Seasonal Variability of Volatilome from Dictyota dichotoma

Dictyotaceae, the large family of brown algae with the genus Dictyota as the richest one among them, produce a significant number of secondary metabolites, like diterpenes. The aim of this study was to investigate the variations in the composition of the volatile organic compounds (VOCs) of the brown alga Dictyota dichotoma collected from the Adriatic Sea. For the first time, both seasonal changes and the impact of air drying were examined. Headspace solid-phase microextraction (HS-SPME) on two fibres with different polarities and hydrodistillation (HD), followed by gas chromatography and mass spectrometry (GC-MS) analysis, was performed on both fresh (FrDd) and air-dried (DrDd) D. dichotoma. The major compounds of HS-FrDd were pentadecane and oct-1-en-3-ol. The percentage of pentadecane in HS-DrDd was increased up to 7.8 times in comparison with HS-FrDd. Principal component analysis (PCA) identified differences between the variability of data among fresh and dried samples over months and clearly dissociated the fresh May samples from the others in the HS-SPME results. The most abundant group of VOCs in HD were terpenes, with diterpenes isopachydictyol A and cembra-4,7,11,15-tetraen-3-ol as the major compounds. Diterpene pachydictyol A was also found and among sesquiterpenes and gleenol was the most abundant. Based on the dominant compound analyses, the PCA showed distinct separation of the fresh and dried samples, indicating similarities between the samples and allowing the establishment of typical VOCs significant for the chemotaxonomy of D. dichotoma.


Introduction
The genus Dictyota, the richest genus of the Dictyotaceae family, contains more than 40 species, but in the Mediterranean in general, only the species D. dichotoma and D. linearis are found [1].
In the course of our investigations towards the bioprospecting of the Adriatic Sea and chemical profiling of secondary metabolites from different marine sources, we investigated the volatile organic compounds (VOCs)-volatilome chemical composition of the brown alga D. dichotoma collected from the Adriatic Sea. There are limited or scarce data on D. dichotoma chemical composition from the Adriatic Sea [12][13][14]. In our previous research [14], sesquiterpenes were found among D. dichotoma headspace VOCs, predominantly germacrene D followed by other cadinenyl (abundant), muurolenyl and amorphenyl structures. The headspace VOCs in D. dichotoma from the Yellow Sea of China were also characterised as sesquiterpenes (predominantly germacrene D) and diterpenes [15]. However, there is a gap in the research on less volatile compounds of D. dichotoma as well as about VOC seasonal variability that could be also important for the algal chemotaxonomy or biological activity. Therefore, the goals of our study are to: (a) isolate the headspace VOCs from fresh D. dichotoma (HS-FrDd) by headspace solid-phase microextraction (HS-SPME) and by hydrodistillation (HD; first-time research); (b) analyse VOCs by gas chromatography and mass spectrometry (GC-MS); (c) investigate the impact of air-drying (first time report) to chemical profiles of dried D. dichotoma obtained by HS-SPME (HS-DrDd) and HD (HD-DrDd); (d) compare the chemical profiles obtained by HS-SPME and HD within the seasons; (e) determine the variability of the VOCs composition from May to September from fresh and air-dried samples, and to elaborate the data by principal component analysis (PCA); (f) provide new insights for the algal chemotaxonomy regarding moderate and less volatile compounds and their seasonal changes, or the impact of air-drying.

Results and Discussion
The VOCs were isolated by two methods (HS-SPME and HD) and analysed by GC-MS in order to identify the full range of present VOCs (headspace, low, moderate and less volatiles compounds) in fresh and air-dried D. dichotoma. For HS-SPME, two fibres were used in accordance with our preliminary research, taking into account the number of identified compounds; this was conducted in order to obtain more complete chemical profiles containing a broad range of headspace compounds. Great variability among HS-FrDd volatiles was found within the months by two different fibres for HS-SPME/GC-MS: divinylbenzene/carboxen/PDMS (I) and polydimethylsiloxane/divinylbenzene (II). To collect the VOCs obtained by HD, the solvent trap was used.

Variations of Headspace Volatilome of Fresh D. dichotoma
The identified HS-VOCs of fresh D. dichotoma (HS-FrDd) within all collection months can be classified into aliphatic compounds (including C 11 -hydrocarbons (dictyopterenes)), terpenes (mainly sesquiterpenes) and C 13 -norisoprenoids. Great variability can be noted in May in comparison to other months (Tables 1 and 2), particularly with respect to the higher abundance of sesquiterpenes. It should be noted that the sea temperature in May was the lowest (20.1 • C) in comparison with other months. Such a low temperature could affect the biosynthesis of secondary metabolites.
The major compound of HS-FrDd for all months was pentadecane, with the highest abundance in July and August (up to 23.41% (I) and up to 11.51% (II)) and the lowest in September (4.92% (I); 2.14% (II)). Pentadecane was reported for the first time as a metabolite of D. dichotoma using nuclear magnetic resonance (NMR) and electron ionization mass (EIMS) spectra [2]. Its unsaturated derivative, pentadec-1-ene, was the most abundant in May (3.98% (I); 3.69% (II)). These two compounds were not present in our previous research of fresh D. dichotoma headspace [14], but the sample was from a different singlepoint collection location. On the other hand, pentadecane was found in the headspace of D. dichotoma from the Yellow Sea of China [15]. Another present higher alkane was heptadecane (up to 3.53% (I) or 2.35% (II)), which was not present in the research of Wang et al. [15]. Tridecanal abundance increased from May to September up to 23.24% (I) or 13.89% (II), as well as (E)-non-2-enal up to 1.97% (I) or 1.80% (II). (E)-Non-2-enal was present in D. dichotoma from the Yellow Sea of China [15] and tridecanal was a minor      Another abundant compound from June to September was oct-1-en-3-ol (up to 19.85% (I); 27.25% (II)) with the lowest percentage in May (0.41% (I); 0.92% (II)). The same trend was noted for other major aliphatic acyclic C 8 -compounds: (E)-oct-2-enal (up to 5.54% (II) in June), (E)-oct-2-en-1-ol (up to 10.74% (I) or 15.97% (II) in September) and octan-1-ol (up to 7.42% (I) or 9.98% (II) in September). The aliphatic C 8 -compounds were found previously in HS of fresh D. dichotoma [14], but not in the research of Wang et al. [15]. Minor percentages of hexanal, nonanal, (E,E)-deca-2,4-dienal and pentadecanal were also periodically found (Tables 1 and 2). Lower aliphatic compounds were not so represented in D. dichotoma from the Yellow Sea of China [15]; however, (E)-hept-2-en-1-ol was the most abundant (4.62%).
Other found aliphatic compounds were dictyopterenes: dictyopterene D' (up to 4.42% (I) or 8.22% (II)) and periodically dictyopterene C' (up to 0.54% (I) or 1.69% (II)). Dictyopterenes as C 11 -hydrocarbons (cyclopropanes and cycloheptadienes) and their derivatives appear to be most abundant in the brown algae of the genus Dictyopteris [16]. Dictyopterene C' (also known as dictyotene) was found as a minor compound in the essential oil of vegetative parts of Dictyopteris [16], and also in freshly released eggs of marine brown alga D. dichotoma, as the substance that attracts spermatozoids [17]. Our previous research [14] also detected dyctopterenes in this alga.

Variations of Headspace Volatilome of Air-Dried D. dichotoma
The predominant compound of HS-DrDd in all months was pentadecane (up to 92.83% (I) or 88.23% (II) in September). Its percentage was increased up to ca. 4.3 or 7.8 times in comparison with Fr-HSDd; this could be connected to fatty acid degradation, particularly palmitic acid, since the results of incubation of palmitic-16-14 C acid in a culture of Nostoc muscorum indicated a direct decarboxylation into heptadecane and pentadecane [19], but also to the evaporation of the most volatile compounds. It is already known [20][21][22] that pentadecane predominates in the brown algae and heptadecane in the red algae, but our research demonstrates that drying could significantly influence the macroalga headspace composition, remarkably increasing the abundance of pentadecane that could be important for D. dichotoma chemotaxonomy.
An array of other HS-VOCs (Tables 1 and 2) was identified in HS-DrDd, but with less abundance including aliphatic compounds, sesquiterpenes and C 13 -norisoprenoids. Only in July and August, there was a slightly increase in other compounds and a decrease in pentadecane in HS-DrDd. In July, benzyl alcohol increased to 2.24% (I) or 2.43% (II) as well as dictyopterene D' to 1.74% (II) and dictyopterene C' up to 2.32% (II). Another major difference in comparison to HS-FrDd is a notable reduction in the number and abundance of sesquiterpenes in HD-DrDd that could have evaporated; this is also interesting for the chemotaxonomy. A selected example of the total ion chromatogram (TIC) from the fresh (HS-FrDd) and dried (HS-DrDd) sample is presented in Figure 1.

Statistical Analysis of the Headspace VOCs
The dominant compounds (>2%) from HS-FrDd and HS-DrDd were subjected to principal component analyses (PCA) to describe the variations among the volatiles in relation to material preparation (fresh or dry) and seasonal effect. The PCA results for headspace volatiles (two different fibres) are shown in Figures 2a-d and 3.
The correlation plot and score plot of the dominant components from fresh samples are shown in Figure 2a,b. The first two PCs described 78.25% of the initial data variability

Statistical Analysis of the Headspace VOCs
The dominant compounds (>2%) from HS-FrDd and HS-DrDd were subjected to principal component analyses (PCA) to describe the variations among the volatiles in relation to material preparation (fresh or dry) and seasonal effect. The PCA results for headspace volatiles (two different fibres) are shown in Figures 2a-d and 3.   . The correlation plot and score plot of the dominant components from fresh samples are shown in Figure 2a,b. The first two PCs described 78.25% of the initial data variability while the remaining PCs each accounted for less than 1% of the total variance. The correlation between certain groups of the compounds was observed (Figure 2a). The highest variable contribution, based on correlation, was observed for oct-1-en-3-ol, octan-1-ol, ß-gurjunene, α-humulene, epi-bicyclosesquiphellandrene, α-amorphene, pentadec-1-ene, germacrene C and δ-cadinene. Dictyopterene D' and pentadecane had the highest values of factor coordinates for the PC2, with the highest variable contributions, based on the correlations. The position of the samples in the multivariate space of the first two PCs is shown in Figure 2b. Scores were arranged in two areas with a clear separation between May samples and other sampling months. The compounds that significantly affect the separation of the May sample from other months belong to the group of sesquiterpene hydrocarbons (β-gurjunene, α-humulene, epi-bicyclosesquiphellandrene, α-amorphene, germacrene D, germacrene C and δ-cadinene) and two sesquiterpene alcohols (dactylol nad epicubenol). They all strongly correlate with each other Figure 2. Correlation loadings (a,c) and score plots (b,d) of the dominant compounds from the headspace volatiles obtained by two different fibres for HS-SPME/GC-MS: divinylbenzene/carboxen/polydimethylsiloxane (I) and polydimethylsiloxane/divinylbenzene (II) of fresh (a,b) and dried (c,d) D. dichotoma samples.  The correlation plot and score plot of the dominant components from dried samples are shown in Figure 2c,d. The first two PCs described 70.36% of the initial data variability. Oct-1-en-3-ol, benzyl alcohol, octan-1-ol, dictyopterene C', and pentadecane showed the highest variable contribution to PC1 and factor-variable correlation. The July (fibre II) samples were separated in the right part of the plot based on the lowest pentadecane content, while for PC2 August sample showed the highest case contribution with the highest pentadecane content. A negative correlation was observed between pentadecane and oct-1-en-3-ol as well as pentadecane and germacrene D.
When the data were analysed together, clear separation was observed between fresh and dried samples. The dried samples appeared in the upper right corner of the plot, while the fresh ones were placed in the lower right corner (Figure 3). Based on the higher content of sesquiterpenes, sesquiterpene alcohols and pentadec-1-ene, as well as the lowest content of pentadecane, the fresh samples from May (both fibres) were segregated in the left part of the plot. There was no clear separation between the other sampling months nor correlation with temperature change over months.

Sample Collection
The samples of Dichtiota dichotoma (Hudson) J.V. Lamouroux, 1809 were collected from May to September 2021. The location of the sampling was off the coast of the island Čiovo (43.493373° N, 16.272519° E) in the Adriatic Sea. All samples were collected from the same lagoon at a depth ranged from 20 to 120 cm. At every sampling, a YSI Pro2030 probe (Yellow Springs, OH, USA) was used to measure the sea temperature (Table 4). In dry samples, aliphatic compounds (heptadecane, pentadecanal, (Z)-octadec-9-en-1-ol) and diterpene alcohol cembra-4,7,11,15-tetraen-3-ol were variables with the highest contribution to PC1, while octan-1-ol and tridecanal, contributed the most to PC2.
PCA of all dominant VOCs obtained by hydrodistillation showed clear separation between fresh and dried samples (Figure 7). There were more similarities in seasonal variation between fresh samples. The dried samples appeared in the left part of the plot. The distribution was along the PC1 axis and was in the relation to aliphatic compounds (oct-1-en-3-one, oct-1-en-3-ol, (E)-oct-2-en-1-ol, octan-1-ol, tridecanal, and heptadecane) and (Z)-octadec-9-enoic acid abundance, while the distribution along the PC2 axis was related to sesquiterpene alcohols (cembra-4,7,11,15-tetraen-3-ol and gleenol) and methyl (all Z) eicosa-5,8,11,14-tetraenoate) abundance in the samples. Similar was observed for the fresh samples, and their vertical distribution was related to high cembra-4,7,11,15-tetraen-3-ol and low gleenol abundance. No correlation was found between the compounds' content change over months and temperature change.

Statistical Analyses
The principal component analysis (PCA) was used to determine relations between the dominant volatiles (>2%) of fresh and dried alga samples analysed by different methods (HS-SPME and HD) [32]. The data gathered for HS-SPME from two fibres for fresh and dried samples were used for the analysis (average percentage of peak area). Similarly, for HD data gathered for fresh and dried algae samples were submitted to PCA. Prior to analyses all data were log transformed. The analyses were carried out using STATISTICA ® (version 13, StatSoft Inc., Tulsa, OK, USA).

Conclusions
Two methods (HS-SPME and HD) were successfully applied to identify the full range of present VOCs (headspace, low, moderate and less volatile compounds) in fresh and air-dried D. dichotoma. The striking differences between the profiles are not that striking considering different extraction mechanisms for HD and SPME (exhaustive distillation + solvent trap vs. partition coefficients based microextraction).
Great variability among HS-FrDd VOCs was found within the months by two different fibres. The major compound of HS-FrDd for all months was pentadecane. The great variability of HS-FrDd is noted in May in comparison to other months, particularly for the higher abundance of sesquiterpenes (derivatives of cadinane and selinane followed by a minor abundance of aromadendrane, caryophyllane and other derivatives). Since selinane and cadinane are two main types of sesquiterpenes in brown algae it is important to be aware of their seasonal variability in D. dichotoma. The percentage of pentadecane in HS-DrDd was increased up to cca. 4.3 or 7.8 times in comparison with Fr-HSDd; this could be connected to fatty acids degradation and it should be considered for D. dichotoma chemotaxonomy. Another major difference in comparison to HS-FrDd is the notable reduction in the number and abundance of sesquiterpenes in HD-DrDd. Based on this reduction and the changes in the abundance of dominant components over months and within the fibres used, the PCA was successfully applied to distinguish differences between the variability of data among fresh and dried samples. PCA segregated the fresh May samples from the others; however, the correlation to sea temperature changes was not observed.
The composition of HD-VOCs was strikingly different in comparison to HS-VOCs from D. dichotoma. The PCA showed clear separation of the fresh and dried samples based on the dominant compound analyses, indicating similarities between the samples and allowing the establishment of typical VOCs significant for the taxonomic group of D. dichotoma. The major compounds in HD-FrDd were diterpenes isopachydictyol A followed by its isomer pachydictyol A. Another abundant diterpene alcohol was cembra-4,7,11,15-tetraen-3-ol. Among sesquiterpenes, gleenol was the most abundant. All these compounds were not found by HS-SPME (due to their lower volatility), indicating the importance of comprehensive screening of VOCs for obtaining reliable chemical profiles for the chemotaxonomy. As expected, the VOCs chemical profile of HD-DrDd was remarkably different from HS-DrDd and it showed similarity to HD-FrDd. The major compounds were similar to HD-FrDd and prenylated-guaiane diterpenes were present. The most striking difference among HD-FrDd and HD-DrDd was the increase (in all months) of (Z)-octadec-9-enoic acid and tetradecanoic acid up and the appearance of (Z)-octadecen-9-al in HD-DrDd. All these compounds indicate oxidation processes during drying.