See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/237077424

Genetic diversity and population structure of Moringa oleifera

Article  in  Conservation Genetics · June 2013

DOI: 10.1007/s10592-013-0503-x

CITATIONS READS

34 2,439

5 authors, including:

Awais Khan Muhammad Jaskani

Cornell University University of Agriculture Faisalabad
120 PUBLICATIONS   1,870 CITATIONS    138 PUBLICATIONS   1,025 CITATIONS   

SEE PROFILE SEE PROFILE

Iqrar Ahmad Khan

50 PUBLICATIONS   304 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Improved potato varieties and water management technologies to enhance water use efficiency, resilience, costeffectiveness, and productivity of smallholder farms in
stress-prone Central Asian environments View project

research publication View project

All content following this page was uploaded by Muhammad Jaskani on 20 May 2014.

The user has requested enhancement of the downloaded file.

Conserv Genet
DOI 10.1007/s10592-013-0503-x

RESEARCH ARTICLE

Genetic diversity and population structure of Moringa oleifera

Umbreen Shahzad • M. Awais Khan •
Muhammad Jaffar Jaskani • Iqrar Ahmad Khan •

Schuyler S. Korban

Received: 19 December 2012 / Accepted: 26 May 2013

Ó Springer Science+Business Media Dordrecht 2013

Abstract Moringa is a genus of the tropical flowering collected in Pakistan; whereas low genetic diversity is
plant family Moringaceae containing 13 diverse species. detected in cultivated accessions obtained from ECHO.
Among the different species, only Moringa oleifera L. is Taken together, these results agree with previous reports
cultivated. This species has great potential in serving as a that M. oleifera is native to the Indo-Pakistan ecological
high-value crop for food, medicinal products, as well as region, and provides sufficient diversity for genetic
fodder for animals, particularly in developing tropical exploration as well as for genetic improvement efforts.
regions of the world. In this study, the genetic diversity and
population structure of world-wide collections of M. ol- Keywords Moringa  Germplasm  Microsatellite
eifera were investigated using DNA markers. A total of 19 markers  Phylogenetic relationships  SSRs  atpB
microsatellite or simple sequence repeat (SSR) markers
along with a partial sequence of the chloroplast gene atpB
were used to study genetic diversity within 161 accessions Introduction
of M. oleifera collected from Asia, Africa, North and South
America, and the Caribbean. On average, 8.3 alleles/per Moringaceae are old-world perennial soft-wood trees that
SSR were amplified in each accession. A total number of are distributed in tropical regions of the world. These trees
158 alleles were detected in 131 accessions collected from indigenous to the western and sub-Himalayan tracts,
the wild in Pakistan and from 30 accessions obtained from including India, Pakistan, Asia Minor, Africa, and Arabia
ECHO (Florida). Observed heterozygosity varied from (Somali et al. 1984), but have now spread to other regions
0.16 to 0.86, with an average of 0.58, while the average of the world, including the Philippines, Cambodia, Central
PIC value was 0.59. Partial sequencing of chloroplast genes America, North and South America, and the Caribbean
of 43 of 161 plants generated mixed patterns. These find- Islands (Anwar et al. 2007). A total of 13 tropical and
ings have demonstrated that there is a large genetic subtropical species of the Moringa genus are known, and of
diversity present in wild collections of M. oleifera these, many are in danger of extinction, including M. ar-
borea, M. borziana, M. longituba, M. rivae, M. ruspoliana,
and M. stenopetala (Stephenson and Fahey 2004). Among
U. Shahzad  M. J. Jaskani  I. A. Khan
Institute of Horticultural Sciences, University of Agriculture, all Moringa species, only M. oleifera L. is cultivated
Faisalabad, Pakistan (Sanchez et al. 2006a).
M. oleifera serves as an important food, and is receiving
M. A. Khan  S. S. Korban (&)
growing attention. Referred to as the ‘natural nutrition of
Department of Natural Resources & Environmental Sciences,
University of Illinois, Urbana, IL 61801, USA the tropics’, leaves, fruits, flowers, and immature pods of
e-mail: [email protected] this tree are highly nutritious, and used in many countries,
particularly in India, Pakistan, Philippines, Hawaii, and in
Present Address:
many countries in Africa (Anwar and Bhanger 2003).
M. A. Khan (&)
International Potato Center (CIP), P.O. Box 1558, Lima 12, Peru Moringa leaves have been reported to be rich sources of
e-mail: [email protected] b-carotene, protein, vitamin C, calcium, and potassium,

123
 Conserv Genet

and serve as sources of natural antioxidants including The aim of this study is to investigate levels of genetic
ascorbic acid, flavonoids, phenolics, and carotenoids (Dil- diversity within and among world-wide accessions of M.
lard 2000; Siddhuraju and Becker 2003). In the Philippines, oleifera, and to detect patterns of genetic structure within
it is referred to as ‘mother’s best friend’ as it is consumed this genome. These studies will be useful for pursuing
to increase a woman’s milk production, and it is sometimes future conservation strategies and for exploiting this
prescribed for treating anemia (Siddhuraju and Becker genetic diversity in breeding and genetic improvement
2003; Estrella et al. 2000). In addition to food uses, programs.
M. oleifera is used for animal feed (Sanchez et al. 2006a).
When supplemented to the diet of dairy animals, M. ol-
eifera leaves improve dry matter intake, digestion, and Materials and methods
milk production, but without affecting smell, taste, or color
of milk (Sanchez et al. 2006b). Moreover, it serves as a Plant material
natural coagulant for treatment of turbid water (Suarez
et al. 2003; Bhatia 2007), as well as a source of phyto- M. oleifera accessions were surveyed, and young leaves
medical compounds (Anwar et al. 2007). were collected from trees growing in different locations
Thus far, there is limited knowledge of available representing 10 districts of Punjab province in Pakistan
genetic diversity present in Moringa species in general (Fig. 1a; Table 1). From each district, five locations were
and of M. oleifera in particular. However, substantial surveyed (distances between locations were *10 km
variation in quantitatively inherited traits has been doc- apart), and within each of these locations, leaves were
umented in natural populations of Moringa in India collected from three different trees.
(Ramachandran et al. 1980). Moreover, limited studies Young leaves were collected, and immediately placed
have been conducted using DNA-based markers to on ice in a cooler, and transferred the same day to the lab
identify and assess diversity among various genotypes of for DNA extraction. Further, young fresh leaves of 30
M. oleifera. Muluvi et al. (1999) and Ulloa (2005) used accessions of M. oleifera were obtained from ECHO
amplified fragment length polymorphism (AFLPs) to (Educational Concerns for Hunger Organization) in Flor-
investigate M. oliefera populations present in Kenya, ida. These accessions were obtained from nine countries
revealing significant differences between regions and (Fig. 1b; Table 2).
populations. Mgendi et al. (2010) identified genetic
variations between cultivated and non-cultivated popula- Genomic DNA extraction and quantification
tions of M. oleifera present in Tanzania using 12 random
amplified polymorphic DNA (RAPD) primers, and simi- Genomic DNA was extracted using CTAB (Doyle and
larly Da Silva et al. (2012) have used RAPD markers to Doyle 1987) extraction protocol with minor modifications
assess the genetic diversity of 16 Moringa accessions as in (Khan et al. 2012). A total of 25 mg of leaf tissue
present in Brazil, concluding that conservation strategies from each accession was ground into a fine powder in
should be adopted for these plants. Recently, Wu et al. liquid nitrogen. The powder was then placed in 50 mL
(2010) have developed 20 microsatellite or simple Falcon tubes containing 15 mL of 2 % CTAB extraction
sequence repeat (SSR) markers for M. oleifera that can buffer [20 mM EDTA, 0.1 M Tris–HCl-pH 8.0, 1.4 M
serve as useful markers for pursuing additional genetic NaCl, 2 % CTAB, plus 1 % b-mercaptoethanol, added
diversity studies. immediately prior to use]. Tubes were incubated at 65 °C
Besides the nuclear genome, chloroplast and mito- for 40 min; inverting each tube after 10 min. Fifteen mL of
chondrial genomes are also widely used in assessing phy- chloroform-isoamylalcohol (24:1) was added to the solu-
logenetic relationships among species in plants. As tion, vortexed for 10 s, and centrifuged at 10,000 rpm for
chloroplasts have more conserved genes, they are particu- 10 min. The supernatant was then transferred to a fresh
larly useful for identification of plants at higher taxonomic tube. Cold isopropanol (-20 °C) was added to the super-
levels (divisions and classes), but are also useful for lower natant (0.7 of total volume of supernatant collected),
taxonomic (genera and species) levels (Logacheva et al. samples were gently mixed, and centrifuged at 10,000 rpm
2007). The chloroplast atpB gene has been extensively for 5 min. DNA pellets were washed twice with 70 %
used to investigate phylogenetic relationships of various ethanol, allowed to dry for approximately 12 h at room
genera and species, such as Lardizabalanceae (Hoot et al. temperature, re-suspended in 100 lL TE buffer solution,
1995) and Capsicum (Walsh and Hoot 2001), among oth- and stored at -20 °C. DNA was quantified using a
ers. Recently, Nock et al. (2011) revised the phylogenetic NanoDrop Spectrophotometer (Nanodrop technologies
relationships among rice species using chloroplast gene Inc., Wilmington, DE) at 230, 260 and 280 nm and on 2 %
sequencing. agarose gel.

123
Conserv Genet

Fig. 1 a A map showing

locations within Pakistan where
Moringa oleifera samples were
collected and used in this study.
Scale is in km and showing
sampling areas. b A world map
showing origins of Moringa
oleifera accessions obtained
from ECHO (Educational
Concerns for Hunger
Organization), Ft. Myers,
Florida, USA. (Color figure
online)

PCR amplification and fragment analysis USA) using the following conditions: initial denaturation at
95 °C for 5 min, 94 °C for 1 min, 35 cycles of 1 min for
A total of 20 SSR markers (Stephenson and Fahey 2004) primer annealing temperatures of 54–56 °C, 1 min at
were used for PCR amplification, but with some modifi- 72 °C; and 8 min at 72 °C for the final extension, holding
cations. An M13 (5-TTTCCCAGTCACGACGTT-3) tail at 4 °C. Following PCR amplification, two to three SSR
was added to forward and reverse primers (Integrated DNA products with a difference of at least 50 bp of expected
Technologies, Coralville, IA). PCR reactions were carried amplicon sizes were multiplexed.
out in a 96-well plate with a final volume of 10.0 lL, An ABI 3730xl sequencer (Applied Biosystem, Inc.
containing 0.1 lL Go Taq Flexi polymerase (5 U) (Pro- Foster City, CA, and USA) at the W. M. Keck Centre at the
mega, Madison, WI, USA), 1.0 lL 59 GoTaq flexi buffer University of Illinois, Urbana-Champaign (USA) was used
(Mg? Free), 0.7 lL MgCl2 (25 mM), 0.25 lL forward for fragment analysis with 15 s injection time as described
primer (10 lM), 0.25 lL reverse primer (10 lM), 0.15 lL by Potts et al. (2012). We used LIZ 500 and LIZ 600 as size
FAM labeled M13 forward primer (10 lM), 0.2 lL dNTP standard, depending on the expected size on SSR fragments
(10 mM each), 6.35 lL H2O, and 1 lL template DNA for estimation of allele sizes.
(20 ng lL-1). The PCR reaction was carried out in a SSR alleles were visualized and scored using GeneM-
Thermoelectron PCR thermal cycler (Thermoelectron, apper version 3.7 (Applied Biosystem, Inc., Foster City,

123
 Conserv Genet

Table 1 A listing of 131 Moringa oleifera accessions sampled from Table 1 continued
10 districts in Punjab, Pakistan used in this study along with their
corresponding ID and location within the district Code no. Plant ID Location atpB

Code no. Plant ID Location atpB District Muzaffargarh

44 MM1 Mehmoodkot H
District Bahawalpur
45 MM2 Mehmoodkot
1 BC1 Chandrani H
46 MM3 Mehmoodkot
2 BC2 Chandrani
47 MT1 Taleeri
3 BC3 Chandrani
48 MT2 Taleeri
4 BI1 Izzat
49 MT3 Taleeri
5 BI2 Izzat
50 MB1 Bukhi
6 BI3 Izzat
51 MB2 Bukhi
7 BR1 Raman H
52 MB3 Bukhi
8 BR2 Raman H
53 MK1 Kenal Mastoi H
9 BF1 Fatto Wali
54 MK2 Kenal Mastoi
10 BF2 Fatto Wali H
55 MK3 Kenal Mastoi
11 BF3 Fatto Wali
56 MS1 Sarai Kotadu
12 BK1 Basti Kheeri H
57 MS2 Kotadu
13 BK2 Basti Kheeri
58 MS3 Kotadu
14 BK3 Basti Kheeri
District Multan
District Rajanpur
59 MNB1 Barar
15 RM1 Basti Machi
60 MNB2 Barar
16 RM2 Basti Machi
61 MNB3 Barar
17 RM3 Basti Machi
62 MNS1 Shuja-abad
18 RH1 Basti Hamoon H
63 MNS2 Shuja-abad H
19 RH2 Basti Hamoon
64 MNS3 Shuja-abad
20 RH3 Basti Hamoon
65 MNP1 Phatak Bagh
21 RB1 Basti Bhanjar H
66 MNP2 Phatak Bagh
22 RB2 Basti Bhanjar
67 MNP3 Phatak Bagh
23 RB3 Basti Bhanjar
68 MNC1 CCRI
24 RA1 Basti Akilpur H
69 MNC2 CCRI H
25 RA1_P2 Basti Akilpur
70 MNC3 CCRI
26 RA2 Basti Akilpur
71 MNA1 Aarapul H
27 RM1 Basti Meeran
72 MNA2 Aarapul
28 RM2 Basti Meeran
73 MNA3 Aarapul
District Hafizabad
District Lodhran
29 HK1 Kot Saleem H
74 LOA1 Arifwala
30 HL1 Lakely H
75 LOA2 Arifwala
District Khanewal
76 LOA3 Arifwala
31 KK1 Kalhory Wala
77 LOM1 Malkan
32 KK2 Kalhory Wala
78 LOM2 Malkan
33 KP1 Pull Akil
79 LOM3 Malkan
34 KP2 Pull Akil H
80 LoR1 Rajapur
35 Ky1 Yousaf Wala
81 LoR2 Rajapur
36 KY2 Yousaf Wala
82 LOK1 Kasaiwala H
37 KY3 Yousaf Wala H
83 LoK2 Kasaiwala
38 KM1 Miyan Chanoo
84 LoK3 Kasaiwala
39 KM2 Miyan Chanoo
85 LoD1 Dunyapur H
40 KM3 Miyan Chanoo H
86 LOD2 Malkan H
41 KA1 Abidabad
District Faisalabad
42 KA2 Abidabad
87 FF1 Forest Nursery
43 KA3 Abidabad

123
Conserv Genet

Table 1 continued CA, USA). A default setting in GeneMapper version 3.7

Code no. Plant ID Location atpB
was used for SSR data analysis, and then alleles at each
locus were manually reviewed and corrected, if necessary,
88 FF2 Forest Nursery for final genotypic scores. SSR markers showing unspecific
89 FF3 Forest Nursery bands and having missing data were removed from further
90 FA1 AARI analysis.
91 FA2 AARI
92 FA3 AARI Genetic diversity and population genetic analysis
93 FS1 Samandri
94 FS2 Samandri Basic genetic features of SSRs and diversity indices for
95 FS3 Samandri accessions were calculated using PowerMarker V3 (Liu and
96 FG1 Gatwala Muse 2004). This includes estimation of heterozygosity,
97 FG2 Gatwala polymorphism information content (PIC), number of
98 FG3 Gatwala alleles, and total number of genotypes for each locus.
99 FJ1 Jaranwala A Bayesian approach implemented in STRUCTURE soft-
100 FJ2 Jaranwala H ware (Pritchard et al. 2000) with an admixture model was
101 FJ3 Jaranwala used to study relationships among different accessions, and
District Layyah the genetic structure in this collection. Prior knowledge of
102 LA1 Ahadnagar geographic regions of accessions was used in the analysis.
103 LA2 Ahadnagar STRUCTURE was run for five times, each with burn-in
104 LA3 Ahadnagar H period of 100,000 iterations and sampling phase of 200,000
105 LI1 Ilyani H replicates assuming 2–10 clusters. The number of sub-
106 LI2 Ilyani
populations was estimated using delta-K values as descri-
107 LI3 Ilyani
bed by Evanno et al. (2005).
108 LK1 Khokran
Amplification and sequencing of the chloroplast gene
109 LK2 Khokran
atpB
110 Lk3 Khokran
111 LW1 Wig H
A total of 43 accessions were selected to represent the
112 LW2 Wig
diversity of the entire population of M. oleifera, based on
113 LW3 Wig
analysis using 19 SSRs, and used to amplify and sequence
114 LS1 Sayyadan H
atpB. The sequence of the chloroplast gene atpB was
115 LS2 Sayyadan
downloaded from NCBI (National Centre for Biotechnol-
116 LS3 Sayyadan
ogy Information) (www.ncbi.nlm.nih.gov/) database. Pri-
District D.G.Khan
mer3V 0.4.0 (Rozen and Skaletsky 2000) was used with
117 DG1 Gilani
default settings to design forward and reverse primers.
118 DG2 Gilani Using 400 bp partial sequence of the atpB gene, the for-
119 DG3 Gilani ward and reverse primers GGCCGTATTGCTCAAATCAT
120 DK1 Khajiwala and TTTCCTCCACGACGATAAGG, respectively, were
121 DK2 Khajiwala designed. The universal M13 primer 5,-AGGGTTTTCC-
122 DK3 Khajiwala CAGTCACGACGTT-3, was appended with the atpB for-
123 DN1 Nawan H ward primer sequence at the 50 -end. These primers were
124 DN2 Nawan synthesized by IDT (Integrated DNA Technologies, Cor-
125 DN3 Nawan H alville, IA).
126 DB1 Bol PCR reactions were carried out as described above for
127 DB2 Bol SSR amplification with FAM labeled M13 primer in a
128 DB3 Bol H 96-well plate in 25 lL volume. The PCR reaction was
129 DJ1 Jaskani carried out in a Thermoelectron PCR thermal cycler using
130 DJ2 Jaskani an annealing temperature of 59 °C for 1 min. The PCR for
131 DJ3 Jaskani every accession was done twice using a 25 lL reaction
Column atpB shows accessions selected, based on diversity analysis volume to increase the quantity of the PCR product for
using 19 SSRs, for further analysis of phylogenetic relationships using later PCR purification. Two sets of PCR for each sample
sequences of the chloroplast atpB gene were pooled together and purified using QIAquick PCR

123
 Conserv Genet

Table 2 A listing of 30 Moringa oleifera accessions along with their Sequence data were imported in Sequencher 5 (Gene
corresponding origin and accession IDs, obtained from ECHO Codes Corporation, Ann Arbor, Michigan), trimmed, and
(Educational Concerns for Hunger Organization), Florida, USA
aligned with default settings according to Han et al. (2009).
Code no. Accession # Source atpB For each accession, both forward and reverse sequences
were assembled using default settings to obtain one con-
South Asia
sensus sequence for each plant.
1 00045-011A PKM-1 Horti, India H
2 03005-031A PKM-2 U. Asmar, India H
Phylogenetic analysis based on atpB
3 91070 Pocha Exports, India
4 03056-031D Trust Hospital, India
Consensus sequences for each of 43 accessions in FASTA
Africa
format were uploaded in molecular evolutionary genetics
5 03052-031D CWS Senegal H
analysis (MEGA 5) (Tamura et al. 2011), and MUSCLE
6 98018 Malawi, Tanzania H
(Edgar 2004) alignment with default options. Aligned
7 03034-031D Msingi, Tanzania
sequences were used for phylogenetic analysis in MEGA 5
8 01088-011D Optima, Tanzania H (Tamura et al. 2011) using Nucleotide as substitution
9 03066-031H Optima, Tanzania model, maximum composite likelihood as method with
10 03055-031D Groves, Mozambique H transitions and transversions, uniform rate as rates among
11 03053-031D Binga Trees, Zimbabwe H sites, and un-weighted pair group method with arithmetic
North America mean (UPGMA) as the statistical model.
12 02099-021D Bradenton, Florida
13 92028-991E ECHO Farm
14 FL92026 Fort-Myers –Florida H Results
15 00099-001D N. Wood, Florida
Central America Features of SSRs and genetic diversity
16 03051-031D Villoria, Belize H
17 01084-011D Mexico H Out of 20 SSRs used in this study, 19 provided good
Caribbean amplification, and were used to assess genetic diversity in
18 01046-011A K. Flanagan, Haiti 161 accessions of Moringa (Table 3). On average, number
19 02055-021H Bohoc, Haiti of alleles over all 19 SSR markers ranged from 6 to 13
20 02073-021H La Gonave, Haiti H (Table 3). A total of 158 alleles were detected in all sam-
21 02057-021H Les Cayes, Haiti ples with a mean number of 8.32 alleles/locus. Gene
22 02056-021H Port Au Prince, Haiti diversity (expected heterozygosity) ranged from 0.29 in
23 03064-031H C. Thede, Haiti MO6 to 0.82 in MO8 with a mean value of 0.64 (Table 3).
24 03065-031H C. Thede, Haiti The observed heterozygosity (Ho) of markers ranged from
25 03067-031H C. Thede, Haiti 0.16 to 0.86 with MO6 being lowest and both MO61 and
26 03068-031H C. Thede, Haiti MO68 the highest, with a mean Ho value of 0.57 for all
27 03069-031H C. Thede, Haiti markers.
28 03070-031H Archai, Haiti H Polymorphic information content values for SSR
29 03071-031H C. Thede, Haiti markers ranged from 0.28 for MO6 to 0.79 for MO8, with
30 02058-021H Titayen, Haiti an overall mean value of 0.59 (Table 3). The most infor-
mative markers were MO13, MO48, MO8, and MO15 with
The column atpB shows accessions selected, based on diversity
PIC values of 0.77, 0.74, 0.79, and 0.79, respectively. A
analysis using 19 SSRs, for further analysis of phylogenetic rela-
tionships using sequences of the chloroplast atpB gene wide range in major allele frequencies was observed, with
a low of 0.26 for MO15 and a high of 0.84 for MO6, with
an average major allele frequency of 0.49 (Table 3). The
purification column (Qiagen Valencia, California, USA) highest number of genotypes detected was 24 with SSR
according to the manufacturer protocols. Purified DNA was marker MO13, while the lowest number of genotypes
quantified using NanoDrop. The final concentration of detected was 11 with SSR markers MO12, MO41, and
purified PCR products was maintained at 20–30 ng/lL for MO45, with an average detectability of 15 genotypes/SSR
sequencing. Purified PCR products were directly marker (Table 3).
sequenced using ABI Prism BigDye Terminator Kit v3.1 Genetic diversity estimates between accessions col-
on ABI 3730xl sequencer (Applied Biosystem, Inc. Foster lected from Pakistan versus those obtained from ECHO
City, CA, USA) as described by Han et al. (2009). were also determined. The number of alleles from the two

123
 Conserv Genet

Table 3 Summary of genetic estimates for 19 SSR markers tested using 161 Moringa oleifera accessions
SSR marker MAF MAF-E MAF-P Ng Ng-E Ng-P Na Na-E Na–P GD GD-E GD-P Ho Ho-E Ho–P PIC PIC-E PIC-P

MO1 0.48 0.83 0.58 12 6 6 7 5 6 0.59 0.30 0.55 0.73 0.27 0.83 0.51 0.28 0.47
MO6 0.84 0.38 0.94 15 11 7 10 7 5 0.29 0.72 0.11 0.16 0.67 0.04 0.28 0.67 0.11
MO8 0.27 0.50 0.22 19 7 17 7 4 7 0.82 0.64 0.81 0.51 0.41 0.53 0.79 0.58 0.78
MO10 0.75 0.50 0.84 13 8 10 7 5 6 0.41 0.65 0.29 0.22 0.50 0.12 0.39 0.59 0.27
MO12 0.44 0.55 0.54 11 7 9 6 5 5 0.68 0.57 0.61 0.80 0.59 0.85 0.62 0.50 0.56
MO13 0.27 0.28 0.32 24 15 9 10 8 6 0.80 0.81 0.74 0.75 0.43 0.83 0.77 0.78 0.69
MO15 0.26 0.72 0.32 12 5 8 9 4 7 0.82 0.43 0.77 0.77 0.27 0.88 0.79 0.37 0.74
MO18 0.47 0.70 0.41 13 7 6 8 5 4 0.67 0.47 0.66 0.77 0.47 0.85 0.62 0.43 0.59
MO41 0.49 0.77 0.49 11 6 8 6 4 5 0.59 0.39 0.58 0.68 0.23 0.78 0.51 0.36 0.49
MO44 0.62 0.39 0.68 10 7 7 7 5 6 0.54 0.70 0.45 0.46 0.18 0.53 0.48 0.65 0.37
MO45 0.47 0.55 0.45 11 6 7 7 4 5 0.64 0.57 0.65 0.43 0.43 0.43 0.58 0.49 0.58
MO46 0.62 0.50 0.74 18 10 10 13 9 7 0.57 0.67 0.43 0.26 0.60 0.18 0.53 0.62 0.40
Mo48 0.29 0.43 0.34 19 9 12 10 6 9 0.78 0.65 0.73 0.62 0.83 0.56 0.74 0.59 0.68
MO55 0.68 0.47 0.79 14 7 10 10 5 7 0.51 0.65 0.36 0.25 0.53 0.18 0.49 0.59 0.35
MO56 0.40 0.53 0.37 20 11 16 8 6 7 0.73 0.65 0.72 0.56 0.43 0.59 0.69 0.62 0.67
MO58 0.46 0.60 0.57 14 6 9 8 5 5 0.68 0.53 0.55 0.54 0.72 0.50 0.63 0.45 0.46
MO61 0.44 0.43 0.54 15 10 6 7 6 5 0.68 0.71 0.57 0.86 0.70 0.90 0.63 0.66 0.49
MO62 0.55 0.63 0.54 15 8 14 7 6 7 0.61 0.57 0.60 0.53 0.57 0.52 0.56 0.54 0.54
MO68 0.48 0.25 0.55 22 15 11 11 9 10 0.69 0.83 0.62 0.86 0.87 0.86 0.65 0.81 0.56
Mean 0.49 0.53 0.54 15 8 10 8 6 6 0.64 0.61 0.57 0.57 0.51 0.58 0.59 0.56 0.52
MAF, Ng, Na, GD, Ho and PIC correspond to major allele frequency, number of genotypes, number of alleles, gene diversity (expected heterozygosity), observed heterozygosity, and
polymorphic information content, respectively. While, suffix –E and –P are estimates for ECHO and Pakistan collections, respectively, while estimates without any suffix correspond to the
combined collections

123
 Conserv Genet

collections was similar in both groups, with some markers Pakistan and ECHO collections, respectively (Table 3).
showing higher numbers of alleles in one collection or the The average number of alleles was 6.3 in the Pakistan
other. For example, in the Pakistan collection, SSR markers collection and 5.6 in the ECHO collection. The heterozy-
MO8, MO15, and MO48 revealed presence of three addi- gosity was also higher (0.57) in the Pakistan collection
tional alleles to those detected in the ECHO collection compared to that (0.51) in the ECHO collection. When the
(Table 3). However, it is noteworthy to point out that the two collections were compared, *75 % of the genetic
number of genotypes for MO8 and MO62 are almost twice variation was attributed to differences among accessions,
in the Pakistan collection compared to those in the ECHO while *25 % of the variation (significant at p = 0.001)
collection. In general, genetic diversity is slightly higher in was due to differences between the two collections.
the ECHO collection, while observed heterozygosity is
slightly higher in the Pakistan collection (Table 3). Over- Population genetic structure
all, the usefulness of markers for differentiating among
genotypes is similar for both collections; however, some M. oleifera accessions collected in Pakistan showed dif-
markers are more informative in one group than the other. ferent genetic identities than those collected from nine
For example, the PIC value for MO6 is 0.67 in the ECHO other countries of the world and maintained at ECHO.
collection and 0.11 in the Pakistan collection; whereas, the Delta-K values (Evanno et al. 2005) calculated using
PIC values for MO15 are 0.74 and 0.37 in the Pakistan and results of STRUCTURE version 2.0 (Pritchard et al. 2000)
ECHO collections, respectively (Table 3). and based on 19 SSRs indicated that those 161 M. oleifera
The average number of genotypes was 9.6 and 8.4 in the accessions were clustered into three groups (Fig. 2a). There
in the Pakistan and ECHO collections, respectively, while is a sharp decrease in delta-K values from K2 to K3, fol-
the total number of alleles was 119 and 108 alleles for the lowed by a plateau at K4. When only two plant populations

Fig. 2 a Graphical plot based A

on delta-K calculated from
Evanno et al. (2005) to estimate
actual number of clusters for
161 Moringa oleifera
accessions used in this study.
Structure (Pritchard et al.2000)
output data based on 19 SSRs
was used for calculating delta-
K. b Assignment of 161
Moringa oleifera accessions
from 10 counties into either two
(K2) or three (K3) clusters using
a Bayesian-based population
genetic structure analysis using
STRUCTURE (Pritchard et al.
2000). Each solid bar represents
a single accession, while
colored areas correspond to
different clusters. Those bars
with multiple colors denote B
admixed genomes of accessions.
(Color figure online)
K2

K3
Ta egal ue

Ind abwe
iq

tan
Zimnzania
Sezamb
Moexico
rida
Flo lize

kis
ia
iti

b
n
Be

Pa
Ha

123
Conserv Genet

(K2) were considered, two groups were identified, one

group (in red) has some of the accessions collected from
Pakistan, while the other group (in green) has all remaining
accessions from Pakistan grouped together with ECHO HL1
accessions (Fig. 2b). At K3 (Fig. 2b), grouping of acces- FS2
sions was more distinct, as suggested by delta-K (Fig. 2a). BC1
Accessions from Pakistan were sub-divided into two RA1

clusters (in green and in orange), while accessions from all FJ2

other nine countries present in the ECHO collection BK1

MNS2
(Fig. 1b; Table 2) grouped together (in purple). The cluster
(in purple) of ECHO accessions was different from those
clusters (in both orange and green) of the Pakistan 0.006 0.005 0.004 0.003 0.002 0.001 0.000

collection. Fig. 3 A phylogenetic tree based on sequences of the chloroplast

Accessions of the ECHO collection showed high geno- atpB gene in 43 Moringa oleifera accessions selected from all 161
mic similarities amongst themselves, but with some accessions used in this study, based on diversity analysis using 19
exceptions. For example, one accession from Florida (Fl SSRs. This analysis was performed using MEGA 5 (Tamura et al.
2011) with nucleotide as a substitution model, maximum composite
92026) and one accession from Tanzania (03066-031H) likelihood as a method with transitions and transversions, uniform
showed 14 and 16 %, respectively, admixture of the gen- rate as rates among sites, and un-weighted pair group method with
ome from one of the clusters (orange) from the Pakistan arithmetic mean (UPGMA) as a statistical model. The solid black-
collection (Fig. 2b). Accessions of the Pakistan collection colored triangle represents collapsed accessions due to high sequence
similarities. Sample codes are in accordance with the geographical
were in either one of two clusters, orange and green, but information listed in Table 1
with few exceptions. Genome compositions of DK1 and
DK2 were similar, 62 and 68 %, respectively, to one of the
clusters (orange), while the remaining portions were shared Discussion
with the second cluster (purple), and similar to genomes of
the ECHO collection. The genome of KM3 revealed con- All SSR markers used in this study have proved to be
tributions from all three clusters at 11 % (purple), 2 % highly informative in assessing genetic diversity present
(orange), and 87 % (green). Moreover, small genome within geographically different populations of M. oleifera.
portions of MB1, MNA1, MNS2, and RA1 appeared to Several previous studies have used SSRs for assessing
have been contributed by other clusters (Fig. 2b). genetic diversity and population structure in tropical tree
species. For example, Dayanandan et al. (1999) and Col-
Phylogenetic analysis based on atpB levatti et al. (2001) have used SSRs to assess population
structure in fragmented populations of tropical trees,
All 43 accessions, selected from a total of 161 accessions, including Carapa guianensis (Meliaceae) and the endan-
sequenced for the partial chloroplast atpB gene exhibited gered tropical tree species Caryocar brasiliense. In this
small variations among the wild collection from Pakistan study, SSR markers have exhibited high levels of poly-
and the cultivated collection from ECHO. On average, a morphism when accessions of the Pakistan and the ECHO
total length of 468 nucleotides with GC content of 43.02 % collections were evaluated. Moreover, high levels of
was obtained for these 43 sequences. Moreover, the num- genetic diversity have been detected in M. oleifera acces-
ber of segregating sites (s) and minimum number of sions. Overall, accessions in the ECHO collection have
mutations (H) were 12. The Theta value (H) for s (per exhibited lower levels of genetic diversity compared to
DNA sequence) was 2.77345, and a variance (with no those of the Pakistan collection. Moreover, these ECHO
recombination) of 1.203 ± 1.097, while the average num- accessions have shown similarities to accessions from a
ber of nucleotide differences (k) was 0.957. Based on single province of Pakistan (Punjab province, Fig. 1a;
phylogenetic analysis, seven accessions including HL1, Table 1) although they have been collected from nine
FS2, BC1, RA1, FJ2, BK1, and MNS2, from different different countries (Fig. 1b; Table 2).
geographical locations of Pakistan, were slightly different Contrary to Wu et al. (2010), higher numbers of alleles
from all other accessions, and were grouped together. The were detected for Moringa SSRs, ranging from 6 to 13
remaining 36 accessions, including all remaining acces- alleles per locus, and within a similar range to that reported
sions from Pakistan together with accessions of ECHO by Dayanandan et al. (1999) for Carapa guianensis, but
were highly similar for the 400 bp partial sequence of the much lower than that reported by Collevatti et al. (2001)
atpB gene (Fig. 3). for Caryocar brasiliense (20–27 alleles per locus). This

123
 Conserv Genet

could be attributed to differences in sampling regions. This Englishman, carried Moringa over to Jamaica earlier in
finding along with higher observed and expected hetero- 1784. Now, Moringa is in cultivation throughout the tropics
zygosity values could be attributed to higher levels of of West Indies, Bermuda (Britton 1918) and in Latin
genetic diversity present in M. oleifera accessions used in America, including Cuba, Haiti, Jamaica, Puerto Rico,
this study. The most informative SSR marker in this study Panama, Belize, Mexico, Brazil, and Venezuela (Correll
was MO8, while SSR marker MO6 was the least infor- et al. 1982). Subsequently, the United States Department of
mative. On average, the numbers of accessions in the Agriculture (USDA) obtained seeds from both Cuba (Bur.
Pakistan collection were higher than those in the ECHO Pl. Indus 1915) and Nicaragua (Bur. Pl. Indus 1923), and
collection (Table 3). Similarly, numbers of alleles and germinating seedling trees were grown as ornamentals in
levels of heterozygosity were higher in accessions of the South Florida and in California. However, it is not clear as
Pakistan collection than those of the ECHO collection. In to whether or not these seeds have been first introduced
addition, the chloroplast atpB gene revealed that seven into Africa from multiple regions of the Indian subconti-
accessions from different geographical regions in Punjab nent, and of the genetic relationships between these intro-
(Pakistan) were somewhat different from all other acces- duced Moringa species (Muluvi et al. 1999; Mgendi et al.
sions used in this analysis (Fig. 3). Taken together, these 2010). Muluvi et al. (1999) have reported that a relatively
findings provided insights into the genetic diversity present small number of accessions introduced from India have
in M. oleifera as well as those genetic relationships among served as the core collection of Moringa populations
accessions grown in different countries. These results present in Africa, thus resulting in low levels of genetic
supported earlier studies reporting that Moringa must have diversity. However, it cannot be ruled out that strong
originated in the northern Indian sub-continent (Kanthara- selection pressure for certain attributes could have also
jah and Dodd 1991; Fahey 2005). However, additional contributed to some of the observed changes. Based on the
accessions from other locations in that region or from other historical background and observed genetic similarities, it
parts of the world should be explored and included in is assumed that accessions of the ECHO collection prob-
future studies to confirm these current findings. Based on ably represent seeds introduced by traders, Indian expa-
the phylogenetic analysis using the chloroplast gene atpB triates, and British colonialists into these countries.
(Logacheva et al. 2007), accessions from Punjab (Pakistan) Moreover, as accessions from India (PKM1, PKM2, 91070,
were more diverse compared to accessions present in the and 03056-031D) are in fact advanced selections origi-
ECHO collection, thus providing further supporting evi- nating from Southern India and show more similarities to
dence to the northern Indian origin of Moringa. accessions from other countries instead of those accessions
It was interesting to note that accessions in the ECHO from Punjab, a northern region of the sub-continent, it is
collection were quite similar despite the fact they have likely that the export pathway of Moringa is likely to have
been collected from nine different countries. This sug- originated from the coastal regions of India, the region with
gested that these accessions must be derived from a com- most movement of goods and people in the nineteenth
mon population or from a few populations. As previously century.
reported, Moringa species have been recently introduced Findings obtained in this study also suggest that multiple
from the Indian subcontinent to other countries, either by ancestral populations of M. oleifera within close geo-
traders or Indian immigrants, and grown in backyards as graphical regions must have been introduced to other parts
vegetables and/or for medicinal purposes. In many African of the world. There are four accessions in the ECHO col-
countries, Moringa is commonly grown by people of lection from India, including PKM1, PKM2, 91070, and
Indian descent, suggesting that it might have been recently 03056-031D (Table 2) that have been developed as
introduced from India, and ECHO had indeed received advanced selections, and have been subsequently distrib-
seeds from these households (Beth Doerr from ECHO, uted as commercial cultivars. Moreover, clustering of
Pers. Comm.). accessions at K = 3 (Fig. 2b) strongly suggests that the
Moringa has been introduced to Kenya from India dur- germplasm of the ECHO collection must have predomi-
ing the early years of the twentieth century (Muluvi et al. nantly come from Africa, via India. However over time,
1999). The germplasm in Kenya and India is rather similar, some genetic changes might have occurred. The Pakistan
and those few changes detected in genomes of this germ- germplasm collected from Punjab shows presence of two
plasm have been attributed to adaptation to new environ- clusters; however, it is difficult to discern clear relation-
ments, as well as to the fast-growing growth habit of ships between geographical origin and genetic diversity for
Moringa (Muluvi et al. 1999). British colonialists also these accessions.
played a role in transporting Moringa as an ornamental tree In conclusion, a selected set of 19 SSRs has clearly
from India to Africa (Berger et al. 1984). Early on, Fawcett differentiated the Moringa germplasm originating from
and Rendle (1910) mentioned that Hinton East, an different regions. Findings in this study demonstrate that

123
Conserv Genet

there is a wide genetic diversity present in Moringa. The Fawcett W, Rendle AB (1910). In: Anonymous (ed) Flora of Jamaica,
diversity present in samples collected from Pakistan containing descriptions of the flowering plants known from the
island, Trustees of the British Museum, London
emphasize that there is great potential for genetic Han Y, Chagné D, Gasic K, Rikkerink EHA, Beever JE, Gardiner SE,
improvement of Moringa to enhance its multi-purpose Korban SS (2009) BAC-end sequence-based SNPs and Bin
uses; therefore, various strategies should be explored to mapping for rapid integration of physical and genetic maps in
exploit this available genetic diversity. apple. Genomics 93:282–288
Hoot SB, Culham A, Crane PR (1995) The utility of atpB gene
sequences in resolving phylogenetic relationships: comparison
Acknowledgments We acknowledge the Higher Education Com- with rbcL and 18S ribosomal DNA sequences in the Lardiza-
mission (HEC) of Pakistan for awarding Umbreen Shahzad an IRSIP balaceae. Ann Missouri Bot Garden 82:194–207
Scholarship, which allowed her to complete the present work at the Kantharajah AS and Dodd WA (1991) Rapid clonal propagation of M.
University of Illinois at Urbana-Champaign. We also like to thank oleifera Lam., using tissue culture. South Indian Hortic
Beth Doerr from the Educational Concerns for Hunger Organization 39:224–228
(ECHO) in Ft. Myers, Florida. Khan MA, Han Y, Zhao YF, Korban SS (2012) A high-throughput
apple SNP genotyping platform using the GoldenGateTM assay.
Gene 494:196–201
References Liu K, Muse S (2004) PowerMarker: new genetic data analysis
software. Version 3.0. Free program available online
Anwar F, Bhanger M (2003) Analytical characterization of Moringa Logacheva MD, Penin AA, Samigullin TH, Vallejo-Roman C,
oleifera seed oil grown in temperate regions of Pakistan. J Ag Antonov AS (2007) Phylogeny of flowering plants by the
Food Chem 51:6558–6563 chloroplast genome sequences: in search of a ‘‘lucky gene’’.
Anwar F, Latif S, Ashraf M, Gilani AH (2007) Moringa oleifera: a Biochemistry 72:1324–1330
food plant with multiple medicinal uses. Phytoth Res 21:17–25 Mgendi MG, Manoko MK, Nyomora AM (2010) Genetic diversity
Berger MR, Habs M, Jahn SAA, Schmahl D (1984) Toxicological between cultivated and non-cultivated Moringa oleifera Lam.
assessment of seeds from Moringa oleifera and Moringa provenances assessed by RAPD markers. J Cell Mol Biol
stenopetala, two highly efficient primary coagulants for domes- 8:95–102
tic water treatment of tropical raw waters. East Afr Med J Muluvi GM, Sprent JI, Soranzo N, Provan J, Odee D, Folkard G,
61:712–716 McNicol JW, Powell W (1999) Amplified fragment length
Bhatia S (2007) Coagulation-flocculation process for POME treat- polymorphism (AFLP) analysis of genetic variation in Moringa
ment using Moringa oleifera seeds extract: optimization studies. oleifera Lam. Mol Ecol 8:463–470
Chem Eng J 133:205–212 Nock CJ, Waters DL, Edwards MA, Bowen SG, Rice N, Cordeiro
Britton NL (1918) In: Anonymous (ed) Flora of Bermuda, C. GM, Henry RJ (2011) Chloroplast genome sequences from total
Scribner’s sons, New York DNA for plant identification. Plant Biotech J 9:328–333
Collevatti RG, Grattapaglia D, Hay JD (2001) Population genetic Potts SM, Han Y, Khan MA, Kushad MM, Rayburn AL, Korban SS
structure of the endangered tropical tree species Caryocar (2012) Genetic diversity and characterization of a core collection
brasiliense, based on variability at microsatellite loci. Mol Ecol of Malus germplasm using simple sequence repeats (SSRs).
10:349–356 Plant Mol Biol Rep 30:827–837
Correll DS, Fawcett P, Correll HB (1982). In: Anonymous, J. Cramer Pritchard JK, Stephens M, Donnelly P (2000) Inference of population
(ed) Flora of the Bahama archipelago: including the Turks and structure using multilocus genotype data. Genetics 155:945–959
Caicos Islands, Vaduz Ramachandran C, Peter K, Gopalakrishnan P (1980) Drumstick
Da Silva AVC, Dos Santos ARF, Da Lédo AS, Feitosa RB, Almeida (Moringa oleifera)—a multipurpose indian vegetable. Econ Bot
CS, Da Silva GM, Rangel MSA (2012) Moringa genetic 34:276–283
diversity from Germplasm Bank using RAPD markers. Tropical Rozen S, Skaletsky H (2000) Primer3 on the WWW for general users
Subtrop Agroecosyst 15:31–39 and for biologist programmers. Meth Mol Biol 132:365–386
Dayanandan S, Dole J, Bawa K, Kesseli R (1999) Population Sanchez N, Ledin S, Ledin I (2006a) Biomass production and
structure delineated with microsatellite markers in fragmented chemical composition of Moringa oleifera under different
populations of a tropical tree, Carapa guianensis (Meliaceae). management regimes in Nicaragua. Agrofor Syst 66:231–242
Mol Ecol 8:1585–1592 Sanchez N, Sporndly E, Ledin I (2006b) Effect of feeding different
Dillard CJ (2000) Phytochemicals: nutraceuticals and human health. levels of foliage of Moringa oleifera to creole dairy cows on
J Sci Food Agric 80:1744–1756 intake, digestibility, milk production and composition. Livestock
Doyle JJ, Doyle JL (1987) A rapid DNA isolation procedure for small Sci 101:24–31
quantities of fresh leaf tissue. Phytochem Bull 19:11–15 Siddhuraju P, Becker K (2003) Antioxidant properties of various
Edgar R (2004) MUSCLE: multiple sequence alignment with high solvent extracts of total phenolic constituents from three
accuracy and high throughput. Nucl Acids Res 32:1792–1797 different agroclimatic origins of drumstick tree (Moringa
Estrella MCP, Mantaring JBV, David GZ, Taup MA (2000) A double oleifera Lam.) leaves. J Agric Food Chem 51:2144–2155
blind randomised controlled trial on the use of malunggay Somali MA, Bajneid MA, Al-Fhaimani SS (1984) Chemical compo-
(Moringa oleifera) for augmentation of the volume of breastmilk sition and characteristics of Moringa peregrina seeds and seeds
among non-nursing mothers of preterm infants. Philippines J oil. J Amer Oil Chem Soc 61:85–86
Pediatrics 49:3–6 Stephenson K, Fahey J (2004) Development of tissue culture methods
Evanno G, Regnaut S, Goudet J (2005) Detecting the number of for the rescue and propagation of endangered Moringa spp.
clusters of individuals using the software STRUCTURE: a germplasm. Econ Bot 58:S116–S124
simulation study. Mol Ecol 14:2611–2620 Suarez M, Entenza J, Doerries C, Meyer E, Bourquin L, Sutherland J,
Fahey J (2005) Moringa oleifera: a review of the medical evidence Marison I, Moreillon P, Mermod N (2003) Expression of a plant-
for its nutritional, therapeutic, and prophylactic properties. Part derived peptide harboring water-cleaning and antimicrobial
1. Trees Life J 1:5 activities. Biotech Bioeng 81:13–20

123
 Conserv Genet

Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S Walsh B, Hoot S (2001) Phylogenetic relationships of Capsicum
(2011) MEGA5: molecular evolutionary genetics analysis using (Solanaceae) using DNA sequences from two noncoding
maximum likelihood, evolutionary distance, and maximum regions: the chloroplast atpB-rbcL spacer region and nuclear
parsimony methods. Mol Biol Evol 28:2731–2739 waxy introns. Int J Plant Sci 162:1409–1418
Ulloa M (2005) Molecular markers learning modules: using molec- Wu J, Yang J, Gu Z, Zhang Y (2010) Isolation and characterization of
ular marker technology in studies on plant genetic diversity. twenty polymorphic microsatellite loci for Moringa oleifera
Crop Sci 45:2676–2677 (Moringaceae). HortSci 45:690–692

123

View publication stats