10 min for the final extension. The amplified fragments were visualized in 1% agarose gel electrophoresis at 120 volts for 30 min. The size of the amplified fragment was estimated with the Pst I marker (Microzone^®). Once electrophoresis ended, the gel was analyzed with a Chemi Genius 2 Bio Imaging System photodocumenter (Syngene^®). The PCR products were purified as described by ^{Kirby (1965}) and sequenced in Macrogen Europe (Amsterdam, The Netherlands). The quality of the sequences was determined by observing the electropherograms with the 4peaks^® Nucleobytes software (Griekspoor and Groothuis, 1994), and U GENE (^{Okonechnikov et al., 2012}) for the consensus sequence. The obtained sequences were compared to the records in the GenBank database of the National Center for Biotechnology Information (NCBI) using the Blastn tool (^{Altschul et al., 1990}). The sequences were stored in the same database.

The seed viability was verified before inoculating the seed, as described by ^{Quezada-Salinas et al. (2013}). For inoculation, the teliospores were adhered to the seed using a 10% aqueous grenetine solution at 20 °C (^{Márquez-Licona et al., 2018}). The seeds (66 of each population, 22 per replication) were immersed in the grenetine solution for 1 min, recovered in a Petri dish and completely covered with teliospores (~40 000 teliospores/seed). The inoculated seeds were dried for 24 h at 22 °C. Seeds of the hybrid Az 41801^®, inoculated and non-inoculated, were used as the control.

Evaluation of the response of native maize. The response of native maize populations to S. reilianum f. sp. zeae infection was evaluated by planting them in a lot at Ejido Cinta Larga, Mixquiahuala, Hidalgo (20° 11’ 24.6” N 99° 14’ 35.2” W; 2030 masl), with semi-dry and temperate climate, in the 2015 and 2016 spring-summer cycles, using the same randomization. In both evaluations, the native maize populations, the plot where the evaluation was conducted, the source of inoculum and the percent of germinated teliospores were stable. The inoculated seed of each population was sown in 3-m long rows and 0.80 m between rows, at 5 cm depth, separated 13.5 cm between seeds. The experimental unit consisted of one 22-plant row with 3 replications. The experiment was established in a completely randomized design. After planting, flood irrigation was applied to the plot (an irrigation sheet of 140 mm), and subsequent irrigation was applied at stages V5, V7, VT, R2 and R5. The lack of moisture between the first and second irrigation increased the percent of head smut incidence, as reported by ^{Matyac and Kommedahl (1985}). As part of the agricultural practices of soil removal after sowing, only one cultivator step was carried out in stage V4. Insect pests and weeds were controlled according to the regional technological package, this is, using chemical products that are sold in the region.

The disease incidence in the populations was recorded 45 days after flowering started. A plant with symptoms in the tassel or the ear, or both, was considered as an infected plant. The disease incidence was determined as the quotient of the number of plants with symptoms and the total of plants in the experimental unit multiplied by 100. Based on the incidence values, the populations were grouped according to the scale proposed by ^{Quezada-Salinas et al. (2017}), which considers the following ranges: 0-10%=highly resistant, 11-25%=moderately resistant, 26-50%=moderately susceptible, 51-75%=susceptible, and 76-100%=highly susceptible. To obtain the disease incidence, a combined analysis of variance of the two years was conducted, where the model also included the State factor, and the genotypes were nested within the State and the year×gen Interaction (state). The base 10 logarithm transformation was applied to the incidence response variable. The tests were conducted under the logarithm transformation to keep the statistical validity of the analysis of variance, but the mean values were expressed in the original units. The comparison of multiple means for years and states was made using the Least Significant Difference method, and Tukey’s Honest Significant Difference method to compare multiple means adjusted by minimum squares (lsmeans), considering a level of 5% significance in both cases. The statistical analyses were conducted using SAS^® Systems for Windows V 9.4 (^{SAS Institute Inc., 2013}) statistical software.

RESULTS AND DISCUSSION

Source of inoculum and seed inoculation. When the inoculum was morphologically characterized, teliospore clusters 73 µm in diameter were observed. The teliospores (10-12 x 11-13.5 µm) were globose-to-subglobose, mainly dark brown in color, with ornamentations. The previously described characteristics are in agreement with those reported by ^{Vánky, 2012} for Sporisorium reilianum f. sp. zeae species. In the molecular characterization of the inoculum, the comparison of the sequences obtained in this study (KY856895, KY856896) using the Blastn tool (^{Altschul et al., 1990}) showed 99% identity with a fragment corresponding to the ITS region of the full genome of S. reilianum f. sp. zeae deposited in the GeneBank (NCBI) by ^{Schirawski et al. (2010}), thus confirming the identity of the inoculated pathogen. Of the teliospores used as a source of inoculum, 48% germinated in acidified PDA medium after 24 h incubation in darkness at 25 °C, thus demonstrating the inoculum viability, which exceeded the viability percent reported in other investigations such as those of ^{Osorio and Frederiksen (1998}), ^{Potter (1914}) and ^{Quezada-Salinas et al. (2013}). In this study, the germinated teliospores formed a four-cell septate basidium, and each cell produced a hyaline, unicellular and subglobose basidiospore, which produced a great number of sporidia that then formed yeast-type colonies, cream in color (Figure 1 A-B), which is in agreement with what ^{Márquez-Licona et al. (2018}) reported.

Evaluation of the response of native maize. In 2015, from March to September the average temperature value was of 17.7 °C, with average rainfall of 67.5 mm, and average relative humidity of 72.6%. In 2016, during the same period, the temperature average value was of 16.8 °C, average rainfall of 101.2 mm and average relative humidity of 74.8%. Thirty days after sowing, when seedlings infection occurs, the average rainfall was 41 mm higher in 2016 than in the previous year. Forty five days after flowering started, plants of the non-inoculated control remained healthy, thus demonstrating the absence of inoculum in the soil, while the plants from inoculated seed showed formation of sori filled with teliospores which totally or partially replaced the tassel and the ear (Figure 1 C-F), as mentioned by ^{Xu et al. (1999}), which caused a marked reduction in pollen and grain production, as described by ^{Ghareeb et al. (2011}).

In 2015, 37 of the native maize populations showed head smut symptoms in at least one of the replications, and the disease incidence in the 55 native maize populations ranged from 0 to 28.8%. Of the 55 native maize populations, 72.7% had incidence ranging from 0 to 10%, 21.8% incidence from 11 to 25%, and 5.5% of the collections from 26 to 50% incidence. In the same evaluation, 100 and 83.3% of the native maize populations from the States of Mexico and Tlaxcala, respectively, were highly resistant to S. reilianum f. sp. zeae,, while populations 19 and 21 from Guerrero, and population 30 from Oaxaca, reached a level of incidence higher than 26%, so they were classified as moderately susceptible. For the 2016 cropping cycle, only 21 of the 55 native maize populations showed head smut symptoms in at least one of their replications, and the disease incidence in the 55 native maize populations ranged from 0 to 22.2%. Out of the 55 native maize populations, 94.5% showed incidences ranging from 0 to 10%, and 5.5% from 11 to 25%. In the same evaluation, 100% of the native populations from the States of Mexico, Tlaxcala and Oaxaca were highly resistant to S. reilianum f. sp. zeae, while populations 19 and 24 from Guerrero and population 16 from Puebla reached a level of incidence ranging from 11 to 25%, so they were classified as moderately resistant (Table 2). In the second evaluation year (2016), the disease incidence in the populations was lower than in the previous year. The same happened to the inoculated control (hybrid AZ 41801^®), where the disease incidence decreased by 28.4% between the first and the second evaluation. This may be attributed to a soil humidity increase due to higher rainfall in the second evaluation year. These results suggest that a humidity increase after sowing affected the fungus ability to infect maize seedling roots, as mentioned by ^{Matyac and Kommedahl (1985}).

Figure 1  Sporisorium reilianum f. sp. zeae. A) Teliospores germinated in PDA 24 h after sowing. B) Monobasidial culture in PDA 5 days after sowing. C) Partial replacement of male inflorescence. D) Symptoms of phyllody in infected plants. E) Male inflorescence atrophy. F) Total replacement of ear by sori filled with teliospores. 

The high values of disease incidence in the inoculated hybrid, compared to the incidence recorded for native maize populations, may be due to the fact that no disease resistance genes have been incorporated into the selection process of the hybrid parent lines, while the genetic diversity present in the native maize collections is broader and can express certain levels of resistance, as mentioned by ^{Fehr (1993}) and ^{Hallauer et al. (2010}). For this reason, keeping maize genetic diversity is essential for future genetic development, since the lack of genetic diversity may compromise the possibility of developing high-yielding materials better adapted to adverse environmental conditions, as well as the development of disease-resistant germplasm (^{Giordani et al., 2019}).

Regarding the combined analysis of variance of head smut incidence in tassel in 55 maize genotypes evaluated for two years, there were highly significant differences in the main effects of year, state and genotype, but not in the year × genotype interaction (Table 3). The results show that the populations (genotype), the origin and the evaluation year influenced the incidence of head smut in maize tassel. Significant statistical differences of the same type were found by ^{Briones-Reyes et al. (2015}) when the authors evaluated the response of the same native maize populations to ear rot caused by Fusarium spp. under natural infection conditions. The differences among the populations and their origin show the phenotypic and genotypic diversity present in the variants of the existing native maize in the states of origin. The significant differences in disease incidence from one evaluation year to another, where the only source of variation was the amount of soil humidity, suggest that there is an inversely proportional relation between soil moisture and disease incidence, which is in agreement with the results obtained by ^{Matyac and Kommedahl (1985}).

The comparison of multiple means for years and states, using the Least Minimum Difference test and the logarithm transformation of the variable response to stem head smut incidence, divided the populations into four defined groups (Table 4). The first group included populations from Tlaxcala and the State of Mexico, which had a lower percent of disease incidence. The second group was formed by populations from Oaxaca and Puebla. The third group included only populations from Guerrero, and the last group was the inoculated control, which had the highest incidence mean in the study.

Table 5 shows the results of the comparison of means of the genotypes nested in the states and through the two years. In this analysis, 8 of the 11 populations of native maize from the State of Mexico were not affected by the pathogen in none of their replications, in both evaluations, while 6 of 13 populations from Guerrero had an intermediate incidence equal to or higher than 10%. The resistance of the maize populations from the State of Mexico could be explained by the fact that the disease was first reported in Amecameca (^{Borlaug, 1946}) in teocintle plants and that the maize crops in the region may have been exposed to the pathogen, thus occurring the coevolution between the pathogen and the host. On the other hand, the maize populations from Guerrero were the most susceptible to the disease, and this result is in agreement with our initial hypothesis, which is based on observations in the field, where a greater disease susceptibility has been detected in hybrids developed using maize lines from the state of Guerrero (Tuxpeño race) (Dr. Fernando Castillo González, personal communication). Conversely, when the same 55 populations were used to evaluate their response to ear rot caused by Fusarium spp., the native maize populations from the State of Mexico showed higher susceptibility to ear rot, as reported by ^{Briones-Reyes et al. (2015}).

Considering the scale used to classify the populations and the mean separation shown in Table 5, it can be observed that 48 of the 56 evaluated genotypes are within the range of 0-10% incidence, and, therefore, they are considered as highly resistant to the disease. Seven of the evaluated genotypes are within the range of 11-25% incidence, so they are considered as moderately resistant to the disease, and only the inoculated control, which was within the range of 51-75% incidence, was classified as susceptible to head smut. Previous results show that the diversity present in the evaluated maize populations would allow the selection of disease-resistant germplasm. Similarly, the variation of the response of the evaluated materials to the disease makes it clear that the type of resistance to head smut is quantitative. This is demonstrated by the selection method (S1 recurrent) used by ^{De León (2020)} to develop open-pollinated varieties resistant to the disease.

The information obtained in the present study is the first exploration of the presence of disease resistance genes in native maize populations, which will allow the disease resistance improvement programs to exclude the populations with a higher incidence percent in both evaluations. The information obtained in this study is meant to encourage the rescue and use of maize native populations, as well as launching improvement programs to develop maize lines genetically resistant to the disease, as ^{Quezada-Salinas et al. (2017}) have done, and the subsequent development of high-yielding open-pollinated varieties and genetically resistant maize varieties, as the open-pollinated variety CP-Vero1 developed by ^{De León (2020)}. Finally, it is suggested to continue evaluating native maize germplasm including a greater number of populations classified by race, conducting simultaneous evaluations under different environments, and molecular studies, to ensure the presence of resistance genes in the populations to be used as possible sources of resistance to head smut.

CONCLUSIONS

There is variation in the response of native maize germplasm to S. reilianum f. sp. zeae. infection depending on its geographical origin. The variation of the populations response to the infection suggests the presence of genes for resistance to the pathogen in the populations that were not affected by the disease. The native maize from Guerrero was more susceptible to the disease, while maize collected in the States of Mexico and Tlaxcala was more resistant. The high-soil humidity levels in the first development stages of the crop reduces the disease incidence.