Profiling Phlorotannins in Brown Macroalgae by ... - Wiley Online Library

Research Article Received: 18 November 2011,

Revised: 11 January 2012,

Accepted: 18 January 2012

Published online in Wiley Online Library

(wileyonlinelibrary.com) DOI 10.1002/pca.2354

Profiling Phlorotannins in Brown Macroalgae by Liquid Chromatography–High Resolution Mass Spectrometry Aaron J. Steevensz,a† Shawna L. MacKinnon,a Rachael Hankinson,a Cheryl Craft,a Solène Connan,b{ Dagmar B. Stengel,b and Jeremy E. Melansona* ABSTRACT: Introduction – Phlorotannins, phenolic compounds produced exclusively by Phaeophyceae (brown algae), have recently been associated with a wide variety of beneficial bioactivities. Several studies have measured the total phenolic content in extracts from various species, but little characterisation of individual phlorotannin components has been demonstrated. Objective – The purpose of this study was to develop a liquid chromatography–mass spectrometry (LC-MS) based method for rapid profiling of phlorotannins in brown algae. Methodology – Phlorotannin-enriched extracts from five phaeophyceaen species were analysed by ultrahigh-pressure liquid chromatography (UHPLC) operating in hydrophilic interaction liquid chromatography (HILIC) mode combined with high resolution mass spectrometry (HRMS). The method was optimised using an extract of Fucus vesiculosus; separation was achieved in less than 15 min. The basic mobile phase enhanced negative-ion electrospray ionisation (ESI), and generated multiply charged ions that allowed detection of high molecular weight phlorotannins. Results – The phlorotannin profiles of Pelvetia canaliculata, Fucus spiralis, F. vesiculosus, Ascophyllum nodosum and Saccharina longicruris differed significantly. Fucus vesiculosus yielded a high abundance of low molecular weight (< 1200 Da) phlorotannins, while P. canaliculata exhibited a more evenly distributed profile, with moderate degrees of polymerisation ranging from 3 to 49. HRMS enabled the identification of phlorotannins with masses up to 6000 Da using a combination of accurate mass and 13 C isotopic patterns. Conclusion – The UHPLC-HRMS method described was successful in rapidly profiling phlorotannins in brown seaweeds based on their degree of polymerisation. HILIC was demonstrated to be an effective separation mode, particularly for low molecular weight phlorotannins. Copyright © 2012 John Wiley & Sons, Ltd. Supporting information can be found in the online version of this article. Keywords: high resolution mass spectrometry; hydrophilic interaction chromatography; brown seaweed; phlorotannins; polyphenols

Introduction Phlorotannins are dehydro-oligomer derivatives of phloroglucinol (1,3,5-trihydroxy benzene) found exclusively in brown seaweeds (Phaeophyceae). As shown in Fig. 1, these polyphenols have been systematically categorised according to bonding type between phloroglucinol units, and/or the presence of additional hydroxyl groups (Amsler and Fairhead, 2006). For instance, fucols consist exclusively of aryl–aryl coupling, while phlorethols consist exclusively of arylether linkages (Fig. 1). Halogenated and sulphated phlorotannins have been reported (Glombitza and Knoss, 1992; Glombitza and Schmidt, 1999). The molecular weight (MW) of these compounds can range from 126 to > 1 105 Da (Parys et al., 2007). Phlorotannins are found in the epidermal cortex of brown seaweeds with concentrations ranging from 0.5 to 20% of the dry weight. However, concentrations vary with habitat, time of harvest, light intensity exposure and nutrient availability in the surrounding waters (e.g. Ragan and Glombitza, 1986; Targett and Arnold, 1998; Toth and Pavia, 2001). Phlorotannins are assumed to be formed biosynthetically via the polyketide pathway (Arnold and Targett, 2002; Amsler and

Phytochem. Anal. 2012

Fairhead, 2006) and to be involved in a number of secondary roles such as chemical defence, protection against UV irradiation, interactions with other organisms or the abiotic environment, as well as being integral components of cell walls (Schoenwaelder and Clayton 1998; Koivikko et al., 2005, 2007; Parys et al., 2007). The

* Correspondence to: J. E. Melanson, National Research Council, Institute for Marine Biosciences, 1411 Oxford Street, Halifax, Nova Scotia, Canada B3H 3Z1. E-mail: [email protected] a

National Research Council of Canada, Institute for Marine Bioscience 1411 Oxford Street, Halifax, Nova Scotia, Canada, B3H 3Z1

b

Botany and Plant Science, School of Natural Sciences, Ryan Institute for Environmental, Marine and Energy Research, National University of Ireland Galway, Galway, Ireland

† Current address: Department of Chemistry and Biochemistry, University of Windsor, 401 Sunset Avenue, Windsor, ON, Canada {

Current address: INTECHMER – CNAM, Digue de Collignon, BP 324, 50103 Cherbourg Cedex, France

Copyright © 2012 John Wiley & Sons, Ltd.

A. J. STEEVENSZ et al. OH

HO

OH

phloroglucinol OH OH

HO HO

OH

HO

OH

OH O

O

OH

O OH OH

OH

O

OH

phlorethol

fuhalol HO

O

OH HO

OH

HO

OH

HO

OH

OH OH O

OH HO

OH

O

HO HO

fucol

OH

O

eckol OH

HO

OH

HO

OH

OH

OH

isofuhalol

O

O

OH

OH

OH OH HO

O

HO

OH HO

HO

HO

HO

O

O

HO

HO

OH

O HO

OH

fucophlorethol

OH

Figure 1. Chemical structures of phloroglucinol and structural classes for six different classes of phlorotannins, adapted from (Amsler and Fairhead, 2006).

previously described variability in both phlorotannin levels and composition has important implications for potential commercial applications that require compound stability (Stengel et al., 2011). Recently, scientists have been intrigued with the wide variety of bioactivities and potential beneficial health effects of phlorotannins, including antioxidant properties (Shibata et al., 2008; Li et al., 2009), anti-allergic effects (Li et al., 2008), antiinflammatory activity (Kim et al., 2009), anti-HIV-1 activity (Artan et al., 2008), anti-carcinogenic activity (Kong et al., 2009), antidiabetic activity (Okada et al., 2004), acting as photochemopreventive agents (Hwang et al., 2006), anti-plasmin inhibitors (Fukuyama et al., 1989), HAase inhibitors (Shibata et al., 2002) and as bactericides (Nagayama et al., 2002). Given the high potential of phlorotannins for use as therapeutics, analytical methods are required for the characterisation of extracts for both discovery phase research and subsequent product development (standardisation and quality control). In many cases phlorotannins are analysed as total phenolics, where the total contents of phenolic compounds are measured by colorimetric assays such as the Folin-Ciocalteu (FC), the Folin–Denis reagent and 2,4dimethoxybenzaldehyde (DMBA) (Stern et al., 1996; Parys et al., 2007). Although these tests are simple to use, they

wileyonlinelibrary.com/journal/pca

provide little information on the chemical composition of the phenolic content. Due to the high complexity of some phenolic extracts, chromatographic techniques are an attractive option for their analyses. However, analyses of brown algal phlorotannins have been hampered by their high susceptibility to oxidation and that they have considerable similarities in physical properties, despite having broad structural and molecular weight diversities (Koivikko et al., 2007; Parys et al., 2007). Reversed phase HPLC is a powerful separation mode, but the high polarity of phlorotannins causes them to elute with little or no retention due to the lack of interaction with the non-polar stationary phase (Koivikko et al., 2007). Normal phase liquid chromatography (NPLC) is more appropriate for retaining polar compounds, and was shown to offer superior performance when compared with reversed phase for the separation of phlorotannins from Fucus vesiculosus (Koivikko et al., 2007). However, interfacing typical NPLC solvents with ESI can be problematic due to reduced sensitivity caused by poor ionisation efficiency in 100% non-polar solvents (Koivikko et al., 2007; Yanagida et al., 2007; Nguyen and Schug, 2008). In addition, poor hydrophilic analyte solubility in NPLC eluents can promote adsorption onto the silica stationary phase, compromising peak shape and resulting in long retention times. Finally, the need for chlorinated solvents in NPLC makes the technique less attractive for high-throughput applications. In contrast to NPLC, hydrophilic interaction liquid chromatography (HILIC) uses a polar stationary phase with aqueous/polar organic mobile phase in which water is introduced as the stronger eluting solvent (generally 5–50%; Alpert, 1990). While the mechanism for separation is still under debate, it was originally proposed that partitioning of analytes occurred between a water-enriched layer of stagnate eluent partially immobilised on a hydrophilic stationary phase and the relatively hydrophobic bulk eluent (Alpert, 1990). Considered a ‘mixed-mode’ separation scheme (Hemstrom and Irgum, 2006), HILIC offers selectivity that is based on electrostatic interactions, dipole–dipole interactions, cation/anion exchange and hydrogen bonding. Despite these potential benefits, very few applications of HILIC for the separation of polyphenolics have been reported (Yanagida et al., 2007). In this study we employ HILIC in combination with highresolution mass spectrometry (HILIC-HRMS) for analysing crude phlorotannin enriched extracts from five different brown seaweed species. Separation was achieved by ultrahigh-pressure liquid chromatography (UHPLC) using a HILIC column with amide functionality. HRMS has emerged as an ideal tool for profiling complex samples due to its sensitivity, mass accuracy and rapid full scanning mode. Mass spectrometry data are collected in a non-targeted fashion, allowing for retrospective data analysis for compounds of interest that may arise in the future.

Experimental Extraction and sample preparation The following five seaweed species were collected: whole thalli of Fucus vesiculosus and Ascophyllum nodosum were collected from Ketch Harbour, Nova Scotia, Canada in October 2008 (vegetative biomass only for both species); samples of F. spiralis were collected from Digby Neck, Nova Scotia, Canada in July 2008 (samples included some reproductive biomass); samples of Saccharina longicruris consisting of stipes and blades were collected from Duncan’s



Phlorotannins in Seaweed by LC-HRMS Cove, Nova Scotia, Canada (July 2008; presence of sori not noted but most biomass vegetative), and whole individuals of Pelvetia canaliculata (vegetative biomass only) from Spiddal, Galway Bay, Ireland, in August 2008. Algal specimens were prepared and species identification was confirmed by phycologists Carolyn Bird and Dr Dagmar Stengel. The collected samples consisting of at least five individuals (> 50 in the case of P. canaliculata) were freeze-dried and ground to a fine powder using a FitzWMill grinder. The dried powder was stirred with 80:20 methanol:water (7.5–13.2:1 solvent:mass) for 30 min, after which the supernatant was decanted and the remaining pellet was extracted three subsequent times (3.4–6.6:1 solvent:mass). The extraction solvent: mass ratios varied based on the degree of swelling of the dried seaweed powder of the different species. The extractants for each species were pooled and the methanol was removed using a Büchi Rotovapor R-114 with a B-480 water heater set at 35 C. The remaining aqueous mixture was defatted three to four times using dichloromethane (1:1 v/v) partitioning, which was conducted in a vial. The resulting aqueous phase was evaporated under vacuum to remove any remaining organic solvent and then freeze-dried using a VirTis Genesis Pilot Lyophilizer with a Wizard 2.0 control system. The aqueous extract containing carbohydrates and phlorotannins was dried and then fractionated using a C18 Sep-Pak cartridge (6 cc; 500 mg), which had been preconditioned with 12 mL of methanol followed by 18 mL of Milli Q (MQ) water. A sample (100–200 mg) of the freeze-dried aqueous fraction was dissolved in 2–3 mL of MQ water (sonicated when necessary) and loaded onto the cartridge, which was then washed with 20 mL of MQ water to elute the salt and carbohydrate components. The phlorotannins were then eluted with 30 mL of methanol. The methanol was removed using a stream of nitrogen and the resulting phlorotannin-enriched fraction was suspended in MQ water, freeze-dried and stored under N2 gas at 80 C. These fractions were dissolved into 75:25 acetonitrile:methanol at 1 mg/mL and 3 mg/mL and filtered using UltrafreeW-MC centrifugal filters with Durapore PVDF membranes (0.45 mm pore size).

Ultrahigh-pressure liquid chromatography conditions Ultrahigh-pressure liquid chromatography was performed using an Accela High Speed LC from Thermo Scientific capable of handling sub-2-mm particle columns up to 15000 psi, equipped with a quaternary pump and autosampler. Separations were achieved using a Waters UPLCW BEH Amide 1.7 mm (2.1 x 100 mm) column maintained at 30 C using 3-mL injections and a flow-rate of 400 mL/min. Mobile phase A was composed of 10.0 mM ammonium acetate adjusted to pH 9.0 with ammonium hydroxide and mobile phase B was acetonitrile. The gradient consisted of an initial hold at 5% mobile phase A for 1 min, followed by a linear gradient to 35% A in 16 min, followed by re-equilibration for 5 min at 5% A, for a total run time of 22 min.

Mass spectrometry conditions Mass spectrometry was performed on a Thermo Scientific Exactive™ bench top mass spectrometer with Orbitrap™ technology capable of sub-ppm mass accuracy and resolution up to 100000. Scans were collected in negative ESI mode over a range of m/z 150–2000 in the ‘ultra-high’ resolution setting (100000, 1 Hz). The instrument was calibrated daily to optimise performance and mass accuracy by infusing a calibration solution at 5 mL/min containing 10 pmol/mL sodium dodecyl sulphate (SDS), 10 pmol/mL sodium taurocholate, 0.001% Ultramark 1621, and 0.01% acetic acid in 50:25:25 acetonitrile:methanol:water. For flow rates of 400 mL/min the heated electrospray ionisation source (HESI) was configured as follows: sheath gas 55 psi, auxillary gas 18 (arbitrary units), sweep gas 0 (arbitrary units), spray voltage 2.7 kV, capillary temperature 350 C, capillary voltage 60 V, tube lens voltage 125 V and heater temperature 300 C. Xcalibur 2.1 software was used for instrument control and data acquisition.


Results and Discussion Selection of liquid chromatography stationary phase Since HILIC was first described as a variant to NPLC in 1990, a variety of polar stationary phases have emerged that are composed of bare silica, or silica phases modified with amino, amide or zwitterionic functional groups (Alpert, 1990; Hemstrom and Irgum, 2006). In this study, a column composed of bridged ethylene hybrid particles with trifunctional amide ligands (BEH amide) was investigated because of its similarity to the Amide-80 column, which successfully separated proanthocyanidins based on their degree of polymerisation (DP) up to pentamers (Yanagida et al., 2007). The BEH amide particles are stable over a broad pH range (2–11). As phlorotannins are weak acids (phloroglucinol pKa1 = 9.01), post-column addition of ammonium hydroxide has been employed in NPLC separations of phlorotannins, presumably to help facilitate negative mode ionisation by increasing the pH (Koivikko et al., 2007). Since the BEH particles are resistant to degradation at elevated pH values, our strategy involved adjusting the mobile phase to pH 9.0, thereby eliminating the complication of post-column addition. As well as offering enhanced ionisation efficiency, partial ionisation of the phlorotannins could also increase chromatographic retention on the amide stationary phase, as previously demonstrated on the BEH amide column with HILIC analysis of organophosphates (McCalley, 2007; Fountain et al., 2010).

Optimisation of liquid chromatography conditions There are no commercially available standards for phlorotannin oligo- or polymers. As a result, the phlorotannin-enriched fraction we prepared from F. vesiculosus extract obtained from whole (vegetative) individuals, which was found to have a high concentration of low mass phlorotannins, was used for optimisation of the separation. The water–acetonitrile gradient was optimised to maximise peak capacity to allow identification of the greatest number of phlorotannin compounds. As detection by HRMS does not require complete separation of all compounds for identification, enhancing the resolution of specific phlorotannins species was not performed. The effect of altering the final aqueous phase content in the mobile phase is shown in Fig. 2, which displays gradients that go from an initial 5% water to final aqueous contents ranging from 50 to 30% over 16 min. These total ion chromatograms (TIC) for F. vesiculosus yielded many peaks pertaining to low mass phlorotannins ranging from DP 3 to 10, tentatively identified based on their accurate masses. In general, decreasing the water content increased the retention times and in some cases gave better baseline resolution, but sensitivity and peak shape were sacrificed. According to our results, a gradient ranging from 5% to 35% aqueous phase yielded a reasonable compromise between run times, peak shape and sensitivity. In an effort to enhance chromatographic selectivity, methanol was investigated as a protic modifier in the mobile phase. Studies comparing non-aqueous HILIC with traditional aqueous-HILIC have demonstrated that elution strengths of protic modifiers are as follows: water > ethyleneglycol > methanol > ethanol, but under certain conditions can be accompanied by distinct effects on chromatographic selectivity (Bicker et al., 2008). As shown in Fig. 3, gradients with 0, 5, 10 and 15% methanol supplementing



A. J. STEEVENSZ et al.

a)

200

8a

4a 5c 5b

120

6a 7a

8b

4b

40

3a

30

5a 3b

20

9a 10a

10

9a 10 a

3b

b)

50

5a

80

6b 7a 6e 6c

6a

9a 10a

3b

106

4a

7b 5b 5a 5c

80

6d 6b

6a 3a

40

4b 3b

5b 5a 5c

3b 3c

20

7a 6 7b 6b 6d e 7c 8a 6a 6c 9a 8b 9 10a

4b

3a

b

x106

c)

160 120

30

10

8b

4b

3a

x106

8a

6d

Intensity

106

Intensity

7b

5c 5b

120

200

4a

40 4a

160

40

7 6b a 7b 6 6c d 8a 6a 6e

4b

3c 3a

b)

200

4a 5c 5b

6e 6d

6c 5a 6b

80

a)

40

x106

160

106

50

7b

6e

7a 6c

140 120 100 80 60 40 20

c)

120

d)

4a

3b

3a

4b

4

5a

5b

6a

8a 7c

9a 8b

7 6d 6e b 8a 7c 7 9 10 6c a 8b a 9b a

6b

4a

100

10a

9b

4

5c

4

5

6

7

8

9

10

11

12

Time (min)

60 5a

40

Figure 2. LC-MS total ion current (TIC) chromatograms of a phlorotannin-enriched Fucus vesiculosus extract demonstrating the effect of altering the final aqueous phase content in the mobile phase over the 16 min gradient: (a) 5 to 50%; (b) 5 to 40%; and (c) 5 to 30%. Peak numbers correspond to the number of phloroglucinol units and the letters correspond to different isomers of the same degree of polymerisation. For situations where multiple peaks are co-eluting, the peak assignment is based on the most intense peak in the MS spectrum.

the aqueous phase did not significantly alter chromatographic selectivity, but baseline resolution is enhanced at the cost of peak shape and peak capacity. Although the addition of up to 15% methanol (Fig. 3d) did generate a separation of some additional structural isomers, the loss of peak shape and sensitivity outweighed any marginal enhancements in chromatographic resolution. Therefore, methanol was not employed as a protic modifier for further studies. The optimised separation that used a water gradient from 5 to 35% is shown in Fig. 4. It should be noted that under all separation conditions tested, larger phlorotannins (> 1200 Da) generated broad peaks that eluted between 10 and 14 min and many co-eluted, despite their moderate abundance in F. vesiculosus and some other seaweed species (see below). This is consistent with previous NPLC studies of F. vesiculosus, whereby it appeared that only smaller phlorotannins were separated but identification was not possible with the UV detector employed (Koivikko et al., 2007). Therefore, it seems that larger phlorotannins are either retained irreversibly or these


x106

80 3

4b

20

3b

3a

4

5

6

5b 5c

4

7

6a

8

9

6d 7 6e b 7c 8a 7 10 6c a 8b 9a 9b a

6b

10

11

12

13

14

Time (min) Figure 3. TIC chromatograms of a phlorotannin-enriched Fucus vesiculosus extract demonstrating the effect of various amounts of methanol supplementation of the aqueous phase over 16 min: (a) 0% methanol; (b) 5% methanol; (c) 10% methanol; and (d) 15% methanol. Peaks are numbered as in Fig. 2.

separation modes are not effective in resolving highly retained compounds. Another possibility for this phenomenon is simply that the isomers for these larger phlorotannins are partially resolved and appear as broad peaks. As the biosynthetic pathway for phlorotannin synthesis is largely unknown and could involve indiscriminant coupling, the number of isomers would increase with size due the increase in variety of branching and bonding type. The net result is probably a combination of both issues, whereby both reduced chromatographic performance and the greater number of isomers contribute to the broad and co-eluting peaks. High resolution mass spectrometry of phlorotannins Mass spectrometry data were collected in a non-targeted fashion, by acquiring full spectrum data in negative ion mode from m/z 150 to 2000. The data were then analysed by searching for the theoretical mono-isotopic masses corresponding to all possible phlorotannins over a broad range of DP. For instance,



Phlorotannins in Seaweed by LC-HRMS

1-

4a

1.1 Relative Abundance

90

1.0 0.9

Intensity (x109)

6d

100

6d

0.8 0.7

80 70 60 6b

50 40

4b

5c

0.4 0.3 0.2

6

7

6.8

7.2

7.6

8.0

8.4

8.8

Time (min)

10a

9

10

11

12

13

14

Time (min) Figure 4. TIC chromatogram of a phlorotannin-enriched Fucus vesiculosus employing the LC conditions determined to be optimal. The regions of the chromatogram where the predominant charge state observed for the phlorotannins are indicated. (Inset) Extracted ion chromatogram of m/z 745.10463, demonstrating the number of phlorotannin isomers derived from six phloroglucinol units. Peaks are numbered as in Fig. 2.

extracting m/z 745.10464 revealed the presence of a phlorotannin with six phloroglucinol units in F. vesiculosus. In this case, at least seven chromatographic peaks were revealed in the extracted ion chromatogram (XIC) shown in Fig. 4 (inset), suggesting that isomers are being separated. If this separation follows a similar mechanism to that of proanthocyanidins separated with the TSKgel Amide 80 column (Yanagida et al., 2007), we would expect retention times of compounds with the same degree of polymerisation to increase as follows: phlorethols < fucophlorethols < fucols, based on the number of free hydroxyl groups. In the future, this could potentially be verified by tandem mass spectrometry experiments, but this was considered unnecessary for the rapid profiling method reported, and thus was not performed. Despite the limited mass range scanned (m/z 2000), larger phlorotannins could be detected as they appeared as multiply charged ions as the number of free hydroxyl groups increased. As shown in Fig. 4, singly charged ions [M H]1 were observed between 5 and 10 min of the chromatogram, followed by doubly charged ions [M 2H]2 between 9 and 12 min, and triply charged ions [M 3H]3 between 11 and 13 min. Typical mass spectra of phlorotannins in F. vesiculosus over this range of charge states are shown in Fig. 5. Figure 5a shows the singly charged spectrum of the hexamer, whose mono-isotopic mass was detected at m/z 745.10753. Figure 5b displays the spectrum of a phlorotannin polymer with DP of 16 at m/z 992.13320, whereby spacing of the 13 C isotope peaks of m/z 0.5 indicates a doubly charged ion. Similarly, Figure 5c shows the spectrum of a phlorotannin with DP of 34 at m/z 1405.18545, whereby spacing of the 13 C isotope peaks of m/z 0.33 indicates a triply charged ion. The spectrum of this larger phlorotannin with 34 phloroglucinol units (C204H138O132; MW = 4701.19 g/mol) demonstrates the need for HRMS in the identification of phlorotannins, as lower resolution mass spectrometers would not be able to resolve the


747.11080

20

m/z

8b

8

746.11080

40

744.0 744.5 745.0 745.5 746.0 746.5 747.0 747.5 748.0

9b

5

60

6c

9a

0.1

80

6a

6e

6c

z = -1 DP = 6 Formula: C36H26O18 MW = 746.11 g/mol

745.10753

0

7b

6a 7a

a)

10

8a 5a5b 6b

0.5

100

0

30 20

0.6

6e

Relative Abundance (%)

1.2

3-

100

b)

992.13320

992.63477

z=-2 DP = 16 Formula: C96H66O48 MW = 1986.27 g/mol

80 993.13637

60 40 993.63727

20 0 991.5 992.0 992.5 993.0 993.5 994.0 994.5 995.0

m/z 100


[M-H]

[M-3H]

2-


[M-2H]

1405.85316

c)

1406.18790

1405.51861

80

z=-3 DP = 34 Formula: C204H138O102 MW = 4218.56 g/mol

1406.52194

60 40

1406.52194

1405.18545

20 0 1405.0

1406.0

1407.0

1408.0

m/z Figure 5. Typical mass spectra of phlorotannins in Fucus vesiculosus: (a) C36H26O18, DP = 6, charge state = 1; (b) C96H66O48, DP = 16, charge state = 2; (c) C204H138O102, DP = 34, charge state = 3.

isotopic peaks to determine the charge state, without which the MW and molecular formula could not be determined. Using this approach for the identification of the phlorotannins, polymers with up to 49 phloroglucinol units were detected, and their masses and associated molecular formula are listed in Supplementary Material Table S1. Typically, experimental masses were matched within 5 ppm, while some low-abundant compounds yielded mass errors approaching 10 ppm. Despite the large phlorotannins detected in this study, they are still considerably smaller than those detected by size exclusion chromatography (SEC) (Ragan and Glombitza, 1986; Parys et al., 2007). However, these are the largest phlorotannins detected by MS to the best of our knowledge, whereby the exact molecular formula and degree of polymerisation was determined. Profiling of brown seaweeds The profiles of the phlorotannin-enriched extracts of five brown seaweed species (P. canaliculata, F. spiralis, F. vesiculosus, A. nodosum, S. longicruris) were compared using a heat map to visualise the phlorotannin compounds identified with observed



A. J. STEEVENSZ et al. masses corresponding to the varying number of phloroglucinol units (Fig. 6). As significantly higher levels of low molecular weight pholorotannins were observed for the extract of F. vesiculosus compared with other species, the heat map is based on a log scale to better accommodate the majority of the data. The numbers within the boxes refer to the number of isomers detected at that particular DP. Interestingly, phloroglucinol units up to 39 could be detected in the four species collected in Nova Scotia (F. spiralis, F. vesiculosus, A. nodosum, and S. longicruris) but 49 units were detected from P. canaliculata. The latter species occupies extreme upper limits of intertidal shores in Ireland and regularly experiences prolonged dehydration for several days during neap tides, in addition to potentially extreme, and sharply fluctuating, temperatures and high solar radiation. By contrast, the heat map of S. longicruris, a large kelp that grows in subtidal regions and is therefore adapted to more stable and less extreme environmental conditions, revealed a simpler phlorotannin profile dominated by phlorotannin DPs of 17. The profiles of the two Fucus species were surprisingly different as that of F. spiralis had a lower content of phlorotannin compounds above 14 DP. However, samples of these two species were collected at different locations, and F. spiralis samples included some reproductive biomass while F. vesiculosus biomass was vegetative. It is thus possible that some of the differences observed in Fig. 6 were due to differences in sampling location, or to exposure and adaptation to different environmental regimes. In addition, differences in the reproductive states of the different seaweed species at time of collection could account for some of the variation observed in Fig. 6. Previous research has shown that phenolic composition also potentially varies within algal thalli (e.g. Connan et al., 2006). In this instance, neither S. longicruris nor P. canaliculata samples collected contained reproductive biomass, but 50

10

40

1

0.1

Phloroglucinol Units

100

for the kelp both stipes and blades were included. Interestingly, although the Pelvetia samples probably were biologically the most homogeneous (no differentiation into main stems/stipes as for the other brown algae examined, and exclusion of reproductive receptacles), extracts exhibited the greatest diversity of different degrees of polymerisation. The development of this profiling method will make it possible to conduct further more detailed analyses of internal, spatial and temporal variability on phlorotannins in macroalgae. Improved analytical methodology in tandem with an advancement of our understanding of the linkages between phenolic composition and bioactivity will be invaluable for future targeted applications of seaweed phenolics.

Supporting Information Supporting information can be found in the online version of this article. Acknowledgements The authors wish to thank Thermo Fisher Scientific for the loan of the Exactive™ mass spectrometer. This is NRC publication no. 54080.

References Alpert AJ. 1990. Hydrophilic-interaction chromatography for the separation of peptides, nucleic acids and other polar compounds. J Chromatogr A 499: 177–196. Amsler CD, Fairhead VA. 2006. Defensive and sensory chemical ecology of brown algae. In Advances in Botanical Research, Vol. 43. Academic Press: London; 1–91.

30

20

Log Scale

10

3

6 5 3 7 7 6 4 4

2 5 3 2 3 8 5 3

6 5 6 6 8 6 3 2

1 2 2 3 4 3 3 2

2 7 4 4 5 7 5 3

Fucus vesiculosus

Ascophyllum nodosum

Fucus spiralis

Saccharina longicruris

Pelvetia canaliculata

Figure 6. Heat map displaying relative abundances based on peak areas of the phlorotannins at each DP over the five brown algal species studied. The numbers in the boxes from DP 3 to 10 correspond to the number of isomer peaks detected at the corresponding mass.




Phlorotannins in Seaweed by LC-HRMS Arnold TM, Targett NM. 2002. Marine tannins: The importance of a mechanistic framework for predicting ecological roles. J Chem Ecol 28: 1919–1934. Artan M, Li Y, Karadeniz F, Lee SH, Kim MM, Kim SK. 2008. Anti-HIV-1 activity of phloroglucinol derivative, 6,6 ’-bieckol, from Ecklonia cava. Bioorgan Med Chem 16: 7921–7926. Bicker W, Wu JY, Lämmerhofer M, Lindner W. 2008. Hydrophilic interaction chromatography in nonaqueous elution mode for separation of hydrophilic analytes on silica-based packings with noncharged polar bondings. J Sep Sci 31: 2971–2987. Connan S, Delisle F, Deslandes E, Ar Gall E. 2006. Intra-thallus phlorotannin content and antioxidant activity in Phaeophyceae of temperate waters. Bot Mar 49: 39–46. Fountain KJ, Xu J, Diehl DM, Morrison D. 2010. Influence of stationary phase chemistry and mobile-phase composition on retention, selectivity, and MS response in hydrophilic interaction chromatography. J Sep Sci 33: 740–751. Fukuyama Y, Kodama M, Miura I, Kinzyo Z, Mori H, Nakayama Y, Takahashi M. 1989. Anti-plasmin Inhibitor. V. Structures of novel dimeric eckols isolated from the brown alga Ecklonia kurome OKAMURA. Chem Pharm Bull 37: 2438–2440. Glombitza K-W, Knoss W. 1992. Sulphated phlorotannins from the brown alga Pleurophycus gardneri. Phytochemistry 31: 279–281. Glombitza KW, Schmidt A. 1999. Nonhalogenated and halogenated phlorotannins from the brown alga Carpophyllum angustifolium. J Nat Prod 62: 1238–1240. Hemstrom P, Irgum K. 2006. Hydrophilic interaction chromatography. J Sep Sci 29: 1784–1821. Hwang H, Chen T, Nines RG, Shin HC, Stoner GD. 2006. Photochemoprevention of UVB-induced skin carcinogenesis in SKH-1 mice by brown algae polyphenols. Int J Cancer 119: 2742–2749. Kim AR, Shin TS, Lee MS, Park JY, Park KE, Yoon NY, Kim JS, Choi JS, Jang BC, Byun DS, et al.. 2009. Isolation and identification of phlorotannins from Ecklonia stolonifera with antioxidant and anti-inflammatory properties. J Agr Food Chem 57: 3483–3489. Koivikko R, Loponen J, Honkanen T, Jormalainen V. 2005. Contents of soluble, cell-wall-bound and exuded phlorotannins in the brown alga Fucus vesiculosus, with implications on their ecological functions. J Chem Ecol 31: 195–212. Koivikko R, Loponen J, Pihlaja K, Jormalainen V. 2007. High-performance liquid chromatographic analysis of phlorotannins from the brown alga Fucus vesiculosus. Phytochem Anal 18: 326–332. Kong CS, Kim JA, Yoon NY, Kim SK. 2009. Induction of apoptosis by phloroglucinol derivative from Ecklonia cava in MCF-7 human breast cancer cells. Food Chem Toxicol 47: 1653–1658. Li Y, Lee SH, Le QT, Kim MM, Kim SK. 2008. Anti-allergic effects of phlorotannins on histamine release via binding inhibition between IgE and Fc epsilon RI. J Agr Food Chem 56: 12073–12080.


Li Y, Qian ZJ, Ryu BM, Lee SH, Kim MM, Kim SK. 2009. Chemical components and its antioxidant properties in vitro: An edible marine brown alga, Ecklonia cava. Bioorgan Med Chem 17: 1963–1973. McCalley DV. 2007. Is hydrophilic interaction chromatography with silica columns a viable alternative to reversed-phase liquid chromatography for the analysis of ionisable compounds? J Chromatogr A 1171: 46–55. Nagayama K, Iwamura Y, Shibata T, Hirayama I, Nakamura T. 2002. Bactericidal activity of phlorotannins from the brown alga Ecklonia kurome. J Antimicrob Chemoth 50: 889–893. Nguyen HP, Schug KA. 2008. The advantages of ESI-MS detection in conjunction with HILIC mode separations: Fundamentals and applications. J Sep Sci 31: 1465–1480. Okada Y, Ishimaru A, Suzuki R, Okuyama T. 2004. A new phloroglucinol derivative from the brown alga Eisenia bicyclis: Potential for the effective treatment of diabetic complications. J Nat Prod 67: 103–105. Parys S, Rosenbaum A, Kehraus S, Reher G, Glombitza KW, Konig GM. 2007. Evaluation of quantitative methods for the determination of polyphenols in algal extracts. J Nat Prod 70: 1865–1870. Ragan MA, Glombitza KW. 1986. Phlorotannins, brown algal polyphenols. In Progress in Phycological Research, Round FE, Chapman DJ (eds). Biopress: Bristol; 129–241. Schoenwaelder MEA, Clayton MN. 1998. Secretion of phenolic substances into the zygote wall and cell plate in embryos of Hormosira and Acrocarpia (Fucales, Phaeophyceae). J Phycol 34: 969–980. Shibata T, Fujimoto K, Nagayama K, Yamaguchi K, Nakamura T. 2002. Inhibitory activity of brown algal phlorotannins against hyaluronidase. Int J Food Sci Tech 37: 703–709. Shibata T, Ishimaru K, Kawaguchi S, Yoshikawa H, Hama Y. 2008. Antioxidant activities of phlorotannins isolated from Japanese Laminariaceae. J Appl Phycol 20: 705–711. Stengel DB, Connan S, Popper ZA. 2011. Algal chemodiversity and bioactivity: Sources of natural variability and implications for commercial application. Biotechnol Adv 29: 483–501. Stern JL, Hagerman AE, Steinberg PD, Mason PK. 1996. Phlorotanninprotein interactions. J Chem Ecol 22: 1877–1899. Targett NM, Arnold TM. 1998. Predicting the effects of brown algal phlorotannins on marine herbivores in tropical and temperate oceans. J Phycol 34: 195–205. Toth GB, Pavia H. 2001. Removal of dissolved brown algal phlorotannins using insoluble polyvinylpolypyrrolidone (PVPP). J Chem Ecol 27: 1899–1910. Yanagida A, Murao H, Ohnishi-Kameyama M, Yamakawa Y, Shoji A, Tagashira M, Kanda T, Shindo H, Shibusawa Y. 2007. Retention behavior of oligomeric proanthocyanidins in hydrophilic interaction chromatography. J Chromatogr A 1143: 153–161.

