Improved QTL analysis of apple volatile compounds by PTR-TOF-MS

In the modern definition of fruit quality, aroma is considered one of the main factors together with appearance, texture and nutritional properties [1]. The importance of aroma is due to its direct influence on the consumer appreciation, since a pleasant aromatic “bouquet” is a fundamental requirement for a high fruit quality standard. The VOC set in apple is biochemically composed by major classes of compounds, such as alcohols, aldehydes, ketones, polypropanoids, sesquiterpens and esters [2]. The latter category (esters), in particular, is the most important class for apple, representing the major VOC class (~ 90%) produced during the apple fruit ripening process. The great impact that the aroma has on the apple fruit marketability, stimulated in the last decade the scientific community to gain knowledge about novel and valuable molecular markers associated with these traits. This became a main objective in order to facilitate apple breeding programs in the selection of the most favourable individuals. In this context, a valuable approach is to identify Quantitative Trait Loci (QTL), thus determining links between genetics and fruit aroma in apple. Dunemann and colleagues [2] carried out a comprehensive QTL mapping assay in apple, identifying QTLs associated to 20 major compounds (alcohols, esters and terpens), and located on 12 linkage groups, by using 150 seedlings of the cross ‘Discovery x Prima’ and a HS-SPMEGC detection equipment. Gas-chromatographic technique is, however, laborious and time consuming, limiting the aroma characterization of large apple collections, such as breeding material. A valuable technological and analytical alternative is offered nowadays by techniques that privilege rapidity over analytical information, and have little sample preparation and no chromatography. The advantage is twofold: on the one hand a broader number of samples can be screened, on the other hand potential artefacts caused by extraction and concentration procedures are minimized [3]. A technique as such is represented by Proton Transfer Reaction - Mass Spectrometry (PTR-MS). In this work the position of a set of QTL associated to VOCs was identified and validated in three different environments, where the progeny ‘Fiesta x Discovery’ was replicated (Wädenswil, Conthey and Cadenazzo), thus extending previous explorative studies [4]. Aroma emission profiles were characterized by a PTR-MS instrument. The QTL-VOC combined analysis performed among these three locations validated the presence of important QTL in genomic regions, two located in the linkage groups 2 and one in linkage group15, respectively, for compounds related to esters (m/z: 41, 43, 57, 61 and 131) and possibly to the hormone ethylene (m/z: 28). The QTL set presented here confirmed that in apple some compounds are highly genetically regulated and stable across environments. In a second study a modern PTR-MS coupled with a time-of-flight mass analyzer (PTR-TOFMS) was employed for a combined QTL-VOC analysis on three different populations, encompassing hundreds of sibling samples. Given the higher resolution of PTR-TOF-MS 259 compared to PTR-MS, the phenotypization of the apple siblings is strongly enhanced, entangling hundreds of peaks for which a sum formula could be assigned [5,6]. In many cases also the underlying compound could be determined, thanks to targeted analyses carried on on a limited number of samples using SPME/GC-MS [7,8]. A very large number of QTL could be identified, going far beyond the confirmation of the QTL found in the first study by PTR-MS. Such QTL were associated to several classes of VOCs constituting the apple aroma profile. The statistical significance of the QTL was also extremely enhanced, reflecting the better phenotypization of the samples provided by PTR-TOF-MS. Employing PTR-TOF-MS therefore represented a major breakthrough in addressing this important problem. We also devoted a strong effort to automatize the whole combined QTL-VOC analysis process by writing specific routines in MATLAB (MathWorks, Natick, USA) and R (R Foundation for Statistical Computing, Vienna, Austria) which allow studying very large populations with a sustanable effort. In general, these findings can be of outmost importance in ongoing breeding programs