Content uploaded by Suthasinee Somyong
Author content
All content in this area was uploaded by Suthasinee Somyong on Jun 07, 2019
Content may be subject to copyright.
32 Chiang Mai J. Sci. 2019; 46(1)
Identifying a DELLA Gene as a Height Controlling
Gene in Oil Palm
Suthasinee Somyong* [a], Kitti Walayaporn [a,b], Nukoon Jomchai [a], Siti Hajar Hassan [a,c],
Tanapong Yodyingyong [d], Chalermpol Phumichai [d], Anek Limsrivilai [e],
Arthorn Saklang [f], Sudprasong Suvanalert [f], Chutima Sonthirod [a],
Laurensia Danis Anggradita [a,h] and Sithichoke Tangphatsornruang [a]
[a] National Center for Genetic Engineering and Biotechnology (BIOTEC), Pathum Thani 12120, Thailand.
[b] Interdisciplinary Graduate Program in Genetic Engineering and Bioinformatics, Kasetsart University,
Bangkok, 10900, Thailand.
[c] Faculty of Industrial Sciences & Technology, Universiti Malaysia Pahang, Gambang, Kuantan,
Pahang 26300, Malaysia.
[d] Department of Agronomy, Faculty of Agriculture, Kasetsart University, Bangkok 10900, Thailand.
[e] Golden Tenera Company Limited, Krabi 81000, Thailand.
[f] CPI Agrotech Company Limited, Chumphon 86140, Thailand.
[g] Sitthiporn Kridakorn Research Station, Kasetsart University, Prachuap Khirikhan 77170, Thailand.
[h] Faculty of Biotechnology, Atma Jaya Catholic University of Indonesia, South Jakarta, DKI Jakarta 12930,
Indonesia.
*Author for correspondence; e-mail: suthasinee.som@biotec.or.th
Received: 7 June 2018
Accepted: 12 September 2018
ABSTRACT
The research in oil palm (Elaeis guineensis Jacq.) yield improvement is still being
done to serve the consumption demand needed for cooking oil, cosmetics and biodiesel
production. Moreover, a harvesting-facilitating trait such as stem height still needs more
improvement to lower labor intensity and harvesting losses of oil palm bunches. This study
aims to understand the genetic control for the stem height trait (HT) and its correlation
with yield component traits such as bunch number (BN) and fresh fruit bunch yield (FFB).
To identify height controlling genes, gene-based markers targeting three candidate genes
including DELLA gene, gibberellin (GA) 2-oxidase and asparagine synthase were designed
and identified the association of their potential polymorphic markers with HT in various oil
palm populations. We identified the DELLA gene (renamed as EgDELLA1), a GA nuclear
repressor, at chromosome 14, highly associated with height in the GT population at p value of
0.0261, and 0.0429 for the two of three times of height phenotype recorded. The strongest
expressions of EgDELLA1 were found in apical meristem and, to a lesser extent in leaf
and fruit development while no expression was detected in male inflorescence. This suggested
that EgDELLA1 is a major component of stem elongation initiating at apical meristem.
In addition, weak to moderate positive correlation was found between HT and BN
(r = 0.215-0.395), and between HT and FFB (r = 0.254-0.499) from the KU and GT
populations. Thus, improvement of semi-dwarf oil palm with higher yield is possible
Chiang Mai J. Sci. 2019; 46(1) : 32-45
http://it.science.cmu.ac.th/ejournal/
Contributed Paper
Chiang Mai J. Sci. 2019; 46(1) 33
1. INTRODUCTION
African oil palm (Elaeis guineensis Jacq.)
is a perennial monoecious species with a
life-span of up to 25-30 years, reaching up to
15-18 meters in height. African oil palm plant
height increased by 45-75 cm (centimeter) per
year [1]. When oil palm ages and grows to
heights more than 2-3 meters, the harvesting
of fruit bunches becomes more challenging.
Oil palm with short stem height is favorable
for bunch harvesting. Therefore, short
stem and high yield are more favorable
characteristics for selection by oil palm
breeders. Several oil palm breeding teams
have been successful at making oil palm
hybrids with slow trunk growth. For example,
interspecific F1 hybrids with height increase
of only 15-25 cm per year were made by
crossing E. guineensis with E. oleifera, which
has a height increase of about 5-10 cm
annually [2]. Intraspecific hybrid with high
yield and annual height increase of about
40 cm was also made as an oil palm genetic
resource [3]. Currently, information about
genes controlling height in oil palm is still
limited.
To date, two published QTL studies of
height were conducted in an F2 population
created from self-pollination of Tenera
clone A43/9 [4] and in two Tenera palm F1
populations, which resulted from crossing
from Dura and Pisifera individuals [5]. Three
QTLs were found on chromosome 10, 14
and 15, that contributed to stem height and
explained 10-21% of phenotype variance
expected (PVE) [4]. Two potential genes,
DELLA gene (GAI1) and GA2 -oxidase
(GA2OX2), which linked the QTL on
chromosome 14, were proposed to play an
important role in height control. In another
QTL study [5], they found one major QTL
(51% of PVE) on LG 5 (chromosome 16)
responsible for stem height in the F1
populations. There are 8 genes in the QTL
region and asparagine synthase related
gene was thought to play an important
role in height control. They found the
significantly higher expression (qRT-PCR)
of the asparagine synthase related gene in
the trunk of dwarf trees, as compared with
the expression in tall trees but they were unable
to determine whether any other of the 8 genes
in this QTL region also control height in
these populations.
Here, we aim to determine the
correlation between height (HT) and yield
component traits such as bunch number
(BN) and fresh fruit bunch yield (FFB) and
identify genes controlling height, targeting
three candidate genes explained above
including the DELLA gene, GA2 -oxidase
and asparagine synthase. Several oil palm
populations differing in age and planting
locations were used in this study. The
correlation between HT, BN and FFB was
first determined in two different populations
with the available trait information.
Next, to identify the genes controlling
height, the height contribution of each
polymorphic locus of each candidate gene
was compared using ANOVA analysis and the
association analysis. To perform the
association by TASSEL analysis, the
population structure was first performed
using STRUCTURE. The RNA expression
but still challenging. Our study provides some information that would be useful for oil palm
variety improvement in the near future.
Keywords: oil palm, DELLA, height, dwarf, bunch number, fresh fruit bunch yield
34 Chiang Mai J. Sci. 2019; 46(1)
of genes with significant association with
height was investigated using RT-PCR in
various tissues.
2. MATERIALS AND METHODS
2.1 Plant Materials
Four oil palm populations and resources
were used in this study, coming from four
different growing regions in Thailand, named
as GT, SIT, KU populations and CPI
resource. The GT population samples were
kindly provided from the Golden Tenera
Company Limited, Krabi, Thailand. This GT
population resulted from 30 crosses of
crossing between 6 female parents and 5 male
parents. The female parents were Dura fruit
type, consisting of A 43/9D, A 1/2D, R 15/
14D, R 8/9D, R 10/1D and R 10/5D while
the male parents were Pisifera fruit type,
consisting of R 9/8P, R 5/21P, R 3/8P, KA
17/2P and R 16/7P. The GT population
contained 180 individuals planted in 2008.
The SIT population samples were kindly
provided by the Sitthiporn Kridakorn
Research Station, Kasetsart University,
Prachuap Khirikhan, Thailand. This SIT
population consisted of 6 Tenera-type
commercial oil palm lines, including
Univanich, Nigeria, Uti, PaoRong, Papua
New Guinea and Ekona. The SIT population
contains 132 individuals (22 individuals
from each commercial line) and was planted
in 2008. The KU population samples were
kindly provided by the oil palm technology
development for commercial bio diesel
industry in newly planted area project
(OPTD), Department of Agronomy, Faculty
of Agriculture, Kasetsart University,
Saraburi, Thailand. This KU population
consisted of 8 Tenera-type commercial oil
palm lines, including Golden Tenera, Surat
Thani 2, Surat Thani 3, Surat Thani 4, Surat
Thani 5, Surat Thani 6, Univanich and Uti.
The KU population contains 144 individuals
(18 individuals from each commercial line)
that were planted in 2010. The fourth oil
palm resource was kindly provided by CPI
Agrotech Company Limited, Chumphon,
Thailand. This CPI oil palm resource
contained 20 oil palm individuals. All
individuals were self-fertilized from an oil
palm named S1-47 (Tenera-type). These oil
palm individuals were planted in 2008.
2.2 Phenotype Collecting
Phenotypes used in this study include
height (HT, cm/palm), bunch number
(BN, bunch number/palm/year) and fresh
fruit bunch yield (FFB, kg (kilogram)/palm/
year). For the GT population, BN and FFB
was recorded in three successive years,
2013 (5 year palm), 2014 (6 year palm) and
2015 (7 year palm) while HT was recorded
from 8 year oil palm in March 2016,
October 2016 and March 2017. For the KU
population, BN and FFB was recorded in one
year, 2015-2016 (6 year oil palm), while HT
was recorded 3 times from 5-6 year palm,
in September 2015, March 2016 and August
2016. For the SIT population, only HT was
recorded over 2 successive six months from
8 year oil palm in February 2016 and
September 2016. Also, HT of the CPI
resource was recorded once on March 2016
in 8 year oil palm.
2.3 Primer Designing, and Genotyping of
SSR and Gene-based Markers
Oil palm leaf samples were subjected
to DNA extraction by using DNA extraction
kits, including DNeasy Plant Mimi Kit
(QIAGEN, Germany) and Invisorb Spin
Plant Mini Kit (STRATEC Biomedical AG,
Germany). Markers used in genotyping
were from both previous studies and newly
designed gene-based markers. The previous
markers included some SSR markers [6],
mEgSSRffb10-8 and mEgACCO-pr2 [7].
Chiang Mai J. Sci. 2019; 46(1) 35
In addition, newly designed gene-based
markers were designed from candidate genes
controlling height, including putative
DELLA protein GAI1 and putative GA2OX2
[4], and asparagine synthase related protein
[5]. New gene-based primers were designed
to have a DNA product of about 150-300
bp (base pairs) covering from the promoter
region (5′ UTR) to the end of genomic
sequences of above candidate genes using
primer3 (http://bioinfo.ut.ee/primer3-0.4.0/
primer3/).
To identify the potential polymorphic
markers controlling height, PCR was
conducted in 2 oil palm groups with the
opposite height score using short and tall
CPI and GT DNA pools as templates. For
the CPI resource, the short oil palms were
pooled from 10 individuals of Dura, Tenera
and Pisifera types, named as CPI 2/8T, CPI
2/10T, CPI 2/21T, CPI 1/25P, CPI 1/26D,
CPI 1/28T, CPI 2/33P, CPI2/31P, CPI2/
30D and CPI 2/27T while the tall oil palms
were pooled from 10 individuals of both
Dura and Tenera types, named as CPI 1/6T,
CPI 1/14D, CPI 1/30D, CPI 1/31 D, CPI
1/34P, CPI 1/38T, CPI 1/40D, CPI2/41T,
CPI 1/41T and CPI 2/44T. For the GT
resource, the short oil palms were pooled
from 10 individuals of Tenera type, named
as T3/28, T2/25, T3/29, T1/27, T2/26,
T1/25, T2/27, T1/26, T2/32 and T6/16.
The tall oil palms were pooled from 10
individuals of Tenera type, named as T5/10,
T5/6, T5/8, T3/10, T5/4, T5/12, T6/7, T5/
14, T2/20 and T5/13. These 20 individuals
from the GT resource were selected from
the shortest and tallest oil palms in the GT
population. After the potential polymorphic
markers controlling height were identified,
they were used for genotyping in the GT,
SIT, KU and CPI populations. The PCR
reaction for the amplification was conducted
as the same as our previous study [7].
2.4 Statistical, Population Structure and
Association Analyses
SPSS 11.5 was used to analyze the
phenotype data, including HT, BN and FFB
in terms of descriptive statistics and the
correlation between the three traits. This
statistics package was also used to analyze
preliminary relationships or association of
polymorphic loci of targeted genes with HT,
BN and FFB, by comparing mean height using
Independent-Samples T Test or One-Way
ANOVA. If significant relationships were
found using the above analysis, the population
structure and association analysis was then
performed. STRUCTURE 2.3.4 (http://
pritchardlab.stanford.edu/structure.html)
and STRUCTURE harvester (http://taylor0.
biology.ucla.edu/structureHarvester/)
were used to analyze population structure
and determine optimal population (K),
respectively. Inferred ancestry of individuals
of optimal K from STRUCTURE output
was used as Q-matrix by setting its value
as covariance in the association analysis of
the targeted markers with HT, BN and
FFB traits, using TASSEL 2.1 (http://www.
maizegenetics.net/#!tassel/c17q9).
2.5 Reverse Transcription Polymerase
Chain Reaction Analysis (RT-PCR) of
Target Genes Controlling Height
To test the expression of various plant
tissues, leaf, stigma, ovary, young and mature
fruit, male flower, and meristem samples were
used. These samples were the same samples
used by our previous study [7] except for the
meristem tissues, which were added for this
study. The meristem was collected from the
top of CPI 1/26D and CPI 1/34P oil palm
trees. These oil palms were two of the 20
samples used in the CPI sources for
identification of markers associated with
height as mentioned above.
Each sample was ground under liquid
36 Chiang Mai J. Sci. 2019; 46(1)
nitrogen to a fine powder using a mortar and
pestle. Total RNA extraction was performed
by using InviTrap® Spin Plant RNA Mini Kit,
Stratec Molecular GmbH (Germany).
The RNA integrity was analyzed on 1%
agarose gel and stained by ethidium bromide
before visualization. The RNA concentration
was measured using NanoDrop®ND-1000.
The total RNA was immediately kept on ice
or at -80 °C for longer use. The RT-PCR
reaction was performed in three replicates for
both target and reference genes using
QIAGEN OneStep RT-PCR Kit, (QIAGEN,
Germany). Cyclophilin 2 (renamed as EgCyp2)
was chosen for the reference gene [8].
The RT-PCR reaction was followed from
our previous study [7].
3. RESULTS
3.1 Details of Height Frequency of the
KU, GT and SIT Populations
In this study, height was recorded
3 times, HT-1, HT-2 and HT-3 for the KU
and GT populations while in SIT population,
it was recorded 2 times, HT-1 and HT-2.
The time interval between each height
recording for all three populations was
6 months. The mean height of the KU
population was the shortest about 114 cm
for HT-1, 136 for HT-2 and 152 cm for
HT-3, with a 19 cm in average increase
within 6 months. The mean height of the
GT population was the tallest of the 3
populations recorded. Its mean height
was 192 cm for HT-1, 231 cm for HT-2
and 263 for HT-3, with a 35.5 cm in
average increase within 6 months. The mean
height of the SIT population was 159 cm for
HT-1 and 175 cm for HT-2, with a 16 cm
increase within 6 months. The height
comparison of the 3 populations is shown
in Figure 1.
Figure 1. Comparison of height recorded over 6 month intervals (HT-1, HT-2 and HT-3) in
years 2015-2016 for 6 year old oil palms of the KU population, in 2016 and 2017 for 8 year
old oil palms of the GT population, and in 2016 for 8 year old oil palms of the SIT population.
Chiang Mai J. Sci. 2019; 46(1) 37
3.2 Correlation Coefficient of Height
with Bunch Number and Fresh Fruit
Bunch Yield of the KU and GT
Populations
The degree of correlation between
height and yield components, including
BN and FFB would help to determine a
target height that could be selected by oil palm
breeders. Only the KU and GT populations
with available HT, BN and FFB trait data
were used to determine the correlations
between the traits. The result of correlation
analysis of the KU population was illustrated
in Table 1.
The height of all 3 records (HT-1, HT-2
and HT-3) was shown to have a weak
to moderate positive correlation at
r = 0.337-0.405 with BN and at r = 0.335-
0.360 with FFB. However, BN was shown
to have a very strong correlation with FFB
at r = 0.828. For the correlation result of the
GT population as shown in Table 2, height
in all HT-1, HT-2 and HT-3 records was
shown to have a weak positive correlation,
at r = 0.215-0.395, with BN and a moderate
correlation at r = 0.254-0.499, with FFB.
In addition, BN was shown to have a strong
correlation with FFB, at r = 0.733
Table 1. Pearson Correlation of HT, BN and FFB of the 144 individuals from the KU
population. Several levels of correlations were found from the three traits including very
strong positive significant correlations (r = 0.893-0.940) for HT-1, HT-2 and HT-3, weak to
moderate positive correlation (r = 0.335-405) between HT, BN and FFB, and very strong
correlation (r = 0.828) between BN and FFB.
** Correlation is significant at the 0.01 level (2-tailed).
HT-1
HT-2
HT-3
BN-2016
FFB-2016
HT-1 HT-2
.940(**)
HT-3
.893(**)
.919(**)
BN-2016
.405(**)
.379(**)
.337(**)
FFB-2016
.359(**)
.360(**)
.335(**)
.828(**)
Table 2. Pearson Correlation of HT, BN and FFB of the 180 individuals from the GT
population. Several levels of correlations were found between the three traits including very
strong positive significant correlation (r = 0.940-0.968) for HT-1, HT-2 and HT-3, weak
positive correlation (r = 0.215-0.395) between HT and BN, moderate positive correlation
(r = 0.254-0.499) between HT and FFB, and weak to strong positive correlation (r = 0.162-
0.733) between BN and FFB.
HT-1
HT-2
HT-3
BN-2013
FFB-2013
BN-2014
FFB-2014
BN-2015
FFB-2015
HT-1 HT-2
.968(**)
HT-3
.940(**)
.953(**)
BN-2013
.390(**)
.392(**)
.386(**)
FFB-2013
.499(**)
.480(**)
.496(**)
.550(**)
BN-2014
.391(**)
.387(**)
.395(**)
.480(**)
.376(**)
FFB-2014
.403(**)
.404(**)
.440(**)
.232(**)
.675(**)
.645(**)
BN-2015
.256(**)
.282(**)
.215(**)
.302(**)
.062
.406(**)
.162(**)
FFB-2015
.254(**)
.288(**)
.264(**)
.185(**)
.384(**)
.260(**)
.470(**)
.733(**)
** Correlation is significant at the 0.01 level (2-tailed).
* Correlation is significant at the 0.05 level (2-tailed).
38 Chiang Mai J. Sci. 2019; 46(1)
3.3 Marker Designing from DELLA,
GA2 -oxidase (GA2OX2) and Asparagine
Synthase Genes, and Identification of the
Potential Markers Controlling Height
Sequences of DELLA gene
(P5_sc00033.v1.gene762) and GA2OX2
(P5_sc00033.v1.gene644) was obtained from
a public database, Malaysian Palm Oil Board
(MPOB) (http://genomsawit.mpob.gov.my/
genomsawit/). The sequence length of the
DELLA gene (renamed as EgDELLA1),
which was 2695 bp containing 1000 bp of
the 5′ UTR region and 1695 bp of expected
transcribed gene, were used for marker
development. Fourteen primers in the size
length of 203-291 bp have been designed
for covering the whole gene. Primers
1-4 (renamed mEgDELLA1-1 to
mEgDELLA1-4) covered the promoter
region while primers 5-14 (renamed
mEgDELLA1-5 to mEgDELLA1-14)
covered the gene region. For GA2OX2
(P5_sc00033.v1.gene644), the sequence
length, which was 2016 bp, containing
1000 bp of the expected promoter region
and 1016 bp of the gene. Ten primer pairs
in the size range of 213-299 bp have been
designed. Primers 1-5 (mEgGA2OX2-1 to
mEgGA2OX2-5) covered the 5′ UTR
region and primers 6-10 (mEgGA2OX2-6
to mEgGA2OX2-10) covered the gene
sequence. For Asparagine synthase gene,
the genomic sequence (1745 bp) [5] were used
to design 11 primer pairs (mEgAPG1-1 to
mEgAPG1-11).
To identify potential polymorphic
markers controlling height, separated pooled
DNA of short and tall CPI resources
and GT resources was used in PCR
amplification. For the CPI resource, height
of 10 short and 10 tall oil palm individuals
was significantly different (p value = <0.001).
The mean height of short individuals was
98.50 cm (range = 75-120 cm) while the mean
height of tall individuals was 192 cm
(range = 175-215 cm) in 8 year palms. For
the GT resource, height of 10 short and
10 tall individuals was also significantly
different (p value = <0.001). The mean height
of the short individuals was 118.4 cm
(range = 100-133 cm) while the mean
height of the tall individuals was 268.8 cm
(range = 256-296 cm) in 8 year palms. As a
result, three polymorphic markers, including
mEgDELLA1-1, mEgDELLA1-11 and
mEgAPG1-4, have high possibility to
control height. No polymorphic marker was
found for GA2OX2. Then, the polymorphic
markers were next PCR individually in
the CPI and GT resources. The amplicons of
mEgDELLA1-1 were located at the 5′ UTR
of the gene sequence expected to be the
promoter region while the amplicons of
mEgDELLA1-11 and mEgAPG1-4 were
located at the coding region of the genes.
These three markers were next genotyped
in the KU, GT and SIT population to
determine their relationship or association
with height.
3.4 Genotyping of mEgDELLA1-1,
mEgDELLA1-11 and mEgAPG1-4, and
Their Statistical Analyses
Genotyping of mEgDELLA1-1 (291
bp) (Forward primer, 5′ TTTTCGTACA
TTCGGCTCTG 3′ and reverse primer,
5′ TATCAAATGGCACCGGCTAT 3′)
was conducted in the GT and KU populations.
Three alleles were amplified, including the
top band (named as Allele A), the middle
band (named as allele B), and the low band
(named as allele C) (Figure 2A). Five
genotypes were found in the GT population
including genotypes, A/A, A/B, A/C, B/B
and B/C while 4 genotypes were found in
the KU population including A/A, A/B,
A/C and B/B. The number of individuals in
each genotype was different. Two genotypes,
Chiang Mai J. Sci. 2019; 46(1) 39
A/A (61 individuals) and A/B (81 individuals)
were found to make up the majority of the
GT population while three genotypes, A/A
(28 individuals), A/B (69 individuals) and
A/C (39 individuals), make up the majority
of the KU population. The sequencing
of mEgDELLA1-1 amplicon products
confirmed that they were parts of
EgDELLA1 with an expected of 291 bp.
Next, ANOVA analysis was performed
in the GT and KU populations. Only in
the GT population, the mean height
of individuals with genotype A/A was
significantly shorter than that of individuals
with genotype A/B, for HT-1, HT-2 and
HT-3 at p value of 0.023, 0.013 and 0.024,
respectively. The height difference between
individuals containing genotype A/A and
A/B was 19 cm for HT-1, 21 cm for HT-2,
and 22 cm for HT-3 (Table 3). No significant
height difference was found between
genotypes A/C, B/B and B/C because the
number of individuals were small as 5-18
palm individuals. This suggests that the
promoter region (mEgDELLA1-1) which
is about 630 bp upsteam from the
EgDELLA1 start codon is responsible
for height difference in this GT oil palm
population. In addition, no significance
was found among genotypes amplified
by mEgDELLA1-1 in the KU population
even though the individuals with genotype
A/A still displayed shorter (3-6 cm) height
than individuals with genotype A/B in all
3 HT records as shown in Table 3.
Genotyping of mEgDELLA1-11 (262
bp) (Forward primer, 5′ TGTTTTCTTGC
AAGCACTGG 3′ and reverse primer,
5′ CCTGACAATCAAGCTGCTCA 3′) was
conducted in the GT and SIT populations.
Two alleles were found including the top
band (named as allele A) and the low band
(named as allele B) (Figure 2B). Two
genotypes were found in the GT and SIT
populations, including genotype B/B and
A/B. The 115 and 65 individuals of GT
population contained genotypes B/B and
A/B, respectively while 70 and 59 individuals
of the SIT population contained genotypes
B/B and A/B, respectively. After the ANOVA
analysis was performed in the populations,
no height difference was found between
genotype B/B and A/B in the GT population
but a significant height difference was
found in the SIT population (p value = 0.026)
only from the HT-1 record. The height
difference between individuals of the SIT
population containing genotype A/B with
genotype B/B of mEgDELLA1-11 was
13 cm for HT-1 (Table 3).
Figure 2. The pattern of PCR product on
4.5% acrylamide gels for three polymorphic
markers, mEgDELLA1-1 amplified in the
GT population (A.), mEgDELLA1-11
amplified in the SIT population (B.) and
mEgAPG1-4 amplified in the KU population
(C.).
Genotyping of mEgAPG1-4 (Forward
primer, 5′ CGGCTTCGTGCTCTACGA
TA 3′ and reverse primer, 5′ ACAGCGGGA
TCTTTCCATC 3′) was conducted only in
the KU population. The presence of one
allele with the expected size of 191 bp was
found from the genotyping as shown in
Figure 2C. Presence and absence of amplicon
40 Chiang Mai J. Sci. 2019; 46(1)
Table 3. Genotyping results of the 3 markers (mEgDELLA1-1, mEgDELLA1-11 and
mEgAPG1-4, mean height details in each allele locus and statistical analysis of the mean height
difference between each genotype in the KU, GT and SIT populations. The regions amplified
by mEgDELLA1-1 are shown to control height in the GT population, amplified by
mEgDELLA1-11 are shown to control height in the SIT population. Regions amplified by
mEgAPG1-4 are shown to control height in some commercial lines (Surat Thani 5, Surat
Thani 5 and Surat Thani 6) of the KU population.
product were both detected, which named
as genotype A/A and O/O, respectively. The
sequencing of mEgAPG1-4 amplicon
product confirmed that it was part of the
asparagine synthase gene sequence with an
expected size of 191 bp. ANOVA analysis was
conducted in only 3 commercial lines, Surat
Thani 2, Surat Thani 5 and Surat Thani 6
because they were shown to represent the
majority of individuals to have this allele
present. Significant height difference was
found between the allele that resulted in
amplicon product and the allele that resulted
in no product, at p value of 0.005 for HT-1,
0.007 for HT-2 and 0.015 for HT-3. The height
difference between individuals with band
producing alleles and individuals with alleles
that do not produce bands was 18 cm for
HT-1, 20 cm for HT-2 and 18 cm for HT-3
(Table 3).
** Highly significant level at p value < 0.01
* Significant level at p value = 0.01-0.05
N/A = non significance, cm = centimeter, SD = standard deviation
Markers
mEgDELLA1-1
mEgDELLA1-11
mEgAPG1-4
Population
GT
population
KU
population
GT-
population
SIT
population
KU
population
(in Surat
Thani 2,
Surat
Thani 5
and Surat
Thani 6
lines)
Allele
loci
A/A
A/B
A/C
B/B
B/C
total
A/A
A/B
A/C
B/B
total
B/B
A/B
Total
B/B
A/B
Total
A/A
(present)
O/O
(absent)
Total
Number of
individuals in
each locus
61
81
12
18
5
177
28
69
39
5
141
115
65
180
70
59
159
18
36
54
Mean height (cm±SD)
of each record
HT-1
181±35
200±36
194±35
187±28
206±31
192±35
110±19
116±23
113±22
129±16
114±22
191±34
194±36
192±35
165±41
152±27
159±36
105±22
123±20
117±22
HT-2
220±40
241±41
234±43
223±29
247±28
231±40
129±24
132±35
139±30
151±19
134±31
231±40
233±41
232±40
181±43
170±29
175±37
126±24
146±24
139±25
HT-3
250±44
272±44
272±51
254±35
274±29
263±44
120±21
124±28
126±26
140±17
124±26
263±45
263±43
263±44
142±30
160±21
154±26
Statistical analysis using
one-way ANOVA
HT-1
N/A
N/A
0.026*
0.005*
HT-2
N/A
N/A
N/A
0.007**
HT-3
N/A
N/A
0.015*
-0.023*-0.013*-0.024*
Chiang Mai J. Sci. 2019; 46(1) 41
3.5 Association of mEgDELLA1-1 with
Height, Bunch Number and Fresh Fruit
Bunch Yield of the GT Population
Three types of data were used for
association analysis using TASSEL including
genotyping data information from targeted
gene-based markers, phenotype data and
the Q-matrix of the optimal K, analyzed by
STRUCTURE. To analyze the GT population
structure (180 individuals), 40 random SSR
markers [6] were used in the analysis.
STRUCTURE was run 3 times at this setting:
Length of Burnin Period = 50,000, number
of MCMC Reps after Burnin = 50,000 and
K setting = 1-10. The optimal K was
determined by using STRUCTURE Harvester
[9]. The optimal K calculated using the
Evanno method [10] was 3, because the
highest Delta K = 2658 was found for this
value. The gene-based markers included
mEgDELLA1-1, which was shown to
associate with height using ANOVA
analysis, mEgDELLA1-11, which was not
associated with height in the GT population
by ANOVA analysis and mEgACCO-pr2
on chromosome 10, which was reported to
control FFB in oil palm [7]. In addition,
mEgSSRffb10-8, the SSR marker, which
was reported to link with BN and FFB QTLs
on chromosome 6 was also included in
the genotyping data [7]. The phenotypes of
this population included HT-1, HT-2, HT-3,
BN-2013, FFB-2013, BN-2014, FFB-2014,
BN-2015 and FFB-2015. After performing
TASSEL analysis using the GLM model
and 10,000 permutations, mEgDELLA1-1
was found to associate significantly with
HT-1 and HT-2 with a p value of 0.0429
and 0.0261, respectively. This marker was
found to nearly associate with HT-3 with a
p value of 0.0571. In addition, mEgACCO-
pr2 was found to significantly associate with
BN-2015 with a p value of 0.0384. Markers,
mEgDELLA1-11 and mEgSSRffb10-8 did
not associate significantly with any traits.
3.6 The RT-PCR Expression of
EgDELLA1 in Various Tissues
The RNA expression of EgDELLA1
was examined via RT-PCR from various
tissues, including leaf, stigma, ovary, young
and mature fruit, and male inflorescence
and meristem tissues. The primer chosen
for RT-PCR was mEgDELLA1-11 (Forward
primer, 5′ TGTTTTCTTGCAAGCACTGG
3′ and reverse primer, 5′ CCTGACAATCAA
GCTGCTCA 3′) with an expected size of
262 bp. EgCyp2 was used as a reference
control. Expression of EgDELLA1 was
found in almost all tissues, except for in
male inflorescence (Figure 3). The highest
expression of this gene was found in meristem
tissues, and to a lesser extent in leaves and
developing fruits. Results suggest that the
gene contributes to height by controlling
stem elongation.
4. DISCUSSION
4.1 Correlations Between Height,
Vegetative and Yield Component Traits
Height increment in oil palm depends
on several factors, including genotypes,
age, planting area, environment, seasoning,
Figure 3. An RT-PCR for expression of
EgDELLA1 in various tissues was amplified
by primer mEgDELLA1-11 (262 bp),
at 35 PCR cycles. EgCyp2 (163 bp) was used
as a reference control. Lanes 1= leaf from
clone B, 2 = 1 day opened stigma from B15/
9D, 3 = ovary from B15/9D, 4 = young fruit
from clone B, 5 = mature fruit from clone B,
6 = male inflorescence from clone B, 7 =
meristem from CPI 1/26 D, 8 = meristem
from CPI 1/34 P.
42 Chiang Mai J. Sci. 2019; 46(1)
and management practices. This study
showed that age, planting area and seasoning
affected height increment in oil palm.
We found that height increment was low in
young age palm and dry seasons, and was
rapidly increasing during a raining season.
In terms of genotypes, several oil palm
breeding teams, such as ASD Costa Rica
Company, Costa Rica, have successfully
transferred the dwarf gene from E. oleifera
to E. guineensis by several backcrosses of
E. guineensis to the interspecific hybrid and
named the resulting offspring COMPACT
lines. The COMPACT lines contained
several characters such as high yield, disease
resistance and dwarf character and short
leaf from E. Oleifera [11] .
A strong positive correlation was also
found between height and other vegetative
traits, including total plant dry weight
(r = 0.71), crop growth rate (r = 0.65),
leaf area (r = 0.90) and leaf number (r = 0.73)
[12]. However, the height correlation with
FFB was weak to moderately positive
correlation in both the GT and KU
populations. The positive correlation between
seed yield and plant height was reported in
many plant species such as wheat [13, 14],
maize [15] and chick pea [16]. However,
plant height also had a negative correlation
with grain yield in wheat [17], mungbean [18]
and basmati rice [19] .
For intraspecific palm varieties, too
short or tall oil palms are not selected for
the commercial lines of E. guineensis.
The reasons may include the tree architecture
of oil palms, which affect the amount of
photosynthetic reserves, which are stored
as glucose and starch in the tree trunk [20].
This suggests that dwarf oil palms with thick
trunks would have higher yield than the
dwarf palm with thinner trunk, due to more
area for storing food reserves. In the same
way, tall oil palms have much trunk volume
for food reserves but their harvesting
difficulty prohibit from becoming preferable
commercial lines. Consequently, improvement
of a dwarf oil palm with yield component
traits is challenging for oil palm breeders.
4.2 Role of a Stem Elongation Controlled
by DELLA Genes
Using the BlastX tool (http://
blast.ncbi.nlm.nih.gov/), EgDELLA1 was
predicted as a putative DELLA protein
RGL1-like of both E. guineensis and Phoenix
dactylifera (E = 0). It also is highly matched
with other DELLA protein RGL1-likes
of other species such as Citrus clementina
(E = 5e-128), Jatropha curcas (E = 7e-118),
Glycine max (E = 5e-116), Vigna angularis
(E = 9e-115), and Solanum lycopersicum (E =
4 e-119), DELLA protein GAI1 of Glycine
soja (E =1e-110), and GRAS transcription
factor of Medicago truncatula (E = 2e-109).
DELLA proteins are members of the GRAS
protein family, which plays an important
role in diverse processes such as signal
transduction, meristem maintenance and
development [21]. Examples of DELLA
genes include GAI, RGA, RGA-LIKE1
(RGL1), RGL2 and RGL3 in Arabodopsis,
SLR1 in rice and SLN1 in barley. DELLA
proteins take part in GA signaling by
acting as GA nuclear negative regulators.
They can inhibit GA-promoted processes
by modulating GA and ABA pathways [22].
Gain-of-function mutations of DELLA
genes reduced GA signaling, resulting in
dwarfing phenotype while loss-of-functions
of DELLA genes increased GA signaling,
resulting in tall and slender phenotypes
[23]. GA promotes growth by repressing
the DELLA protein via proteosomal
degradation. GA-dependent GID1-DELLA
complex formation, resulting in DELLA
recognition and ubiquitylation by SCFSLY1,
leads to DELLA proteolysis via the 26S
Chiang Mai J. Sci. 2019; 46(1) 43
proteosome. Mutations that occur in any
proteins in the complex formation can affect
the DELLA degradation, such as a mutation
in SLY protein (one F-box subunit of
SCF that catalyze polyubiquitylation of
DELLA protein) that leads to DELLA not
be degraded, resulting in inhibited GA
response [24]. In addition, mutations within
the N-terminal DELLA domain, which is
essential for GA-dependent ribosomal
degradation, or mutations out of the domain
lead to GA insensitivity [25].
Stem growth of palms is involved the
formation of wide stem base without
internodal elongation like in cereal species
such as rice and wheat [26]. Palms have one
terminal growing point at the apex of the
stem, containing apical meristem where
the primary stem growth begins. The highest
expression of EgDELLA1 was found in
apical meristem tissues, and to a lesser extent,
leaves and fruit development. No expression
was found in male inflorescence, suggesting
that the gene is highly expressed in the
active tissues, meristem, which is important
for stem elongation and leaf initiation in oil
palm. This expression corresponds to the
expression of dwarfing genes in wheat,
Rht A1, Rht B1 and Rht D1 [27] and the
expression of OsGAI in rice [28] where they
are highly expressed in stem internodes,
which are important for stem elongation,
compared to expression in other tissues
such as peduncle and peduncle node in
wheat or root and leaf blade in rice. In
addition, DELLA genes were also involved
in modulating flower development and
seed germination.
The polymorphic promoter region of
EgDELLA1 was found to be associated
with variation in stem height in the GT
population. This suggests that this region
may affect the expression of the EgDELLA1
gene. High expression of EgDELLA1 is
linked to reduced height while low expression
of EgDELLA1 results in an increase in
GA response, which leads to tall height.
Some evidence has shown that high
expression of DELLA genes reduced height.
For example, over expression of Rose
DELLA gene, RoDELLA in transgenic
polargonium led to reduced growth,
associated with an increase in node and
branch number and also a delay in
flower and root formation [29]. Moreover,
overexpression of rice SLR1-like 1 (SLRL1)
in normal rice plants induced a dwarf
phenotype [30].
In conclusion, we identified EgDELLA1
as controlling some height variation in
some oil palm genotypes. We propose
that the DELLA genes are GA-negative
regulators which play a crucial role in
controlling height in some oil palm
genotypes but GA-biosynthesis enzymes
may play more crucial roles in other oil
palm genotypes as well. It is possible that
several genes in both GA signaling, GA
biosynthesis and other pathways such as
reserve assimilation, may contribute to
height variation in some genotypes.
Consequently, improvement of dwarfing oil
palm with higher yield is still challenging
because both traits are complex with
several genes and environmental factors
controlling them.
ACKNOWLEDGEMENTS
This research was fully supported by
the National Center for Genetic Engineering
and Biotechnology (BIOTEC), the National
Science and Technology Development
Agency (NSTDA), Thailand, under the
Research Initiative program (grant number:
P-15-51492).
44 Chiang Mai J. Sci. 2019; 46(1)
REFERENCES
[1] Barcelos E., Rios S.D.A., Cunha R.N.V.,
Lopes R., Motoike S.Y., Babiychuk E.,
Skirycz A. and Kushnir S., Front. Plant Sci.,
2015; 6: 1-16.
[2] Corley R.H.V. and Tinker P.B., The Oil
Palm, Oxford:JohnWiley&Sons, 2003.
[3] Rajanaidu N., Kushairi A., Rafii M.,
Din M., Maizura I. and Jalani B.,
Advances in Oil Palm Research, Kuala Lumpur
: Malaysian Palm Oil Board, 2000.
[4] Pootakham W., Jomchai N.,
Ruang-areerate P., Shearman J.R.,
Sonthirod C., Sangsrakru D.,
Tragoonrung S. and Tangphatsornruang
S., Genomics, 2015; 105: 288-295.
[5] Lee M., Xia J.H., Zou Z., Ye J.,
Rahmadsyah, Alfiko Y., Jin J., Lieando
J.V., Purnamasari M.I., Lim C.H.,
Suwanto A., Wong L., Chua N.H. and
Yue G.H., Sci. Rep., 2015; 5: 8232.
[6] Billotte N., Marseillac N., Risterucci A.M.,
Adon B., Brottier P., Baurens F.C.,
Singh R., Herran A., Asmady H.,
Billot C., Amblard P., Durand-Gasselin
T., Courtois B., Asmono D., Cheah S.C.,
Rohde W., Ritter E. and Charrier A.,
Theor. Appl. Genet., 2005; 110: 754-765.
[7] Somyong S., Poopear S., Sunner S.K.,
Wanlayaporn K., Jomchai N., Yoocha T.,
Ukoskit K., Tangphatsornruang S. and
Tragoonrung S., Mol. Genet. Genomics,
2016; 291: 1243-1257.
[8] Yeap W.C., Loo J., Wong Y. and
Kulaveerasingam H., Plant Cell Tiss. Org.
Cult., 2014; 116: 55-66.
[9] Earl D.A. and vonHoldt B.M., Conserv.
Genet. Resour., 2012; 4: 359-361.
[10] Evanno G., Regnaut S. and Goudet J.,
Mol. Ecol., 2005; 14: 2611-2620.
[11] Alvarado A. and Henry J. ASD Oil Palm
Papers, A Biannual Publication of ASD Costa
Rica, Agricultural Services and Development,
2015.
[12] Agyei-Dwarko D., Ofori K. and Kaledzi
P. D. , Agriculture, 2012, 47: 8946-8949.
[13] Jamali K.D. and Ali S.A., Pak. J. Bot.,
2008; 40: 1805-1808.
[14] Khan A.A., Alam M.A., Alam M.K.,
Alam M.J. and Sarker Z.I., Bangladesh J.
Agric. Res., 2013; 38: 219-225.
[15] Halidu J., Abubakar L., Lzge U.A.,
Ado S.G., Yakubu H. and Haliru B.S.,
J. Plant Breed. Crop Sci., 2015; 7: 9-17.
[16] Joshi P. and Yasin M., Indian J. Appl. Pure
Biol., 2015; 30: 97-100.
[17] Villareal R.L., Rajaram S. and Del Toro
E., J. Agron Crop Sci., 1992; 168: 289-297.
[18] Gul R., Khan H., Mairaj G., Ali S.,
Farhatullah and Ikramullah, Sarhad J.
Agric., 2008; 24: 37-42
[19] Ratna M., Begum S., Husna A., Dey S.R.
and Hossain M.S., Bangladesh J. Agric. Res.,
2015; 40: 153-161.
[20] Legros S., Mialet-Serra I., Caliman J.P.,
Siregar F.A., Clement-Vidal A. and
Dingkuhn M., Ann. Bot., 2009; 104:
1171-1182.
[21] Bolle C., Planta, 2004; 218: 683-692.
[22] Eckardt N.A., Plant Cell, 2007; 19: 2970.
[23] Hauvermale A.L., Ariizumi T. and
Steber C.M., Plant Physiol., 2012; 160:
83-92.
[24] McGinnis K.M., Thomas S.G., Soule J.D.,
Strader L.C., Zale J.M., Sun T.P. and
Steber C.M., Plant Cell, 2003; 15:
1120-1130.
[25] Willige B.C., Ghosh S., Nill C., Zourelidou
M., Dohmann E.M.N., Maier A. and
Schwechheimer C., Plant Cell, 2007; 19:
1209-1220.
Chiang Mai J. Sci. 2019; 46(1) 45
[26] Corley R.H.V. and Tinker P.B.H., The Oil
Palm World Agriculture Series Vol. 5,
John Wiley & Sons, 2015.
[27] Pearce S., Saville R., Vaughan S.P.,
Chandler P. M., Wilhelm E.P., Sparks
C.A., Al-Kaff N., Korolev A., Boulton
M.I., Phillips A.L. Hedden P., Nicholson
P. and Thomas S.G., Plant Physiol., 2011;
157: 1820-1831.
[28] Ogawa M., Kusano T., Katsumi M. and
Sano H., Gene, 2000; 245: 21-29.
[29] Hamama L., Naouar A., Gala R., Voisine
L., Pierre S., Jeauffre J., Cesbron D.,
Leplat F., Foucher F., Dorion N. and
Hibrand-Saint Oyant L., Plant Cell Rep.,
2012; 31: 2015-2029.
[30] Itoh H., Shimada A., Ueguchi-Tanaka M.,
Kamiya N., Hasegawa Y., Ashikari M.
and Matsuoka M., Plant J., 2005; 44:
669-679.
