The increased dry matter yield in Lamina and Stipe (Table S1) in summer are not surprising, given their higher photosynthetic activity during this period. The high levels of iodide content observed (Fig. 3) were expected as Laminariales are known to accumulate iodide from sea water (Küpper et al. 2007) and it may be incorporated into iodo-metabolites including hormone-like components, microbial repellents, antifouling agents or oxidative protectants (Almeida et al. 2001). A protective role for iodide might explain its accumulation in the outermost tissues that most interact with surrounding seawater, such as the bark-derived samples (SCP and MBL SCP; Fig. 3). This is the first report of the relative amounts of iodide in different tissues and co-products using MS-based methods. Indeed, other work at the Hutton has confirmed iodide can reach mM levels in water and methanol extracts of MBL SCP (results not shown).
Mannitol content observed (Fig. 3) agrees with its proposed role as the main carbon storage sugar (Lee 2008). It accumulates in tissues with active cell division, such as in the blade and meristem, with highest concentrations during the active photosynthetic period, when light is most available, from late spring, summer to early autumn (April to September) (Schiener et al. 2015). This justifies its higher concentration in S Lamina. However, mannitol is spread through other tissues due to active transport through the sieve cells or trumpet hyphae cells and is the main osmoregulatory metabolite in the Laminariales (Shao et al. 2014).
Glycine betaine (Fig. 3) was also present in all tissues as may be expected of this proposed osmoregulatory component (Blunden et al. 2010). The highest levels were in the MBL SCP co-product in W. A component putatively identified as the betaine trigonelline (m/z 138 NEG), which has also been noted before in Laminaria species and has been identified as compound of interest for development of seaweed extracts for crop improvement (Blunden et al. 2010). This was also found in all tissues (Fig. 3) but with lowest levels in MBL SCP and highest levels in Stipe in W.
One of the “specific” Lamina metabolites was maltitol (m/z 383 NEG), or mannitol linked to a glucose (Fig. 4) which could be the first building block of laminarin (Kadam et al. 2015). Lamina had the most “tissue-specific” metabolites, with 5 in NEG mode and 2 in POS mode (Fig. 4 and see supplementary data; Fig. S2). This may relate to the central role of this tissue in photosynthesis, storage of fixed carbon and reproduction (ØDegaard et al. 1998; Sjøtun et al. 1998; Lee 2008; Schiener et al. 2015). Some components were clearly highest in the Lamina samples, e.g., dialanine and the eisenin-like peptide, but others, such as citric acid and the unknowns, were abundant in Lamina but also present in other tissues and co-products (see supplementary data; Fig. S2).
Six putative metabolites were particularly abundant in MBL SCP, 5 in POS mode and 1 in NEG, including ectoine as shown in Fig. 4. Others included the related compound hydroxyectoine, but also isobutyl-piperidone, 2-hydroxy-N-(7-methyloctyl) acetamide, butyryl carnitine, and N-propyl hexanamide (in POS mode) and benzene dicarboxylic acid (in NEG mode), (see supplementary data; Fig. S3).
SCP had 4 abundant components, with the abundance of piperidine dicarboxylic acid shown in Fig. 4 but pyroglutamate, diacylglycerol-phosphocholine; and an unidentified metabolite also appeared abundant, if not specific to this co-product (supplementary data Fig. S4). Piperidine dicarboxylic acid has been identified in the red seaweed, Porphyra tenera (Kawauchi et al. 1978) but the role of this imino acid derivative is not known.
The Stipe had only 1 “specific” metabolite noted in POS mode, the putative picolinic acid derivative shown in Fig. 4.
Although MBL LCP shared many metabolites with the other samples, there were 5 putative metabolites that were more abundant in MBL LCP. Three of these were detected in NEG mode (glyceraldehyde, Fig. 4, succinic acid and ketoglutarate, see supplementary data Fig. S5), and 2 in POS mode (glutamic acid and aminobutyric acid, see supplementary data Fig. S5). Aminobutyric acid is an amino acid normally derived from glutamic acid, which in terrestrial plants has a role in stress regulation (Shelp et al. 2017; Rashmi et al. 2018). The relative abundance of compounds in MBL LCP could reflect its origin as a H2O rather than a MeOH extract. Nevertheless, some H2O fractions of other samples were also analysed by HILIC-MS and these showed similar profiles to the MeOH extracts (results not shown).
Fucoxanthin (FX) was much more apparent in POS mode, and this preferential ionisation has previously been reported for this compound (Rajauria et al. 2017). It was present in all samples apart from MBL LCP, which probably reflects that this fraction was a water extract (Fig. 5). Fucoxanthin plays an important role as an energy transferring pigment via fucoxanthin-chlorophyll-protein complexes in the thylakoid (Rajauria et al. 2017). The clearly higher content in W samples (Fig. 5) was unexpected, because one might expect fucoxanthin to accumulate during the summer, when the seaweed would have greatest need for photosynthetic activity. However, others have noted higher fucoxanthin levels in L. digitata in winter and spring (Heffernan et al. 2016). Other metabolites such as trihydroxy-octadecenoic acid, proline, penicinoline E, and an unidentified component, also showed different abundances between W and S samples (Fig. S6).
Some metabolites showed changes in abundance between W and S (Fig. S7) but generally only in certain tissues/ co-products. For example, m/z 202 (RT 1129-NEG; glycyl-lysine) was higher in the winter but only in the Stipe sample (Fig. S7). Sulphosalicylate (SulSA; m/z 217; RT 1964-NEG) was higher in winter but only in Stipe and SCP (Fig. S7). Proline betaine (m/z 144; RT 0988-POS) was higher in summer in SCP but higher in winter in MBL SCP (Fig. S7). Glutamine (m/z 147) showed higher levels in the summer in Stipe and LCP. Indeed, many of the other putatively identified metabolites showed this seasonal variation of levels in certain tissues or co-products.
Eisenin is a peptide of pyroglutamate-glutamate-alanine identified in the brown seaweed, Eisenia bicyclis with immunological activity amongst other reported bioactivities (Kojima et al. 1993). Although detectable in both negative and positive modes (Tables 1 and 2), this tripeptide was most readily discriminated from the major co-eluting component glycine betaine in negative mode. It was abundant in the Lamina in summer over winter but absent in the MBL SCP fraction (Fig. S7). It has not been previously identified in L. hyperborea or other Laminaria spp. A component was noted with MS properties that match with a Eisenin-like peptide but of pyroglutamate, glutamate, and valine, which had a similar tissue and co-product distribution (see Fig. S2). These peptides and the amino acids, glutamic acid, glutamine, possibly pyroglutamate could contribute to umami flavour and contribute to the flavour of other Laminaria spp. consumed as Kombu (Pereira 2016). Glutamic acid is a key amino acid metabolite in growth, development, and adaptation to environmental stress in plants (Qiu et al. 2020) which might explain why it is mainly found in MBL LCP rather than being mostly accumulated in a particular tissue (see supplementary data Fig. S4). MBL LCP is the soluble fluid of the L. hyperborea peeled stipe tissue and may be similar to what might be carried out by Laminariales sieve cells (Lee 2008).
As could be expected, the positive mode MS profiles were dominated by nitrogenous compounds, which ionise more readily in this mode (Table 2). These included the quaternary ammonium betaines (proline betaine, glycine betaine, aminobutyric acid betaine, trigonelline and butryl carnitine), amino acids (carnitine, glutamic acid, glutamine, proline, threonine, and pyroglutamate), peptides (dialanine, glycyl lysine (in 2 isoforms), leucyl-glycyl-glycine and the Eisenin peptides) and a range of other amino-derivatives.
The various betaines have previously been identified in L. hyperborea and are thought to function as osmoregulatory components but also to contribute to the biostimulant effectiveness of seaweed extracts on plants (Blunden et al. 2010). As would be expected for components involved in osmoregulation of deep-water seaweeds, they were generally present in most tissues and co-products and varied only slightly between S and W. However, aminobutyric acid betaine appeared to be more abundant in MBL LCP, especially in the winter (Fig. S5).
Notably, the metabolic profiles of MBL SCP and SCP differed substantially although they differed only in their storage conditions. SCP was peeled from the stipe in the laboratory and was frozen and freeze-dried immediately. On the other hand, MBL SCP was prepared at MBL and stored for some time at 4°C after peeling, before transport to the James Hutton Institute and freeze-drying. Remarkably, two major components, ectoine and hydroxyectoine, were present in MBL SCP but absent in SCP (Fig. 4 & S3). Although ectoine has been noted in a previous LC-MS study of brown seaweed samples from Turbinaria species (Stranska-Zachariasova et al. 2017), ectoine and hydroxyectoine were first noted as osmoregulatory metabolites produced by the halotolerant marine bacteria, Ectothiorhodospira halochloris (Peters 1990), and subsequentially noted in other halotolerant species. Ectoine and hydroxyectoine act as osmoregulatory components in these bacteria and have also been considered as high-value cosmetic products (Czech et al. 2018).
The reduced levels of other components in MBL SCP compared to SCP, e.g.: mannitol, sulphosalicylate, pyroglutamate, diacylglycero-phosphocholine; triacylglycerol, fucoxanthin and piperidine dicarboxylic acid (Fig. 3–5) plus the presence and accumulation of known bacterial metabolites (e.g., ectoine and hydroxyectoine) suggests substantial metabolic transformation of the SCP tissue by microflora. This is also supported by the identification of 2-hydroxy-N-(7-methyloctyl) acetamide (m/z 202), which has been noted as a metabolite of Streptomyces species (Alshaibani et al. 2017) and accumulated in MBL SCP (Fig. S3). The presence of these microbial metabolites in certain samples suggests that they might arise from the endemic microflora. For example, penicinoline E, which was noted in many tissues and co-products (Fig. S6), was originally isolated from a marine-derived fungus (Penicillium spp.) (Gao et al. 2012). Also, there was a putative picolinic acid derivative which was much more abundant in Stipe and effectively absent from SCP (peeled bark from the stipe; Fig. 4). If this compound originated from endemic microflora, one might expect it to be present in the outer tissues of the bark rather than from the inner stipe tissues. Most research has focused on surface microflora in seaweeds (Singh and Reddy 2014) but specific endophytic microflora (e.g., Streblonema spp., Ectocarpus sp., Laminariocolax tomentosoides; Bolbocoleon spp., Acrochaete spp., Enteromorpha spp and Mikrosyphar zosterae) are known to invade the intercellular spaces of tissues of L. hyperborea, and also other seaweeds (e.g., Chondrus ocellatus) (Ellertsdóttir and Peters 1997; Ogandaga et al. 2017). It may be relevant that many of these putative microbial metabolites were more abundant in winter samples (Fig. 4), which as beach-cast material with apparent damage due to storm events, may have allowed microflora invasion.
Many of the putative metabolites were identified as components with central roles which would be expected to be present in seaweeds, e.g., gallic acid, succinic acid, malic acid, ketoglutarate and citric acid (see supplementary data Fig. 5, S2 and S5). Gallic acid might perform a structural role in the cell-wall, but it could also have a potential defensive role against herbivores and bacteria, or as antioxidant protection against UV radiation and oxidative stress (Mekinić et al. 2019) This would fit with its higher levels in Lamina (Fig S2), the main photosynthetic tissue. Its presence in SCP and Stipe could reflect a role as a precursor for secondary metabolites, as in plants (Muir et al. 2011) (Fig S2). Citric acid, succinic acid and ketoglutarate are central metabolic intermediates of the citric acid cycle/dark respiration, normally occurring in photosynthetic tissues (Raven and Beardall 2003; Lee 2008; Saxena et al. 2017). However, citric acid was more abundant in Lamina, while succinic acid and ketoglutarate were mainly present in MBL LCP (Fig. S5). As previously mentioned, MBL LCP reflects the soluble fluids of the L. hyperborea peeled stipe tissue. Therefore, succinic acid and ketoglutarate might have originated in Lamina, but are not accumulated there and are instead transported throughout the remaining seaweed thallus.
The peak with m/z 343 (RT 1529-Fig. 4) was most abundant in the Lamina and identified as maltitol, or mannitol-glucose, by exact mass formula and MS2 data. This could represent the first intermediate in the biosynthesis of M-chain laminarin oligosaccharides and ultimately laminarin polymers (Kadam et al. 2015). Lamina MeOH extracts and H2O extracts gave very similar MS traces in negative mode, with the mannitol peak dominant in both (Fig. 6). When the MS data was searched at m/z values equivalent to sequential addition of glucose (G; C6H10O5; + 162) units to mannitol (M), specific peaks could be discerned. The first in this series was m/z 343 maltitol (MG; C12H23O11; Table 1) but then a series of oligosaccharides were noted with predicted molecular formula consistent with laminarin oligomers, the largest being MG4 (Fig. 6). Oligosaccharides of larger chain length may be present, but their abundance diminished with chain length, so they were not detectable in these samples. This example highlights that the MS data can be mined for the presence of lower intensity peaks that may reveal further metabolic diversity. However, further targeted analyses would need to be performed to confirm the identity and levels of these new compounds and gain a better understanding of their variation across the samples and seasons.
There were several major components that could not be putatively identified even though they yielded MS data which gave predicted molecular formula with low errors (< 2 ppm). The probable reason for this number of unknowns (3 in positive mode and 8 in negative mode) is the paucity of metabolomic studies in seaweeds. Basically, these compounds have not yet been identified. Available databases are skewed towards mammalian, human and microbial metabolites, which means that the compounds are unlikely to be identified unless further research is undertaken. In future, the MS data could be subjected to other data handling methods to discern other lower intensity peaks or components that are partly hidden by other more abundant peaks. The MS data is rich and non-targeted analysis using various data-handling software techniques (e.g., using XCMS data deconvolution and use of multivariate statistics, e.g., McDougall et al. (2014), could compare the abundance of all peaks and pick out those which showed similar trends of tissue abundance, e.g., abundant in “lamina” or “stipe and LCP”) to those identified peaks shown above. The XCMS technique can also automatically fit the MS data to formula by systematically and interactively examining a wider range of elements, which could identify new classes of compounds.
The methods established here could be used in future work to track the seasonal metabolome between different tissues, and across a full year. Applying these methods to different potential biotopes of L. hyperborea, or other species, would bring a better understanding of the variation in metabolic processes associated with geolocation and help identify new components and new commercial opportunities.