Genetical metabolomics in apples (Malus x domestica Borkh)

The aim of this thesis was finding genes that control the production of potentially health beneficial metabolites in apple fruits. The approach was genetic mapping of secondary metabolites such as phenolic compounds in an F1 progeny, leading to the detection of genetic loci that controlled these metabolites. At these genetic loci candidate genes were identified, using the whole genome sequence of apple, and it was investigated whether the expression of these candidate genes in the F1 progeny correlated with the metabolite levels. The cultivated apple (Malus x domestica Borkh) is among the most diverse and ubiquitously cultivated fruit species. It belongs to the family of Rosaceae which includes many commercial fruit species such as pear, strawberry, cherry, peach, apricot, almond, black cherry, and crab apple. Apple has a haploid chromosome number of 17. It is a self-incompatible and highly heterozygous crop. The breeding is further hampered by the long juvenile period which makes breeding in this crop a very slow process. The saying “An apple a day keeps the doctor away” has encouraged many researchers to search for the “magic” ingredients found in apple. Due to the beneficial role of apple phenolics, it is also called as a “new agrochemical crop”. Apple possesses many health beneficial properties for human beings as it is a rich source of phenolic compounds.It has been associated with reducing the risks of certain diseases such as cancers, particularly prostate, liver, colon, and lung cancers, cardiovascular diseases, coronary heart diseases, asthma, type-2 diabetes, thrombotic stroke, and ischemic heart disease. The second chapter of this thesis describes the construction of genetic linkage maps of the parents of a segregating population derived from the cross between the cultivars ‘Prima’ and ‘Fiesta’. For this purpose the already available linkage maps, as described in this chapter, were made denser by inclusion of 240 Diversity Array Technology (DArT) markers. Thus the total number of markers for ‘Prima’ and ‘Fiesta’ integrated map reached to 820. DArT-markers are hybridization based dominant DNA-markers. DArT provides a high-throughput whole genome genotyping platform for the detection and scoring of hundreds of polymorphic loci without any need for prior sequence information. This is the first report on DArT in horticultural trees. Genetic mapping of DArT markers in two mapping populations and their integration with other marker types showed that DArT is a powerful high throughput method for obtaining accurate and reproducible marker data, at low cost per data point. This method appears to be suitable for aligning the genetic maps of different segregating populations. Sequencing of the marker clones showed that they are significantly enriched for low copy, gene rich regions. Chapter 3 describes metabolic diversity of Malus. Wild germplasm was compared to advanced breeding selections and to the segregating F1 population from the cross between the cultivars ‘Prima’ and ‘Fiesta’. The metabolic profiles were analyzed by means of liquid chromatography-mass spectrometry (LC-MS). LC-MS is an analytical chemistry technique that combines the physical separation capabilities of liquid chromatography with the mass analysis capabilities of mass spectrometry. This resulted in the detection of 418 putative metabolites in the peel and 254 in the flesh. Fruits from 23 wild species, eight advanced selections and the segregating F1 population were analyzed. The data were subjected to Principle Components Analysis (PCA). Variance analysis of the first PC showed that genetic variation accounted for 96.6 % in peel and 97.4 % in flesh of the total metabolic variation. Technical variation accounted for 1.4 % and 0.8%, while environmental variation accounted for 2.0% and 1.8% in peel and flesh respectively. The genetic variation between wild genotypes was very large, compared to the advanced selections and the F1 progeny. Only 8 % of the genetic variation of the first principle component was captured by the advanced selections. This indicates strong genetic erosion during breeding. This genetic erosion was mainly caused by reduction of the levels of several flavonoids including catechin, epicatechin and procyanidins. PCA of the F1 progeny of the ‘Prima’ x ‘Fiesta’ cross showed a clear 3:1 Mendelian segregation of metabolites. These metabolites were 4.2 fold less in both peel and flesh in progeny that had inherited the recessive alleles of a gene at the top of Linkage Group16 (LG16) from the heterozygous parents. We found a separate group of 11 metabolites in peel and 12 in flesh. These metabolites were putatively identified as glycosylated forms of b-glycols: R-octane-1, 3-diol and its unsaturated form R-5-(Z)-octene-1, 3-diol which have a potential role in controlling infection by microorganisms and influence the aroma of some ciders. The levels of these metabolites were up to 50 fold more abundant in some progeny compared to both parents. Genetic mapping showed that this strong increase was caused by one locus at the top of LG8, in progeny that had inherited only the recessive alleles of that locus from the heterozygous parents. This research illustrates not only the strong genetic erosion in apple breeding regarding metabolic diversity, and strong reduction of flavonoids in some progeny, but also shows that inbreeding can lead to a strong increase of metabolites that were present at much lower levels in both parents and advanced selections. This loss and gain of metabolites was especially observed in case of accumulation of recessive alleles during inbreeding. The genetic factors controlling metabolite composition were studied in more detail in Chapter 4. We investigated the genetic factors of the quantitative variation of these potentially beneficial compounds (Chapter 3, 4), by combining the genetic maps (Chapter 2) with the LC-MS data for thesegregating F1 population from the cross ‘Prima’ x ‘Fiesta’. This resulted into metabolite quantitative trait loci (mQTLs). When using the software MetaNetwork, 669 significant mQTLs were detected: 488 in the peel and 181 mQTLs in the flesh. Four linkage groups (LGs) i.e. LG1, LG8, LG13 and LG16 were found to contain mQTL hotspots, mainly regulating metabolites that belong to the phenylpropanoid pathway. These include various metabolites i.e. sinapate hexoside, coumaroyl hexoside, phloridzin, quinic acids, phenolic esters, kaempferol glycosides, quercetin glycosides, cyanidin pnetoside, flavan-3-ols (catechin, epicatechin), and procyanidins. The genetics of annotated metabolites was studied in more detail using MapQTL®. It was found that quercetin conjugates had mQTLs on LG1 and LG13. The most important mQTL hotspot with the largest number of metabolites was, however, detected at the top of LG16: mQTLs for 32 peel-related and 17 flesh-related phenolic compounds. The metabolites that mapped in the mQTL hotspot on LG16 all belong to the phenylpropanoid pathway of secondary metabolites. These compounds showed a monogenic Mendelian inheritance in a 3:1 segregation ratio. Procyanidins dimer II was used as a representative of the numerous compounds that mapped at the LG16 mQTL hotspot. By means of graphical genotyping of this monogenic trait, a genetic window could be made in which the gene that caused the mQTL hotspot should reside. We located structural genes involved in the phenolic biosynthetic pathway, using the genetic map together with the published whole genome sequence of apple. The structural gene leucoanthocyanidin reductase (MdLAR1) was detected in the mQTL hotspot window on LG16, as were seven transcription factor genes. To our knowledge, this is the first time that a QTL analysis was performed on such a high number of metabolites in an outbreeding plant species. The expression of the candidate genes found in the mQTL window on LG16 was studied and discussed in Chapter 5. qPCR was used for this purpose and it was found that the expression of only the structural gene MdLAR1 was strongly positively correlated with the metabolite procyanidin dimer II content. Neither the expression profiles of other structural genes of the phenylpropanoid pathway, the transcription factor genes at the mQTL hotspot, nor of transcription factor genes outside the mQTLs hotspot, showed any significant correlation with the procyanidin dimmer II content that mapped at the mQTL hotpot. This indicates that MdLAR1 was the gene, which caused this mQTL hotspot (Chapter 5). The progeny that had inherited one or two copies of the dominant alleles (Mm, MM) showed on the average a 4.4 and 11.8 fold higher expression level of MdLAR1 respectively, compared to the progeny that had inherited the recessive alleles only (mm). This led to a 4.0 fold increase of procyanidin dimer II level at the ripe stage. Strikingly, at the mQTL hotspot at the top of LG16, there is also a locus that controls acidity of the ripe fruits. However, the dominant alleles for acidity appeared to be in repulsion to the dominant alleles for high metabolite levels (Chapter 6). This shows that acidity is controlled by another gene than the metabolite levels. The combination of the genetic position based on the whole apple genome sequence, annotation of potential genes, and expression profiling indicated that the malic acid transporter gene MdALMT2 was responsible for the clear differences in malic acid content and pH in mature apple fruits of the segregating F1 population. The genetic inheritance of at least one dominant allele (MaMa/Mama) of this gene sufficed for a three-fold increase of the malic acid concentration and a reduction of the pH from 4 to 3 in ripe apples, compared to the presence of only the lower expressed recessive allele (mama). This malic acid transporter gene is located at the top of LG16.Malic acid is the predominant organic acid associated with the pH in apple fruits. It is synthesized in the cytoplasm and transported into the cell vacuole. The concentration of malic acid in the cell vacuole determines the pH of the cell. pH is very important for the overall taste of many fruits, including apple, and has profound effects on the organoleptic quality of apples. The pH of mature apples was genetically mapped on LG16 in the segregating population from the cross ‘Prima’ x ‘Fiesta’. To our knowledge, this is the first time that the genetic segregation of the pH in apple is assigned to a specific gene. Further, this gene has not been reported yet in conjunction to pH of apples or other fruits. After cloning of the MdALMT2 gene, it can be used for, proof of principle, influencing the acidic of existing varieties either by silencing this gene in more acidic cultivars or by inserting this gene into the low acidic cultivars. Another step would be to develop an allele specific molecular marker for selection (Marker Assisted Selection) of the acidity of fruits already at seedling stage, five years before the trees carry fruits. In another study, a dominantly mutated allele of the transcription factor gene MdMYB10,including its upstream promoter, coding region and terminator sequence, was introduced by transformation into apple, strawberry and potato plants. The dominantly inherited mutant allele of MdMYB10 from apple induces anthocyanin production throughout the plant, also at the early stage after transformation. The aim was to determine whether MdMYB10 could be used as a visible selectable marker for plant transformation as an alternative to chemically selectable markers, such as kanamycin resistance. After transformation, the color of calli, shoots and well-growing plants were evaluated. Red and green shoots were harvested from apple explants and examined for the presence of the MdMYB10 gene by PCR analysis. Red shoots of apple explants always contained the MdMYB10 gene but not all MdMYB10 containing shoots were red. Strawberry plants transformed with the MdMYB10 gene showed anthocyanin accumulation in leaves and roots. No visible accumulation of anthocyanin could be observed in potato plants grown in vitro, even the ones carrying the MdMYB10 gene. However, acid methanol extracts of potato shoots or roots carrying the MdMYB10 gene contained up to four times higher anthocyanin content than control plants. Therefore, anthocyanin production as a result of the dominant MdMYB1010 gene can be used as a selectable marker for apple, strawberry and potato transformation, replacing kanamycin resistance gene such as nptII. We reported this MdMYB10 as a cisgenic selectable marker gene for apple transformation (Chapter 7). The results from all experimental chapters have been discussed in a broader sense in the general discussion (Chapter 8). The future prospectives and potential challenges in the genetical metabolomics are also highlighted. The approaches we developed in the current thesis could be used not only for developing potentially a more healthy and improved apple but can also be applied for the genetical metabolomics studies in other important crops.