The cacao Criollo genome v2.0: an improved version of the genome for genetic and functional genomic studies.

BACKGROUND: Theobroma cacao L., native to the Amazonian basin of South America, is an economically important fruit tree crop for tropical countries as a source of chocolate. The first draft genome of the species, from a Criollo cultivar, was published in 2011. Although a useful resource, some improvements are possible, including identifying misassemblies, reducing the number of scaffolds and gaps, and anchoring un-anchored sequences to the 10 chromosomes. METHODS: We used a NGS-based approach to significantly improve the assembly of the Belizian Criollo B97-61/B2 genome. We combined four Illumina large insert size mate paired libraries with 52x of Pacific Biosciences long reads to correct misassembled regions and reduced the number of scaffolds. We then used genotyping by sequencing (GBS) methods to increase the proportion of the assembly anchored to chromosomes. RESULTS: The scaffold number decreased from 4,792 in assembly V1 to 554 in V2 while the scaffold N50 size has increased from 0.47 Mb in V1 to 6.5 Mb in V2. A total of 96.7% of the assembly was anchored to the 10 chromosomes compared to 66.8% in the previous version. Unknown sites (Ns) were reduced from 10.8% to 5.7%. In addition, we updated the functional annotations and performed a new RefSeq structural annotation based on RNAseq evidence. CONCLUSION: Theobroma cacao Criollo genome version 2 will be a valuable resource for the investigation of complex traits at the genomic level and for future comparative genomics and genetics studies in cacao tree. New functional tools and annotations are available on the Cocoa Genome Hub ( http://cocoa-genome-hub.southgreen.fr ).

Creation of BAC genomic resources for cocoa ( Theobroma cacao L.) for physical mapping of RGA containing BAC clones.

We have constructed and validated the first cocoa ( Theobroma cacao L.) BAC library, with the aim of developing molecular resources to study the structure and evolution of the genome of this perennial crop. This library contains 36,864 clones with an average insert size of 120 kb, representing approximately ten haploid genome equivalents. It was constructed from the genotype Scavina-6 (Sca-6), a Forastero clone highly resistant to cocoa pathogens and a parent of existing mapping populations. Validation of the BAC library was carried out with a set of 13 genetically-anchored single copy and one duplicated markers. An average of nine BAC clones per probe was identified, giving an initial experimental estimation of the genome coverage represented in the library. Screening of the library with a set of resistance gene analogues (RGAs), previously mapped in cocoa and co-localizing with QTL for resistance to Phytophthora traits, confirmed at the physical level the tight clustering of RGAs in the cocoa genome and provided the first insights into the relationships between genetic and physical distances in the cocoa genome. This library represents an available BAC resource for structural genomic studies or map-based cloning of genes corresponding to important QTLs for agronomic traits such as resistance genes to major cocoa pathogens like Phytophthora spp ( palmivora and megakarya), Crinipellis perniciosa and Moniliophthora roreri.

A new cacao linkage map based on codominant markers: development and integration of 201 new microsatellite markers.

A linkage map of cacao based on codominant markers has been constructed by integrating 201 new simple sequence repeats (SSR) developed in this study with a number of isoenzymes, restriction fragment length polymorphisms (RFLP), microsatellite markers and resistance and defence gene analogs (Rgenes-RFLP) previously mapped in cacao. A genomic library enriched for (GA)(n) and (CA)(n) was constructed, and 201 new microsatellite loci were mapped on 135 individuals from the same mapping population used to establish the first reference maps. This progeny resulted from a cross between two heterozygous cacao clones: an Upper-Amazon Forastero (UPA 402) and a Trinitario (UF 676). The new map contains 465 markers (268 SSRs, 176 RFLPs, five isoenzymes and 16 Rgenes-RFLP) arranged in ten linkage groups corresponding to the haploid chromosome number of cacao. Its length is 782.8 cM, with an average interval distance between markers of 1.7 cM. The new microsatellite markers were distributed throughout all linkage groups of the map, but their distribution was not random. The length of the map established with only SSRs was 769.6 cM, representing 94.8% of the total map. The current level of genome coverage is approximately one microsatellite every 3 cM. This new reference map provides a set of useful markers that is transferable across different mapping populations and will allow the identification and comparison of the most important regions involved in the variation of the traits of interest and the development of marker-assisted selection strategies.

Identification of QTLs related to cocoa resistance to three species of Phytophthora.

This study aimed to compare the genetic control of cacao resistance to three species of Phytophthora: Phytophthora palmivora, Phytophthora megakarya and Phytophthora capsici. The study was conducted on 151 hybrid progenies created in Côte d'Ivoire and grown in a green-house in Montpellier. Phytophthora resistance was screened by leaf-test inoculation with two different strains per species. Selection of the best individuals for resistance to P. palmivora at a 10% selection rate, would lead to a genetic progress of 47% in the disease evaluation for this species and a genetic progress of 42% and 21% for the two other species. A genetic map with a total length of 682 cM was built with 213 markers, 190 AFLPs and 23 microsatellites. QTLs were identified using composite interval mapping. QTLs were found located in six genomic regions. One of these was detected with five strains belonging to the three Phytophthora species. Two other regions were detected with two or three strains of two different species. Three additional QTLs were detected for only one species of Phytophthora. Each QTL explained between 8 to 12% of the phenotypic variation. For each strain, between 11.5% to 27.5% of the total phenotypic variation could be explained by the QTLs identified. The identification of multiple QTLs involved in resistance to Phytophthora offers the possibility to improve durability of resistance in cocoa by a possible cumulation of many different resistance genes located in different chromosome regions using marker-aided selection.

Cacao domestication II: progenitor germplasm of the Trinitario cacao cultivar

Cacao (Theobroma cacao L.) has been cultivated in Central America since pre-Columbian times. The type of cacao cultivated in this region was called Criollo; cacao populations from the Amazon basin were called Forastero. The type of Forastero most commonly cultivated until 1950 was named Amelonado. Historical data show Trinitario cacao to have originated in Trinidad, resulting from natural hybridisation between Criollo and Amelonado Forastero. Doubts persist on the source of the Amelonado Forastero involved in the origin of Trinitario; the Amelonado parent may have come from the Lower Amazon, the Orinoco or the Guyanas. Most of the cacao cultivated worldwide until 1950 consisted of Criollo, Trinitario and Amelonado. From the early 1950s, Forastero material collected in the Upper Amazon region during the 1930s and 1940s began to be employed in breeding programmes. To gain a better understanding of the origin and the genetic basis of the cacao cultivars exploited before the utilisation of germplasm collected in the Upper Amazon, a study was carried out using restriction fragment length polymorphism and microsatellite markers. Trinitario samples from 17 countries were analysed. With molecular markers, it was possible to clearly identify three main genotypes (represented by clones SP1, MAT1-6 and SIAL70) implicated in the origin of most Trinitario clones.

Cacao domestication II: progenitor germplasm of the Trinitario cacao cultivar.

Cacao (Theobroma cacao L.) has been cultivated in Central America since pre-Columbian times. The type of cacao cultivated in this region was called Criollo; cacao populations from the Amazon basin were called Forastero. The type of Forastero most commonly cultivated until 1950 was named Amelonado. Historical data show Trinitario cacao to have originated in Trinidad, resulting from natural hybridisation between Criollo and Amelonado Forastero. Doubts persist on the source of the Amelonado Forastero involved in the origin of Trinitario; the Amelonado parent may have come from the Lower Amazon, the Orinoco or the Guyanas. Most of the cacao cultivated worldwide until 1950 consisted of Criollo, Trinitario and Amelonado. From the early 1950s, Forastero material collected in the Upper Amazon region during the 1930s and 1940s began to be employed in breeding programmes. To gain a better understanding of the origin and the genetic basis of the cacao cultivars exploited before the utilisation of germplasm collected in the Upper Amazon, a study was carried out using restriction fragment length polymorphism and microsatellite markers. Trinitario samples from 17 countries were analysed. With molecular markers, it was possible to clearly identify three main genotypes (represented by clones SP1, MAT1-6 and SIAL70) implicated in the origin of most Trinitario clones.

Mapping QTL for yield components, vigor, and resistance to Phytophthora palmivora in Theobroma cacao L.

Quantitative trait loci (QTL) mapping for agronomic traits was carried out in cocoa (Theobroma cacao L.). Regions of the genome involved in yield, vigor, and resistance to Phytophthora palmivora were identified. Three heterozygous clones, one upper Amazon Forastero (IMC78) and two Trinitario (DR1 and S52), were crossed with the same male parent, a lower Amazon Forastero (Catongo), known to be highly homozygous. Observations were made on progeny over nine consecutive years. One to three QTL related to yield were detected in each of the three populations, located on chromosomes 1, 2, 4, 5, 9, and 10. They explained between 8.1 and 19.3% of the phenotypic variation and showed various levels of repeatability. In IMC78, the QTL detected on chromosome 5 was the most repeatable over years. The QTL for the average individual pod weight on chromosome 4 was the most significant with an LOD of 17.3 and an R2 of 43.7. QTL related to these traits were identified in the same region of the genome in clones of different genetic groups. This suggests that molecular markers can be used to improve cocoa varieties.

Mapping quantitative trait loci for bean traits and ovule number in Theobroma cacao L.

Quantitative trait loci (QTL) mapping for bean traits and the number of ovules per ovary was carried out in cocoa (Theobroma cacao L.) using three test-cross progenies derived from crosses between a lower Amazon Forastero male parent (Catongo) and three female parents: one upper Amazon Forastero (IMC78) and two Trinitario (DR1 and S52). RFLP (restriction fragment length polymorphism), microsatellite, and AFLP (amplified fragment lengthpolymorphism) markers were used for mapping. Between one and six QTL for bean traits (length, weight, and shape index) and one and four QTL for the number of ovules per ovary were detected using composite interval mapping (CIM). Individual QTL explained between 5 and 24% of the phenotypic variation. QTL clusters were identified on several chromosomes, but particularly on chromosome 4. QTL related to bean traits were detected in the same region in both Trinitario parents and in a close region in the upper Amazon Forastero parent. In reference to a previous diversity study where alleles specific to Criollo and Forastero genotypes were identified, it was possible to speculate on the putative origin (Criollo or Forastero) of favorable QTL alleles segregating in both Trinitario studied.

Cacao domestication I: the origin of the cacao cultivated by the Mayas.

Criollo cacao (Theobroma cacao ssp. cacao) was cultivated by the Mayas over 1500 years ago. It has been suggested that Criollo cacao originated in Central America and that it evolved independently from the cacao populations in the Amazon basin. Cacao populations from the Amazon basin are included in the second morphogeographic group: Forastero, and assigned to T. cacao ssp. sphaerocarpum. To gain further insight into the origin and genetic basis of Criollo cacao from Central America, RFLP and microsatellite analyses were performed on a sample that avoided mixing pure Criollo individuals with individuals classified as Criollo but which might have been introgressed with Forastero genes. We distinguished these two types of individuals as Ancient and Modern Criollo. In contrast to previous studies, Ancient Criollo individuals formerly classified as 'wild', were found to form a closely related group together with Ancient Criollo individuals from South America. The Ancient Criollo trees were also closer to Colombian-Ecuadorian Forastero individuals than these Colombian-Ecuadorian trees were to other South American Forastero individuals. RFLP and microsatellite analyses revealed a high level of homozygosity and significantly low genetic diversity within the Ancient Criollo group. The results suggest that the Ancient Criollo individuals represent the original Criollo group. The results also implies that this group does not represent a separate subspecies and that it probably originated from a few individuals in South America that may have been spread by man within Central America.

Ascertaining maternal and paternal lineage within Musa by chloroplast and mitochondrial DNA RFLP analyses.

In banana, the maternal transmission of chloroplast DNA and paternal transmission of the mitochondrial DNA provides an exceptional opportunity for studying the maternal and paternal lineage of clones. In the present study, RFLP combined with hybridization of heterologous mitochondrial and chloroplastic probes have been used to characterize 71 wild accessions and 131 diploid and 103 triploid cultivated clones. In additon to Musa acuminata and Musa balbisiana, other species from the four Musa sections were studied to investigate their contribution to the origin of cultivated bananas. These molecular analyses enable the classification of the Musa complex to be discussed. Results ascertain relationships among and between the wild accessions and the mono- and interspecific diploid and triploid bananas, particularly for the acuminata genome. Parthenocarpic varieties are shown to be linked to M. acuminata banksii and M. acuminata errans, thus suggesting that the first center of domestication was in the Philippines - New Guinea area.

Genetic mapping of resistance factors to Phytophthora palmivora in cocoa.

Phytophthora palmivora causes pod rot, a serious disease on cocoa widespread throughout the producing regions. In order to ascertain the genetic determination of cocoa resistance to P. palmivora, a study was carried out on two progenies derived from crosses between a heterozygous, moderately resistant Forastero clone, T60/887, and two closely related and highly susceptible Forastero clones, one completely homozygous, IFC2, and one partially heterozygous, IFC5. The cumulative size of both progenies was 112 individuals. Plants were subjected to natural and artificial inoculation of P. palmivora in C te d'Ivoire. The genetic maps of T60/887 and of IFC5 were constructed using amplified fragment length polymorphism (AFLP) markers and microsatellites. The map of T60/887 comprised 198 markers assembled in 11 linkage groups and representing a total length of 793 cM. The map of IFC5 comprised 55 AFLP markers that were assembled into six linkage groups for a total length of 244 cM. Ratio of rotten over total number of fruit under natural infection was measured for each tree over two harvests. Artificial inoculations were performed on leaves and pods. These tests were weakly correlated with the pod rot rate in the field. Five quantitative trait loci (QTLs) of resistance were detected for T60/887 but none were common between the three traits measured. Stability and reliability of the experimental procedures are discussed and revealed the difficult use of these artificial tests on adult trees for a good prediction of field resistance.

Diploid Musa acuminata genetic diversity assayed with sequence-tagged microsatellite sites.

The sequence-tagged microsatellite site (STMS) discrimination potential was explored using nine microsatellite primer pairs. STMS polymorphism was assayed by nonradioactive urea-polyacrylamide gel electrophoresis. Genetic relationships were examined among 59 genotypes of wild or cultivated accessions of diploid Musa acuminata. The organization of the subspecies was confirmed and some clone relationships were clarified.

Nonradioactive sequence-tagged microsatellite site analyses: a method transferable to the tropics.

Utilization of existing isozyme analysis facilities to detect sequence-tagged microsatellite site (STMS) polymorphism or any simple sequence repeat (SSR) variation is described. Different parameters concerning the difficulties in transferring molecular techniques to less sophisticated laboratory infrastructures (i.e. tropical outstations) are discussed (e.g. reproducibility, efficacy, precision). Nonradioactive STMS analysis is bound to foster collaborative research between "biodiversity" and "biotechnology" centers.

Use of competitive PCR to assay copy number of repetitive elements in banana.

Banana is one of the most important subtropical fruit crops. Genetic improvement by traditional breeding strategies is difficult and better knowledge of genomic structure is needed. Repeated sequences are powerful markers for genetic fingerprinting. The method proposed here to determine the copy number of nuclear repetitive elements is based on competitive reverse transcription-polymerase chain reaction and can also be used for quantifying cytosolic sequences. The reliability of this method was investigated on crude preparations of total DNA. Variations due to the heterogeneity of crude DNA extracts showed that a single locus reference is needed for accurate quantification. A mapped microsatellite locus was used to normalize copy number measurements. Copy number assay of repetitive elements using this method clearly distinguishes between the two banana subspecies investigated: Musa acuminata spp. banskii and M. acuminata spp. malaccensis. Two repetitive sequence families, pMaCIR1115 and pA9-26, were assayed that cover up to 1% of the M. acuminata genome. Their copy number varied up to six fold between the two subspecies. Furthermore, sequence quantification showed that mitochondrial genomes are present in crude leaf-extracted banana DNA at up to 40 copies per cell.

RFLP mapping in cultivated sugarcane (Saccharum spp.): genome organization in a highly polyploid and aneuploid interspecific hybrid.

Sugarcane cultivars are polyploid, aneuploid, interspecific hybrids between the domesticated species Saccharum officinarum and the wild relative S. spontaneum. Cultivar chromosome numbers range from 100 to 130 with approximately 10% contributed by S. spontaneum. We have undertaken a mapping study on the progeny of a selfed cultivar, R570, to analyze this complex genome structure. A set of 128 restriction fragment length polymorphism probes and one isozyme was used. Four hundred and eight markers were placed onto 96 cosegregation groups, based on linkages in coupling only. These groups could tentatively be assembled into 10 basic linkage groups on the basis of common probes. Origin of markers was investigated for 61 probes and the isozyme, leading to the identification of 80 S. officinarum and 66 S. spontaneum derived markers, respectively. Their distribution in cosegregation groups showed better map coverage for the S. spontaneum than for the S. officinarum genome fraction and occasional recombination between the two genomes. The study of repulsions between markers suggested the prevalence of random pairing between chromosomes, typical of autopolyploids. However, cases of preferential pairing between S. spontaneum chromosomes were also detected. A tentative Saccharum map was constructed by pooling linkage information for each linkage group.

Mitochondrial DNA diversity in wild and cultivated sorghum.

Cultivated sorghum (Sorghum bicolor ssp. bicolor) is classified into five main races on the basis of spikelet morphology. Isozyme analyses provided new insight into the genetic diversity of sorghum and revealed marked geographic grouping, while nuclear restriction fragment length polymorphisms showed racial differentiation and intraguinea race differentiation. Wild sorghum is diploid or tetraploid and African sorghum (S. bicolor ssp. arundinaceum) is classified into four races, that are considered to be progenitors of cultivated sorghum. We performed mitochondrial DNA analyses to compare the diversity of wild and cultivated sorghum and to study the genetic origin of guinea margaritiferum. The same overall patterns were obtained with the different phenogram construction techniques. Our results confirmed the specificity of guinea margaritiferum and demonstrated the presence of two genetic entities within this subrace. Another guinea group was also noted, which corresponded to Asian guinea roxburghii. In wild sorghum, the arundinaceum race appeared to be homogenous, while the verticilliflorum race was separated into two groups, one of which was associated with the arundinaceum race. The diversity observed in cultivated forms was found to be encompassed within the wild pool, except for one guinea margaritiferum group. There did not seem to be any particular relationship between wild races and cultivated races.

A genetic linkage map of Theobroma cacao L.

A linkage map of the cocoa genome comprising 193 loci has been constructed. These loci consist of 5 isozymes, 101 cDNA/RFLPs, 4 loci from genes of known function, 55 genomic DNA/RFLPs and 28 RAPDs. A population of 100 individuals derived from a cross between two heterozygous genotypes was used. Segregation analyses were performed with the JoinMap program. Ten linkage groups, which putatively correspond to the ten gametic chromosomes of cocoa, were identified. The map covers a total length of 759 cM with a 3.9 cM average distance between 2 markers. A small fraction (9%) of the markers deviated significantly from the expected Mendelian ratios.

Maximum-likelihood models for mapping genetic markers showing segregation distortion. 1. Backcross populations.

A maximum-likelihood approach is used in order to estimate recombination fractions between markers showing segregation distortion in backcross populations. It is assumed that the distortions are induced by viability differences between gametes or zygotes due to one or more selected genes. We show that Bailey's (1949) estimate stays consistent and efficient under more general assumptions than those defined by its author. This estimate should therefore be used instead of the classical maximum-likelihood estimate. The question of detection of linkage is also discussed. We show that the order of markers on linkage groups may be affected by segregation distortion.