MOLECULAR IDENTIFICATION AND GENETIC RELATIONSHIPS AMONG COFFEE SPECIES ( COFFEA L ) INFERRED FROM ISSR AND SRAP MARKER ANALYSES

The identification and genetic relationships of 23 coffee species and one coffee-related species Canthium diccocum were studied using ISSR and SRAP markers. The average polymorphism information content of SRAP primers (0.81) was lower than ISSR primers (0.86), whereas the average resolving power of the SRAP primers (9.74) is higher than the ISSR primers (8.64). The genetic similarity among the species ranged from 0.30 to 0.89 using ISSR and 0.11 to 0.90 using SRAP marker systems. Based on marker analysis, all twenty three coffee species were clustered into two major groups. Both the markers amplified species-specific fragments and are useful in genetic diversity analysis of coffee.


INTRODUCTION
The genus Coffea L. belongs to the Rubiaceae family and comprises 103 species (Davis et al. 2006).The majority of coffee species occurs naturally in Africa, Madagascar and the Mascarenes, predominantly restricted to humid evergreen forest, but some species are found in seasonally dry deciduous forest and/r bush land (Maurin et al. 2007).The importance of coffee as an agricultural commodity relies mainly upon two varieties, Coffea arabica and C. canephora, which contribute about 65% and 35% of total production, respectively.In spite of the commercial and social importance of the genus, the genetic relationship between the majorities of coffee species is not extensively studied and their taxonomic status is poorly understood.Understanding the genetic relationship between coffee species is not only important for resolving taxonomic ambiguity but also important for the genetic improvement program.Conventionally, morphological descriptors such as growth habit, leaf type, floral characteristics and fruit morphology are used to characterize the various species.However, developing morphological descriptors for any particular species/cultivar has severe limitations as these characteristics are influenced by environmental conditions.In contrast to the morphological markers, DNA-based marker techniques are more efficient, precise and reliable for discriminating between closely related species and cultivars.In previous studies, genetic diversity and phylogenetic relationships between different coffee species was carried out by using both Random Amplified Polymorphic DNA (RAPD) analysis as well as Restriction Fragment Length Polymorphism of chloroplast and mitochondrial DNA (Berthou, 1983;Cross et al. 1998;Lashermes et al. 1993;Lashermes et al. 1996;Orozco-Castillo, et al. 1996).Recently, Davis et al. (2006) published a detailed annotated taxonomic conspectus of the genus Coffea and included five indigenous coffee species from India, of which C. bengalensis, C. travancorensis, C. wightiana are placed under the genus Psilanthus and an-other two species, C. khasiana and C. jenkinsii, are placed under the genus Nostolachma, both under Coffeeae.Based on the analysis of the internal transcribed spacers (ITS) of nuclear DNA, Lashermes et al. (1997) observed limited sequence divergence between Psilanthus and Coffea and concluded that Psilanthus should not be recognized as a separate genus from Coffea.Based on the comparative sequence analysis of plastid DNA, Cross et al. (1998) concurred with Lashermes et al. (1997) about the close relationship between Coffea and Psilanthus although their tree topology shows an unresolved relationship between two species of Psilanthus (P.mannii and P. ebracteolatus) studied by them and Coffea.However, both these studies did not include a representative of closely related genera for comparison.Further studies by Andreasen et al. (1999), Andreasen and Bremer (2000) and Davies et al. (2007) showed that Psilanthus is nested within Coffea and stressed the importance of species level studies to resolve the phylogenetic relationship.Maurin et al. (2007) observed that the separation of Coffea and Psilanthus based on morphological features is not convincing and suggested the inclusion of molecular data from other species of Psilanthus to establish the relationship between these two genera.Based on the morphological features, Nostolachma is placed under Coffeeae (Davis et al. 2007) although molecular data from additional species may prove to be useful.Although various researchers have studied the taxonomic and phylogenetic relationship between various coffee species, there are no reports yet available for the species-specific markers that are useful for distinguishing various species.In this paper, we have used two PCRbased molecular markers, Inter Simple Sequence Repeats (ISSR) and Sequence Related Amplified Polymorphism (SRAP), in developing species specific markers for 23 coffee species available in coffee germplasm in India.To our knowledge, both these markers are used in coffee for developing species diagnostic markers for the first time.The main objectives of the study are to generate a molecular database and identification tag for each species that may be useful for conservation and systematic studies as well as coffee breeding programs.

Plant materials
Leaf material was collected from 10 individuals of twenty three coffee species (Table 1) including five indigenous coffee species from India (C. bengalensis, C. travancorensis, C. wightiana, C. khasiana and C. jenkinsii), maintained at the germplasm plot of Central Coffee Research Institute, Chikmagalur, Karnataka and Regional Coffee Research Station Thandigudi, Tamil Nadu.Mature seeds were also collected and grown in earthen pots under net house conditions.The leaf materials were used for DNA isolation.

DNA extraction
Genomic DNA was extracted from fresh young leaves using a modified CTAB method.About 200 mg of leaf tissue was ground to a fine powder in liquid nitrogen, it was transferred to a 30 ml tube containing 5 ml preheated extraction buffer (2% CTAB (w/v), 100mM Tris-HCL (pH 8.0), 25mM EDTA, 2M Nacl, 0.1 % beta-mercaptoethanol).The tubes were incubated at 60 0 C for 1 h with occasional shaking.After incubation, the tubes were cooled to room temperature and centrifuged at 6000 rpm for 20 min.The supernatant was transferred to a new tube and extracted twice with equal volumes of chloroform-isoamyl alcohol (24:1).The supernatant was transferred to 2 ml tubes, precipitated with 0.7 vol of isopropanol at room temperature for 30 min., and then centrifuged at 8000 rpm for 20 min at 4° C. The pellet formed after centrifugation was washed with 75% (v/v) ethanol for 10 min and dissolved in 60 µl of Tris-EDTA (1-10mM).The concentration of DNA was measured using 0.8% agarose gel stained with ethidium bromide as well as by UV spectrophotometry at 260 nm and a 280 nm.The resuspended DNA was then diluted in sterile distilled water to 10ng/ µl concentration for use in amplification reactions.

ISSR, SRAP analysis
A total of 20 ISSR primers from the University of British Columbia and 60 SRAP primer combina-tions (8 forward primers and 14 reverse primers) synthesized by Sigma, India, were initially screened to determine the suitability of each primer for the study.Primers were selected for further analysis based on their ability to detect clear and distinct polymorphic amplification products within the species of Coffea.12 ISSR primers and 21 SRAP primer combinations that had a high level of polymorphism and the best readability were used for PCR amplification (Table 2).The PCR reaction was carried out in a palm cycler (Corbett Research).
The PCR products of both ISSR and SRAP were run on 2% (w/w) agarose gels containing 0.5 µg ethidium bromide/ml in 1X TAE buffer and then visualized and photographed using the UV-transilluminator (SYNGENE) and documented using the Gene Snap software program.Both ISSR and SRAP PCR reactions were repeated at least twice to confirm the reproducibility of each PCR band.

Data analysis
The ISSR and SRAP amplified bands were scored for the presence (1) or absence (0).The total number of bands, the distribution of bands across all species, polymorphic bands, species-specific bands and average number bands per primer were calculated.The value of each primer was assessed using two indices; PIC, which is the same as a diversity index (Botstein et al. 1980;Milbourne et al. 1997) and resolving power (Rp) (Prevost and Wilkinson, 1999).PIC or DI was estimated as PIC= (1-p 2 i )/n, where n is the number of band positions analyzed in all the species, p i is the frequency of the banding pattern.The resolving power of a primer is Rp = I b where I b (band informativeness) takes the value of 1-[2x (0.5-p)] and p is the ratio of six species sharing the band.A pairwise similarity matrix was constructed using the Dice similarity coefficient (Sneath and Sokal, 1973).The relationship between the species was displayed as a dendrogram constructed using NTSYS -PC 2.10e software (Rohlf, 1995) based on Unweighted Pair Group Method using Arithmetic averages (UPG-MA).Statistical support of the clusters was assessed by means of 1000-bootstrap replicates.

ISSR polymorphism and species identification
Out of the 20 ISSR primers initially screened, 12 primers were found to be polymorphic and produced clear and reproducible amplification patterns.These 12 primers could produce 185 distinct reproducible bands across the 24 species with an average of 15.58 per primer (Table 3).The size of the amplified products ranged from 100 to 2900 bp.Of the total 185 amplified bands, 174 (93.06%) were polymorphic, with an average of 14.5 polymorphic fragments per primer.All the ISSR primers except UBC-881 and UBC-855 showed 100% polymorphism.The primers UBC-881 and UBC-855 showed 8.3% and 90.90% of polymorphism respectively.The average poly-morphism percentage recorded with ISSR primers is 93.06.The resolving power of the 12 ISSR primers tested ranged from 4.08 (UBC-840) to 22.17 (UBC-881) with an average of 8.64.Similarly, the polymorphism information content (PIC) of the 12 ISSR primers ranged from 0.083 (UBC-881) to 0.963 (UBC-840) with average of 0.84.
Out of the 12 ISSR primers, only one ISSR polymorphic primer (UBC 842) could discriminate all the species independently (Fig. 1).Twelve ISSR primers generated 40 unique fragments of which 8, 7 and 5 fragments are obtained in C. khasiana, C. jenkinsii and C. arabica, respectively.Further, the ISSR primer UBC-836 has amplified species diagnostic fragments in five different species (Table 4).The five Indian coffee species generated 19 unique fragments of which maximum numbers of eight unique fragments were generated by C. khasiana, whereas only one unique fragment was generated by both by C. travancorensis and C. wightiana.The genetic similarity derived from the data of the ISSR marker analysis varied from 0.27 between C. bengalensis and C. liberica to 0.89 between C. canephora and C. congensis (Table 5).Among the Indian species, C. travancorensis showed maximum genetic similarity with C. wightiana, based on the ISSR marker analysis.Furthermore, Canthium diccocum, which is used as a reference species, showed maximum genetic similarity (0.65) with C. racemosa.
The dendrogram based on ISSR data was constructed by UPGMA analysis, grouping all of the coffee species into two major clusters (Fig. 2).The

SRAP polymorphism and species identification
Sixty SRAP primer combinations were initially screened against 23 coffee species along with the ref-erence species, C. decorum, of which 21 primer combinations are found to be highly polymorphic and produce clear amplification patterns.These 21 primers could produce 336 distinct scorable bands with  10 primer combinations showed 100% polymorphism.The resolving power (RP) of 21 SRAP primer combinations ranged from 3.83 (Me4-Em1) to 14.58 (Me3-Em3) with an average of 9.74.Similarly, the polymorphism information content (PIC) or the genetic diversity of 21 SRAP primer combinations ranged from 0.61 (Me1-Em2) to 0.96 (Me1-Em12, Me3-Em11) with an average of 0.82.
Only three SRAP primer pairs could discriminate all the species independently, although none of these primers could amplify species-specific fragments for all the 23 species (Fig. 3).20 SRAP primer combinations generated 75 unique fragments in 23 coffee species out of which 5 Indian species generated 30 unique fragments.Among the Indian species, the maximum number of unique fragments was generated by C. jenkinsii (16) followed by C. khasiana (8) and the least number of unique fragments was in C. travancorensis (1) which could be used as a unique fingerprinting tools.The genetic similarity derived from the data of the SRAP marker analysis varied from 0.11 between C. wightiana and C. congensis to 0.90 between C. canephora and C. canephora var.laurentii (Table 6).Among the Indian species, maximum similarity was observed between C. travancorensis and C. wightiana and the least genetic similarity was obtained between C. khasiana and C. jenkinsii based on the SRAP marker analysis.
The UPGMA clustering algorithm from the SRAP analysis grouped the 23 coffee species and C. diccocum into two major clusters (Fig. 4).The first major cluster was represented by C. wightiana and the second major was divided into two minor clusters of which the first minor cluster consisted of C. jenkinsii alone.The second minor cluster divided in to two sub-minor clusters.The first sub-minor cluster included C. congensis, C. canephora varieties, C. arabica and C. liberica, whereas the second sub-minor cluster further dived into two groups.The first group included C. zanguebariae, C. racemosa, C. abeokutae, C. dewevrei, C. dewevrei var. aruwimiensis, C. stenophylla, C. arnoldiana and C. dewevrei var. excelsa, whereas the second group included three Indian species viz.C. bengalensis, C. travancorensis, C. khasiana along with Canthium diccocum.A strict consensus tree based on both ISSR and SRAP data were constructed (Fig. 5).This dendrogram shows maximum similarity to ISSRbased dendrogram except that the position of C. arnoldiana and C. dewevrei var.excelsa was taken over by C. zanguebariae and C. racemosa, respectively, in the first major sub-cluster.The genetic similarities obtained using both ISSR and SRAP data has revealed the highest similarities (0.90) between C. canephora and C. canephora var.laurentii and the lowest similarities (0.36) between C. liberica and C. bengalensis (Table 7).

DISCUSSION
Analysis of crop genetic diversity is very important for breeding and conservation programs, and molecular markers offer an approach to unveil the genetic diversity among different species and cultivars based on nucleic acid polymorphisms.In this study, both ISSR and SRAP marker systems were simultaneously used to investigate the genetic diversity among 23 coffee species available in the germplasm collection of India.The results showed that both ISSR and SRAP markers were suitable for genetic diversity analysis in coffee by amplifying several species specific diagnostic markers.Species specific ISSR primers were generated in eucalyptus (Balasaravanan et al. 2006) and oak (Carvalho et al. 2009).In the present study, species-diagnostic ISSR markers were identified in 12 coffee species including all the indigenous coffee species from India except C. travancorensis (Table 4).Similar problems were experienced while developing species-specific ISSR markers in eucalyptus (Balasaravanan et al. 2006).This could be due to the occurrence of a high divergence among the particular species population and/or the low number of ISSR primers used in this study.In contrast to the ISSR markers, SRAP is more effective and amplified species specific markers in all the coffee species except C. zanguebariae, and C. dewevrei var.aruwimiensis (Table.4).
Both genetic factors and selection pressure influence genetic diversity (Sun and Lin 2003).Since various coffee species have their origin in different agro-climatic zones in their native habitat, with varying selection pressure during the course of evolution, it is therefore not surprising to find wide polymorphism among different coffee species.The percentage of polymorphic bands detected by SRAP primer combinations (96.12%) was higher compared to ISSR (93.06%).Similarly, the number of polymorphic bands detected by SRAP primer combinations (15.38) was higher than that obtained by the ISSR primer (14.5).This indicated that SRAP markers are more efficient than ISSR markers in genetic diversity analysis of coffee.
The average PIC of the SRAP primers (0.82) was less than that of the ISSR primers (0.84), whereas the average RP of the SRAP primers (9.74) was higher than that of the ISSR primers (8.64).The difference in PIC and RP values among ISSR and SRAP markers is expected because both markers differ in their operational principle and each marker targets a different region of the genome.This is evident from the UPGMA clustering of the coffee species using ISSR and SRAP markers.While both the C. travancorensis and C. wightiana from the southern peninsular India showed maximum similarity between themselves (Table 5) and placed close to each other in the dendrogram (Fig. 2) using ISSR markers, using a SRAP marker, C. travancorensis shared maximum similarities with C. bengalensis (Table 6) and placed closely in the dendrogram (Fig. 4).Interestingly, the genetic similarities observed between various indigenous coffee species with other coffee species using an ISSR marker varied from 0.3 to 0.6 and 0.1 to 0.6 using SRAP markers (Tables 5 and 6).This difference could be explained by the fact that ISSR markers target the region within the microsatellite repeats whereas SRAP markers preferentially detect polymorphism in coding sequences which are usually conserved among closely related cultivars and species with low mutation rate.Therefore, simultaneous use of different types of molecular markers may be useful in generating "all-sided" information.
In conclusion, this study provides information on genetic relatedness not only among different coffee species but also on their relatedness with the indigenous coffee species which are either placed under the genus Psilanthus or Nostolachma.The present study has clearly demonstrated a close relationship between Indian indigenous coffee species and other wild coffee species and supports the contention of their inclusion in the genus Coffea rather than placing them under the genus Psilanthus or Nostolachma, as suggested by previous workers (Cross et al., 1998;Lashermes et al. 1997;Andreasen et al. 1999, Andreasen and Bremer, 2003and Davies et al. 2007).Further understanding of the level and partitioning of genetic variation within the species would provide an important input in designing appropriate breeding exercises and conservation strategies.

Fig. 1 .
Fig.1.The extent of polymorphism observed among various coffee species listed in table 1 by ISSR primer 842 Lanes M: Molecular ruler (Gene ruler 100 bp ladder plus).

Fig. 2
Fig.2Dendrogram generated using unweighted pair group method with arithmetic average (UPGMA) analysis showing relationships among different coffee species using ISSR data.

Fig. 3
Fig.3 The extent of polymorphism observed among various coffee species listed in Table 1 by SRAP primer combinations Me3 and Em12Lanes M: Molecular ruler (Gene ruler 100 bp ladder plus).

Fig. 4
Fig.4Dendrogram generated using the unweighted pair group method with arithmetic average (UPGMA) analysis showing relationships among different coffee species using SRAP data.

Fig. 5
Fig.5 Dendrogram generated using the unweighted pair group method with arithmetic average (UPGMA) analysis showing relationships among different coffee species using ISSR and SRAP data.

Table 1 .
Coffee species used in the study with their code, origin and conservation status

Table 2 .
Primer sequences used for SRAP analysis of coffee species

Table 3 .
Polymorphism obtained by RAPD, ISSR and SRAP analysis in coffee speciesТable 4. Species-diagnostic ISSR and SRAP markers in coffee species