Volatile profile of bee bread

Bee bread is one of the least studied bee products. In this study, ten bee bread samples were characterized using palynology and HS–SPME–GC–MS (headspace solid-phase microextraction gas chromatography-mass spectrometry). In total, over one hundred different volatile components were identified, belonging to different chemical groups. Only ten common components were detected in all the samples. These volatiles were ethanol, methylene chloride, ethyl acetate, acetic acid, α-pinene, furfural, nonane, nonanal, n-hexane and isovaleric acid. Several other components were commonly shared among various bee bread samples. Over sixty detected compounds have not been previously reported in bee bread. The analysis required a mild extraction temperature of 40 °C, as higher temperatures resulted in the Maillard reaction, leading to the production of furfural. The profile of volatile compounds of the tested bee pollen samples was complex and varied. Some relationships have been shown between botanical origin and volatile organic compound profile.

bee pollen grains undergo fermentation due to the presence of different microorganisms, mainly lactic acid bacteria.Both fermentation and digestive enzymatic biotransformations cause pollen and nectar to break down, forming a nutrient-rich food rich in carbohydrates, amino acids, unsaturated fatty acids, vitamins, and minerals 1,9,10 .These processes are needed to destroy the external layer of pollen exine made of resistant sporopollenin.The fermented bee pollen grains, known as bee bread are regarded as being more bioavailable than crude bee pollen, and contain more vitamin K as well as polyunsaturated fatty acids 9,11 .Their high nutritional value allows them to serve not only as the primary source of nutrients for the growth of bees, as it was naturally intended, but also makes them a valuable food source for human consumption.While essential nutrients are responsible for nutritional benefits, minor components such as different phytochemicals, including phenolic acid, anthocyanins, volatile compounds, and carotenoids, might evoke some beneficial biological functions.Some authors have claimed that pollen and bee bread possess numerous potential therapeutic activities, such as anti-tumor, anti-inflammatory, antimicrobial, and dyslipidemia-correcting effects 1,4,5,[12][13][14] .These claims, however, have no solid evidence in the scientific literature, and more research is needed to support them.Bee bread and bee pollen are considered to be rich sources of antioxidants due to the high content of phenolic compounds.In fact, they are considered to be richer sources of antioxidants than honey 4,9,12,15 .
It can be assumed the chemical composition of bee bread can vary depending on the types of pollen and nectar collected by bees (thus on botanical and indirectly-geographical origin), the microbiota of the bees, and the duration of the fermentation process.Fermentation results in the production of many volatile compounds.Some of the volatile compounds found in bee bread include alcohols, aldehydes, esters, and ketones 16,17 .In addition to the volatile compounds produced by the fermentation process and other digestive biotransformations occurring in the stored grains, bee bread may also contain other volatile compounds of botanical origin-that are found in the pollen and nectar that the bees collect.These compounds can vary depending on the types of plants the bees visit, and they may include terpenoids, flavonoids, and other plant-derived compounds.These compounds contribute to the pleasant aroma of bee bread and may evoke some physiological effects when consumed.The BB volatilome is the profile of volatile organic compounds that are emitted by bee bread.Its investigation can provide important information about BB composition, flavor, and aroma, as well as safety and potential health effects.The volatile profile of bee bread has been hardly studied and is poorly understood.Previous studies have focused on honey, and to a lesser extent, propolis or unstored bee pollen.According to our best knowledge, there are only two articles in the literature that have explored the volatiles of Apis mellifera bee bread using solid phase microextraction from headspace (HS-SPME).The first study was conducted by Kaškoniene et al. 17 , who studied only one bee bread sample from Lithuania, along with honey samples of various floral origins.The second study was done by Starowicz and coworkers 16 , who characterized volatiles and sensory profiles of beeswax, bee bread, bee pollen, and honey from Poland.The authors of this study also analyzed only one sample of bee bread.One another related study was conducted by Montaser et al. who used gas chromatography coupled mass spectrometry (GC-MS) with headspace injection (HS) to analyze three bee bread samples of Egyptian origin.
As a result, the volatile profile of only few bread samples have been recorded in existing scientific literature.This is not sufficient to characterize this bee product, and further studies are needed to understand its composition in terms of volatile compounds.Therefore, the goal of our research was to analyze the volatile compounds emitted from various bee bread samples using solid phase microextraction from headspace (HS-SPME) and gas chromatography with mass spectrometry (GC-MS).The samples were collected in Eastern Europe and were identified according to their botanical origin.

Bee bread samples
Bee bread samples were taken from the Apis mellifera hives by professional beekeepers.Samples 1 to 10 were obtained in 2020 from different locations in Ukraine. Figure 1 indicates the origin place of the samples.
The samples were packed into polyethylene containers and kept in the freezer (minus 18 °C).

Palynological analysis
The analyses were outsourced and performed at a professional laboratory specializing in the palynological analysis of bee products (HoneyLab, Puławy, Poland).To prepare microscopic preparations for palynological analysis, 10-40 mL of distilled water was added to each bee bread sample (from 5 to 18 g), depending on the weight of the sample.The mixture was shaken vigorously for 2 min, and shaking was repeated several times over the course of 6 h.The resulting homogeneous suspension was used to prepare a smear on the microscope slide.After drying the smear, the slide was covered with a coverslip with a drop of glycerol gelatin 18 .Palynological analysis of two smears was performed for each bee pollen sample, counting over 300 pollen grains and classifying them, if possible, to family, genus, species, or type of structure.Counting 300 pollen grains in one preparation allows one to obtain results that are representative of the entire sample 19 .The arithmetic mean was calculated from the two replicates, and its result was converted into the percentage of each type of pollen.

Headspace solid phase microextraction
The volatile organic compounds were extracted and enriched using solid phase microextraction from the headspace (HS-SPME) of the bee bread sample.The mass of the sample was 2.0 g.The raw bee bread was placed in 20 mL N20 crimped vials, sealed with aluminum caps with silicone/PTFE septa.No solvent or salts were added.
The samples were agitated and incubated for 15 min at a temperature of 40 °C.The following HS-SPME extraction lasted for 45 min (40 °C, constant agitation) and was performed using a manual holder and 1 cm long Supelco 50/30 μm StableFlex DVB/CAR/PDMS fiber (Merck, Darmstadt, Germany).After extraction, the SPME fiber was manually desorbed in the injector port of the gas chromatograph.

Gas chromatography-mass spectrometry (GC-MS)
The analyses were performed using a Hewlett Packard gas chromatograph (HP 6890 series GC) with a 5973mass detector (Agilent, Santa Clara, USA).The analytes were separated using ZB-5HT capillary column (5% diphenyl-and 95% dimethylpolysiloxane stationary phase, 30 m of length, an inner diameter of 0.32 mm, and a film thickness of 0.25 μm; Phenomenex Inc., Torrance, CA, USA).The carrier gas was helium with a constant flow of 2 ml/min.The oven was programmed as follows: initial 40 °C, kept for 5 min, then gradually increased by 3 °C/min to 180 °C, then changed by 15 °C min to 280 °C, and finally held for 1 min.The injector was set to 250 °C in spitless mode.After five minutes the injector was vented.
The GC-MS interface temperature was set to a temperature of 300 °C, the MS ion source temperature was 230 °C, and the scan range was 30-550 m/z.

Analyte identification
The chromatograms were analyzed using MSD Chemstation (Agilent) and NIST MS Search software (version 2.7) (NIST, Maryland, USA).The data was analyzed using LRI created by the authors 20 with the additional help of AMDIS software (NIST) in case of coelutions.The volatile analytes were identified by comparing the mass spectra data with the NIST 11 and NIST 14 libraries and by checking their retention index with literature values (using the Retentify tool and NIST Chemistry WebBook, SRD 69).The LRIs were calculated using an n-alkanes standard mixture from Supelco, which was also subjected to SMPE extraction before the analysis (Merck).The relative abundance was calculated by dividing the peak area of a particular analyte by the total peak area on the chromatogram (excluding time 0-3 min and peaks of silanes leaking from SPME fiber).

Statistical analysis
All the analyses were carried out in triplicate, and the results were expressed as means.Standard deviations and coefficient of variation were calculated for each set of results.Due to the size of the dataset, this data is available in the supplement.
The statistical analyses were performed using MS Excel (Microsoft, Redmont, WA, USA) and the online platform MetaboAnalyst 6.0 (https:// www.metab oanal yst.ca, Canada).The obtained data were analyzed using principal component analysis (PCA) and hierarchical cluster analysis (HCA, Ward method using Euclidean distance).These two classification techniques (PCA and HCA) were used to identify any clusters in the data and examine differences between the analyzed bee bread samples.

Analysis of volatile organic compounds of bee bread
The volatilomes of the studied samples were highly complex and varied.Tables 2, 3, and 4 in the manuscript, as well as Table S1 in the supplement present the result of the VOCs analysis.Altogether, 107 compounds were identified, belonging to different chemical groups including terpenoids (monoterpenes-13, oxygenated monoterpenes-10, sesquiterpenes-2), aldehydes (10), alcohols (7), ketones (6), carboxylic acids (11), esters (13), lactones (4), nitriles (3), sulfides (3) as well as alkanes (18), alkenes (3), and one representative of phenylpropanoid group.Six of them were also furan derivatives.Only ten components were detected in all of the samples: ethanol, ethylene chloride, ethyl acetate, acetic acid, α-pinene, furfural, nonane, nonanal, n-hexane, and isovaleric acid.Acetic acid, dimethyl disulfide, furfural, nonane, and nonanal were also observed in bee bread by Starowicz et al. 16 and Kaškoniene et al. 17 .Isovaleric acid was detected by the first given authors only.However, due to strong adsorption on the (5%-phenyl)-methylpolysiloxane column, this analyte might be difficult to observe by the second authors if occurred at a lower level.Ethanol and dimethyl sulfide were observed by the latter authors only.Three of the above-discussed components, namely acetic acid, furfural, and nonanal, were also observed by Montaser et al. in Trifolium bee bread of Egyptian origin.Acetic acid was found to be one of the most dominant components in this study, as well as in studies of cited above authors.It was also observed by Bakour et al. 21who determined acetic acid in the Romanian bee bread using HPLC-DAD at a level of 10.7 g/kg.Like ethanol, acetic acid derives as the product of biochemical transformations conducted by bee bread microorganisms.It is also abundant in the volatilome of bee pollen, which is the closest bee product to bee bread.Karabagias et al. 22 found acetic acid to contribute to seven percent of the volatiles of Greek bee pollen samples.Prdun et al. 23 analyzed multiple bee pollen samples of Croatian origin and found this carboxylic acid in only six of 21 analyzed bee pollen samples.Acetic acid is not generally found in the honey aroma, or it is found at much lower levels.For example, Kaškoniene et al. 17 did not find acetic acid in any of the fifteen tested different honey samples of Lithuanian origin.Similarly, Makowicz et al. 24 analyzed 15 samples of various honey from Poland and did not detect acetic acid in any of them.However, Starowicz et al. 16 found acetic acid not only in bee bread and bee pollen but also in honey (at about four percent).Acetic acid is also present in the aroma of propolis.Kamatou et al. 25 found it in South African propolis samples at varied concentrations, ranging from traces up to almost 62% in the headspace volatiles.Similarly, Cheng et al. 26 found a significant fraction of acetic acid (ranging from 11 to 60%) in Chinese propolis volatiles.
We hypothesize that the presence of acetic acid and alcohol in bee bread is not connected to the botanical origin of bee pollen.The decrease in pH of stored bee pollen is caused by lactic and acetic acid fermentation, as well as alcoholic fermentation, which helps prevent the spoilage of BB.Bee pollen transforms BB through lactic acid fermentation primarily carried out by bacteria such as Pseudomonas spp., Lactobacillus spp., Bacillus spp., and yeasts such as Saccharomyces spp. 27.Yeasts acquire energy via the conversion of various sugars into ethanol and carbon dioxide 28 .Lactic acid bacteria (LAB) constitute a diverse group of bacteria characterized by the production of lactic acid (LA) as the major metabolic end product of carbohydrate fermentation.LAB generate energy through substrate-level phosphorylation following two metabolic pathways for hexose fermentation, i.e., homofermentative and heterofermentative.The first pathway is based on glycolysis followed by the production of LA, whereas the second one, known as the pentose phosphate pathway, is characterized by the production of carbon dioxide, and ethanol or acetate in addition to LA 29 .
The presence of a large amount of methylene chloride is puzzling, but certain as the blank SPME in the same environment, using the same consumables, did not result in the appearance of a dichloromethane peak.
Chromatographic profiles of volatile organic compounds (VOCs) of studied bee bread samples were highly varied, both in terms of qualitative composition and the intensity of detected peaks.As mentioned earlier, only www.nature.com/scientificreports/ten compounds were found in all of the tested samples, but several other components were also commonly shared among various BB samples.Namely, acetone, sulcatone (6-methyl-5-hepten-2-one), methyl octanoate, and limonene were identified in nine out of eleven samples, while 2-methylbutanoic acid, p-cymene, butyrolactone, decane, hexanal, n-heptane, and caryophyllene were detected in eight samples.Furthermore, benzaldehyde was indicated in seven samples.Other volatile compounds were found in fewer bee bread samples.
Table 3.The HS-SPME-GC-MS abundances [%] of volatile organic compounds from the studied bee bread samples.The most volatile components (0-3.5 min) were excluded.In the right columns, there is a mark if the analyte has been already detected in the bee bread so far.Published compositions of bee bread of L Lithuanian origin 17 , P Polish origin 16 , E Egyptian origin The results of VOC analyses are split into two separate tables (Tables 2 and 3) due to the fact there were big, mostly unresolved peaks at the beginning of the chromatogram, which could not be precisely integrated.Therefore, integration was inhibited from 0 to 3 min.Moreover, big peak areas of solvent front peaks would overwhelm the % abundancies, and further peak contributions would not be as readable (due to small % contributions).
To the best of our knowledge, over sixty detected compounds have not been previously reported in bee bread.Table 3 on the right side encompasses columns that indicate whether the analyte has been previously detected in this bee product.Among the most abundantly occurring volatiles in the studied samples that have not been already reported, there are α-pinene, α-thujene, γ-terpinene, limonene, p-cymene, α-terpinene, sabinene, β-pinene, and methyl hexanoate.Most of these are monoterpenes and their presence can be attributed to the plant material from which bee bread was made, as these are common specialized plant metabolites.
Apart from acetic acid and methylene chloride, the main components (ranked according to the sum of relative abundancies) were α-pinene, benzaldehyde, limonene, isovaleric acid, and sulcatone (Table 4).α-Pinene is a common plant monoterpene, most abundantly present in the essential oils of many coniferous tree species, but popular also in other plant species.It was present in all studied Ukrainian samples.Except for samples BB3 and www.nature.com/scientificreports/BB7, it occurred at a significant level in the volatile fraction.In some samples (BB2, BB5, BB6, BB8-10) it was over 16%.The high abundance (~ 64%) of α-pinene in the volatile profile of the BB2 sample can be explained by its high abundance in the Helianthus essential oil.This BB sample contained 61% of Helianthus pollen.The essential oils of this plant contain monoterpene hydrocarbons, in particular α-pinene (49-59%), sabinene (2-17%), β-pinene (3-6%), and limonene (4-7%) 30 .BB8 sample also contained a high abundance of α-pinene that could be related to its content in essential oils in air-dried samples of aerial parts of Solidago canadensis L. (Asteraceae).Eight local invasive populations of Solidago canadensis L. from different countries contained α-pinene as a major component of essential oil (13-52%) 31 .Moreover, α-pinene (11-29%) was among the major components of the essential oils of four Solidago spp. of Lithuanian origin (S. gigantea, S. canadensis, S. niederederi, and S. virgaurea) 32 .We can relate its content and botanical origin due to its popularity in nature, therefore, we cannot use α-pinene as a specific marker of botanical origin for functional purposes.
The second-ranked component, benzaldehyde, is naturally occurring in many fruits and vegetables and has an aroma role with notes of almond, nuttiness, and stone fruit.It is one of the examples of volatile phenols, which are formed mainly in the shikimic acid pathway.It was not abundant in most of the samples except for sample BB7, which was distinct and contained benzaldehyde at almost 80% level.The palynological composition of this sample was varied and Salix pollen was the most dominant (about 20%), then; Rubus (18%), Frangula (15.5), and Prunus (13.4%) pollen grains were also abundant.Based on the obtained results, we do not see a correlation between the botanical origin of bee pollen and the presence of benzaldehyde.Benzaldehyde has already been detected in bee bread aroma by Kaškoniene et al., who reported it at a low level of 0.9%, and Starowicz et al. 16 , who found it at a higher level of 5.55%.It has been reported as an important constituent of bee products' aroma in numerous studies.For example, Kaškoniene et al. 17 found it at contribution levels of 1.1-21.4%,and Starowicz et al. 16 at abundance of 52.39% in multiflorous honey.It is also common in the volatile profile of bee pollen (0-40.3%) 16,22,23, propolis (0.2-18.2%) 25,26 , and beeswax (4.3%) 16 .
Another common volatile compound occurring in the tested bee bread samples was limonene.Similar to α-pinene, it is plant monoterpene that is derived from methylerythritol phosphate (MEP) and mevalonate (MVA) pathways.We may assume that its presence is only due to the botanical origin of the bee bread samples.While it has not been reported so far in bee bread, it has been indicated in some other bee products.It is occasionally reported in honey volatiles 17,33,34 .Additionally, it can occur also in the volatilomes of bee pollen 22,35 and propolis 26,36 .11 samples of BP out of 14 from the Baltic region contained limonene 37 .In our study nine samples of BB contained limonene.We suppose that this monoterpene also is a nonspecific component of BP and BB 37 .
Isovaleric (3-methylbutanoic) and 2-methylbutanoic acids were among the most abundant volatile carboxylic acids in the tested BB samples, along with acetic and butanoic acids.Starowicz et al. 16 observed 2/3-methylbutanoic acid in bee bread (1.2%) and multiflorous honey samples (14.9%).It is unclear whether the authors determined both 3-methylbutanoic and 2-methylbutanoic acids as coelution or presented the results as 2/3-methylbutanoic acid due to uncertainty in the identification of the correct isomer.Isovaleric acid has been also detected in the honey of cashew and marmeleiro origin 38,39 .2-and 3-methylbutanoic acids have been indicated by Pasini et al. 34 in six of ten studied buckwheat honey samples collected from Italy and Eastern Europe.The authors concluded that 3-methylbutanoic acid, along with a specific phenolic pattern, is a potential identifying characteristic of buckwheat honey.2-and 3-methylbutanoic acids were also determined in the HS-SPME volatile profiles of linden, honeydew, and especially buckwheat honey samples 40 .These compounds were also assigned by the authors as buckwheat honey markers.Among our tested bee bread samples, only four of them-BB2, BB3, BB8, and BB10 had Fagopyrum spp.pollen identified at levels of 2.2, 12.8, 8.3, and 0.9%, respectively.BB3, which had a higher percentage of buckwheat pollen, did not have the higher contribution of the discussed carboxylic markers.Therefore, we may hypothesize that 2-and 3-methylbutanoic acids are not necessarily distinct markers of buckwheat origin.They might be present in the bee bread samples as by-products of microorganisms' metabolism.Furthermore, out of 14 samples of bee pollen from the Baltic region, 11 contained 2-methylbutanoic acid 37 .
Another significant and frequent volatile component of bee bread is sulcatone (6-methyl-5-hepten-2-one).This hydrophobic unsaturated ketone can be found in all eukaryotes, ranging from yeast to humans.It is a commonly reported odor compound that is secreted with sweat by the human body and in different animals may play the role of alarm or attractant pheromone 41 .Sulcatone has been reported in numerous plants and essential oils from citronella, lemon grass, and palmarosa.It is a crucial component for the flavor profile of several fruits, including tomato, guava, and papaya 42 where it is synthesized during carotenoid metabolism.Sulcatone has a flavor that can be described as a combination of apple, peach, and musty notes.Its presence enhances the sweet and fruity taste of numerous fruits 43,44 .It has been also detected in flower volatiles 45,46 .However, Ma et al. indicated that sulcatone does not act as a bee attractant but rather as a bee repellent.Nevertheless, it is a significant volatile and aroma contributor in our tested bee samples.It has been so far detected by Starowicz et al. in Polish bee bread (at a level of 4.4%) and bee pollen (0.7%).It has been also determined by Karabagias et al. in Greek bee pollen samples at levels of 6.1-7.9% and by Kaškoniene et al. in Lithuanian bee pollen at levels of 2.2-2.5%.Whether its origin is botanical or it is a product of bee metabolism or further biochemical processes occurring in bee bread, is unclear.
In our studies, nonanal was detected in all of the studied samples of bee bread.Nonanal was also identified as a component of bee bread by the previously cited authors 10,16,17 .Moreover, it was detected in all 14 tested bee pollen samples from the Baltic region 37 .It was suggested that some aldehydes and ketones are formed by the oxidation of fatty acids, particularly linoleic and linolenic 17 .According to Dąbrowska 47 , a significant increase in the concentration of nonanal and decanal in surface waters was recorded during the plant vegetation period.All of this can indicate that nonanal is a popular component of the plant kingdom and therefore, it is a nonspecific component of bee bread.
Furfural has been detected in the volatile profile of all tested samples.Its presence was most probably caused by secondary reactions.Furfural is an important intermediate compound of the Maillard reaction of amino acids and reducing sugars to create brown melanoidins 48,49 .While heated, bee bread samples were gently "baked" and the furfural appearance had come from the Maillard reaction.When a higher temperature was used the bee bread was browned and stuck together (Fig. 2), in addition, furfural content raised significantly in the volatile fraction, from about 2% up to 44% (Table 5).When left overnight at 70 °C, the bee bread was burnt.Therefore, even though SPME extraction was more efficient in higher temperatures, higher peaks of analytes could be noticed or even new peaks detected, especially those with higher boiling points, the mild temperature of 40 °C was chosen for performing the extractions.
Apart from furfural, the volatile profile extracted at the higher temperature was richer in higher-boiling analytes such as long-chained hydrocarbons, fatty acids, and their esters (Table 5).

Statistical analysis
To discover natural groupings in the data and examine differences between the analyzed bee bread samples, statistical methods were implemented.Principal component analysis (PCA) was performed using mean-centered data, including 84 variables (detected analytes).The loadings and score plot are presented in Fig. 3a and b, respectively.The principal component (PC) 1 represents around 59%, and PC2 represents about 27% of the variance in the data.PC1 is mostly influenced by two chemical entities found in the bee bread volatilome: benzaldehyde, contributing positively to the PC1, and α-pinene, contributing negatively to PC1 (Fig. 3a).Regarding PC2, both benzaldehyde and α-pinene contribute negatively, while sulcatone, methyl hexanoate, undecane, and nonane contribute positively.
The most distinct sample is BB7, characterized by a completely different composition of the volatilome compared to the rest of the samples.This sample is positioned on the right in the PCA score plot due to its high abundance of benzaldehyde and only a minor content of α-pinene.Another apart sample is BB2.It is located in the lower-left quadrant of the plot due to the high abundance of α-pinene in the volatile profile-almost 64%, meaningfully more than the other samples.Samples BB5-6 and BB8-10 are clustered (as "green group") in the center of the PCA plot (scores close to 0).These several samples are relatively similar, particularly when considering α-pinene at levels of 16-25%.
The samples B1, B3, and B4, located above the green cluster in the score plot, are characterized by smaller α-pinene content in the volatilome, and substantial amounts of methyl hexanoate, undecane, nonane, and additionally in the case of BB1 and BB3-sulcatone.
Figure 4 presents the relationships and grouping of the ten bee bread samples based on Euclidean distance and the Ward method.The results of hierarchical cluster analysis (HCA) are in line with the results of the PCA and confirm that the most distinct sample is BB7.It is mostly characterized by a high abundance of benzaldehyde.Other samples, such as BB1, BB2, BB4, BB6, and BB8 also contained some benzaldehyde.Based on the obtained results, we do not see a correlation between the botanical origin and the presence of benzaldehyde.As previously mentioned, BB7 contained pollen from Prunus, Salix, Frangula, and Rubus, while BB4 has neither of them but a lot of pollen from Melilotus and Ambrosia.The origin of benzaldehyde in bee products is poorly understood, and it is unclear if it is produced by plants or through other biochemical processes in the bee products or both.Benzaldehyde is widely present among volatile compounds and is probably phylogenetically one of the oldest compounds, as it is produced by more than 50% of plant families studied for their volatile profiles.Additionally, insects and non-insect arthropods also produce benzaldehyde 50 .It is also worth noting that benzaldehyde is used in beekeeping as a bee repellent for easy access to the hive to extract honey without being stung 51,52 .
The second distant sample, according to HCA, is BB2.This sample contained mainly Helianthus pollen grains and a higher fraction of α-pinene in the volatilome-about 64%, significantly more than the other studied objects.
The clusters formed in the dendrogram for samples BB5, BB6, and BB8-BB10 (green cluster) and the cluster for samples BB1, BB3, and BB4 are also in line with the result of the PCA.The analysis of ten bee bread samples suggests some relationship between the volatiles existing and the botanical origin of the sample, as the indicated α-pinene resulting from Helianthus, Solidago, and Ambrosia pollen.However, this is a popular and non-specific monoterpene that exists in many other plant species.While attempts have been made to find volatile markers for botanical origin, the complex nature of bee product volatilome, the occurrence of these volatiles in multiple bee products of different origin, and the encountered natural variance may make it difficult or impossible to draw definitive conclusions.As discussed earlier, previous studies 34,40 have identified 2-and 3-methylbutanoic acids as potential markers for buckwheat origin, but our results showed that www.nature.com/scientificreports/these compounds were present in almost all tested bee bread samples of various origins.Therefore, any proposed marker would need to be validated using multiple samples of bee products of various botanical and geographical origins.According to Makowicz et al., high-performance thin layer chromatography (HPTLC) profiling of lipophilic extracts of honey may be more suitable for botanical origin authentication.The obtained chromatograms had the potential to create a type of identifier similar to a barcode, which could be used to distinguish between honey samples without requiring the identification of individual components 24,53 .Nevertheless, our analyses have revealed the volatile components present in bee bread-one of the least studied and understood bee products.Many of the volatiles we detected had never been previously identified in bee bread before, greatly expanding our knowledge of its volatilome.Further research is needed to characterize the volatilomes of bee bread from different origins, explore potential relationships between its origin and composition, and gain a better understanding of biochemical processes occurring in the bee bread resulting in the presence of volatile components contributing to its pleasant aroma.

Conclusions
In this study, ten bee bread samples from Eastern Europe were analyzed using solid phase microextraction from the headspace followed by analysis using gas chromatography coupled with mass spectrometry.The bee bread samples were of various botanical origins.A wide range of volatile components, comprising different chemical groups, were identified, over one hundred in total.The analyses revealed significant differences in the volatile profiles of studied bee bread samples.Despite the fact that most samples contained compounds from common chemical classes of identified compounds, such as hydrocarbons, terpenoids, alcohols, ketones, aldehydes, carboxylic acids, esters, lactones, nitriles, and sulfides, their qualitative and quantitative compositions differed from one another.Only ten components were found in all samples analyzed, including ethanol, ethylene chloride, ethyl acetate, acetic acid, α-pinene, furfural, nonane, nonanal, n-hexane, and isovaleric acid.Other components such as acetone, sulcatone, methyl octanoate, limonene, 2-methylbutanoic acid, p-cymene, butyrolactone, decane, and caryophyllene were commonly detected in various bee bread samples.The presence of furfural was enforced when using elevated temperature for SPME extraction, and mild temperature conditions were required to prevent alterations in the volatile profile.This may be explained by the fact that the Maillard occurs at elevated temperatures resulting in the production of furfural and brown melanoidins.
The obtained results allowed to characterize the volatile profiles of the various bee bread samples.However, it was impossible to pinpoint the exact volatile compounds that could be used as indicators of any botanical origin.Nevertheless, our results help to characterize the bee bread composition, one of the least studied bee bread products.Until our work, only the composition of a few samples has been studied and published, indicating www.nature.com/scientificreports/ the need for further and more in-depth characterization of this unique bee product.We discovered over sixty different components that have not been so far identified in bee bread.

Figure 1 .
Figure 1.Ukraine.The place of origin of the studied samples.

Figure 2 .
Figure 2. The influence of temperature on the appearance of bee bread (sample BB10).

Figure 4 .
Figure 4. Dendrogram of the studied bee bread samples.

Table 1 .
Palynological analysis of the tested bee bread samples.

Table 2 .
Rough assessments of the most volatile compounds in the studied bee bread samples.S small, M medium, B big, VB very big, d detected (due to coellution impossible to evaluate amount), tr traces.*Octane coellution.

Table 4 .
Top volatile components as % abundancies (peak of silanes and a front of the chromatogram excluded).

Table 5 .
The influence of the temperature on the volatilome of bee bread (BB10); nd-not detected.