No Variation at 29 Microsatellites in the Genome of Jatropha curcas.

Genetic variation in the genome of a given species is the basis for natural selection and genetic improvement through selective breeding. We applied 29 microsatellites located on 11 linkage groups to study genetic variation in 276 accessions of J. curcas collected from nine locations in five countries in South America, Asia and Africa to initiate a breeding program. To our surprise, we did not detect any genetic diversity at all 29 microsatellites loci. All the 276 accessions were homozygous at all loci and shared the same genotype at each locus, suggesting no microsatellite variation in the genome of Jatropha curcas. This result is quite unusual, and may have a profound influence on the breeding strategies and genome study of this species.

Introduction

Jatropha curcas (Euphorbiaceae) is oil-bearing tree species tree and grows naturally in countries of the equatorial Americas 1. It has been introduced to other tropical countries in Africa and Southeast Asia (Indian, Indonesia, Malaysia and China) by Portuguese sailors 400 years ago 1. J. curcas is considered a potential bioenergy tree. It does not compete with food crops for fertile land resources 2. Therefore, it was introduced rapidly in various rural programmes with the expectation of producing biodiesel sustainably and profitably 3,4. Genetic variation is the basis for natural selection and genetic improvement through selective breeding. Genetic variation can be measured at DNA level 5. Microsatellites are one of the most preferred co-dominant DNA markers for studying genetic diversity 6 because of their abundance in a given genome, high polymorphism and the ease of genotyping with automatic DNA sequencers. In J. curcas, microsatellites have been recently developed 7,8 and used in constructing a linkage map 8, and studying genetic diversity within and among populations 7,9,10. Results of studies on genetic diversity of J. curcas are controversial 11. Some reported no genetic variation at microsatellite loci 8,10,12, while some reported limited variation 13,14.

To initiate a breeding program for genetic improvement of J. curcas, we collected seeds from nine locations in five countries in South America, Asia and Africa since 2006 and 2007 (Table 1). The collected seeds were germinated in Singapore, and 276 trees were planted and managed under the same condition in our experimental farm in 2007. We observed substantial variations of phenotypic traits, such branch number, tree height and seed yield 10. We isolated DNA from fresh leaves of each tree using a plant DNA isolation kit (Qiagen, Hilden, Germany). DNA was examined in 1% agarose gels to ensure sufficient quality and quantity for PCR. We selected 29 microsatellites from over 1000 markers developed in our lab 8 because they were able to amplify DNA from J. curcas and the related species J. integerrima. Among the 29 microsatellites, 21 named with Jatr- were developed from a J. curcas individual, while the remaining 8 named with Jcuint- were cloned from a hybrid (J. curcas X J. integerrima) individual. Sixteen markers were already mapped onto 11 linkage groups of jatropha 8, and the remaining 13 had not been mapped yet (Table 2). We applied the reference families, which were used to construct the first generation linkage map of jatropha 8, to examine the inheritance patterns of alleles at the 29 loci. All microsatellites were genotyped using the DNA sequencer ABI3730xl (Applied Biosystems, CA, USA) as described previously 8.

Table 1

Sampling locations of Jatropha curcas, Jatropha integerrima and their hybrid (J. curcas x J. integerrima) on three continents

Population	Species	Location	Country/Continent	Number
Jcu01	J. curcas	Java	Indonesia/Asia	32
Jcu02	J. curcas	Sumatera	Indonesia/Asia	48
Jcu03	J. curcas	Yunnan	China/Asia	48
Jcu04	J. curcas	Sichuan	China/Asia	32
Jcu05	J. curcas	Tamil Nadu	India/Asia	32
Jcu06	J. curcas	Sulawesi	Indonesia/Asia	8
Jcu07	J. curcas	Suriname	Suriname/South America	32
Jcu08	J. curcas	Tanzania	Tanzania/Africa	32
Jcu09	J. curcas	Mixed	Not known	12
Jint01	J. integerrima	Singapore	Singapore/Asia	1
Hybrid	J. cu. x J. int.	Singapore	Singapore/Asia	1

Table 2

Genotypes of 276 J. curcas individuals, one J. integerrima individual and one hybrid (J. curcas x J. integerrima) individual in 29 microsatellite loci*

LG	Locus	Genbank Accession no.	J. curcas	J. integerrima	J. curcas x J. integerrima
1	Jcuint091	JF502118	215/215	189/189	189/215
2	Jatr691	EU664836	142/142	173/177	142/173
3	Jatr901	EU664920	159/159	163/169	159/163
3	Jcuint221	JF502150	394/394	373/375	373/394
4	Jcuint034	JF502095	338/338	334/338	334/338
4	Jcuint298	JF502166	229/229	225/227	225/229
5	Jcuint061	JF502107	358/358	340/340	340/358
5	Jcuint192	JF502143	214/214	196/208	196/214
6	Jatr152	EU664742	192/192	188/204	188/192
6	Jatr190	EU664749	95/95	75/75	75/95
7	Jatr220	EU664757	223/223	223/235	223/235
8	Jatr679	EU664757	171/171	186/186	171/186
9	Jcuint079	JF502114	162/162	145/171	145/162
10	Jatr153	EU664743	260/260	242/242	242/260
11	Jatr154	JF502114	462/462	454/454	454/462
11	Jatr229	EU664743	151/151	161/163	151/163
UN	Jatr161	EU664646	250/250	234/246	234/250
UN	Jatr211	EU664755	283/283	259/283	259/283
UN	Jatr219	KC934904	356/356	362/362	356/362
UN	Jcuint240	KC934905	393/393	435/435	393/435
UN	Jatr194	KC934906	126/126	114/114	114/126
UN	Jatr221	EU664758	133/133	146/146	133/146
UN	Jatr030	EU664686	178/178	170/170	170/178
UN	Jatr079	KC934907	195/195	204/206	195/206
UN	Jatr093	KC934908	87/87	89/89	87/89
UN	Jatr095	KC934909	185/185	185/185	185/185
UN	Jatr102	EU664717	390/390	382/394	382/390
UN	Jatr103	KC934910	139/139	135/137	135/139
UN	Jatr105	KC934911	315/315	319/325	315/319

*: All the 276 J. curcas trees showed the same genotypes at each of the 29 microsatellite loci. LG: Linkage group where the markers were mapped. UN: Unmapped. Primer sequences of mapped microsatellites can be found in the paper of Wang et al (2011) 8, while the primer sequences of unmapped markers were submitted to GenBank. The numbers (e.g. 215/215) under each species are the sizes of alleles in base pairs (bp) at each locus.

To further examine microsatellite variation, we blasted the sequences of the 29 selected microsatellites against the draft genome sequence of a J. curcas individual from Japan 15. We also examined the reproduction patterns of J. curcas using artificial pollination. We tested whether the female flowers developed into fruits by covering the female flowers with paper bags to prevent pollinations and by pollinating the female flowers with pollens from individuals of J. curcas or closely related species J. integerrima. If seeds were obtained, we germinated the seeds and genotyped them with the microsatellites to see whether the offspring obtained alleles from the father side.

All the 29 microsatellites amplified specific PCR products in J. curcas and J. integerrima (Table 2). Surprisingly, all the 276 J. curcas individuals were identical and homozygous in all 29 loci, whereas in the individuals from the species J. integerrima and the hybrid, most microsatellites were heterozygous and polymorphic (Table 2 and Figure 1). These results suggest that (A) there is no genetic variation in the 29 microsatellite loci in J. curcas, (B) technical limitations prevent detection of genetic variation; or (C) the J. curcas accessions are cloned from a single tree recently, or (D) J. curcas is haploid or double haploid or (E) J. curcas is capable of apomixes. However, we can rule out the possibilities B-E. As mentioned above, we used one (i.e. ABI3730xl sequencer) of the most advanced genotyping platforms for genotyping microsatellites. The markers showed Mendelian inheritance in the two reference families used for constructing the linkage map of jatropha 8, thus excluding the possibility of errors in genotyping. The 276 trees used in this study were collected from five countries on three continents, and some seeds were directly collected from the wild. Therefore, it is impossible that they were cloned from single individuals recently. Even if all individuals evolved from a single ancestor, after the first introduction 400 years ago (ca. 400 generations), microsatellites should have shown some degree of variation. This is because microsatellites usually show a higher degree of mutation (10-4/generation) than other regions (10-7/generation) of the genome 6. Karyotype analysis showed that J. curcas was diploid, containing 11 chromosome pairs 16, thus excluding the possibility of it being haploid. When the female flowers were covered with paper bags to prevent pollination, the female flowers aborted, whereas using artificial pollination, the female flowers developed into fruits. Microsatellite genotyping showed that the offspring inherited an allele at each locus from the father side (Table 2 and Figure 1). These data indicate that J curcas can reproduce sexually. Thus, our results suggest that there is no microsatellite variation among the 276 individuals.

Fig 1

No microsatellite variation in Jatropha curcas, but some variation in one Jatropha integerrima and two hybrid (J. curcas x J. integerrima) individuals at microsatellite loci Jatr079 and Jatr190. 1-3: thee Jatropha curcas individuals, 4: one J. integerrima individual and 5-6: two hybrid (J. curcas x J. integerrima) individuals. The numbers of each peak are the sizes of allele in base pairs (bp).

Furthermore, in the linkage map of jatropha 8, all 216 mapped microsatellites and 290 SNPs were homozygous in the J. curcas parent, while they were heterozygous in the two hybrid parents. By blasting the sequences of the 22 microsatellites isolated from J. curcas against the draft genome sequence of an individual from Japan 17, we did not find any difference in the number of repeat unit (Figure 2), indicating that there is no microsatellite variation between our J. curcas individual and the individual from Japan. By blasting the eight microsatellites isolated from the hybrid individual against the J. curcas genome sequence, we detected variation of both sequence and repeat length (Figure 2) in 7 of the eight microsatellites. Only in the locus Jcuint221, there was no repeat length difference, probably the sequence of Jcuint221 originated from J. curcas. The sequence similarity of the seven microsatellites ranged from 89% for Jcuint091 to 97% for Jcuint079, while the number of repeat unit changed by 1-10 (Figure 2), indicating that there is microsatellite variation between J. curcas and J. integerrima and within J. integerrima. Yi et al. 10 used 12 SSR primer pairs developed by Sudheer et al 7 to genotype 162 J. curcas samples from five countries. They did not detect any polymorphism and found that all the individuals were homozygous, sharing the same genotype in each locus. In a recent study 12 using 28 microsatellites, no microsatellite variation was detected in J. curcas accessions collected from many places outside Mexico, strongly supporting that there are no microsatellite variation in the genome of J. curcas.

Fig 2

Alignment of DNA sequences of two microsatellite markers (Jatr161 and Jcuint221) against the genome sequence of a J. curcas individual from Japan. At the locus Jatr161 isolated from J. curcas, there is not repeat and sequence variance, whereas at the locus Jcuint221 isolated from the hybrid (J. curcas x J. integerrima), both DNA sequences and the number of repeats are different.

The phenomenon of no microsatellite variation, but high phonotypic variation in J. curcas is quite unusual and worthy further studying. Epigenetic variations may play an important role in the phenotypic difference 10. Examining epigenetic patterns across the whole genome and in many individuals of J. curcas using whole genome bisulfite sequencing 18 may shield new insights into how the epigenetic variations contribute to the phenotypic variation. In fact, understanding more about the genetic and epigenetic variations in J. curcas and the mechanism underlying phenotypic variations are surely helpful in designing breeding programs to improve the oil yield of J. curcas for sustainable and profitable production of biodiesel.