Saponins from Pea Ingredients to Innovative Sponge Cakes and Their Association with Perceived Bitterness.

Pea-based ingredients are increasingly being used in foods because of their nutritional, functional and environmental benefits. However, their bitter taste is not appreciated by consumers. Saponins have been reported to be bitter in whole pea flour (PF) but not in the purified ingredients obtained from it, such as pea protein isolate (PPI) and pea starch (PS). In addition, the evolution of saponins in cooked foods made from these ingredients and their relationship to bitter flavor has not been investigated. This study, therefore, explored the presence of two bitter saponins, ßg and Bb, in whole pea flour (PF) and a composite flour reconstructed from the two main fractions (PS + PPI). In addition, it investigated the impact of baking on the chemical state of these compounds in a sponge cake. Finally, the sensory impact of the baking process on the perceived bitterness of cakes made with these two pea flours was also evaluated. High-Performance Liquid Chromatography-High-Resolution Mass Spectrometry (HPLC-HRMS) was used to identify and quantify pea saponins in the flours and cakes, and a descriptive sensory analysis was obtained by a trained panel to assess sensory differences in bitterness. Our results showed marked differences in saponin concentration and composition among the pea ingredients studied. Concentrations were highest in PPI (1.497 mg·g-1 dry matter), with 98% of saponin Bb. PS had the lowest saponin concentration (0.039 mg·g-1 dry matter, with 83% Bb), while 0.988 mg·g-1 dry matter was quantified in PF, with only 20% Bb and 80% ßg. This research also highlighted the thermal degradation of saponin ßg to Bb in sponge cakes during baking at 170 °C. However, at a sensory level, these chemical changes were insufficient for the impact on bitterness to be perceived in cakes made with pea flour. Moreover, baking time significantly reduced the bitter flavor in cakes made with the composite flour (PS + PPI).

Applicability of pea ingredients in baked products: Links between formulation, reactivity potential and physicochemical properties.

This study aimed to evaluate the applicability of purified pea ingredients (starch and protein isolate) by assessing their potential to form volatile compounds during the different steps of sponge cake development compared to pea flour and wheat flour. While pea flour was highly susceptible to lipid oxidation during batter beating, the combination of purified pea starch and pea protein yielded significantly fewer oxidation markers with known green-beany off-odors. This was due more to the inactivation of lipoxygenase during flour fractionation than to differences in batter structure. However, fractionated ingredients were highly prone to participating in the Maillard reaction and caramelization during baking, leading to a more complex mixture of pyrazines, Strecker aldehydes and furanic compounds with potential malty and roasted notes compared to cakes based on pea flour or wheat flour. These findings confirm that using purified pea fractions can create high-quality products with an attractive composition.

In vitro digestion of protein and starch in sponge cakes formulated with pea (Pisum sativum L.) ingredients.

This study investigated the in vitro digestion of purified pea fractions (protein isolate and starch) in sponge cakes when compared to unrefined pea flour and to the whole wheat flour and purified maize starch commonly used in the food industry. Proteins in the wheat cake were hydrolysed more rapidly than those in cakes made with either pea flour or a combination of pea proteins and purified starch. In absolute terms, however, more readily bioaccessible protein was released from these pea cakes (by around 40%). By contrast, cakes containing wheat flour or maize starch were more susceptible to amylolysis compared to those based on pea starch in the form of the purified ingredient or whole flour. This could be attributed to a higher proportion of amylose and resistant starch in the pea cakes as well as structural characteristics that might have decelerated enzyme-substrate interactions. Interestingly, similar digestion patterns were observed regarding the purified pea ingredients and unrefined whole pea flour. It was therefore concluded that pea ingredients, and particularly the less purified and thus more sustainable whole pea flour, are promising plant-based alternatives for use in gluten-free baked products.

From flours to cakes: Reactivity potential of pulse ingredients to generate volatile compounds impacting the quality of processed foods.

This study investigated the impact of substituting wheat with pulse flours (lentil, chickpea, lupin, green and yellow pea) on reactivity during different steps of sponge cake development. Pulses exhibited a greater ability to generate volatiles with probable odor activity. Batter beating initiated lipid oxidation which depended on lipoxygenase activity and the fatty acid profile of the flours. Among the pulses, pea batters were richest in oxidation markers whereas lupin was least reactive, probably due to thermal pre-treatment. Baking triggered caramelization and Maillard reactions, notably with the pulse products which were particularly enriched in pyrazines and furanic compounds. Principle component analysis revealed that pea cakes were associated with oxidation markers that typically possess green-beany flavors, while Maillard markers known to impart nutty, roasted notes were assigned to lentil and chickpea cakes. These findings highlight the importance of ingredient type and its pre-processing in the development of quality-related markers for gluten-free products.

Quality-driven design of sponge cake: Insights into reactivity, furan mitigation and consumer liking.

This work highlights the importance of considering reactivity into the quality-driven design of heat-treated foods, which should cover the mitigation of process-induced contaminants and the improvement of the sensory properties of the foodstuff. The joint effects of formulation and baking conditions on reactivity and several quality aspects (i.e. volatile generation, physical properties, sensory and consumer tests), followed by product optimization (i.e. consumer liking and furan mitigation) were studied. While key markers are affected by all factors and their interactions, the effect of sugar and whole egg are the clearest. Furan would be predominantly generated from glucose via caramelization and/or Maillard reaction, whereas the formation of Strecker aldehydes and lipid oxidation products would be favoured by precursors in whole egg. Formulations with a low glucose content, baked at low temperatures/short times lead to optimal products. Egg-based ingredient content may be set according to preference or by applying different optimization approaches.

How ingredients influence furan and aroma generation in sponge cake.

A wide range of compounds can be formed during thermal processing of food, some of which are relevant for aroma (e.g., furfural), while others are of great health concern (e.g., furan). This paper presents the study of formulation as affecting the simultaneous generation of furan and furfural, along with other aroma quality markers, in sponge cake by means of headspace trap/GC-MS. Ingredients were screened according to their category (fat, salt, sugar, egg-based). Glucose-containing formulation resulted in the highest content of furan and furfural (12.5 ±â€¯0.5 ng g-1 and 9.2 ±â€¯0.2 µg g-1, dry basis, respectively), while their lowest amount was found in the egg-white recipe (3.1 ±â€¯0.1 ng g-1 for furan and 0.287 ±â€¯0.078 µg g-1 for furfural, dry basis). The latter also related negatively to all studied compounds. This work will be useful for developing novel strategies to deliver safe foods with appealing organoleptic attributes.

Multivariate optimization of headspace trap for furan and furfural simultaneous determination in sponge cake.

Furan, a possibly carcinogenic compound to humans, and furfural, a naturally occurring volatile contributing to aroma, can be both found in thermally treated foods. These process-induced compounds, formed by close reaction pathways, play an important role as markers of food safety and quality. A method capable of simultaneously quantifying both molecules is thus highly relevant for developing mitigation strategies and preserving the sensory properties of food at the same time. We have developed a unique reliable and sensitive headspace trap (HS trap) extraction method coupled to GC-MS for the simultaneous quantification of furan and furfural in a solid processed food (sponge cake). HS Trap extraction has been optimized using an optimal design of experiments (O-DOE) approach, considering four instrumental and two sample preparation variables, as well as a blocking factor identified during preliminary assays. Multicriteria and multiple response optimization was performed based on a desirability function, yielding the following conditions: thermostatting temperature, 65°C; thermostatting time, 15min; number of pressurization cycles, 4; dry purge time, 0.9min; water / sample amount ratio (dry basis), 16; and total amount (water + sample amount, dry basis), 10g. The performances of the optimized method were also assessed: repeatability (RSD: ≤3.3% for furan and ≤2.6% for furfural), intermediate precision (RSD: 4.0% for furan and 4.3% for furfural), linearity (R2: 0.9957 for furan and 0.9996 for furfural), LOD (0.50ngfuran gsample dry basis-1 and 10.2ngfurfural gsample dry basis-1), LOQ (0.99ngfuran gsample dry basis-1 and 41.1ngfurfural gsample dry basis-1). Matrix effect was observed mainly for furan. Finally, the optimized method was applied to other sponge cakes with different matrix characteristics and levels of analytes.

Furan quantification in bread crust: development of a simple and sensitive method using headspace-trap GC-MS.

To study reactivity in bread crust during the baking process in the pan, we followed furan mainly resulting from Maillard and caramelisation reactions in cereal products. Furan quantification is commonly performed with automatic HS-static GC-MS. However, we showed that the automatic HS-trap GC-MS method can improve the sensitivity of the furan quantification. Indeed, this method allowed the LOD to be decreased from 0.3 ng g(-1) with HS-static mode to 0.03 ng g(-1) with HS-trap mode under these conditions. After validation of this method for furan quantification in bread crust, a difference between the crust extracted from the bottom and from the sides of the bread was evident. The quantity of furan in the bottom crust was five times lower than in the side crust, revealing less reactivity on the bottom than on the sides of the bread during the baking process in the pan. Differences in water content may explain these variations in reactivity.

Lipid oxidation in baked products: impact of formula and process on the generation of volatile compounds.

This paper investigates the effect of ingredients on the reactions occurring during the making of sponge cake and leading to the generation of volatile compounds related to flavour quality. To obtain systems sensitive to lipid oxidation (LO), a formulation design was applied varying the composition of fatty matter and eggs. Oxidation of polyunsaturated fatty acids (PUFA) and formation of related volatile compounds were followed at the different steps of cake-making. Optimised dynamic Solid Phase Micro Extraction was applied to selectively extract either volatile or semi-volatile compounds directly from the baking vapours. We show for the first time that in the case of alveolar baked products, lipid oxidation occurs very early during the step of dough preparation and to a minor extent during the baking process. The generation of lipid oxidation compounds depends on PUFA content and on the presence of endogenous antioxidants in the raw matter. Egg yolk seemed to play a double role on reactivity: protecting unsaturated lipids from oxidation and being necessary to generate a broad class of compounds of the Maillard reaction during baking and linked to the typical flavour of sponge cake.

Analysis of human male armpit sweat after fenugreek ingestion: Characterisation of odour active compounds by gas chromatography coupled to mass spectrometry and olfactometry.

In this study, the strong "maple-syrup" odour which appears after fenugreek ingestion was investigated. Headspace solid-phase microextraction (HS-SPME) was applied to extract volatile odourant compounds from human male armpit sweat samples. Two male volunteers were considered who have similar diet; they had to ingest fenugreek infusion over the same period of time. The HS-SPME extracts obtained were then analysed by gas chromatography coupled either to mass spectrometry (GC-MS) or flame ionisation detection and olfactometry (GC-O). In that latter case, a panel of eight assessors was used, and the detection frequency methodology was applied. A total of 44 compounds could be identified in sweat samples, with a wide range of chemical structures, some of them being reported for the first time in human armpit sweat. Eight compounds appearing only after fenugreek ingestion could be identified: 2,5-dimethylpyrazine, ß-pinene, 3-octen-2-one, camphor, terpinen-4-ol, 4-isopropyl-benzaldehyde, neryl acetate and ß-caryophyllene. Due to their odourant notes, such compounds should be responsible for the strong "maple-syrup" odour present in sweat after fenugreek ingestion. GC-O confirmed the role of some odourant compounds in the "maple-syrup" odour of sweat, especially 2,5-dimethylpyrazine which was the best perceived odour. Among these eight compounds, some of them were previously reported in the fenugreek seeds, namely ß-pinene, 3-octen-2-one and camphor.

Role of pulp in flavor release and sensory perception in orange juice.

This work elucidates the role of suspended solids in sensorial perception and flavor release in orange juice. The coarsest pulp (insoluble particles with a diameter of >2 microm) accounted for two major physicochemical effects in orange juice samples: it retained large amounts of aroma compounds, including terpenes and aldehydes, and modified the rheological properties of the juice matrix. These phenomena strongly affected the chemical composition of the vapor phase in the juice samples. On the other hand, orange juice cloud (finest insoluble particles with a diameter of <2 microm) also showed a strong retention effect on ethyl butanoate or hexanal, probably due to the occurrence of molecular interactions with cloud macromolecules. The amount and the size of the suspended solids critically modified not only the texture perception but also the odor and the overall flavor perception, including the "freshly squeezed" and the "artificial flavor" descriptors. The addition of a natural pulp to low-pulp juices increases the fresh orange juice character, a finding that is explained by both physicochemical (fresh pulp contains high amounts of key aroma compounds, including acetaldehyde and mono- and sesquiterpenes) and cognitive effects, mainly due to the tactile properties of the pulp.

Solid phase microextraction (SPME) of orange juice flavor: odor representativeness by direct gas chromatography olfactometry (D-GC-O).

The sensorial quality of solid phase microextraction (SPME) flavor extracts from orange juice was measured by direct gas chromatogrphy-olfactometry (D-GC-O), a novel instrumental tool for evaluating odors from headspace extracts. In general, odor impressions emerging from SPME extracts poorly resembled that of the original orange juice. In an attempt to improve the sensorial quality of extracts, sample equilibration and exposure times were varied on Carboxen/polydimethylsiloxane (CAR/PDMS) and divinylbenzene/Carboxen/polydimethylsiloxane (DVB/CAR/PDMS) SPME fibers. Best sensorial results were obtained with the DVB/CAR/PDMS fiber exposed for the shortest time; a trained panel of eight assessors judged its odor as the most representative of the reference orange juice. The analysis of odor active compounds by classical GC-O accounted for odor characteristics revealed by D-GC-O. A principal component analysis (PCA) was applied on SPME and headspace extracts using flavor recoveries as variables. Interestingly, PCA discriminated samples according to their odor representations described by D-GC-O analysis. This paper provides the first comprehensive methodology to "smell" SPME extracts and "evaluate" their sensorial quality. This method will enable future investigations to further improve SPME performance.

Distribution of volatile compounds in the pulp, cloud, and serum of freshly squeezed orange juice.

The quantitative distribution of volatile compounds in the pulp, cloud, and serum of a freshly squeezed orange juice (cv. Naveline) was measured. Juice monoterpene and sesquiterpene hydrocarbons were primarily recovered from the pulp (74.0 and 87.2%, respectively) and cloud (7.3 and 14.9%, respectively). Esters and monoterpene alcohols were mainly found in the serum (90.4 and 84.1%, respectively). Long chain aliphatic aldehydes tend to concentrate in the pulp. The relative proportions of individual volatile compounds were similar in the pulp and cloud. Pulp and cloud alcohol insoluble residues exhibited similar compositions; half of them are made of nonwall proteins, and the rest are made of cell wall materials. Pulp and cloud total and neutral lipids had similar fatty acids distributions, although the cloud was much richer in total lipids than the pulp. No relationship was found between the retention of aroma compounds in the pulp or cloud and their AIR and lipid content or composition.