Microsatellite markers in analysis of resistance to coffee leaf miner in Arabica coffee

The objective of this work was to analyze coffee (Coffea arabica) genotypes resistant to the coffee leaf miner (Leucoptera coffeella) using microsatellite markers. Sixty-six loci were evaluated, of which 63 were obtained from the Brazilian Coffee Expressed Sequence Tag (EST) database. These loci were amplified in bulks of individuals from F5 progenies of 'Siriema' (C. arabica x C. racemosa) resistant and susceptible to the insect, in eight samples of C. racemosa, and in a F6 population of 'Siriema' with 91 individuals segregating for resistance to the leaf miner. Polymorphisms were verified for two simple sequence repeat (SSR) loci in bulks of the susceptible progenies. The two polymorphic alleles were present in around 70% of the susceptible genotypes in F5 and in approximately 90% of the susceptible individuals in F6. However, the polymorphic EST‐SSR markers among populations contrasting for resistance to leaf miner were not correlated to the evaluated characteristics. SSR markers show inter‐ and intraspecific polymorphism in C. arabica and C. racemosa.


Introduction
The coffee leaf miner [Leucoptera coffeella (Guérin-Mèneville)] is considered the main pest in Brazilian coffee plantations.This microlepidoptera reduces the photosynthetic area of leaves, causing large losses in coffee production (Souza et al., 1998).Up to the present, commercial cultivars of Arabica coffee (Coffea arabica L.) registered in Brazil are susceptible to the coffee leaf miner, making chemical control of this pest indispensable.The use of resistant cultivars is an advantageous alternative for the control of biotic stresses.This strategy promotes a greater competitiveness for farmers, since it allows greater yields by reducing losses caused by the attack of pests and diseases, production costs and pollution of the environment by the application of agricultural chemicals.
The Coffea genus has 103 species (Davis et al., 2006) with 2n=22 chromosomes.The exception is C. arabica, an allotetraploid, with 44 chromosomes, of which 22 came from C. eugenioides S. Moore and the rest from C. canephora Pierre ex A. Froehner (Lashermes et al., 1999).C. arabica has a narrow genetic base due to the small quantity of seeds originally introduced in Brazil associated with the autogamous nature of this species.In addition, the predominance of mass selection and pedigree methods in breeding strategies has also contributed to reduce genetic diversity in C. arabica.
The resistance to L. coffeella in Coffea spp. was identified in the Instituto Agronômico (IAC) coffee breeding program in a fertile plant derived from a natural cross between C. racemosa Ruiz & Pav. and the Blue Mountain cultivar of C. arabica.Hybrid populations obtained from the combination of backcrossing and pedigree methods are established in the field, but these populations are still segregating for this trait (Guerreiro Filho, 2006).
The leaf miner resistance introduced from C. racemosa is controlled by two dominant complementary genes, Lm1 and Lm2 (Guerreiro Filho et al., 1999).However, the difficulty in stabilizing this trait in the genotypes indicates the need for more studies.So far, molecular studies of leaf miner resistance did not help to determine the nature of this resistance in Arabica coffee plants (Mondego et al., 2005;Pinto et al., 2007).
Microsatellite or simple sequence repeat (SSR) markers are based on repetitive DNA sequences and occur in the genome of all organisms.These markers are highly variable, and multiallelic genetic loci stand out for codominance and high content of polymorphism information (Tautz & Renz, 1984;Gupta & Varshney, 2000;Morgante et al., 2002).Microsatellites can be scored by many methods, which include derivation from enriched genomic libraries and screening of cDNA libraries from public databases, such as GenBank, from related species and from EST (expressed sequence tag) databases.EST-SSR have some intrinsic advantages in that they are quick to elucidate -by electronic sorting -, abundant, unbiased in repeat type and present in gene-rich areas (Scott, 2001).The Brazilian Coffee EST database has more than 200 thousand DNA sequences (Vieira et al., 2006).
The objective of this work was to analyze C. arabica genotypes resistant to the coffee leaf miner (Leucoptera coffeella) using microsatellite markers.

Materials and Methods
The genetic materials were obtained from the experimental fields of Coromandel and Boa Esperança, MG, Brazil, of the Fundação Procafé and from the Centro de Café of Instituto Agronômico (IAC), Campinas, SP, Brazil.Molecular analyses were done at the Laboratório de Análise de Sementes e Biotecnologia of the Universidade Federal de Lavras, Lavras, MG, Brazil.
After the C. arabica x C. racemosa cross, the material passed through two backcrossings with C. arabica at IAC, and the segregant IAC 1195-5-6-2 was selected.C1195-5-6-2-119, with high resistance to coffee leaf miner, was selected by the Procafé breeding program and crossed with 'Catimor UFV 417' to incorporate rust resistance.From this cross, segregant 842 was selected, with resistance to rust and to leaf miner.Plants 842-1 and 842-4, designated 'Siriema', were advanced in their generations by the pedigree method, and currently are in the F 5 generation (Carvalho et al., 2008).
The 'Siriema' F 5 progenies were in the field for five years under evaluation for tolerance to the coffee leaf miner.The F 6 population was characterized at the IAC greenhouse (Guerreiro Filho et al., 1999).Lesion reaction was observed in leaf disks infested by L. coffeella, according to four classes: resistant, plants with point lesions; moderately resistant, plants with small threadlike lesions; moderately susceptible, plants with large threadlike lesions; and susceptible, plants with large irregular lesions.
All the primers were screened in the C. racemosa and F 5 progenies.The polymorphic oligonucleotides were used to amplify DNA fragments of each individual of the F 5 progenies and of the F 6 individuals of the population.Genomic DNA was extracted according to Ferreira & Grattapaglia (1996) by the CTAB 2% method.Final reaction conditions were 1X PCR buffer (Tris-HCl, 200 μmol mL -1 ; pH 8.4; KCl, 500 μmol mL -1 ); 2 μmol mL -1 of MgCl 2 ; 400 nmol mL -1 of dNTPs; 0.3 nmol mL -1 of each primer; 0.5 U of DNA Taq polymerase; and 30 ng of DNA.The PCR cycles were: 94 o C, for 3 min, followed by 30 cycles of 94°C for 1 min; annealing temperature from 55 to 57ºC for 1 min, in touchdown, 72 o C, for 1 min, and final extension, at 72 o C, for 5 min.Amplified products were separated through polyacrylamide gel electrophoresis.
Number and size of SSR amplification products were characterized.Qualitative analyses based on the presence or absence of SSR alleles were performed for characterization of polymorphism between segregating materials.
In silico EST analyses performed on leaves and fruits of C. canephora (Poncet et al., 2004) showed that the trinucleotide microsatellites are also the most abundant, followed by the di-and hexanucleotides, representing 34.34, 25.75 and 22.04%, respectively.The significant presence of mononucleotides was also reported.
In comparison with some of the other approaches, EST-SSR may be slightly less polymorphic than genomic library-derived SSR, as there is pressure for sequence conservation in gene regions, reducing polymorphism (Scott, 2001).Selection of EST-SSRs with tri-and dinucleotide motifs was preferred to increase the probability of polymorphic EST markers being identified.Out of the 35 EST-SSR loci randomly selected, fragment amplification was detected in 27.Among these, eight loci were polymorphic, 18 were monomorphic, and, in the others, amplification was not observed or they showed diffuse bands (Table 1).
A total of 63 alleles were amplified and their sizes varied from 120 to 350 bp.These alleles showed amplified fragments with the expected size.The exception was LEG1 that amplified an allele with 600 bp that could be related to an intron.In most loci, two alleles were detected, but loci with three or four alleles were also found, which is expected for tetraploid species, such as C. arabica.
The LEG9, LEG11, LEG12, LEG13, LEG22, LEG26, LEG28 and LEG32 loci were polymorphic only among C. racemosa accessions and C. arabica progenies.Therefore, genetic variability is expected between these two species.The search for polymorphic SSR loci associated with a desirable trait is complex due to the limited number of coffee SSRs available.In species with a narrow genetic base, such as C. arabica, the probability of finding polymorphism is low (Baruah et al., 2003;Vieira et al., 2006).Genetic diversity among coffee species identified by EST-SSR markers can be used as a gene source in breeding programs, with potential for other studies of genetic mapping, in the identification of cultivars or in pre-breeding by the genotyping of germplasm bank accessions.
Polymorphism among the susceptible and resistant genotypes to the coffee leaf miner was observed in loci Sat 225 and Sat 229.The Sat 259 locus was monomorphic in all the evaluated genotypes.The amplification profile of the Sat 225 locus corresponds to a 300 bp allele (Figure 1).This allele was present in both susceptible and resistant bulks and also in C. racemosa accessions.As the C. racemosa parent no longer exists, a bulk of accessions from the IAC germplasm bank was used to represent the genetic diversity of this species.However, no polymorphic alleles associated with resistance to coffee leaf miner were detected.Another allele with approximately 270 bp was verified in the susceptible progenies (S4, S8, S15, S21, S24, S38, S42 and S48).For the Sat 229 primer, a 190 bp polymorphic allele was observed in the individuals of the S4, S8, S15, S21, S38 and S42 bulks.
After screening, bulks were opened and both Sat 225 and Sat 229 were evaluated in individuals from the F 5 Coromandel progenies.The Sat 225 amplification pattern was the same as that shown in the bulks.The 270 bp allele, potentially associated with susceptibility to the coffee leaf miner, was present in 73% of the susceptible individuals and absent in 91% of the resistant individuals.Likewise, the 190 bp polymorphic allele from locus Sat 229 was observed in 70% of the susceptible individuals and absent in 94% of the resistant ones.According to Carvalho et al. (2008), a 30 to 40% of susceptibility to L. coffeella in 'Siriema' F 5 progenies is commonly observed in the field.
The F 6 population was characterized by the evaluation of lesions in the infested leaf disks.A total of 37 individuals were classified as resistant and 54 as susceptible to the insect.Molecular characterization shows that the 270 bp polymorphic allele from the Sat 225 locus was found in 91% of the susceptible genotypes and was absent in 36% of the resistant ones.The 190 bp polymorphic allele from locus Sat 229 was observed in 91% of the susceptible individuals and in 48% of the resistant ones.Although a high amplification frequency is observed in susceptible genotypes, both 270 bp (Sat 225) and 190 bp (Sat 229) alleles were observed in resistant genotypes.However, no significant correlation among the molecular markers and the presence of lesions in coffee leaves infested by L. coffeella was detected.
Herrera et al. ( 2009) evaluated the Sat 225 and Sat 229 loci in a F 2 segregating population resulting from a cross between the susceptible C. arabica cultivar Caturra and the resistant female parental hybrid derived from the cross between C. arabica and C. canephora.These authors observed the occurrence of a variable amplification pattern during the evaluated years, which may be associated with the infestation pressure of the pathogen.Genetic analysis indicates that the two loci could be associated with leaf-rust resistance and that the additive effect would be essential for the expression of resistance.
In the present work, alleles associated with resistance to the coffee leaf miner were not identified.The frequency of plants with resistance in segregating populations has not increased, remaining near 35%, although some families show a higher percentage.The genetic control of resistance to the coffee leaf miner appears to be more complex.Therefore, other factors, besides the control determined by the two genes, as mentioned by Medina Filho (1977) and Guerreiro Filho et al. (1999), may be involved.
The use of markers as tools to assist coffee breeding programs may contribute to save time for the development of new cultivars.However, the coffee molecular markers presently available have shown limited polymorphism.Several strategies for the development of new coffee molecular markers, such as diversity arrays technology (DArTs) (Wittenberg et al., 2005) and single nucleotide polymorphism (SNPs), identified through comparative genomics or re-sequencing, using new generation sequencing, will certainly contribute to the development of more informative markers, which may be used to map valuable traits in a more precise manner.

Conclusions
1. SSR markers show inter-and intraspecific polymorphism in Coffea arabica and C. racemosa.
2. No polymorphic alleles are associated with resistance to coffee leaf miner in the evaluated SSR loci.

Table 1 .
EST-SSR loci selected from the EST Coffee Genome database.

Table 2 .
EST SSR primers designed from contigs available in the coffee EST database.