Gabriela Raquel Amaral Soliz^1,2, Aline Gabrielle A. Carvalho^1,2, Hygor Marcos Ribeiro de Souza^1,2, Alexandre Guedes Torres^1,2, Maria Fernanda C. Guimarães^1,2, Thamizy M. S. Macedo^3,4, Emmanuela P. P. Azevedo^2,3

Pumpkin flowers (Cucurbita spp.) represent a promising matrix with gastronomic, ornamental, and nutritional potential, adding commercial value and offering an income source for family-based agriculture. As an underexplored source, their metabolic profile reveals distinct compositions of bioactive compounds, highlighting their potential as functional ingredients. In this study, the chemical profile of two variants of landrace Cucurbita genus pumpkin flowers (C. moschata and C. maxima), one commercial C. moschata and one biofortified C. moschata variety with beta-carotene all from Recife-PE were analyzed via GC-MS and LC-HRMS-based untargeted metabolomic. The four freeze-dried flower samples (15 mg) provided by Embrapa were vortexed for 30 s with 500 µL of methanol:water (80:20, v/v). The mixture was taken to an ultrasonic bath for 15 min, followed by incubation at 4°C for 15 min and centrifugation at 9000 ×g for 15 min. A 120 µL aliquot of the supernatant was collected for LC-HRMS analysis. The pellet remaining after supernatant removal was dried under vacuum (Speed Vac) and reconstituted in 1000 µL methyl tert-butyl ether. After vortexing and centrifugation, 600 µL of the nonpolar phase was collected, dried under nitrogen flow, and stored at −20°C before the silylation derivatization for GC-MS analysis. In this work, a total of 85 metabolites were annotated using two analytical platforms: 75 were detected by LC-HRMS and 10 by GC-MS. The results from this study highlighted the richness of phenolic compounds, amino acids, and carboxylic acids in pumpkin flowers. In addition, multivariate analysis was applied. The PCA score plot from LC data showed a clear separation between the commercial and landrace C. maxima groups, and an overlap between the landrace C. moschata and biofortified variety groups. The influence of the plant\'s position in the plant-bed (middle row or lateral ends) on the chemical profile of the flowers was also investigated. The PCA score plot showed a group overlap indicating a lack of difference in the metabolic integrity of the species concerning the planting position. The GC-MS analysis enabled the absolute quantification of 10 metabolites. Among these, fatty acids were the predominant class. Interestingly, palmitic acid, stearic acid, and alfa-linolenic acid were found at higher levels in the Commercial and landrace C. maxima groups compared to the other groups. These findings support and are consistent with those obtained using LC-HRMS. In summary, the use of two complementary analytical platforms covered a wide range of metabolites from pumpkin flowers from agronomically important varieties in the Agreste region of Pernambuco. The multivariate analysis allowed discrimination commercial and landrace C. maxima varieties. This result appears scientifically sound, as the commercial and landrace C. maxima samples belong to different species. In addition, the planting cultivation position in the plant-bed has not influenced the metabolic integrity of the species.

Agradecimentos: Financial support: FAPERJ, CAPES, CNPq, and FACEP.