Identifying F2 Oil Palm (Elaeis guineensis Jacq.) Trees for Their Dura, Pisifera and Tenera Types using Fruit Morphology and SSR Markers

Oil palm (Elaeis guineensis Jacq.) is among the globally most important oil crops. Oil palm trees can be classified into dura, pisifera and tenera types based on fruit morphology. The dura with homozygous genotype (sh⁺sh⁺⁾ has a thick shell (2-8 mm) enclosing the endosperm or kernel and has a small proportion of oil–bearing mesocarp (35-70% of fruit weight). The pisifera (sh^-sh^-) has no shell but a ring of fibers encloses the kernel. The kernel of pisifera fruit may be absent, so the oil-bearing mesocarp may constitute the whole fruit. It is used as a male plant due to female-sterile palms. The tenera (sh⁺sh^-) is a hybrid of dura and pisifera. It has a thin shell (0.5-4 mm), a ring of fibers enclosing the kernel, as well as large oil-bearing mesocarp content of 60-95% (Beirnaert and Vanderweyen, 1941; Hardon, 1976). Tenera is planted as a commercial plant to extract oil, due to it giving the highest yields.
       
DNA markers associated with shell thickness can be used to identify fruit traits at the seedling stage of oil palm. Plant breeders can classify F₁ plants of a cross between tenera and tenera oil palm into the types dura, pisifera and tenera using these markers. DNA markers can aid in the identification of tenera form for growing as a commercial crop. From the 1970s onward, the most common DNA markers used for classification of oil palm types have been restriction fragment length polymorphisms (RFLPs), random amplified polymorphic DNAs (RAPDs) and amplification fragment length polymorphisms (AFLPs), but later on simple sequence repeats (SSRs) have become the most popular choice. One RFLP marker, pOPgSP1282, is tightly linked to the shell thickness locus (Mayes et al., 1977). Three RAPD markers, Y20-1180, R11-1282 and T19-1046 are associated with the shell thickness locus (Moretzsohn et al., 2000). The R11-1282 and T19-1046 on both sides of the Sh⁺ gene are in linkage group 4. These two RAPDs were able to identify pisifera and tenera plants with only 1% error rate in the calls. The four RAPD markers, P12-600, P10-700, P6-650 and P19-800 indicate pisifera, two markers, P7–700 and P10-1000, indicate dura and one marker, P28-1100, indicates tenera (Sathish and Mohankumar, 2007). An AFLP marker E-AGG/M-CAA132 is closely linked to the Sh gene (Billotte et al., 2005). The SSR loci mEgCIR008 and mEgCIR1772 could be used to identify the tenera hybrid and to distinguish it from dura and pisifera parents, but these are specific to some families (Thawaro and Te-chato, 2009; Thawaro and Te-chato, 2010). The mEgCIR3428, mEgCIR3519 and mEgCIR0874 markers were used to identify Surat Thani 1-8 oil palm hybrid varieties (Urairong et al., 2017).
              
Currently, SSR markers, also known as microsatelliteDNA markers, are the most used markers for identification of F₁ hybrids, genetic map construction genetic, genetic diversity analysis, QTL discovery and marker-assisted selection of oil palm or other crops, because of their low cost and convenience. An SSR marker locus possesses alleles consisting of tandem repeats of 2–5 nucleotide DNA core units, such as (TC)_n, (CCG)_n or (ACAG)_n, where the index “n” indicates the number of repeats of the given allele. The nucleotide sequences flanking a microsatellite are highly conserved within genotypes of the same species, thus allowing PCR primers to be developed. Polymerase chain reaction (PCR) amplification of SSRs requires a unique forward and reverse primer pair to amplify the intervening SSR alleles for all genotypes (Abdullah et al., 2011; Billotte et al., 2005; Kavya et al., 2019; Sao et al., 2015; Sari et al., 2019; Thawaro and Te-chato, 2009; Thawaro and Te-chato, 2010).
       
The objective of this study was to categorize a mixed F₂ population of oil palm into the types dura, pisifera and tenera using fruit morphology as well as SSR markers.

Plant material
 
A mixed F₂ population of oil palm was identified for the types dura, pisifera and tenera. Sixty seven F₂ plants were selected from one thousand eighty F₂ plants. They derived from F₁ tenera hybrids, which were collected from different oil palm plantations in Southern Thailand. One good performance bunch was selected from each plantation and four to six seeds per selected bunch were used for plantation. They were grown at Klong Hoi Khong Research Station, Faculty of Natural Resources, Prince of Songkla University, Songkhla in 1989 (Yongyut, 2002).
 
Sample collection and DNA extraction
 
Ten fruit were sampled from a bunch of each F₂ plant. These fruit were categorized into the types dura, pisifera and tenera based on shell thickness, proportion of oil-bearing mesocarp and ring of fibers enclosing the kernel.

Young leaves of F₂ plants were collected and put into a plastic bag, then put on ice and later frozen and stored at -20oC. DNA from oil palm leaf tissue was extracted following the N-Cetyl-N, N, N-trimethylammonium bromide method (CTAB) of Doyle (1990). It was slightly modified. Leaf tissue of 0.2 g was ground to fine powder in the presence of liquid nitrogen using mortar and pestle. The powder was then transferred into a 1.5 ml Oak Ridge tube, mixed with 600 ml of extraction buffer (CTAB 2%, 5 M NaCl, 100 mM Tris pH 8.0, 20 mM EDTA pH 8.0, 1.4 M NaCl and 10 mM merchapto-ethanol) and incubated at 65°C for 20 min, then 300 ml potassium acetate was added and the sample was frozen at -20°C for 30-60 min. The mixture was centrifuged at 12,000 rpm for 30 min at 4°C to separate the leaf residues. Seven hundred µl of chloroform/isomyl alcohol (24:1 v/v) was then added. The supernatant was collected into a new tube and mixed with two volumes of isopropanol in order to precipitate the DNA. The DNA was separated by centrifugation at 12,000 rpm for 30 min. After discarding the supernatant, the DNA pellet was washed twice with 70% ethanol containing 10 mM ammonium acetate, dried and dissolved in TE buffer (10mM Tris and 1mM EDTA pH 8.0). RNase treatment was then carried out by adding RNase (50 mg.ml^-1) and incubating at 37°C for 20 min. Ammonium acetate (7.5 M) and two volumes of absolute ethanol were added. The DNA was pelleted by centrifuging at 12,000 rpm for 5 min. After washing with 70% ethyl alcohol (500 µl) two times for 5 min each time, the sample was dried, then 20 µl of TE buffer was added. DNA concentration and purity of these samples was measured using a spectrophotometer at wavelengths 260 and 280 nm.
 
SSR analysis and primer screening
 
SSR analysis of genomic DNA was carried out using nine SSR markers for oil palms. These markers were selected for proximity to the reported map positions of Sh locus as well as to distinguish the tenera hybrid from dura and pisifera parents (Billotte et al., 2001; Billotte et al., 2005; Thawaro and Te-chato, 2010; Abdullah et al., 2011). The primers are listed in Table 1. Amplification of genomic DNA was done according to the protocol of Billotte et al., (2005). The PCR reaction mix consisted of 25 ng of genomic DNA, 10 mMTaq buffer, 0.2 µM of each primer (paired forward and reverse primers), 1.5 mM MgCl₂, 200 μM of each of the four dNTPs and 1 unit of Taq polymerase. PCR amplifications were carried out on a thermocycler (TC-XP-G, Japan) using the following program: denaturation at 94°C for 1 min; 35 cycles  of 94°C for 30 s, 52 to 58°C (depending on the primer used) for 1 min and 72°C for 2 min; and a final elongation step at 72°C for 8 min; then storage at -20°C. An equal volume of loading buffer (98% formamide, 0.025% bromophenol blue and 0.05% xylene cyanol) was added to the amplified products, following denaturation at 94°C for 5 min. The products were analyzed on 7.5% (w/v) denaturing polyacrylamide electrophoresis gel. Silver staining was conducted according to the protocol of Bassam et al., (1991). Bands for each plant were used to classify them into the three types: dura, pisifera and tenera.
 

Identification oil palm type using fruit morphology
 
Sixty seven F₂ oil palm trees were classified based on fruit morphology into twenty four dura, nine pisifera and thirty four tenera trees. The fruit forms of dura, pisifera and tenera oil palm are shown in Fig 1. The mean shell thickness of dura type was 3.17 mm with a standard deviation of 0.48 mm, ranging from 2.10 to 4.27 mm. The mesocarp in fruit by weight of dura type was 55.80%, with a standard deviation of 6.32% and ranging from 46.76 to 70.76%. The pisifera type had no shell and the mesocarp was 77.59% of fruit by weight, with a standard deviation of 20.09% and a range from 41.68 to 100%. The mean shell thickness of tenera type was 1.31 mm with a standard deviation of 0.46 mm and range from 0.79 to 2.79 mm. The mesocarp in fruit was 77.81% by weight, for tenera type, with a standard deviation of 5.31% and range from 65.40 to 86.26% (Table 2). The tenera had a ring of fibers enclosing the kernel, but the dura had no such ring of fibers. The tenera had thicker shell and more mesocarp per fruit than the dura. The mesocarp per fruit of some pisifera type was below 100% because they had a kernel in the middle of fruit. These results are similar to those reported in Beirnaert and Vanderweyen (1941) and in Hardon (1976).
 

 

 
Identification oil palm type using SSR markers
 
A total of nine SSR markers were firstly screened for the classification of fruit forms using a sample of seven dura, seven pisifera and seven tenera trees. The presence or absence of bands and the total number of bands amplified were used for selecting informative and specific markers. The results showed that only two markers, MF233033 and MF233056, could be used for identifying three types. Based on the specificity of amplification, these two markers were used to identify each type as distinct from the other types of oil palm trees. Fig 2 presents the SSR profiles for the dura, pisifera and tenera types generated with MF233033 and MF233056 markers. The MF233033 marker identified the tenera types with 200 bp bands except genotypes 3, 4, 11, 27, 40, 46, 62, 71, 72, 78, 86 and 89. The MF233056 marker identified the pisifera type with 230 bp bands but only nine plants were identified in this research. The dura types did not show those bands of two markers. The reason could be variation of shell thickness and mesocarp or the complex heredity background of each plant. Dura, pisifera and tenera types are able to identify with an accuracy 100% when the specificity of markers are used to identify only parents and off-springs (Abdullah et al., 2011; Chotewattanasak and Eksomtramage, 2008; Thawaro and Te-chato, 2009; Thawaro and Te-chato, 2010). 
 

The results suggest that fruit morphology can be used to identify F₂ oil palm trees into dura, pisifera and tenera types. Two markers, MF233033 and MF233056 could be used to identify the pisifera type as well as not clearly identify dura and tenera types. In the future the more DNA markers or more oil palms need to be studied to identify pisifera, dura and tenera types.

This work was supported by the government budget of Prince of Songkla University (NAT550033S). We are grateful to Assoc. Prof. Dr. Charassri Nualsri for SSR analysis at her Lab. Finally, we would like to thank Klong Hoi Khong Research Station, Faculty of Natural Resources, Prince of Songkla University in Thailand for support plant material. The copy-editing service of RDO/PSU and helpful comments of Assoc. Prof. Dr. Seppo Karrila are gratefully acknowledged.