Introduction

Citrus fruits are grown in the tropical and subtropical regions of the world. In México, citrus agro-industry represents one of the most important, generating an economic outflow of more than 375 million dollars (^{SIAP, 2016}). Genetic improvement of this crop is a routine activity carried out in several countries with the aim of improving the quality of the fruit or to achieve tolerance to biotic and/or abiotic stress (^{Tozlu et al., 1999}; ^{Mendoza-Rodríguez et al., 2001}; ^{Machado et al., 2011}; ^{Omura and Shimada, 2016}).

Species of the genus Citrus present polyembryony and apomixis, generating disturbances in the process of sexual reproduction between vegetable organisms, in which multiple embryos are found within a seed or where several embryos derived from the pollen receptor of the parent are originated from nuclear tissue surrounding the zygotic embryo (^{Koltunow et al., 1996}; ^{Kepiro and Roose 2007}). Most commercial citrus worldwide propagate as grafted trees with the variety of interest in a rootstock (^{Wutscher and Hill, 1995}; ^{Khan and Kender, 2007}).

Most of these rootstocks are apomictic, so if it is required to maintain genetic homogeneity this condition may be advantageous as a process of clonal multiplication. In this way, uniform plants can be produced from seeds at low cost. On the other hand, when it is desirable to generate variability by genetic recombination for obtaining hybrids, with resistance to diseases for example, then apomixis represents an obstacle, since the plants are genetically identical to the pollen receptor parent (^{Khan and Kender, 2007}). For the identification of zygotic plants in citrus, phenotypic markers have been used, which are based mainly on leaf morphology.

Furthermore, isozyme tests and the use of flow cytometry have also been used as tools to perform this type of analysis (^{Anderson et al., 1991}; Ruíz et al., 2000; ^{Viloria et al., 2005}); however, these techniques may present some limitations. Nowadays, the use of molecular biology has allowed to expand the tools used for the genetic identification of plant species through the use of molecular markers, among which are the ISSR (Inter Simple Sequence Repeat), RAPD (Random Amplification of Polymorphic DNA) and SSR (Simple Sequence Repeat) (^{Bastianel et al., 1998}; ^{Golein et al., 2011}; ^{Yildiz et al., 2013}; ^{Mondal and Saha, 2014}; ^{Mondal et al., 2015}).

Turkey has been reported in the efficient use of AG14 and TAA03 oligonucleotide of SSR markers to identify and eliminate nucellar individuals in hybrid populations resulting from crosses between tangerin varieties (C. reticulata), oranges (Citrus sinensis), grapefruit (C. paradisi) (^{Yildiz et al., 2013}). Recently in India four SSR oligonucleotides with repeats have been used AG: CCSM13, CCSM17, CCSM18 and CCSM147 to identify zygotic and nucellar plants in a population of C. reticulata (^{Mondal et al., 2015}).

In Mexico, the species of C. macrophylla and C. volkameriana are commonly used as rootstocks for propagation and establishment of Mexican lime trees. In 2004, ^{Andrade-Rodríguez et al. (2004)} conducted a study to determine the zygotic or nucellar origin in C. volkameriana using RAPD markers, with satisfactory results. Similarly, in the case of Mexican lime (C. aurantifolia), RAPD markers have been successfully used, especially the decamer OPH15 (^{Mondal and Saha, 2014}); without there being reports in which other type of molecular markers, such as SSRs, have been used.

In the INIFAP breeding program, crosses between Mexican limes (C. aurantifolia) and Italian limes (C. limon), are routinely performed since it has been observed that the progeny exhibits some tolerance to HLB bacterial disease (Huanglongbing or yellow dragon disease), which causes one of the main phytosanitary problems affecting productivity in the Pacific region of México. Although there are many reports to identify zygotic citrus plants using molecular markers, in México no studies using SSR markers have been reported to identify hybrid plants of C. aurantifolia. In order to establish changes in the hybrids derived from controlled pollination between C. aurantifolia var. “Colimex”×C. limón var. “Rosenberg” and its reciprocal, the present research was carried out using SSR markers.

Materials and methods

Analysis of SSR markers

Parents C. aurantifolia and C. limon were analyzed with eight sets of oligonucleotides described by ^{Kijas et al. (1997)} (Table 1) in order to identify those that generate differential polymorphisms in both species. The selected oligonucleotides were used to analyze the two progeny populations of C. aurantifolia×C. Lemon and its reciprocal.

Table 1 SSR oligonucleotides used to detect polymorphisms in C. C. aurantifolia and C. limon.  

Amplification reactions by PCR were performed in a volume of 15 μL containing 7.5 μL of RedTaq ReadyMixTM (Sigma Aldrich), 5 μL of genomic DNA (20 ng μ L^-1), 1 uL of each oligonucleotide and 0.5 µL molecular grade water. Amplification conditions for PCR were performed in a MultiGene Labnet thermocycler and were described by ^{Kijas et al. (1997)} with some modifications: initial denaturation at 94 °C for 5 min followed by 35 cycles of 94 °C for 1 min, 55 °C for 30 s and 72 °C for 1 min. The final extension was 72 °C for 10 min.

Vertical polyacrylamide gel electrophoresis was used under non-denaturing conditions using the “PerfectBlue” double gel system (PeQlab). 1 μL of molecular weight marker of 25 and 50 base pairs (bp) (Bioline) and 3 μL of the PCR product of each sample was loaded. The gels were prepared in a 30 mL volume containing 6% acrylamide/bis-acrylamide (29:1), TBE 1 X buffer, 140 μL of 25% APS, 35 μL of TEMED and sterile distilled water, then run at 200 V for 50 min. Finally stained with silver solution according to ^{Sanguinetti et al. (1994)} with some modifications: 10 min in fixing solution (10% ethanol and 0.5% acetic acid), 1 rinse with distilled water, 10 min in staining solution (0.2% AgNO₃), one rinse with distilled water, 10 to 15 min in developer solution (3% NaOH and 0.5% formaldehyde) and a final rinse with distilled water. The gels were visualized in a transilluminator with white light for analysis.

Results and discussion

Eight oligonucleotides were evaluated for differential bands in C. aurantifolia and C. limon parents used in this paper (Table 1). The TAA45 and cAGG09 oligonucleotides were selected because they allowed the amplification of polymorphic bands of easy interpretation; four of the oligonucleotides amplified mostly non-informative and low resolution monomorphic bands and therefore were not used for the detection of hybrids in the populations analyzed. TAA41 initiator generated very diffuse bands that were not recorded. On the other hand, the TAA52 primer did not allow the amplification of some segment of DNA in the analyzed samples (Figure 1).

Figure 1 Oligonucleotides assessment of SSR markers from parents DNA. C= Citrus aurantifolia var. “Colimex”.R= Citrus limon var. “Rosenberg”. MPM= 25 and 50 bp molecular weight marker (Bioline).  

The oligonucleotides of the TAA series described by ^{Kijas et al. (1997)} have been widely used in the detection of hybrids between crossings of several citrus species such as grapefruit (C. paradisi), trifoliate orange (C. trifoliata), sweet orange (Citrus sinensis) and tangerine (C. reticulata) (Ruíz et al., 2000; ^{Ahmad et al., 2012}; ^{Yildiz et al., 2013}; ^{Mondal et al., 2015}). However, there are no reports of its use in Mexican lime (C. aurantifolia) and Italian lemon (C. limon), so these results are the first report on these species to determine its potential as hybrid detectors for genetic improvement programs.

In populations of 40 and 43 individuals from the crosses of C. aurantifolia var. “Colimex”×C. limon var. “Rosenberg” and its reciprocal, respectively, a total of 52 hybrids were detected with the TAA45 and cAGG09 oligonucleotides (Table 2). The TAA45 marker amplified two polymorphic alleles, one present in “Colimex” and absent in “Rosenberg” of approximately 100 bp and another present in “Rosenberg” and absent in “Colimex” of 110 bp. The latter fragment on some gels was visualized as a single band; however, they were three polymorphic bands with very similar molecular weights, reason why sometimes they did not separate. Figure 2a shows the hybrids detected with the TAA45 marker (722, 724, 736, 742 and 828), which also amplified other polymorphic bands that were not taken into account for the identification of hybrids due to their low resolution.

^*= no amplificó o presentó barrido el carril del gel.

Table 2 Zygotic or nucellar origin of C. aurantifolia×C. limon populations and its reciprocal by SSR analysis using two pairs of oligonucleotides. C= zygotic; N= nucelar; I= indeterminate.  

Figure 2 Electrophoretic profiles obtained with the TAA45 (A) and cAGG09 (B) oligonucleotides. MPM: 25 bp molecular weight marker (Bioline). ♀= pollen receiving plant. ♂= pollen donor plant. C= C. aurantifolia var.“Colimex”. R= C. limon var. “Rosenberg”. N= nucellar origin. C= zygotic origin (hybrids). The numbers correspond to different progeny samples  

The cAGG09 primer amplified four polymorphic bands in the analyzed populations, with approximate molecular weights of 90, 100, 110 and 170 bp. In both populations, with this marker, a non-informative 80 bp monomorphic band was amplified. The “Colimex” hybrids were identified by 90 and 170 bp alleles (Figure 2b, lanes 722, 724, 736, 738 and 742), which are present in “Rosenberg” (pollen donor) and absent in “Colimex” (pollen receptor).

Similarly, the “Rosenberg” hybrids were identified by 105 and 115 bp alleles, which are present in “Rosenberg” and absent in “Colimex”. In this sense, it is considered that a hybrid can be determined by identifying a single codominant locus for which the parents do not share alleles, i.e., crosses of the configuration aa×bb, ab×cc or ab×cd.

Also dominant markers, such as RAPD and ISSR, can be used if the pollen receptor is 00 (absent band) and the pollen donor is 11 (homozygous band present) (^{Kepiro and Roose, 2007}). However, confirmation with a second locus is always desirable to confirm the hybrid condition of the individual, which was performed in this research using the TAA45 and cAGG09 markers, whose polymorphisms of each progenitor segregated into the offspring according to the corresponding population.

SSR markers are a useful tool for the study of genetic diversity of citrus by its effectiveness and information generated (^{Amara et al., 2011}; ^{Biswas et al., 2011}). Its use for detecting hybrids (Ruíz et al., 2000; ^{Oliveira et al., 2002}; ^{Ahmad et al., 2012}; ^{Yildiz et al., 2013}; ^{Mondal et al., 2015}) is more practical, effective and advantageous because the researches that used RAPD and ISSR (^{Bastianel et al., 1998}; ^{Andrade-Rodríguez et al., 2004}; ^{Golein et al., 2011}) usually get a lot of uninformative monomorphic bands that may hinder the analysis of the data. The RAPD and ISSR, being dominant markers do not allow to differentiate heterozygous individuals unlike the SSR that are codominant markers.

In addition, RAPDs exhibit low reproducibility, as small modifications in the technique such as DNA template concentration, degree of purity and fidelity of the DNA polymerase enzyme, among other factors, may alter the pattern of amplified DNA fragments generated for a sample. Additionally, RAPD are not locus-specific, so electrophoretic band profiles can not be interpreted in terms of loci and alleles and fragments of similar size may not be homologous (^{Martínez et al., 2010}; ^{Kumari and Thakur, 2014}).

In this paper bidirectional crossbreeding of C. aurantifolia×C. limon were performed, yielding 42.5% of hybrid plants when “Colimex” was used as pollen receiver, and 81.4% when “Rosenberg” was used for the same purpose (Table 2). Grapefruits, oranges and tangerines have more than 90% of nucellar embryos, depending on the variety. In Mexican lemon, the proportion of nucellar seedlings is 78%, while for Italian lemons it is 32-33% (^{Frost and Soost, 1968}). The results of this research showed slightly lower percentages, 57.5 and 18.6% of nucellar embryos in “Colimex” and “Rosenberg”, respectively. These lower values than previously reported could be due to nutritional factors, as well as genotype-environment interaction (^{Kepiro and Roose 2007}).

In this research, individuals with zygotic origin were considered as hybrids, since the pollinations were controlled, avoiding self-fertilization events. Both markers, cAGG09 and TAA45 coincide in 80 and 83.72% in plants determined as hybrid for C. aurantifolia×C. limon populations and its reciprocal. Samples in which both markers did not coincide with the zygotic or nucellar origin of the progeny were considered the result of cAGG09 oligonucleotide because it amplified well-defined polymorphic bands.

On the other hand, the effectiveness of isoenzymes and SSR markers for detecting hybrids has been compared, concluding that molecular markers are more efficient for this purpose. In this same research they used the oligonucleotides used in this paper: TAA41 and TAA45 obtaining polymorphisms which identified the zygotic origin in populations of C. reticulata×C. sinensis and Poncirus trifoliata self-pollination (Ruíz et al., 2000). Although in this research the use of the TAA41 primer amplified diffuse non-informative bands patterns. In another study, they combined the use of morphological markers assisted by SSR molecular markers of the CCSM series to detect hybrids derived from C. reticulata and C. sinensis (^{Oliveira et al., 2002}).

However, they only confirmed the hybrid status of the plants detected by morphology, so the possible existence of more individuals with a sexual origin that could be detected by the SSR markers is doubted. Meanwhile ^{Ahmad et al. (2012)} used the initiators TAA15, TAA27, TAA33 and others from the CCSM series to detect hybrids in three F1 populations of several orange and tangerine crosses, detecting 23 and 5 hybrids of the “NARC 05-18”דTarocco” and NARC 05-17דSanguinello” crosses respectively. In addition, they identified 35 more hybrids of “Kinnow”דTarocco” with the CCSM147 marker. ^{Yildiz et al. (2013)} used the primers TAA01, TAA3, TAA41, TAA45, TAA52, CAC33 and cAGG09 to detect hybrids in several populations using tangerines as pollen receptor and crossing them with oranges and grapefruit.

Of the 500 individuals constituting the progenies of the various crosses, primers AG14 and TAA3 were the most effective for identifying zygotic individuals. The “Fremont” and “Robinson” tangerines produced 36.91 and 31.09% of nucellar plants, respectively. Finally, ^{Mondal et al. (2015)} recently made use of TAA15, TAA27 and TAA33 oligonucleotides for detecting hybrids of C. reticulata and reported that none of these primers allowed the identification of sexual origin in the evaluated progeny, so they resorted to the initiators of the MRCC series for identification purposes.

Although SSR markers provide useful information for the identification of hybrid plants in different citrus species, for this purpose it is necessary to implement techniques such as diversity array technology (DArT), single nucleotide polymorphism (SNP), or high resolution melt (HRM), however, these are currently very expensive.

Conclusions

SSR markers allowed the efficient identification of zygotic and nucellar plants in progeny populations of Mexican lime “Colimex” and Italian lemon “Rosenberg”. With the use of the cAGG09 primer, the amplification of 4 polymorphic alleles, sufficient to determine the sexual origin of Mexican lemon plants, was identified; however, confirmation with a second locus is always desirable to confirm the hybrid condition of individuals.

A total of 52 hybrid in the populations from crosses C. aurantifolia (♀)×C. limon (♂) and C. limon (♀)×C. aurantifolia (♂) were identified. Additionally, identified hybrids may be infected with HLB in field conditions to determine the level of tolerance to the disease.