NMR Characterization and Evaluation of Antibacterial

Phytochemistry 164 (2019) xxx–xxx
Contents lists available at ScienceDirect
Phytochemistry
journal homepage: www.elsevier.com/locate/phytochem
NMR characterization and evaluation of antibacterial and antiobiofilm T

activity of organic extracts from stationary phase batch cultures of five
marine microalgae (Dunaliella sp., D. salina, Chaetoceros calcitrans, C. gracilis
and Tisochrysis lutea)
Ma José Iglesiasa,∗, Raquel Soengasa, Ian Probertb, Emilie Guilloudb, Priscillia Gourvilb,
Mohamed Mehiric, Yuly Lópezd, Virginio Cepasd, Ignacio Gutiérrez-del-Ríoe,
Saúl Redondo-Blancoe, Claudio J. Villare, Felipe Lombóe, Sara Sotod, Fernando López Ortiza,∗∗
a
Área de Química Orgánica, Research Centre CIAIMBITAL, Universidad de Almería, Carretera de Sacramento s/n, Almería, 04120, Spain
b
Roscoff Culture Collection, FR2424 Station Biologique de Roscoff (Sorbonne Université / CNRS), 29680, Roscoff, France
c
Institut de Chimie de Nice, UMR CNRS 7272, Université Nice Sofia Antopolis, 06103, Nice, France
d
Barcelona Institute for Global Health (ISGlobal)-Hospital Clinic-Universitat de Barcelona, Carrer Rosselló 132, 08036, Barcelona, Spain
e
Research Group BIONUC, Departamento de Biología Funcional, Área de Microbiología, University of Oviedo, Oviedo, Principality of Asturias, Spain. IUOPA (Instituto
Universitario de Oncología del Principado de Asturias), ISPA (Instituto de Investigación Sanitaria del Principado de Asturias), Spain
ARTICLE INFO ABSTRACT
Keywords: The chemical composition of five marine microalgae (Dunaliella sp., Dunaliella salina, Chaetoceros calcitrans,
Dunaliella Chaetoceros gracilis and Tisochrysis lutea) was investigated through nuclear magnetic resonance (NMR) spec-
Chaetoceros troscopic study of the soluble material obtained by sequential extraction with hexane, ethyl acetate (AcOEt) and
Tisochrysis methanol of biomass from stationary phase cultures. Hexane extracted the major lipids present in the microalgae
NMR
during the stationary phase of growth, which correspond to storage lipids. Triacylglycerols (TGs) were the only
Metabolite identification
storage lipids produced by Dunaliella and Chaetoceros. In contrast, T. lutea predominantly stored polyunsaturated
Antibacterial
Antibiofilm long-chain alkenones, with sterols also detected as minor components of the hexane extract. The molecular
structure of brassicasterol was determined in T. lutea and the presence of squalene in this sample was also
unequivocally detected. Monogalactosyldiacylglycerols (MGDGs) and pigments were concentrated in the AcOEt
extracts. C. calcitrans and D. salina constituted an exception due to the high amount of TGs and glycerol pro-
duced, respectively, by these two strains. Chlorophylls a and b and β-carotene were the major pigments syn-
thesized by Dunaliella and chlorophyll a and fucoxanthin were the only pigments detected in Chaetoceros and T.
lutea. Information concerning the acyl chains present in TGs and MGDGs as well as the positional distribution of
acyl chains on the glycerol moiety was obtained by NMR analysis of hexane and AcOEt extracts, with results
consistent with those expected for the genera studied. Fatty acid composition of TGs in the two Dunaliella strains
was different, with polyunsaturated acyl chains almost absent in the storage lipids produced by D. salina. Except
in C. calcitrans, the polar nature of soluble compounds was inferred through the relative extraction yield using
methanol as the extraction solvent. Glycerol was the major component of this fraction for the Dunaliella strains.
In T. lutea 1,4/2,5-cyclohexanetetrol (CHT) and dimethylsulfoniopropionate (DMSP) preponderated. CHT was
also the major polyol present in the Chaetoceros strains in which DMSP was not detected, but prominent signals
of 2,3-dihydroxypropane-1-sulfonate (DHSP) were observed in the 1H NMR spectra of methanolic extracts. The
presence of DHSP confirms the production of this metabolite by diatoms. In addition, several other minor
compounds (digalactosyldiacyglycerols (DGDGs), sulphoquinovosyldiacylglycerols (SQDGs), amino acids, car-
bohydrates, scyllo-inositol, mannitol, lactic acid and homarine) were also identified in the methanolic extracts.
The antibacterial and antibiofilm activities of the extracts were tested. The AcOEt extract from C. gracilis showed
a moderate antibiofilm activity.
∗
Corresponding author. Área de Química Orgánica, Universidad de Almería. Carretera de Sacramento s/n, 04120 Almería, Spain.
∗∗
Corresponding author. Área de Química Orgánica, Universidad de Almería. Carretera de Sacramento s/n, 04120 Almería, Spain.
E-mail addresses: [email protected] (M.J. Iglesias), [email protected] (F.L. Ortiz).
https://doi.org/10.1016/j.phytochem.2019.05.001
Received 9 January 2019; Received in revised form 11 April 2019; Accepted 6 May 2019
0031-9422/ © 2019 Elsevier Ltd. All rights reserved.
M.J. Iglesias, et al. Phytochemistry 164 (2019) xxx–xxx
1. Introduction identifying and quantifying the most abundant metabolites present in

biological material. The suitability of NMR techniques for algae analysis
Microalgae are photosynthetic organisms that display high levels of was underlined for more than two decades (Pollesello et al. 1992). In
bio- and chemo-diversity due to the wide range of habitats in which contrast to convectional chromatographic methods that require further
they live and the ability to survive in competitive environments. The fractionation and derivatization for full lipid characterization, the in-
estimated total number of microalgal species ranges from 200,000 to formation contained in a single NMR spectrum covers a wide range of
800,000, but only about 35,000 have been formally described and the compounds. Nonetheless, reports on the analysis of microalgae lipid
number that have been comprehensively examined from the structural- fraction by NMR are limited (Beal et al., 2010; Nuzzo et al., 2013;
chemical point of view is extremely lower (Guiry, 2012; Mobin and Kumar et al., 2014; Sarpal et al., 2015, 2016a, 2016b, 2016c). Simi-
Alam, 2017). Characterization of the chemical composition of micro- larly, metabolomic studies of algal species are also scarce (Kim et al.,
algae is important in order to obtain taxonomic information (Lang 2006; Gupta et al., 2013; Akhter et al., 2016; Azizan et al., 2018;
et al., 2011) and also to promote rational use of this important natural Chauton and Størseth, 2018; Réveillon et al., 2018). The high chemo-
resource. The unique composition (notably of lipids) of microalgae diversity of these organisms makes metabolite identification a complex
means they have enormous potential for applications in important and challenging issue. In a very recent work, Azizan et al. (2018) have
biotechnological fields such as biofuel production, human nutrition, tried to correlate the 1H metabolic profiling of Chaetoceros calcitrans
aquaculture and the discovery of novel active compounds (Wikfors and soluble material obtained using different solvents with antioxidant ac-
Ohno, 2001; Guiry, 2012; Bhattacharjee, 2016; Matos, 2017; Chia et al., tivity. They reported the identification of 29 metabolites. However,
2018). Furthermore, the rapid growth rate potentially facilitates large- some structural assignments are not adequately supported by the ex-
scale biomass production with manipulation of cultivation conditions perimental data provided, which gives rise to inconsistencies in che-
being a tool to increase the production of relevant metabolites. Mi- mical composition. Moreover, the attempts to make a semiquantifica-
croalgae can thus be considered one of the most promising sources for tion based on variables deduced from the assignment of signals with
new products and applications, but their commercial exploitation is contribution from several compounds to a specific metabolite impair
limited mainly due the time and cost required to develop new products, reliability to the conclusions obtained.
as well as issues of market demand and consumer acceptation. In the present study, biomass from stationary phase batch cultures
Marine organisms are the subject of widespread interest in the field of five marine microalgae species, Dunaliella sp. and Dunaliella salina
of drug discovery, being widely considered as a promising source of (Chlorophyta), Chaetoceros calcitrans and Chaetoceros gracilis
compounds with unique chemical structures and biological activities in (Bacillariophyta) and Tisochrysis lutea (Haptophyta), was sequentially
comparison with terrestrial metabolites (Blunt et al., 2015; Jaspars extracted using hexane, ethyl acetate and methanol as solvents. The
et al., 2016; Romano et al., 2017). To date, marine natural products microalgae studied here constitute a selection of species with existing
have mainly been applied in anticancer chemotherapy. In contrast, the industrial interest. Dunaliella is cultivated for the production of carotene
investigation of antibacterial activity of marine extracts is less devel- and glycerol, with β-carotene obtained from D. salina being the first
oped (Mayer et al., 2013; Shannon and Abu-Ghannam, 2016). However, valuable product commercially produced from a microalgae.
there is a pressing need to develop new antibiotics due to the dramatic Chaetoceros and Tisochrysis lutea are used extensively as food source in
increase of multi-resistant bacterial strains as a result of the overuse of aquaculture. Lipid composition and, in particular, the content in the
existing antibiotics. Furthermore, the formation of bacterial biofilms on polyunsaturated fatty acids (PUFAs) arachidonic acid (ARA, C20:4 ω-
medical devices and implants constitutes a major problem being asso- 6), docosahexaenoic acid (DHA, C22:6 ω-3) and eicosapentaenoic acid
ciated, in many cases, with persistent infections. Lipids are well-known (EPA, C20:5 ω-3), are important factors for the nutritional evaluation of
for antibacterial activity (Thormar, 2011) and the high lipid content an algae species used as food for cultivated marine organisms (Volkman
and unique lipid composition of microalgae make them strong candi- et al., 1989; Wikfors and Ohno, 2001; Da Costa et al., 2016). The
dates as a source for compounds with antibacterial and antibiofilm chemical composition of extracts from these microalgal species was
activity. In contrast to higher plants, many algae, in particular marine analyzed by means of 1H and 13C NMR spectroscopy. Metabolite
species, are able to synthetise very long chain polyunsaturated fatty identification was carried out through the acquisition of 2D NMR ex-
acids (PUFAs). The antibacterial properties of microalgal fatty acids periments and by comparison with information in the literature. This
against a range of Gram-positive and Gram-negative bacteria that in- analysis provides wide valuable information concerning the chemical
clude the multidrug-resistant Staphyloccoccus aureus have been reported constituents of the selected strains. A total of 34 metabolites of diverse
(Ohta et al., 1995; Zheng et al., 2005; Desbois and Mearns-Spragg, nature were identified that allowed differentiation at the genus and
2009; Ruffell et al., 2016). On the other hand, the production of anti- even species level. The antibiotic activity and biofilm inhibition ability
quorom sensing compounds for the cyanobacteria Anabaena sp. and of the extracts were also tested.
Lyngbya majuscula has been reported (Romero et al., 2008; Dobretsov
et al., 2010). Moreover, an efficient inhibition on biofilm formation by 2. Results and discussion
clinical significant bacteria has been found for extracts of Spirulina
platensis (LewisOscar et al., 2017) and for two species of the Leptocy- 2.1. Compositional study
lindrus genus (Lauritano et al., 2016).
Solvent extraction is the first step in metabolite isolation (Sarker The distribution of the soluble material between extracts (Fig. 1)
et al., 2006). The metabolome comprises numerous compounds that suggests that relatively polar compounds predominated except for C.
strongly vary in chemical nature, solubility and concentration. Meta- calcitrans for which the relative amount of compounds extracted with
bolite composition is inevitably biased by the extraction procedure and hexane was the highest.
a single solvent is not sufficient for the analysis of all metabolites. Se- NMR analyses showed that extract composition was determined by
quential extraction with solvents of increasing polarity allows a pre- the characteristics of the solvent. For each solvent the information
liminary separation of the metabolites present in the original material, contained in a spectrum covers a wide range of compounds without
which facilitates the characterization of the extracts and downstream need of further separation and/or derivatization. This implies a diffi-
isolation of given compounds. Once soluble material is obtained, effi- culty in the identification of secondary metabolites because their sig-
cient analytical methods are required for comparative and screening nals can be masked by the signals due to the major compounds present
studies. In this context, NMR spectroscopy is a powerful tool in meta- in the complex mixture. Nevertheless, identification of several minor
bolic studies (Fan and Lane, 2016; Deborde et al., 2017). NMR spec- components was possible through of the combination of 1D and 2D
troscopy is a rapid, non-destructive, high-throughput method for NMR experiments. The metabolites identified in the five microalgae
2
Fig. 2. 1H NMR spectrum of the hexane extract from Dunaliella sp. measured in
Fig. 1. Distribution of the soluble material (percentage of to the total soluble CDCl3. A: CH3 in saturated acyl chains (SAFAs) and ω-9, ω-6 unsaturated acyl
material recovered, dry basis). chains (UFAs). B: CH3 in ω-3 PUFAs. C: -(CH2)n-. D: CH2 β to C]O, except to Δ4
acyl chains. E: Allylic CH2, except CH2 in β to C]O of Δ4 acyl chains. F: CH2 α
strains studied are summarized in Table 1 and their chemical structures to C]O, except Δ4 acyl chains. G: CH2 α and β to C]O in Δ4 acyl chains. H:
Bisallylic CH2. I: Olefinic.
and 1H and 13C NMR assignments are given in the Supplementary
Material (Fig. S1 and Table S1, respectively).
4.29 and 4.15 ppm for the sn1 and sn3 protons, respectively, and at δ
2.1.1. Non- and semi-polar metabolites 5.26 ppm for the sn2 proton (Fig. 2). The characteristic signals of acyl
Storage lipids accumulated by microalgae during the stationary chains of fatty acids attached to the glycerol backbone resonated in the
phase were the major components of the hexane extracts. δ 3.00–0.50 ppm region (signals A-I in Fig. 2). Chemical shift of protons
Triacylglycerols (TGs) were the only storage lipids present in the at the α-position to the carbonyl group, signals F and G centered at δ
Dunaliella and Chaetoceros strains (Fig. 2, S2 and S3, Supplementary 2.31 and 2.39 ppm respectively, indicated that the acyl chains were
Material). The δ 5.3–3.0 ppm region of the 1H NMR spectra for these esterified (Satyarthi et al., 2009; Nieva-Echevarría et al., 2014). This
extracts showed the signals due to the glycerol backbone in TGs at δ feature was also confirmed in the 13C NMR and HMBC spectra that
Table 1
Summary of metabolites identified by NMR spectroscopy in the extracts from the five marine microalgae strains analyzed.
Chlorophyta Bacillariophyta Haptophyta
Dunaliella Chaetoceros T. lutea
sp. salina calcitrans gracilis
Storage lipids TG TG TG TG PULCA TG (minor)

Membrane lipids MGDG MGDG MGDG MGDG MGDG
DGDG DGDG DGDG DGDG DGDG
SQDG SQDG SQDG SQDG SQDG
Acyl chains SAFA SAFA SAFAa SAFAa SAFA
OL OL MUFAb MUFAb OL
HDTA HDTA EPA EPA LO
LO LO ARA ARA ALA
ALA ALA C16:3 ϖ-4 C16:3 ϖ-4 SDA DHA
Pigments Pheo a (a’) Pheo a (a’) Pheo a (a’) Pheo a (a’) Pheo a (a’)
Pheo b (b’) Pheo b (b’) Fucoxanthin Fucoxanthin Fucoxanthin
β-Carotene β-Carotene
Osmolytes Glycerol Glycerol CHT CHT CHT
Homarine DHPS DHPS DMSP
Scyllo-inositol Scyllo-inositol Scyllo-inositol
Homarine
Amino acids Ala Ala Ala Ala Ala
Tyr Tyr Tyr Tyr Tyr
Carbohydrates β-Glc β-Glc β-Glc
α-Glc α-Glc α-Glc
chrysolaminarinb chrysolaminarinb (1 → 3, 1 → 6)-β-D-glucanb
Others Lactate Lactate FFA
Squalene
Brassicasterol
Mannitol
TG: triacylglycerols. MGDG: monogalactosyldiacylglycerols. DGDG: digalactosyldiacylglycerols. SQDG: sulphoquinovosyldiacylglycerols. PULCA: polyunsaturated
long-chain alkenones FFA: free fatty acids. Pheo: pheophytin. SAFA: saturated fatty acid. MUFA: monounsaturated fatty acid. OL: oleic acid. LO: linoleic acid. HDTA:
hexadeca-4,7,10,13-tetraenoic acid. ALA: α-linolenic acid. SDA: stearidonic acid. ARA: arachidonic acid. EPA: eicosapentaenoic acid. DHA: docosahexaenoic acid.
CHT: 1,4/2,5-cyclohexanetetrol. DHPS: 2,3-dihydroxypropane-1-sulfonate. DMSP: dimethylsulfoniopropionate. Glc: glucose. Ala: alanine. Tyr: tyrosine.
c. Tentative assignment.
a
Tentatively: myristic (C14:0) and palmitic acids (C16:0).
b
Tentatively: palmitoleic acid (C16:1).
3
Fig. 3. NMR spectra and metabolites identified in extracts from T. lutea. (a) Expansion of the δ 2.50–0.65 ppm region of the 1H NMR spectrum measured in CDCl3, of
the hexane extract. (b) Comparison between the proportions of TGs, PULCAs and brassicasterol obtained from the NMR analysis of the hexane extract and those
reported by Da Costa et al. (2016) using GC. (c) Expansion of the HMBC spectrum measured in CD3OD of the AcOEt extract in the δH 2.5–1.9 ppm and δC 170-
180 ppm regions. Signals 1, 2, 4 and 5: brassicasterol (18, 26 and 27, 28, 19 and 21, 22 and 23). Signals 3, 6, 7, 9, 10 and 11: PULCAs (terminal CH3 overlapped with
terminal CH3 of acyl chains other that ω-3, CH3CH2C]O, CH2 (long-chain) β to C]O, allylic trans, CH3C]O, CH3CH2C]O overlapped with CH2 α to C]O of long-
chain methyl, ethyl ketones and DHA). Signal 8: squalene.
showed the carbonyl resonances of the acyl chains in the range of δ concentration in the mixture could be estimated through the methyl
174-172 ppm (Gunstone and Seth, 1994). Other minor components protons of the CH3CH2CO moiety that resonated at δ 1.05 ppm isolated
present in these extracts were sterols identified through the signals that from other terminal CH3 groups (peak 6 in Fig. 3a). The trans-geometry
appear in the δ 0.64–0.52 ppm range assigned to the methyl group 18 of the carbon-carbon double bonds in PULCAs was determined through
(Fig. 2). Chlorophylls and carotenoids give rise to characteristic signals the chemical shifts of the allylic protons that resonated at
in the high δ region (δ 11.5–5.5 ppm) of 1H NMR spectra (Sobolev δH = 1.96 ppm/δC = 32.1 ppm (peak 9 in Fig. 3a). The methylene
et al., 2005). Pigments were clearly identified in the NMR analyses of protons in β-position with respect to the carbonyl group of PULCAs
AcOEt extracts (see below). However, in the hexane extracts signals in resonated at slightly lower chemical shift than those in acyl chains of
this region are of very low intensity (Fig. S2), which highlights the poor TGs (peak 7, Fig. 3a). The HSQC spectrum of the hexane extract showed
ability of this solvent for extracting this type of compounds. the corresponding correlations with carbon atoms at δ 24 and 25 ppm
The 1H NMR spectrum of the hexane extract from T. lutea showed a for the proton signals at δ 1.56 ppm (CH2 in β-position with respect to
clear difference with respect to the chlorophytes and the diatoms in the the carbonyl group in PULCAs) and δ 1.61 ppm (CH2 in β-position with
composition of the storage lipids. Although TGs are present, the most respect to the carbonyl group in acyl chains), respectively. Further-
intense signals in this extract were due to polyunsaturated long-chain more, this spectrum also showed the contribution in the signal at δ
alkenones (PULCAs), in particular methyl alkenones (Fig. 3) (O'Neil 1.61 ppm of methyl groups at δC 16.1 and 17.9 ppm that may arise from
et al., 2014). Biosynthesis of PULCAs is characteristic of microalgae of squalene (peak 8, Fig. 3a). A close inspection of the HSQC and HMBC
the haptophyte order Isochrysidales to which T. lutea belongs (Volkman spectra, together with data reported in the literature (Borchman et al.,
et al., 1980; Marlowe et al., 1984; Patterson et al., 1994; Eltgroth et al., 2013), allowed confirmation of the presence of this compound in the
2005). These compounds are believed to be synthesized in the chlor- hexane extract. Signals labeled as 1, 2, 4 and 5 in Fig. 3a were assigned
oplast and then accumulated in cytosolic lipid bodies for storage to brassicasterol based on their chemical shifts and the correlations
(Patterson et al., 1994). The main components have a length chain in found in the HSQC and HMBC spectra, which also allowed the assign-
the range C37-C39 with two to four double bonds with an unusual trans- ment of some additional signals of this cell membrane sterol (Table S1,
geometry occurring at intervals of 7 carbon atoms (Rechka and Supplementary Material) (Weete et al., 1985; Sun et al., 2014). The
Maxwell, 1987). presence of brassicasterol is a distinctive feature of Tisochrysis strains
Polyunsaturated long-chain methyl alkenones in the strain of T. with respect to the closely related species Isochrysis galbana that con-
lutea studied here were characterized by the singlet at δ 2.13 ppm (peak tains epibrassicasterol (Patterson et al., 1994).
10 in Fig. 3a) that showed a cross peak in the HSQC at δ 30.0 ppm and The relative proportions of PULCAs, TGs and brassicasterol esti-
correlated in the HMBC spectrum with a carbonyl carbon at δ mated from the 1H NMR spectra of the hexane extract were comparable
209.2 ppm. Ethyl alkenones were also detected and their relative to the data reported for T. lutea (strain CCAP 927/14) by Da Costa et al.
4
Fig. 4. Expansion of the 1H NMR spectrum, measured in CD3OD, of the C.

calcitrans AcOEt extract in the δ 4.60–3.35 pm region showing characteristic
signals of SQDGs, MGDGs and TGs.
(2016) using GC (Fig. 3b). These authors also found minor amounts of
free fatty acids (FFA). FFA were clearly identified in the AcOEt extract
through the resonances due to methylene protons in α-position with
respect to the carbonyl group (Fig. 3c). This similitude in chemical
composition reflects the phylogenetic proximity between the strains
(Bendif et al., 2013).
In general, structural membrane lipids, mainly MGDGs, were the
major lipids identified in AcOEt extracts (Figs. S5–S7). The C. calcitrans
strain was the exception. As can be observed in Fig. 4 the 1H NMR
spectrum of the AcOEt extract for this strain showed a preponderance of
TGs over MGDGs, whereas for the other extracts resonances due to
storage lipids were almost absent. This result indicates that accumula-
tion of TGs in C. calcitrans was much higher than in the other strains. As
a result, under the experimental conditions used their complete ex-
traction in the hexane fraction was not possible.
Although MGDGs preponderate among structural membrane lipids,
inspection of the 2D NMR spectra allowed the detection of minor
amounts of DGDGs and SQDGs. The 1H and 13C chemical shifts for
these compounds (Table S1) were similar to published data (Logvinov
et al., 2015).
Pigments were concentrated in AcOEt extracts and two types of
pigments were identified in the microalgae studied: chlorophylls de-
tected as pheophytins due to the pheophytinisation process during ex-
traction (Pollesello et al., 1992) and carotenoids. Chlorophylls (Chl)
and chlorophyll derivatives can easily be identified by 1H NMR spec-
troscopy due to the isolated proton resonances between δ 11.2 and
8.5 ppm (Abraham and Rowan, 1991). Thus, the singlets at δ 9.53, 9.40
and 8.55 ppm that correlated in the HSQC spectrum with carbon re-
sonances at δ 104.6, 97.8 and 93.2 ppm, respectively, were assigned to
pheophytin a (Pheo a). Pheophytin b (Pheo b) was identified by the
resonances at δH 11.16/δC 188.2 ppm and δH 10.05 and 9.77 ppm
(Fig. 5a). Minor amounts of the a’ and b’ isomers were also clearly Fig. 5. Expansions of the 1H NMR spectra measured in CD3OD of AcOEt ex-
identified (Table S1). tracts from (a) Dunaliella sp. (δ 11.50–5.50 ppm region) and (b) C. calcitrans (δ
Among chlorophylls, Chl a is the major and most important com- 7.20–5.80 ppm region).
ponent found in phothosyntetic organisms, whereas Chl b is only found
in green lineage. In agreement with this, the 1H NMR spectrum of the and correlated in the HMBC spectrum with the carbonyl signal at δ
AcOEt extract for the five studied strains showed the presence of Pheo a 197.9 ppm. On the other hand, the singlet at δ 6.05 ppm (H8′, Fig. 5b)
(a’), whereas Pheo b (b’) was only found in the Dunaliella strains. showed a correlation in the HSQC spectrum with the carbon resonance
The ability of Dunaliella to adapt to a broad range of environmental at δ 103.5 ppm, as well as correlations in the HMBC spectrum at δ 202.2
conditions through accumulation of β-carotene is well known, whereas and 117.0 ppm that confirmed the allenic nature of this proton. Addi-
the xanthophyll fucoxanthin is the main carotenoid present in diatoms tional correlations found in the HMBC spectrum corresponding to
and haptophytes (Peng et al., 2011; Kuczynska et al., 2015; Mulders C10′(δ 128.6 ppm) and C19’ (δ 14.2 ppm) supported this assignment.
et al., 2013; Serive et al., 2017). The terpenic chain of carotenoids The absence of acetylation at C3 was indicated by the chemical shifts of
appeared between δ 6.7 and 6.0 ppm. The unambiguous identification the proton and carbon at this position δH = 5.38 ppm/δC = 68.2 ppm.
of β-carotene in the non-polar fraction of Dunaliella sp. and D. salina Other proton signals of this compound appeared overlapped with those
was possible based on the analysis of the 2D NMR spectra in combi- of MGDGs, the major components of the ethyl acetate extract. Not-
nation with reported data (Tsukida et al., 1981). The allenic nature of withstanding, the complete assignment of fucoxanthin was possible
fucoxanthin and the presence in its carbon skeleton of conjugated through the analysis of the HSQC and HMBC spectra (Table S1). Fu-
carbonyl, epoxide and acetyl groups facilitated the NMR identification coxanthin was also the only carotenoid found in the haptophyte T. lutea.
(Fig. 5b) (Englert et al., 1990). The olefinic proton H10 of fucoxanthin Besides structural membrane lipids and pigments, glycerol was
appeared in the 1H NMR spectrum as a highly deshielded doublet (δ clearly detected in the AcOEt extracts of Dunaliella strains (Fig. 6). The
7.13 ppm, J = 11.1 Hz) due to the conjugation with the carbonyl group relative amount of glycerol present in the AcOEt extracts from the two
5
Fig. 6. Expansion of the 1H NMR spectra measured in CD3OD of AcOEt extracts

from Dunaliella sp. (top) and D. salina (bottom) in the δ 4.50–3.40 ppm region.
strains was different. Thus, it was the major component of the D. salina
AcOEt extract, whereas in Dunaliella sp. MGDGs preponderated
(MGDG:glycerol 1:1.1 and 1.7:1 for D. salina and Dunaliella sp. AcOEt
extracts, respectively). This compound is a photosynthetic product of
Dunaliella and is also formed by the degradation of starch, the storage
product in this microalga (Ben-Amotz and Avron, 1973). The lack of a
rigid cell wall is a morphological characteristic of species in this genus
and the formation and degradation of glycerol constitutes the main
mechanism for the regulation of the osmotic pressure within the cells.
2.1.2. Fatty acid distribution Fig. 7. Expansions of the 13C NMR spectrum of the AcOEt extract from
Fatty acid composition from microalgae has been extensively ana- Dunaliella sp. measured in CDCl3: (a) carbonyl and (b) olefinic region.
lyzed by chromatographic techniques (Yao et al., 2015). The content of
fatty acids may vary between species of the same genus and even be- Table 2
tween isolates of a single species. However, the distribution pattern of 13C NMR chemical shifts of linolenate (ALA, C18:3) and hexadeca-4,7,10,13-
fatty acids provides, in general, a clear differentiation between phyla, tetraenate (HDTA, C16:4) in MGDG. Chemical shifts are reported with respect
classes and genera of microalgae. In Chlorophyta, the frequency of ARA to CDCl3 (δ 77.16 ppm).
(C20:4 ω-6), EPA (C20:5 ω-3) and DHA (C22:6 ω-3) fatty acids has been
ALA HDTA
reported to be very low (Lang et al., 2011). The genus Dunaliella is not
an exception. Although the content and proportion of fatty acids de- C1 173.86 (sn1) 172.87 (sn2)
pend on specific growth conditions (nutrient levels, temperature, light C2 34.27 34.23
C3 24.96 Allylic
intensity, salinity) (Abd El-Baky et al., 2004; Roleda et al., 2013; Kim
C4 29.34 Olefinic
et al., 2015), the pattern is common for different species in this genus C5 29.28
with the ω-3 PUFAs hexadeca-4,7,10,13-tetraenoic acid (HDTA, C16:4) C6 29.24
and α-linolelic acid (ALA, C18:3) as the major components (Volkman C7 29.75
et al., 1989; Viso and Marty, 1993; Zhukova and Aizdaicher, 1995). The ω-1 14.42 14.42
Allylic
high proportion of these two fatty acids is considered distinctive for
ω-2 20.70 20.71
marine microalgae of the class Chlorophyceae within the division Other 27.36 (C8) 22.78 (C3)
Chlorophyta (Viso and Marty, 1993; Zhukova and Aizdaicher, 1995). In Bisallylic
contrast, diatoms are an important source of EPA and ARA. Another 25.77 25.77a
25.68
abundant PUFA found in these microalgae is the ɷ-4 fatty acid C16:3,
Olefinic
but the amount of C18 fatty acids was very low (Volkman et al., 1989; ω-3 132.12 132.22
Viso and Marty, 1993; Zhukova and Aizdaicher, 1995; Dunstan et al., ω-4 127.26 127.15
1994; Miller et al., 2014). In C. calcitrans C16:3 ɷ-4 fatty acid seems to Other 130.39; 129.70; 128.82; 128.61; 128.46; 128.39; 128.03; 127.95; 127.91;
be present predominantly in the glycolipid fraction, whereas EPA is 127.72
stored as TG (Miller et al., 2014). a

This signal represents the three bisallylic carbons present in the chain.
NMR analysis of the hexane and ethyl acetate extracts from the
strains studied here were consistent with the reported fatty acid dis-
confirmed by the HMBC spectrum through correlation between the
tributions. Furthermore, from the analysis of the 13C NMR and 2D 1H,
carbonyl resonance and the sn1 protons of the glycerol backbone in
13C HMBC spectra the positional distribution of the acyl chains on the
MGDGs (δ 4.40 and 4.20 ppm). Analogously, the carbonyl resonance at
glycerol moiety can be obtained. Composition of storage lipids can
δ 172.87 ppm was assigned to HDTA esterified to the sn2 position of the
differ from that of membrane lipids (Millar et al., 2000) and the NMR
glycerol in MGDGs based on correlation with protons at δ 5.30 ppm and
approach also allows their discrimination.
2.38 ppm. This result is in agreement with the classification of Duna-
Inspection of the δ 3.0–0.7 ppm region of the 1H NMR spectra of the
liella microalgae as “16:3 plants” in which the galactolipids are mainly
ethyl acetate extracts from Dunaliella sp. and D. salina (Fig. S5) in-
composed by one polyunsaturated acid of 18 carbon atoms and one
dicated the same distribution of acyl chains in MGDGs. Analysis of the
polyunsaturated acid of 16 carbon atoms (Thompson, 1996).
13C NMR spectrum of the Dunaliella sp. AcOEt extract (Fig. 7, Table 2)
The fatty acid composition of TGs in the two strains of Dunaliella
confirmed that the main ω-3 PUFAs produced by this microalgae are
was, however, clearly different (Fig. S2). Polyunsaturated acyl chains
the expected C18:3 and C16:4. The carbonyl signal with the highest
were almost absent in the storage lipids produced by D. salina. For this
chemical shift was assigned to the α-linolenate (ALA) attached to the
strain SAFA and MUFA in approximately 1:1 ratio preponderated. In
sn1 position of glycerol in MGDGs by comparison with published data
contrast, acyl chains attached to glycerol backbone of TGs in Dunaliella
(Sakano et al., 2005). The esterification position of this chain was
6
Fig. 9. Expansions of the 13C NMR spectrum of the hexane extract from C.
calcitrans measured in CDCl3: (a) carbonyl and (b) olefinic regions.
glycerol in TGs (Figs. 9 and S3). The shape of the signal due to the
allylic protons (Fig. S3) and the intensity of the olefinic carbons with
the major signals at ca. δ 130 and 129 ppm corresponding to the C10
and C9 carbon atoms of MUFA (Fig. 9b) showed that monounsaturated
Fig. 8. Expansions of the 13C NMR spectrum of hexane extract from Dunaliella acyl chains preponderated over polyunsaturated acyl chains. Splitting
sp. measured in CDCl3: (a) carbonyl region and (b) the sn2 and sn1,3 regions. of these olefinic carbons into two signals with a separation of 0.02 ppm
for C9 and 0.01 ppm for C10 was clearly observed. These two signals
correspond to the presence of monounsaturated acyl chains attached to
sp. were more complex as deduced from the higher number of signals in
the sn1,3 and sn2 position of glycerol in TGs, δ 129.86 (C9)/130.17
the carbonyl region of the 13C NMR spectrum of the hexane extract
(C10) and 129.84 (C9)/130.16 (C10) ppm, respectively (Gunstone,
(Fig. 8a). The assignment of the signal at δ 172.29 ppm to HDTA is
1990; Diehl et al., 1995) (Fig. 9b). In light of the major fatty acids found
confirmed by the correlation in the HMBC spectrum with the proton
in diatoms, these signals might be due to palmitic (C16:0), myristic
resonance at δ 2.39 ppm (CH2 in α and β positions with respect to C]
(C14:0) and palmitoleic (C16:1, ω-7) acids (Volkman et al., 1989; Viso
O). However, in this case the HMBC spectrum did not provide un-
and Marty, 1993; Zhukova and Aizdaicher, 1995; Dunstan et al., 1994;
ambiguous information about the position (sn1,3 or sn2) of esterifica-
Miller et al., 2014). The preponderance of both saturated and mono-
tion. Fortunately, inspection of the glyceryl region of the 13C NMR
unsaturated acyl chains was clearly established based on the lowest
(Fig. 8b) allowed verification that esterification of HDTA occurs ex-
field signals of the 13C NMR spectrum (Fig. 9a, Table S1).
clusively at the sn2 position. In fact, the deshielding of the signal as-
EPA and ARA can easily be identified by 1H NMR spectroscopy
signed to the sn2 carbon esterified with HDTA (0.23 ppm) was con-
through the resonance of the methylenic protons in β-position with
sistent with the increase in chemical shift between Δ4 and other acyl
respect to the carbonyl group (Fig. S6). These protons are also in β-
chains reported in the literature (Sacchi et al., 1993).
position to a double bond and thus resonates ca. 0.12–0.08 ppm
The other signals for the carbonyl carbons in TGs fell in two groups
downfield from those of other acyl chains (Vidal et al., 2012). The lack
separated by ca. 0.4 ppm corresponding to acyl chains esterified to the
in the 1H and 13C NMR spectra of characteristic signals for Δ4 (δH
sn1,3 and sn2 positions of the glycerol (Gunstone, 1990; Diehl et al.,
2.42–2.36 ppm) and ALA (δC 132.1 and 127.3 ppm) acyl chains allowed
1995) (Fig. 8a). Three signals were clearly observed in the sn2 region
assignation the triplet at δ 0.95 ppm to the terminal methyl group of
that were assigned by comparison with literature data (Sacchi et al.,
EPA. The integration of this signal and that corresponding to the me-
1993; Salinero et al., 2012) to carbonyl carbons of saturated (δ
thylenic protons in β-position with respect to the carbonyl group pro-
173.02 ppm), oleyl (OL, C18:1) and linoleyl (LO, C18:2, δ 172.99 ppm)
vided an estimation of the relative proportion of EPA and ARA. The
and linolenyl, ALA, (δ 172.97 ppm) acyl chains (Fig. 8a). Although the
values obtained indicated that this strain of C. gracilis does not accu-
effect of the Δ9 double bond in LO and ALA on the carbonyl chemical
mulate ARA as TGs. In the case of C. calcitrans and taking into account
shift is the same, some spectral features support their presence: proton
that TGs preponderate in both the hexane and the AcOEt extracts, the
resonance for the bisallylic methylene in LO at δ 2.76 ppm and the
results suggest that this strain stores, as TG, approximately twice as
characteristic signals for the olefinic ω-3 and ω-4 carbons of ALA at δ
much EPA as ARA.
132.08 and 127.25 ppm, respectively. Taking into account the propor-
In the AcOEt extracts ɷ-4 acyl chains were also unambiguously
tion of ω-3 acyl chains in this microalgae, linolenyl ester should be the
identified in both strains of Chaetoceros and thus they should be mainly
main contributor to the δ 172.97 ppm signal. The carboxyl and glycerol
present as MGDGs (Fig. S6). Miller et al. (2014) also found a high
regions of the 13C NMR spectra (Fig. 8a and b, respectively) high-
amount of C16:3 ɷ-4 in a GC analysis of the fatty acid distribution in the
lighted the non-symmetric structure of the triacylglycerols.
glycolipid fraction of C. calcitrans. The high field region of the 1H NMR
In the Chaetoceros strains, the NMR analysis of the hexane extracts
spectrum of the ethyl acetate extract showed the presence of a triplet at
revealed that SAFA and MUFA are the major acyl chains attached to the
7
Fig. 11. Expansion of the carbonyl region of the 13C NMR measured in CDCl3
of the hexane extract from T. lutea.
clearly identified in the 1H NMR spectrum by the characteristic re-

sonance of the allylic protons (Fig. 3a). Furthermore, the carbonyl re-
sonance of this acyl chain esterified to the sn1,3 positions of the gly-
cerol in TGs (δ 173.41 ppm) besides those of SAFA attached to the sn1,3
and sn2 positions (δ 173.44 and 173.03 ppm, respectively) were the
most prominent signals in the δ 174-172 ppm region of the 13C spec-
trum (Fig. 11). The two carbonyl resonances at the lowest chemical
shift separated by 0.40 ppm indicated the presence of DHA attached to
both the sn2 (δ 172.29 ppm) and sn1,3 (δ 172.69) positions of glycerol.
The correlation with the proton signal at δ 2.39 ppm in the HMBC
spectrum confirmed this assignment. Stearidonic acid (C18:4 ω-3) was
also identified through the carbonyl resonances at δ 172.82 (sn2) and
173.22 (sn1,3) ppm (Fig. 11). It has been reported that in T. lutea LO
and ALA contribute ca. 4% and 6%, respectively to total fatty acid
content (Da Costa et al., 2016). Minor signals characteristic of these
acyl chains were also detected in the NMR spectra. On the other hand,
the lack of EPA corroborates again the phylogenetic separation between
T. lutea and I. galbana (Liu and Lin, 2001).
2.1.3. Polar metabolites

Fig. 10. Expansions of 2D NMR spectra of AcOEt extracts from C. calcitrans
Although the most polar lipids (DGDGs and SQDGs) were detected
measured in CDCl3: (a) COSY spectrum showing the correlations of the methyl
in the AcOEt extract, the polarity of this solvent is not sufficient for
and methylene protons of ω-4 acyl chains; (b) HSQC spectrum (δH
2.15–1.15 ppm and δC 36-8 ppm regions).
their complete extraction. As a result, they appeared also in the me-
thanolic extracts. However, in all cases osmolytes were the major
compounds present in these extracts. Thus, the most prominent signals
δ 0.91 ppm that correlated in the HSQC spectrum with a carbon at δ in the 1H NMR spectra of methanolic extracts from Dunaliella corre-
13.95 ppm. The assignment of this signal to the terminal methyl group sponded to glycerol (Fig. S8). Glycerol was not detected in the other
of ω-4 acyl chains was confirmed through the correlations found in the three strains in which 1,4/2,5-cyclohexanetetrol (CHT) was the major
COSY spectrum. Thus, the triplet at δ 0.91 ppm correlated with the polyol (Fig. 12). This compound was characterized by the proton sig-
multiplet centered at δ 1.38 ppm that, in turn, also showed correlation nals at δ 3.73 and 1.83 ppm that showed correlations in the HSQC
with allylic protons at δ 2.10 ppm (Fig. 10a). Furthermore, a second spectrum at δ 71.6 (CH) and 35.3 (CH2) ppm, respectively (Maras et al.,
correlation between the signal at δ 1.38 ppm and the resonance of 1998; Kobayashi et al., 2007; Garza-Sánchez et al., 2009). Scyllo-in-
methylene protons in β-position to the carbonyl group (δ 1.60 ppm) ositol was also identified in the Chaetoceros and T. lutea strains (Fig. 12).
indicated the presence of acyl chains of stearidonic acid (SDA) (Dais The six methinic groups in this stereoisomer of myo-inositol are
et al., 2015). Biosynthesis of SDA in Chaetoceros strains was confirmed equivalent and appeared at δH 3.21 ppm and δC 75.5 ppm. This cyclitol
by the HSQC spectrum (Fig. 10b). Thus, the multiplet at δ 1.38 ppm has been described in the haptophyte Pavlova sp. (Kobayashi et al.,
corresponding to a CH2 group showed cross peaks at about δ 23 and 2007) and in the diatom Phaeodactylum tricornutum (Ford and Percival,
24 ppm, in the typical range of non-allylic ω-2 and C4 carbon re- 1965). Cyclitols are organic osmolytes, i.e. small solutes that maintain
sonances of unsaturated acyl chains, respectively. Signals at δ 0.91 and cell volume and fluid balance and regulate plant responses to multiple
1.38 ppm have been found in NMR analyses of fish oil supplements and stresses (Yancey, 2005). Thus, in chromalveolates (Chaetoceros, Pavlova,
they have been assigned also to ω-4 acyl chains but with a trans geo- Boekelovia, Nitzschia) 1,4/2,5-cyclohexanetetrol acts as an osmor-
metry of the double bond. However, no unequivocal support for this egulator and its proportion increases when the algae are subjected to
assignment was given (Dais et al., 2015). Detection of trans-fatty acid salinity stress (Fujii et al., 1995; Kobayashi et al., 2007; Garza-Sánchez
chains can be achieved through the chemical shifts of allylic carbons. et al., 2009). In contrast, changes in the amount of scyllo-inositol in
The trans-geometry of the double bond increases the carbon chemical response to external salinity has not been observed in Pavlova sp.
shift of the allylic methylenic carbons that resonate in the δ (Kobayashi et al., 2007). Furthermore, mannitol, one of the most
32.7–32.5 ppm range (Sacchi et al., 1997). There was no correlation in abundant natural polyols, was also identified in the methanolic extract
the HSQC spectrum between the allylic protons and carbon resonances from T. lutea (Fig. 12). This compound can play various important
in the expected range for methylene groups adjacent to a trans double physiological roles such as carbon storage and protection against en-
bond (Fig. 10b). vironmental stress. In alkenone producing strains, both alkenones and
Despite the overlap between the proton resonances of the acyl mannitol seems to be the major carbon/energy storage compounds
chains in TGs and PULCAs, acyl chains corresponding to the major fatty (Obata et al., 2013; Tsuji et al., 2015).
acids reported in T. lutea could be determined by NMR. Oleic acid was Dimethylsulfoniopropionate (DMSP) (Obando et al., 2016) was,
8
Fig. 12. 1H NMR spectra of the MeOH extracts measured in CD3OD from the diatom C. calcitrans (top) and the haptophyte T. lutea (bottom).
together with CHT, the most abundant metabolite found in the T. lutea pseudonana and the marine bacteria Ruegeria pomeroyi, these authors
methanolic extract (Fig. 12). The function of DMSP as an osmoprotector concluded that DHSP released by the diatom is subsequently used as a
in microalgae is well known and its production seems to be taxon-de- major substrate for the bacteria underlying the important role of this
pendent, with Dinophyceae and Prymnessiophyceae being major sources sulfonate as a reservoir of carbon and sulfur in the ocean.
of this compound (Keller et al., 1999). DMSP can also act as an anti- Minor amounts of the quaternary ammonium compound homarine,
oxidant and predator repellent (Yancey, 2005). Furthermore, DMSP, unambiguously identified by 1H NMR through the aromatic protons
produced by phytoplankton in response to different environmental resonating in the low-field region of the spectrum (Affeld et al., 2006;
stresses, is the primary precursor of dimethylsulfide (DMS) and thereby Gebser and Pohnert, 2013) (Figs. 12 and S8), were found in the me-
has important roles in the biogeochemistry and ecology in marine mi- thanolic extracts from T. lutea and Dunaliella sp. but this metabolite was
crobial communities (Kiene et al., 2000). not detected in either D. salina nor Chaetoceros strains.
DMSP was not detected in the studied strains of Chaetoceros. On the other hand, for the five studied strains, the 1H NMR spectra
However, the presence of a methylene attached to a sulfonyl group in of the methanolic extracts also showed relatively weak signals of the
the 1H NMR spectrum of their methanolic extracts was deduced from amino acids alanine (Ala) and tyrosine (Tyr). Carbohydrates were also
the double doublets at δ 3.05 (J = 14.1 Hz, 4.2 Hz) and 2.90 detected except in the Dunaliella strains, in which amounts of lactate
(J = 14.1 Hz, 7.9 Hz) ppm (Fig. 12) that showed a correlation in the were in turn found (Fig. S8). The occurrence of carbohydrates was
HSQC spectrum with the carbon resonance at δ 55.1 ppm. This signal shown by the presence of a set of doublets due to anomeric protons in β-
correlated in the HSQC-TOCSY spectrum with another methylene group (δ 4.67–4.37 ppm, J ≈ 8 Hz) and α-hexapyranoses (δ 5.16–5.10 ppm,
at δH 3.63 and 3.58 ppm/δC 66.2 ppm and a methine group at δH J ≈ 3.7 Hz) (Fig. 12). The two epimers of D-glucose were un-
4.12 ppm/δC 69.7 ppm. These NMR data are consistent with those re- ambiguously identified through the doublets at δ 4.48 ppm (J = 7.8 Hz)
ported for 2,3-dihydroxypropane-1-sulfonate (DHSP) identified as one and 5.11 ppm (J = 3.7 Hz) corresponding to the anomeric protons of β-
of the metabolites formed during the bacterial degradation of sulfo- and α-glucose, respectively, by comparison with a sample of standard
quinovose, the headgroup of SQDG (Roy et al., 2003). DHSP has been glucose measured in methanol-d4.
also identified in cells of the fresh-water diatom Navicula pelliculosa Glucose is the major sugar reported in the polysaccharide fraction of
(Busby, 1966). More recently, Durham et al. (2015) reported the oc- C. calcitrans (Brown, M.R., 1991). This sugar arises from the β-(1 → 3)-
currence of significant intracellular concentrations of this compound in glucan named chrysolaminarin that is the main storage product of
diatoms. From the study of co-cultures of the diatom Thalassiosira photosynthetic CO2 fixation in diatoms. Chrysolaminarin is a relatively
9
short-chained glucose polymer very soluble in water/methanol that is salina extracts clearly differs from the reported activity using pressur-
included in the low molecular weight fraction (Granum and Myklestad, ized liquid extraction (Herrero et al., 2006) or sub- and supercritical
2001). The HSQC spectrum showed correlations between the proton fluid extraction (Mendiola et al., 2008). This microbial activity was
signals in the β-anomeric region with carbons in the range corre- mainly attributed to compounds related to carotenoid degradation (α-
sponding to non-reducing sugars (δ 104.5 ppm). The glycosidic linkage and β-ionone, β-cyclocitral) as well as derived from galactolipid and
between positions 1 and 3 was deduced from the correlation found in chlorophyll catabolism (free fatty acids and phytol) due to the extrac-
the HMBC spectrum between the anomeric protons with carbon signals tion method and conditions used. As discussed above, the precursors of
at δ 83.3 and 87.0 ppm. These data are in agreement with the presence such compounds (β-carotene and galactolipids) were found in extracts
of β-(1 → 3)-glucans in the methanolic extract and may correspond to from D. salina but compounds arising from their degradation were not
different glucopyranosyl units of these polysaccharides (Størseth et al., detected. These results suggest, therefore, that D. salina, in the culture
2004, 2006). On the other hand, the presence of non-reducing sugars conditions used in the present work, is not able to biosynthesize anti-
with α configuration other than α-glucose was revealed by the corre- bacterial compounds itself but the appropriate treatment of the extracts
lations, in the HSQC spectrum, between the doublets at δ 5.14 could transform the original secondary metabolites into potential an-
(J = 3.6 Hz) and 5.12 (J = 3.8 Hz) and carbons at δ 93.4 and 93.6 ppm, timicrobial substances.
respectively. Furthermore, the signal at δ 5.14 ppm showed a correla- The potential of the studied extracts to inhibit biofilm was also
tion in the HMBC with a carbon resonance at δ 84.2 ppm suggesting a tested against nine different microorganisms causing clinical infections.
glycosidic linkage in a position different to the anomeric one. However, They include species of the three major groups involved in human in-
the complexity of the mixture precludes an unambiguous assignment fections, Gram-positive (S. aureus, Coagulase-negative Streptococcus
for this signal. (CNS) and S. epidermidis), Gram-negative (E. coli, Klebsiella pneumonia,
In the methanolic extract of T. lutea the anomeric protons of β- E. cloacae and Pseudomonas aeruginosa) and fungi (C. albicans and C.
hexapyranoses other than β-glucose appeared between δ 4.41 and parapsilosis). The strains were selected on the basis of their resistance
4.37 ppm as a set of overlapping doublets with coupling constants of ca. profile towards clinical treatments (Table 3) and ability to form bio-
8 Hz (Fig. 12, Table S1). These signals showed correlations, in the films.
HSQC spectrum, with carbon resonances at δ 103.7 ppm and in the Only the ethyl acetate extract from C. gracilis presented moderate
COSY spectrum with protons at δ 3.25 ppm. In turn, the HMBC spec- antibiofilm activity against Candida species showing a > 4 fold-de-
trum showed two correlations between these anomeric protons and crease in OD in comparison with the value obtained in the positive
carbon signals at δ 75.9 and 69.4 ppm. These results indicate the pre- control, and presenting MBIC values of 16 mg/L and 32 mg/L for C.
sence of units of β-glucose with a glycosidic linkage between the po- albicans and C. parapsilosis, respectively (Table 4). The difference be-
sitions 1 and 6. Although the nature of carbohydrates in T. lutea is not tween these values and those obtained for the same extract from C.
well known (Garnier et al., 2016), the results suggest the presence of calcitrans should be underlined. Based on the NMR data for these two
highly branched (1 → 3, 1 → 6)-β-D-glucan probably similar to that re- extracts, the composition of pigments and MGDGs is expected to be the
ported for Isochrysis galbana with potential antitumor activity same for both strains. However, in C. calcitrans TGs were the major
(Sadovskaya et al., 2014). components of the AcOEt extract. These compounds, according to the
MBIC values of the hexane extract, do not show antibiofilm activity.
2.2. Bioactivity studies Recently, Lauritano et al. (2016) reported antibiofilm activity against S.
epidermis in two marine Bacillariophyta microalgae of the Chaetocer-
The antimicrobial activity of crude extracts from microalgae has otales order (to which Chaetoceros belongs). The activity was shown to
usually been attributed to the presence of compounds with well-known be dependent on culture conditions but no compositional data of the
activity (Herrero et al., 2006; Mendiola et al., 2007, 2008). The com- extracts were given. Although the biofilm inhibition ability found in C.
positional study of the five strains investigated here (Table 1) showed gracilis in our study was low, it must be stressed that the bioassays were
the presence in these microalgae of several compounds with potential carried out on crude extracts, i.e, a very complex mixture of metabo-
antibacterial and, possibly also, antibiofilm activities. For example, acyl lites. Diatoms from the order Chaetocerotales would appear to be good
chains of unsaturated fatty acids (palmitoleic, OL, LO, ALA, DHA, EPA) candidates for further investigation of potential antibiofilm activity.
that have been reported to have antibiotic effect (Ohta et al., 1995;
Zheng et al., 2005; Desbois and Mearns-Spragg, 2009; Ruffell et al.,
3. Conclusions
2016) were found. In invertebrates, homarine acts as an antipredatory
and antifouling agent. It has been reported (Slattery et al., 1997) that
NMR analyses of sequential hexane, ethyl acetate and methanol
the soft coral Gersemia antarctica releases a mixture of organic com-
pounds that confer a high antibacterial activity to the seawater around
Table 3
it, homarine being responsible for most of such bioactivity. On the other
Selected strains for antibiofilm assays and their resistance profile.
hand, the carotenoid fucoxanthin has been found associated with the
surface of the macroalgae Fucus vesiculosus acting as an inhibitor of the Strain Resistance Profile
settlement of bacteria on its surface (Saha et al., 2011).
Gram-positive
In the present study, however, all extracts were inactive in the an- S. aureus RPG2A VA, LEV, CIP DO, TOB, RDS
tibiotic bioassays for the three microorganisms tested (Staphylococcus S. epidermidis FG013 AK, VA, LEV, CIP TOB, RDS
aureus, Escherichia coli and Candida parapsilosis), as no growth inhibition CNS S13 AK,VA, LEV, CIP DO, TOB, RDS
were generated when culturing these pathogens in the presence of di- Gram-negative
E. coli 15101 LEV, CIP, TOB
verse increasing concentrations of the corresponding extracts (see sec- K. pneumoniae AT AK,VA, LEV, CIP, TOB
tion 4.3). Two human and animal pathogenic bacteria and one fungus E. cloacae 6163 AK,VA, LEV, CIP, TOB
with different resistance profiles were used: clinical isolates of E. coli AR P. aeruginosa AB VA, LEV, CIP,RD
(resistant to ampicillin, erythromycin, kanamycin and tetracycline), S. Fungi
C. albicans 162288680 FCA, VOR, CAS
aureus S54F9 (resistant to erythromycin and kanamycin) and C. para-
C. parapsilosis SMI1416 AmB, FCA, VOR
psilosis SMI416 (resistant to fluconazole). The presence of these re-
sistance profiles may explain the difficulties in finding novel anti- AK, amikacyn; VA, vancomycin, LEV, Levofloxacin; CIP, ciprofloxacin; DO,
microbial compounds against these three pathogens. The lack of doxycycline; TOB, tobramycin; RD, rifampicin; AmB, amphotericine; FCA, flu-
antimicrobial activity against E. coli and S. aureus found here for the D. conazole; VOR, voriconazole; CAS, caspofungin.
10
Table 4 The diatom (Chaetoceros calcitrans and C. gracilis) and haptophyte

Minimal biofilm inhibitory concentrations (MBICs) in mg/L obtained. (Tisochrysis lutea) strains showed some compositional similarities. They
Dunaliella Chaetoceros Tisochrysis lutea all produced the pigments chlorophyll a and fucoxanthin and, in both
cases, 1,4/2,5-cyclohexanetetrol (CHT) was the major polyol bio-
sp. salina calcitrans gracilis synthesized. Unlike chlorophyte strains (Dunaliella spp.), signals of
carbohydrates were detected. For example, the two epimers of D-glu-
Hexane C. albicans 512 32 256 256 64
C. parapsilopsis > 1024 128 > 1024 256 1024 cose were unambiguously identified. The results indicated, however,
E. coli > 1024 1024 > 1024 1024 1024 that glucose arose from structurally different glucans, probably chry-
CNS > 1024 256 > 1024 256 512 solaminarin in the diatoms and a highly branched (1 → 3, 1 → 6)-β-D-
E. cloacae 512 256 > 1024 512 512 glucan in the haptophyte. Homarine and mannitol were only detected
K. pneumoniae > 1024 1024 > 1024 1024 1024
in T. lutea, but the most significant difference between T. lutea and
S. aureus > 1024 512 > 1024 512 512
S. epidermidis > 1024 1024 > 1024 1024 1024 Chaetoceros was without doubt the presence in the latter of DHSP. This
P. aeruginosa > 1024 1024 > 1024 1024 1024 compound seems to be specifically produced by diatoms, together with
AcOEt C. albicans 128 256 > 2048 16 64 CHT, constituting the major component of the Chaetoceros methanolic
C. parapsilopsis > 4096 64 > 2048 32 256
extracts. In T. lutea methanolic extract CHT and DMSP predominated.
E. coli > 4096 512 > 2048 512 1024
CNS > 4096 256 > 2048 128 1024 None of the analyzed extracts showed antibiotic activity. Promising
E. cloacae > 4096 1024 > 2048 256 512 moderate antibiofilm ability was found for the ethyl acetate extract
K. pneumoniae > 4096 512 > 2048 256 512 from C. gracilis.
S. aureus > 4096 512 > 2048 512 1024
S. epidermidis > 4096 1024 > 2048 512 1024
4. Experimental
P. aeruginosa > 4096 1024 > 2048 1024 1024
MeOH C. albicans 256 512 64 256 32
C. parapsilopsis > 1024 64 > 1024 256 1024 4.1. Culture and extraction
E. coli > 1024 512 > 1024 1024 512
CNS > 1024 256 > 1024 256 1024
The five strains used in this study were RCC5 (Dunaliella sp.),
E. cloacae 512 1024 64 1024 512
K. pneumoniae > 1024 256 > 1024 512 512
RCC3579 (D. salina), RCC1811 (Chaetoceros calcitrans), RCC 5953 (C.
S. aureus > 1024 512 > 1024 512 256 gracilis), and RCC1349 (Tisochrysis lutea) from the Roscoff Culture
S. epidermidis > 1024 1024 > 1024 1024 1024 Collection, France (www.roscoff-culture-collection.org). Studied cul-
P. aeruginosa > 1024 1024 64 1024 1024 tures were unialgal, non axenic transferred under laminar flow hood.
Cultures were transferred gradually from 20 mL ventiled flask
(Starlab, CC7682-4825) up to 250 mL, 1L, 4L round-bottomed glass
extracts from stationary phase batch cultures of five marine microalgae
flasks. Growth media was prepare with Roscoff seawater (pH = 8.2,
strains (Dunaliella sp., Dunaliella salina, Chaetoceros calcitrans,
salinity 33‰), filtered on 0.22 μm filter (Millipore, GSWP09000) then
Chaetoceros gracilis and Tisochrysis lutea) provided wide-ranging and
autoclaved. Autoclaved natural seawater was added, under hood, with
valuable information concerning the chemical composition of these
f/2 culture medium supplements (Guillard, 1975). Cultures were
strains. As expected, extract composition was determined by the char-
maintained at 20 °C with a light intensity of ca. 150 μE m−2s−1 pro-
acteristics of the solvents used. In general, storage lipids were selec-
vided by daylight fluorescent tubes (Philips Master TL5 H0 49W/865),
tively extracted with hexane, whereas membrane lipids predominated
with a photoperiod of 14 h light:10 h dark. Cultures were subjected to
in AcOEt (MGDGs) and methanol (DGDGs and SQDGs) extracts.
constant bubbling with sterile-filtered air. Cultures growth rate was
Pigments were concentrated in the AcOEt extracts and osmolytes were
monitored every day by flow cytometry (FACSCanto, Becton-Dick-
the major components of methanolic extracts. Minor components such
inson). After 7–14 days of growth, depending on each culture, cultures
as sterols and squalene, or carbohydrates, amino acids and lactic acid
reach stationary phase. In mid-stationary phase (2 days), cultures were
were also detected in hexane and methanol extracts, respectively.
harvested by centrifugation (7500 rpm for 30 min) and the cell pellets
TGs were the only storage lipids produced by Dunaliella and
were flash-frozen in liquid nitrogen, freeze-dried, then maintained at
Chaetoceros strains whereas T. lutea stored mainly PULCAs. The detec-
−80 °C until extraction.
tion of brassicasterol as the only sterol in T. lutea and the lack of acyl
Prior to extraction, freeze-dried cell pellets were first disrupted by
chains of EPA reflected the genetic differentiation between this species
grinding with a clean ceramic mortar to produce a fine powder.
and Isochrysis galbana.
Disrupted cells were first extracted with 6 mL of hexane (3 × 2 mL in
In accordance with the classification of Dunaliella as a “16:3 plant”,
glass tubes) under ultrasound (240 W, 50 kHz) for 10 min, then cen-
the major MGDGs identified in Dunaliella sp. and D. salina were com-
trifuged at 4600 rpm for 15 min, and the supernatant recovered in a
posed by one acyl chain of ALA (C18:3) and another of DHTA (C16:4)
pre-weighed 8 mL glass vial with a Teflon cap (the 3 different super-
attached to the sn1 and sn2 of the glycerol backbone, respectively.
natants of about 2 mL each were pooled in this vial). The solvent was
However, the two strains exhibited clear differences in the fatty acid
then gently evaporated in a rotavapor and the vial weighed. The re-
composition of TGs. In D. salina there was practically no incorporation
sidual cell pellets in the 3 tubes were kept for subsequent extraction
of PUFAs in TGs. β-Carotene, glycerol, derivatives of chlorophyll a and
steps. This procedure was repeated twice: first with 6 mL of AcOEt, then
b, amino acids and lactic acid were found in both strains. Glycerol was
with 6 mL of methanol as the solvent. Prior to downstream analyses,
the major component of the methanolic extract and also of the AcOEt
each of the extracts was resuspended in the relevant solvent.
extract in D. salina. This feature indicated a much higher production of
glycerol in this strain than in Dunaliella sp. On the other hand, homarine
4.2. NMR analysis
that was unambiguously detected in the methanolic extract of T. lutea
was only detected in Dunaliella sp. of the two Dunaliella strains.
All 1H (600 MHz) and 13C (150.9 MHz) NMR spectra were recorded
C. calcitrans was characterized by high TG production. The two
on a Bruker Avance III HD 600 MHz NMR (14.0 T) spectrometer
Chaetoceros strains showed the same fatty acid composition of TGs with
equipped with a QCI-P CryoProbe™ (proton-optimized quadruple re-
a clear preponderance of SAFA and MUFA. EPA and ARA were detected
sonance NMR ‘inverse’ probe). Hexane extracts were dissolved in
in both TGs and MGDGs, whereas ω-4 acyl chains were only present in
0.5 mL of CDCl3 containing the internal standard tetramethylsilane
the glycolipid fraction. Furthermore, Chaetoceros strains stored, as TGs,
(TMS). For the ethyl acetate and methanol extracts, the samples were
more EPA than ARA.
prepared by dissolving the extracts in 0.5 mL of MeOD-d4. 1H NMR
11
spectra of the CDCl3 samples were recorded using the following para- dilutions ranges were between 8.65 and 0.54 mg/mL (for hexane), 3.05
meters: relaxation delay 1 s; 90° pulse width (P1) of 6.64 μs; number of to 0.19 mg/mL (for ethyl acetate) and 17.75 to 1.11 mg/mL (for me-
scans 128; chemical shift range 0–15 ppm. 1H NMR spectra of the thanol). 50 μL of the corresponding microorganism suspension
MeOD-d4 samples were recorded with solvent suppression using a g- (5 × 105 cfu/mL) in Mueller-Hinton 2x, were added to all wells in the
noesy1d pulse sequence, with presaturation during relaxation delay and microtitre row. Negative controls (wells without microorganism) and
mixing time. The parameters were: relaxation delay 5 s; number of positive controls (wells without extract) were also included.
scans (NS) = 128; chemical shift range 0–15 ppm. 13C NMR spectra Microdilution plates for antibiotic assays were incubated overnight
were recorded with ∼12,000 scans using a relaxation delay time of 2 s, at 37 °C and then all wells were replicated to fresh solid medium: MSA
90° pulse width of 11.7 μs, and sweep width of 0–200 ppm. Parameters (mannitol salt agar) for S. aureus, EMB (eosin methylene blue) for E.
for the 2D NMR spectra were as follows. For gCOSY: 2 s relaxation coli, and Sabouraud agar for C. parapsilosis. Bacteriostatic/fungistatic
delay, 256 increments, 64 transients and sweep width of 10 ppm in both activities (growth inhibition in microdilution but normal growth in
dimensions. A sine bell weighting function (SSB = 0) was applied prior replicated plate) and bactericidal/fungicidal activities (absence of
to Fourier transformation in both dimensions (final matrix of growth in both microdilution and replicated plates) are detected with
1024 × 1024 data points). For gHSQC–TOCSY: phase sensitive via this method.
echo-antiecho mode, 2 s relaxation delay, 256 increments, 64 tran- Biofilm inhibition of these extracts was assayed using E. coli,
sients, a mixing time of 80 ms, and a sweep width of 12 and 200 ppm Klebsiella pneumoniae, Enterobacter cloacae, Pseudomonas aeruginosa, S.
for the 1H and 13C dimensions, respectively. A squared cosine aureus, Coagulase-negative Streptococcus, S. epidermidis, C. albicans and C.
weighting function (SSB = 2) was applied prior to Fourier transfor- parapsilosis. The bacteria were cultured overnight at 37 °C (30 °C in the
mation in both dimensions (1024 × 1024 data points). For multiplicity case of E. coli) in TSB culture medium (Gram-positive bacteria); M63 (E.
edited gHSQC: phase sensitive mode using the echo-antiecho technique, coli); LB (the remaining Gram-negative bacteria); or Yeast Nitrogen
1.5 s relaxation delay, 256 increments, 64 transients and a sweep width Base (Candida sp.). 50 μL of the relevant culture medium were added to
of 12 and 200 ppm for the 1H and 13C dimensions, respectively. each well of the microtitre plate. 50 μL of the microalgae extract in a
Spectra were processed using a squared cosine weighting function known concentration was added to the first well (at four times the
(SSB = 2) prior to Fourier transformation in both dimensions concentration to be evaluated, e.g. to evaluate the minimal inhibitory
(1024 × 1024 data points). For gHMBC: magnitude mode, 2 s relaxa- concentration, MIC, in a range of 4-0.125 mg/mL, an initial con-
tion delay, 128 increments, 128 transients and a sweep width of 12 and centration of 16 mg/mL must be prepared). After mixing, 50 μL were
200 ppm for the 1H and 13C dimensions, respectively. Spectra were transferred into the next well and so on, resulting in serial dilutions.
processed using a sine bell weighting function (SSB = 0) prior to For each strain grown overnight, a 0.5 McFarland inoculum was
Fourier transformation in both dimensions (2048 × 1024 data points). prepared (using a nephelometer) and diluted in Falcon tubes containing
For z-COSY: phase sensitive using the States-TPPI technique, 2 s re- 10 mL of the appropriate culture medium (1 tube per strain). 50 μL of
laxation delay, 256 increments, 64 transients and a sweep width of bacterial inoculum were added to all of the microtitre wells, except the
10 ppm in both dimensions. Spectra were processed using a squared well used as a negative control. In each experiment, a positive control
cosine weighting function (SSB = 2) prior to Fourier transformation in (culture medium plus bacteria) and a negative control (only culture
both dimensions (1024 × 1024 data points). Standard Bruker software medium) were used. Ten dilutions of each extract were assayed. The
(TopSpin 3.6) was used for the acquisition and processing of NMR concentrations ranged from 4096 mg/L to 8 mg/L or from 1024 mg/L to
spectra. 2 mg/L depending of the initial amount of the extract.
Integration of selected resonances was performed using the decon- The plate was covered with an adhesive foil and Parafilm and in-
volution algorithm present in MNova 10.1 software after removal of cubated at 37 °C (30 °C for E. coli) for 48 h. Following the incubation
residual solvent signal and normalization to the total spectral area. period, the supernatant was discarded and the wells gently washed 1 to
Deconvoluted signals for specific metabolites were then divided by the 2 times with 1% PBS. The biofilm was fixed for approximately 20 min at
number of hydrogen in order to obtain its relative molar content (mol 60 °C to evaporate all of the PBS in each well. 200 μL of 2% violet
%). crystal (VC) were added to each well and incubated for 10 min at room
temperature. The VC was discarded and wells were washed 1 to 2 times
4.3. Bioactivity assays with 1% PBS. The stained biofilm was then completely dried for
40 min at 60 °C. Finally, 200 μL of 33% acetic acid were added to each
Antibiotic assays were performed against a Gram-positive bacterium well to resuspend the biofilm. The absorbance was read at 580 λ in a
(Staphylococcus aureus), a Gram-negative bacterium (Escherichia coli) spectrophotometer always discounting the value obtained in the ne-
and a fungus (Candida parapsilosis). The three pathogens were in- gative control.
oculated in 5 mL Mueller-Hinton medium from glycerol stocks, and Minimal biofilm inhibitory concentration (MBIC) was defined as the
incubated overnight at 37 °C with agitation at 250 rpm. After OD600 lowest extract concentration that inhibited biofilm formation.
spectrophometric assessment, 5 × 105 cfu/mL of each pathogen were
used for each antibiotic assay. Acknowledgements
A mixture of 5:95 DMSO:H2O was added to each dry microalgal
extract, and the vial was sonicated until total resuspension. Serial two- Funding from the European Union's Horizon 2020 research and
fold dilutions of the resuspended extract were carried out in a micro- innovation program under grant agreement no. 634588 (NOMORFILM)
titre plate (one row per extract) with 50 μL of sterile Milli-Q water pre- is gratefully acknowledged.
added to all wells. The RCC5 tested dilutions ranges were between 14.4
and 0.9 mg/mL (for hexane), 9.45 to 0.59 mg/mL for ethyl acetate and Appendix A. Supplementary data
8.85 to 0.55 mg/mL for methanol. The RCC1811 tested dilutions ranges
were between 12.85 and 0.8 mg/mL (for hexane), 2.7 to 0.17 mg/mL Supplementary data to this article can be found online at https://
(for ethyl acetate) and 4.55 to 0.28 mg/mL (for methanol). The doi.org/10.1016/j.phytochem.2019.05.001.
RCC3579 tested dilutions ranges were between 26.5 and 1.65 mg/mL
(for hexane), 6.55 to 0.4 mg/mL (for ethyl acetate) and 19.98 to References
1.24 mg/mL (for methanol). The RCC5953 tested dilutions ranges were
between 33 and 0.2 mg/mL (for hexane), 8.85 to 0.55 mg/mL (for ethyl Abd El-Baky, H.H., El Baz, F.K., El-Baroty, G.S., 2004. Production of lipids rich in omega 3
acetate) and 17.02 to 1.06 mg/mL (for methanol). The RCC1349 tested fatty acids from the halotolerant alga Dunaliella salina. Biotech 3, 102–108.
12
Abraham, R.J., Rowan, A.E., 1991. Nuclear magnetic resonance spectroscopy of Skeletonema costatum (Bacillariophyceae). J. Phycol. 37, 772–782.
Chlorophyll. In: Scheer, Hugo (Ed.), Chlorophylls. CRC Press, Boca Raton u.a, pp. Guillard, R.R.L., 1975. Culture of phytoplankton for feeding marine invertebrates. In:
797–834. Smith, W.L., Chanley, M.H. (Eds.), Culture of Marine Invertebrate Animals. Springer,
Affeld, S., Wägele, H., Avila, C., Kehraus, S., Köning, G.M., 2006. Distribution of homarine Boston, MA, pp. 29–60.
in some opisthobranchia (gastropoda: Mollusca) bonn. Zool. Beitr. 55, 181–190. Guiry, M.D., 2012. How many species of algae are there? J. Phycol. 48, 1057–1063.
Akhter, M., Majumdar, R.D., Firtier-McGill, B., Soong, R., Liagati-Mobarhan, Y., Simpson, Gunstone, F.D., Seth, S., 1994. A study of the distribution of eicosapentaenoic acid and
M., Arhomditsis, G., Scmidt, S., Heumann, H., Simpson, A.J., 2016. Identification of docosahexaenoic acid between the α and β glycerol chains in fish oils by 13C-NMR
aquatically available carbon from algae through solution-state NMR of whole 13C- spectroscopy. Chem. Phys. Lipids 72, 119–126.
labelled cells. Anal. Bional. Chem. 408, 4357–4370. Gunstone, F.D., 1990. 13C-NMR spectra of some synthetic glycerol esters alone and as
Azizan, A., Ahamad Bustamam, M.S., Maulidiani, M., Shaari, K., Ismail, I.S., Nagao, N., mixtures. Chem. Phys. Lipids 56, 195–199.
Abas, F., 2018. Metabolite profiling of the microalgal diatom Chaetoceros calcitrans Gupta, V., Thakur, R.S., Reddy, C.R.K., Jha, B., 2013. Central metabolic processes of
and correlation with antioxidant and nitric oxide inhibitory activities via 1H NMR- marine macrophytic algae revealed from NMR based metabolome analysis. RSC Adv.
based metabolomics. Mar. Drugs 16, 154. https://doi.org/10.3390/md16050154. 3, 7037–7047.
Beal, C.M., Webber, M.E., Ruoff, R.S., Hebner, R.E., 2010. Lipid analysis of Neochloris Herrero, M., Ibáñez, E., Cifuentes, A., Reglero, G., Santoyo, S., 2006. Dunaliella salina
oleoabundans by liquid state NMR. Biotechnol. Bioeng. 106, 573–583. microalga pressurized liquid extracts as potential antimicrobials. J. Food Prot. 69,
Ben-Amotz, A., Avron, M., 1973. The role of glycerol in the osmotic regulation of the 2471–2477.
halophilic alga Dunaliella parva. Plant Physiol. 51, 875–878. Jaspars, M., De Pascale, D., Andersen, J.H., Reyes, F., Crawfrod, A.D., Ianora, A., 2016.
Bendif, El M., Probert, I., Schoreder, D.C., de Vargas, C., 2013. On the description of The marine biodiscovery pipeline and ocean medicines of tomorrow. J. Mar. Biol.
Tisochrysis lutea gen. nov. sp. nov. and Isochrysis nuda sp. nov. in the Isochrysidales, Assoc. U. K. 96, 151–158.
and the transfer of Dicrateria to the Prymnesiales (Haptophyta). J. Appl. Phycol. 25, Keller, M.D., Kiene, R.P., Matrai, P.A., Bellows, W.K., 1999. Production of glycine betaine
1763–1776. and dimethylsulfoniopropionate in marine phytoplankton. II. N-limited chemostat
Bhattacharjee, M., 2016. Pharmaceutically valuable bioactive compounds of algae. Asian cultures. Mar. Biol. 135, 237–248.
J. Pharmaceut. Clin. Res. 9, 43–47. Kiene, R.P., Linn, L.J., Bruton, J.A., 2000. New and important roles for DMSP in marine
Blunt, J.W., Copp, B.R., Keyzers, R.A., Munro, M.H.G., Prinsep, M.R., 2015. Marine microbial communities. J. Sea Res. 43, 209–224.
natural products. Nat. Prod. Rep. 32, 116–211. Kim, S.W., Ban, S.H., Ahn, C.Y., Oh, H.M., Chung, H., Cho, S.H., Park, Y.M., Liu, J.R.,
Borchman, D., Yappert, M.C., Milliner, S.E., Smith, R.J., Bhola, R., 2013. Confirmation of 2006. Taxonomic discrimination of cyanobacteria by metabolic fingerprinting using
the presence of squalene in human eyelid lipid by heteronuclear single quantum proton nuclear magnetic resonance spectra and multivariate statistical analysis. J.
correlation spectroscopy. Lipids 48, 1269–1277. Plant Biol. 49, 271–275.
Brown, M.R., 1991. The amino-acid and sugar composition of 16 species of microalgae Kim, S.H., Ahn, H.M., Lim, S.R., Hong, S.J., Cho, B.K., Lee, H., Lee, C.G., Choi, H.K., 2015.
used in marineculture. J. Exp. Mar. Biol. Ecol. 145, 79–99. Comparative lipidomic profiling of two Dunaliella tertiolecta strains with different
Busby, W.F., 1966. Sulfopropanedial and cysteinolic acid in the diatom. Biochim. growth temperatures under nitrate-deficient conditions. J. Agric. Food Chem. 63,
Biophys. Acta 121, 160–161. 880–887.
Chauton, M.S., Størseth, T.R., 2018. HR-MAS NMR spectroscopy of marine microalgae. In: Kobayashi, Y., Torii, A., Kato, M., Adachi, K., 2007. Accumulation of cyclitols functioning
Webb, G.A. (Ed.), Modern Magnetic Resonance, second ed. Springer, Dordrecht, pp. as compatible solutes in the haptophyte alga Pavlova sp. Phycol. Res. 55, 81–90.
1927–1935. Kuczynska, P., Jemiola-Rzeminska, M., Stralka, K., 2015. Photosynthetic pigments in
Chia, S.R., Chew, K.W., Show, K.W., Yap, Y.J., Ong, H.C., Ling, T.C., Chang, J.S., 2018. diatoms. Mar. Drugs 13, 5847–5881.
Analysis of economic and environmental aspects of microalgae biorefinery for bio- Kumar, R., Bansal, V., Patel, M.B., Sarpal, A.S., 2014. Compositional analysis of algal
fuels production: a review. Biotechnol. J. https://doi.org/10.1002/biot.201700618. biomass in a Nuclear Magnetic Resonance (NMR) tube. J. Algal Biomass Utln. 5,
Da Costa, F., Petton, B., Mingant, C., Bourgaran, G., Rouxel, C., Quéré, C., Wikfors, G.H., 36–45.
Soudnat, P., Robert, R., 2016. Influence of one selected Tisochrysis lutea strain rich in Lang, I., Hodac, L., Froedl, T., Feussner, I., 2011. Fatty acid profiles and their distribution
lipids on Crassostrea gigas larval development and biochemical composition. patterns in microalgae: a comprehensive analysis of more than 2000 strains from the
Aquacult. Nutr. 22, 813–836. SAG culture collection. BMC Plant Biol. 11, 124/1–124/16.
Dais, P., Misiak, M., Hatzakis, E., 2015. Analysis of marine dietary supplements using Lauritano, C., Andersen, J.H., Hansen, E., Albrigtsen, M., Escalera, L., Esposito, F.,
NMR spectroscopy. Anal. Methods 7, 5226–5238. Helland, K., Hanssen, K.Ø., Romano, G., Ianora, A., 2016. Bioactivity screening of
Deborde, C., Moing, A., Roch, L., Jacob, D., Rolin, D., Giraudeau, P., 2017. Plant meta- microalgae for antioxidant, anti-inflammatory, anticancer, anti-diabetes, and anti-
bolism as studied by NMR spectroscopy. Prog. Nucl. Magn. Reson. Spectrosc. bacterial activities. Frot. Mar. Sci. 3, 68/1–68/12.
102–103, 61–97. LewisOscar, F., Nithya, C., Bakkiyaraj, D., Arunkumar, M., Alharbi, N.S., Thajuddin, N.,
Desbois, A.P., Mearns-Spragg, A., 2009. A fatty acid from the diatom Phaeodactylum 2017. Biofilm inhibitory effect of Spirulina platensis extracts on bacteria of clinical
tricornutum is antibacterial against diverse bacteria including multi-resistant significance. Proc. Natl. Acad. Sci. India B Biol. Sci. 87, 537–544.
Staphylococcus aureus (MRSA). Mar. Biotechnol. 11, 45–52. Liu, C.P., Lin, L.P., 2001. Ultrastructural study and lipid formation of Isochrysis sp.
Diehl, B.W.K., Heling, H., Riedl, I., Heinz, E., 1995. 13C-NMR analysis of the positional CCMP1324. Bot. Bull. Acad. Sin. (Taipei) 42, 207–214.
distribution of fatty acids in plant glycolipids. Chem. Phys. Lipids 77, 147–153. Logvinov, S., Gerasimenko, N., Esipov, A., Denisenko, V.A., 2015. Examination of the
Dobretsov, S., Teplitski, M., Alagely, A., Gunasekera, S.P., Paul, V.J., 2010. Malyngolide structures of several glycerolipids from marine macroalgae by NMR and GC-MS. J.
from the cyanobacterium Lyngbya majuscula interferes with quorum sensing circuitry. Phycol. 51, 1066–1074.
Environ. Microbiol. Rep. 2, 739–744. Maras, A., Erden, M., Seçen, H., Sütbeyaz, Y., 1998. One-pot synthesis from 1,4-cyclo-
Dunstan, G.A., Volkman, J.K., Barrett, S.M., Leroi, J.M., Jeffrey, S.W., 1994. Essential hexadiene of ( ± )-1,4/2,5-cyclohexanetetrol, a naturally occurring cyclitol deriva-
polyunsaturated fatty acids from 14 species of diatom (Bacillariophyceae). tive. Carbohydr. Res. 308, 435–437.
Phytochemistry 35, 155–161. Marlowe, I.T., Brassell, S.C., Eglinton, G., Green, J.C., 1984. Long chain unsaturated
Durham, B.P., Sharma, S., Luo, H., Smith, C.B., Amin, S.A., Bender, S.J., Dearth, S.P., Van ketones and esters in living algae and marine sediments. Org. Geochem. 6, 135–141.
Mooy, B.A.S., Campagna, S.R., Kujawinski, E.B., Armbrust, E.V., Moran, M.A., 2015. Matos, Â.P., 2017. The impact of microalgae in food science and technology. J. Am. Oil
Cryptic carbon and sulfur cycling between surface ocean plankton. Proc. Natl. Acad. Chem. Soc. 94, 1333–1350.
Sci. Unit. States Am. 112, 453–457. Mayer, A.M.S., Rodríguez, A.D., Taglialatela-Scafati, O., Fusetani, N., 2013. Marine
Eltgroth, M.L., Watwood, R.L., Wolfe, G.V., 2005. Production and cellular localization of pharmacology in 2009-2011: marine compounds with antibacterial, antidiabetic,
neutral long‐chain lipids in the haptophyte algae Isochrysis galbana and Emiliania antifungal, anti-inflammatory, antiprotozoal, antituberculosis, and antiviral activ-
huxleyi. J. Phycol. 41, 1000–1009. ities; affecting the immune and nervous systems, and other miscellaneous mechan-
Englert, G., Bjǿrnland, T., Liaaen-Jensen, S., 1990. 1D and 2D NMR study of some allenic isms of action. Mar. Drugs 11, 2510–2573.
carotenoids of the fucoxanthin series. Magn. Reson. Chem. 28, 519–528. Mendiola, J.A., Torres, C.F., Toré, A., Martín-Álvarez, P.J., Santoyo, S., Arredondo, B.O.,
Fan, T.W.M., Lane, A.N., 2016. Applications of NMR spectroscopy to systems biochem- Señoráns, F.J., Cifuentes, A., Ibáñez, E., 2007. Use of supercritical CO2 to obtain
istry. Prog. Nucl. Magn. Reson. Spectrosc. 92–93, 18–53. extracts with antimicrobial activity from Chaetoceros muelleri microalga. A correlation
Ford, C.W., Percival, E., 1965. The Carbohydrates of Phaeodactylum tricornutum. Part I. with their lipidic content. Eur. Food Res. Technol. 224, 505–510.
Preliminary examination of the organism, and characterisation of low molecular Mendiola, J.A., Santoyo, S., Cifuentes, A., Reglero, G., Ibáñez, E., Señoráns, F.J.J., 2008.
weight material and of a glucan. J. Chem. Soc. 7035–7041. Antimicrobial activity of sub- and supercritical CO2 extracts of the green alga
Fujii, S., Nishimoto, N., Notoya, A., Hellebust, J.A., 1995. Growth and osmoregulation of Dunaliella salina. Food Prot 71, 2138–2143.
Chaetoceros muelleri (Bacillariophyceae) in relation to salinity. Plant Cell Physiol. 36, Millar, A.A., Smith, M.A., Kunst, L., 2000. All fatty acids are not equal: discrimination in
759–764. plant membrane lipids. Trends Plant Sci. 5, 95–101.
Garnier, M., Bougaran, G., Pavlovic, M., Berard, J.B., Carrier, G., Charrier, A., Le Grand, Miller, M.R., Quek, S.-Y., Staeher, K., Nalder, T., Packer, M.A., 2014. Changes in oil
F., Lukomska, E., Rouxel, C., Schreiber, N., Cadoret, J.P., Rogniaux, H., Saint-Jean, content, lipid class and fatty acid composition of the microalga Chaetoceros calcitrans
B., 2016. Use of a lipid rich strain reveals mechanisms of nitrogen limitation and over different phases of batch culture. Aquacult. Res. 45, 1634–1647.
carbon partitioning in the haptophyte Tisochrysis lutea. Algal Res 20, 229–248. Mobin, S., Alam, F., 2017. Some promising microalgal species for commercial applica-
Garza-Sánchez, F., Chapman, D.J., Cooper, J.B., 2009. Nitzschia ovalis (Bacillariophyceae) tions: a review. Energy Procedia 110, 510–517.
mono lake strain accumulates 1,4/2,5 cyclohexanetetrol in response to increased Mulders, K.J.M., Weesepoel, Y., Lamers, P.P., Vincken, J.P., Martens, D.E., Wijffels, R.H.,
salinity. J. Phycol. 45, 395–403. 2013. Growth and pigment accumulation in nutrient-depleted Isochrysis aff. galbana
Gebser, B., Pohnert, G., 2013. Synchronized regulation of different zwitterionic meta- T-ISO. J. Appl. Phycol. 25, 1421–1430.
bolites in the osmoadaption of phytoplankton. Mar. Drugs 11, 2168–2182. Nieva-Echevarría, B., Goicoechea, E., Manzanos, M.J., Guillén, M.D., 2014. A method
Granum, E., Myklestad, S.M., 2001. Mobilization of β‐1,3‐glucan and biosynthesis of based on 1H NMR spectral data useful to evaluate the hydrolysis level in complex
amino acids induced by NH4+ addition to N‐limited cells of the marine diatom lipid mixtures. Food Res. Int. 66, 379–387.
13
Nuzzo, G., Gallo, C., d'Ippolito, G., Cutignano, A., Sardo, A., Fontana, A., 2013. microalgal biomass by NMR spectroscopic and GC–MS techniques. Anal. Bioanal.
Composition and quantitation of microalgal lipids by ERETIC 1H NMR method. Mar. Chem. 407, 3799–3816.
Drugs 11, 3742–3753. Sarpal, A.S., Teixeira, C.M.L.L., Silva1, P.R.M., da Costa Monteiro, T.V., da Silva1, J.I., da
O'Neil, G.W., Knothe, G., Williams, J.R., Burlow, N.P., Culler, A.R., Corliss, J.M., Cunha, V.S., Daroda, R.J., 2016a. NMR techniques for determination of lipid content
Carmichel, C.A., Reddy, C.M., 2014. Synthesis and analysis of an alkenone-free bio- in microalgal biomass and their use in monitoring the cultivation with biodiesel
diesel from Isochrysis sp. Energy Fuels 28 2877-2683. potential. Appl. Microbiol. Biotechnol. 100, 2471–2485.
Obando, C.Z., Linossier, I., Kervarek, N., Zubia, M., Turquet, J., Faÿ, F., Rehel, K., 2016. Sarpal, A.S., Sharma, B.K., Scott, J., Kumar, R., Sugmaran, V., Chopra, A., Bansal, V.,
Rapid identification of osmolytes in tropical microalgae and cyanobacteria by 1H HR- Rajagopalan, N.K., 2016b. Compositional analyses of oil extracts of microalgae bio-
MAS NMR spectroscopy. Talanta 153, 372–380. masses by NMR and chromatographic techniques. J. Anal. Bioanal. Sep. Tech. 1,
Obata, T., Schoenefeld, S., Kranhnert, I., Bergmann, S., Scheffek, A., Ferni, A.R., 2013. 17–41.
Gas-Chromatography Mass-Spectrometry (GC-MS) based metabolite profiling reveals Sarpal, A.S., Costa, I.C.R., Teixeira, C.M.L.L., Filocomo, D., Candido, R., Silva, P.R.M.,
mannitol as a major storage carbohydrate in the coccolithophorid alga Emiliania Cunha, V.S., Daroda, R.J., 2016c. Investigation of biodiesel potential of biomasses of
huxleyi. Metabolites 3, 168–184. microalgaes chlorella, Spirulina and Tetraselmis by NMR and GC-MS techniques. J.
Ohta, S., Shiomi, Y., Kawahima, A., Aozasa, O., Nakao, T., Nagate, T., Kitamura, K., Biotechnol. Biomater. 6, 1000220/1–000220/15.
Miyata, H., 1995. Antibiotic effect of linolenic acid from Chlorococcum strain HS-101 Satyarthi, J.K., Srinivas, D., Ratnasamy, P., 2009. Estimation of free fatty acid content in
and Dunaliella primolecta on methicillin-resistant Staphylococcus aureus. J. Appl. oils, fats, and biodiesel by 1H NMR spectroscopy. Energy Fuels 23, 2273–2277.
Phycol. 7 212-127. Serive, B., Nicolau, E., Bérard, J.B., Kaas, R., Pasquet, V., Picot, L., Cadoret, J.P., 2017.
Patterson, G.W., Tsittsa-Tsardis, E., Wikfors, G.H., Galdu, P.K., Chitwood, D.J., Harrison, Community analysis of pigment patterns from 37 microalgae strains reveals new
D., 1994. Sterols and alkenones of Isochrysis. Phytochemistry 35, 1233–1236. carotenoids and porphyrins characteristic of distinct strains and taxonomic groups.
Peng, J., Yuan, J.P., Wu, C.F., Wang, J.H., 2011. Fucoxanthin, a marine carotenoid pre- PLoS One 12, e0171872/1–e0171872/35.
sent in brown seaweeds and diatoms: metabolism and bioactivities relevant to human Shannon, E., Abu-Ghannam, N., 2016. Antibacterial derivatives of marine algae: an
health. Mar. Drugs 9, 1806–1828. overview of pharmacological mechanisms and applications. Mar. Drugs 14, 81/
Pollesello, P., Toffanin, R., Murano, E., Paoletti, S., Rizzo, R., Kvam, B.J., 1992. Lipid 1–81/23.
extracts from different algal species: 1H and13C-NMR spectroscopic studies as a new Slattery, M., Hamann, M.T., McClintock, J.B., Perry, T.L., Puglisi, M.P., Yoshida, W.Y.,
tool to screen differences in the composition of fatty acids, sterols and carotenoids. J. 1997. Ecological roles for water-borne metabolites from Antarctic soft corals. Mar.
Appl. Phycol. 4, 315–322. Ecol. Prog. Ser. 161, 133–144.
Rechka, J.A., Maxwell, J.R., 1987. Unusual long chain ketones of algal origin. Adv. Org. Sobolev, A.P., Brosio, E., Gianferri, R., Segre, A.L., 2005. Metabolic profile of lettuce
Geochem. 13, 727–734. leaves by high‐field NMR spectra. Magn. Reson. Chem. 43, 625–638.
Réveillon, D., Tunin-Ley, A., Grondin, I., Othmani, A., Zubia, M., Bunet, R., Turquet, J., Størseth, T.R., Hansen, K., Skjermo, J., Krane, J., 2004. Characterization of a β-d-(1 → 3)-
Culioli, G., Briand, J.F., 2018. Exploring the chemodiversity of tropical microalgae glucan from the marine diatom Chaetoceros mülleri by high-resolution magic-angle
for the Discovery of natural antifouling compounds. J. Appl. Phycol. https://doi.org/ spinning NMR spectroscopy on whole algal cells. Carbohydr. Res. 339, 421–424.
10.1007/s10811-018-1594-z. Størseth, T.R., Kirkvold, S., Skjermo, J., Reitan, K.I., 2006. A branched β-d-(1→3,1→6)-
Roleda, M.Y., Slocombe, S.P., Leakey, R.J.G., Day, J.G., Bell, E.M., Stanley, M.S., 2013. glucan from the marine diatom Chaetoceros debilis (Bacillariophyceae) characterized
Effects of temperature and nutrient regimes on biomass and lipid production by six by NMR. Carbohydr. Res. 341, 2108–2114.
oleaginous microalgae in batch culture employing a two-phase cultivation strategy. Sun, Y., Lin, Y., Cao, X., Xiang, L., Qi, J., 2014. Sterols from Mytilidae show anti-aging and
Bioresour. Technol. 129, 439–449. neuroprotective effects via anti-oxidative activity. Int. J. Mol. Sci. 15, 21660–21673.
Romano, G., Costantini, M., Sansone, C., Lauritano, C., Ruocco, N., Ianora, A., 2017. Thompson Jr., G.A., 1996. Lipids and membrane function in green algae. Biochem. Bioph.
Marine microorganisms as a promising and sustainable source of bioactive molecules. Acta 1302, 17–45.
Mar. Environ. Res. 128, 58–69. Thormar, E., 2011. Lipids and Essential Oils as Antimicrobial Agents. John Wiley & Sons,
Romero, M., Diggle, S.P., Heeb, S., Cámara, M., Otero, A., 2008. Quorum quenching Chichester.
activity in Anabaena sp. PCC7120: identification of AiiC, a novelAHL-acylase. FEMS Tsuji, Y., Yamazaki, M., Suzuki, I., Shiraiwa, Y., 2015. Quantitative analysis of carbon
Microbiol. Lett. 280, 73–80. flow into photosynthetic products functioning as carbon storage in the marine coc-
Roy, A.B., Hewlins, M.J.E., Ellis, A.J., Harwood, A.J., White, G.F., 2003. Glycolytic colithophore. Emiliania huxleyi. Mar. Biotechnol. 17, 428–440.
breakdown of sulfoquinovose in bacteria: a missing link in the sulfur cycle. Appl. Tsukida, K., Saiki, K., Sugira, M., 1981. Structural elucidation of the main cis-β-carotenes.
Environ. Microbiol. 69, 6434–6441. J. Nutr. Sci. Vitaminol. 27, 551–561.
Ruffell, S.E., Müller, K.M., McConkey, B.J., 2016. Comparative assessment of microalgal Vidal, N.P., Manzanos, M.J., Goicoechea, E., Guillén, M.D., 2012. Quality of farmed and
fatty acids as topical antibiotics. J. Appl. Phycol. 28, 1695–1704. wild sea bass lipids studied by 1H NMR: usefulness of this technique for differ-
Sacchi, R., Medina, I., Aubourg, S.P., Giuducuanni, I., Paolillo, L., Addeo, F., 1993. entiation on a qualitative and a quantitative basis. Food Chem. 135, 1583–1591.
Quantitative high-resolution 13C NMR analysis of lipids extracted from the white Viso, A.C., Marty, J.C., 1993. Fatty acids from 28 marine microalgae. Phytochemistry 34,
muscle of Atlantic tuna (Thunnus alalunga). J. Agric. Food Chem. 41, 1247–1253. 1521–1533.
Sacchi, R., Addeo, F., Paolillo, L., 1997. 1H and 13C NMR of virgin olive oil. An overview. Volkman, J.K., Egliton, G., Corner, E.D.S., Forseberg, T.E.V., 1980. Long-chain alkenes
Magn. Reson. Chem. 35, S133–S145. and alkenones in the marine coccolithophorid Emiliania huxleyi. Phytochemistry 19,
Sadovskaya, I., Souissi, A., Souissi, S., Grard, T., Lencel, P., Greene, C.M., Duin, S., 2619–2622.
Dmitrenok, P.S., Chizhov, A.O., Shaskov, A.S., Usov, A.I., 2014. Chemical structure Volkman, J.K., Jeffrey, S.W., Nichols, P.D., Rogers, G.I., Garland, C.D., 1989. Fatty acid
and biological activity of a highly branched (1 → 3,1 → 6)-β-d-glucan from Isochrysis and lipid composition of 10 species of microalgae used in mariculture. J. Exp. Mar.
galbana. Carbohydr. Polym. 111, 139–148. Biol. Ecol. 128, 219–240.
Saha, M., Rempt, M., Grosser, K., Pohnert, G., Werinberger, F., 2011. Surface-associated Weete, J.D., Kulifaj, M., Montant, C., Nes, W.R., Sancholle, M., 1985. Distribution of
fucoxanthin mediates settlement of bacterial epiphytes on the rockweed Fucus vesi- sterols in fungi. II. Brassicasterol in Tuber and Terfezia species. Can. J. Microbiol. 31,
culosus. Biofouling 27, 423–433. 1127–1130.
Sakano, Y., Mutsuga, M., Tanaka, R., Suganuma, H., Inakuma, T., Toyoda, M., Goda, Y., Wikfors, G.H., Ohno, M., 2001. Impact of algal research in aquaculture. J. Phycol. 37,
Shibuya, M., Ebizuka, Y., 2005. Inhibition of human lanosterol synthase by the 968–974.
constituents of Colocasia esculenta (Taro). Biol. Pharm. Bull. 28, 299–304. Yancey, P.H., 2005. Organic osmolytes as compatible, metabolic and counteracting cy-
Salinero, C., Feás, X., Mansilla, J.P., Seijas, J.A., Vázquez-Tato, M.P., Vela, P., Sainz, M.J., toprotectants in high osmolarity and other stresses. J. Exp. Biol. 208, 2819–2830.
2012. 1H-nuclear magnetic resonance analysis of the triacylglyceride composition of Yao, L., Gerde, J.A., Lee, S.L., Wang, T., Harrata, K.A., 2015. Microalgae lipid char-
cold-pressed oil from Camellia japonica. Molecules 17, 6716–6727. acterization. J. Agric. Food Chem. 63, 1773–1787.
Sarker, S.D., Latif, Z., Gray, A.I., 2006. Natural Products Isolation, second ed. Humana Zheng, C.J., Yoo, H.Y., Kim, Y.H., Kim, W.G., 2005. Fatty acid synthesis is a target for
Press, New Jersey. antibacterial activity of unsaturated fatty acids. FEBS Lett. 579, 5157–5162.
Sarpal, A.S., Teixeira, C.M.L.L., Silva, P.R.M., Lima, G.M., Silva, S.R., Monteiro, T.V., Zhukova, N.V., Aizdaicher, N.A., 1995. Fatty acid composition of 15 species of marine
Cunha, V.S., Daroda, R.J., 2015. Determination of lipid content of oleaginous microalgae. Phytochemistry 39, 351–356.
14

NMR Characterization and Evaluation of Antibacterial

Uploaded by

Copyright:

Available Formats

NMR Characterization and Evaluation of Antibacterial

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

NMR Characterization and Evaluation of Antibacterial

Uploaded by

Copyright:

Available Formats

Phytochemistry 164 (2019) xxx–xxx

Contents lists available at ScienceDirect

NMR characterization and evaluation of antibacterial and antiobiofilm T

ARTICLE INFO ABSTRACT

1. Introduction identifying and quantifying the most abundant metabolites present in

Dunaliella Chaetoceros T. lutea

sp. salina calcitrans gracilis

Storage lipids TG TG TG TG PULCA TG (minor)

Fig. 4. Expansion of the 1H NMR spectrum, measured in CD3OD, of the C.

Fig. 6. Expansion of the 1H NMR spectra measured in CD3OD of AcOEt extracts

stored as TG (Miller et al., 2014). a

clearly identified in the 1H NMR spectrum by the characteristic re-

2.1.3. Polar metabolites

Table 4 The diatom (Chaetoceros calcitrans and C. gracilis) and haptophyte

You might also like