Planta (1998) 206: 300±307
In-situ identi®cation of major metabolites in the red alga Gracilariopsis lemaneiformis using high-resolution magic angle spinning nuclear magnetic resonance spectroscopy Anders Broberg1, Lennart Kenne1, Marianne PederseÂn2 1
Department of Chemistry, Swedish University of Agricultural Sciences, P.O. Box 7015, S-750 07 Uppsala, Sweden Department of Botany, Stockholm University, S-106 91 Stockholm, Sweden
2
Received: 2 February 1998 / Accepted: 9 April 1998
Abstract. The content of low-molecular-weight compounds in the red alga Gracilariopsis lemaneiformis [(Bory) Dawson, Acleto, et Foldvik] has been analysed in-situ using high-resolution magic angle spinning (HRMAS) nuclear magnetic resonance (NMR) spectroscopy. The major heteroside was shown to be ¯oridoside, but digeneaside and iso¯oridoside were also detected in the alga. Other major components were isethionic acid and the amino acids taurine and citrulline. The results from the HR-MAS NMR analysis were con®rmed with high-resolution NMR spectroscopy, high-resolution fast atom bombardment mass spectrometry (FABMS) and GC-MS, on material isolated from the studied alga, but also on authentic samples. Key words: Digeneaside ± Floridoside ± Gracilariopsis ± High-resolution magic angle spinning ± Isethionic acid ± Iso¯oridoside
Introduction In many studies of metabolism, a severe drawback is the need to disintegrate the organism of study in order to perform the analysis of the metabolites. During sample preparation, modi®cations of the components of interest can occur. Thus a direct non-destructive method, which allows analysis without isolation of the components, is highly desirable. In red algae the low-molecular-weight glycosides of D-galactose and D-mannose, i.e. ¯oridoside [a-D-gala-
Abbreviations: COSY correlation spectroscopy; FABMS fast atom bombardment mass spectrometry; HMQC heteronuclear multiple quantum coherence; HR-MAS high-resolution magic angle spinning; NMR nuclear magnetic resonance; TOCSY total correlation spectroscopy Correspondence to: L. Kenne; E-mail:-
[email protected]; Fax: 46 (18) 673477
ctopyranosyl-(1 ® 2)-glycerol], iso¯oridoside [a-D-galactopyranosyl-(1 ® l)-glycerol] and digeneaside [a-Dmannopyranosyl-(1 ® 2)-D-glycerate] are an important group of metabolites. Floridoside is normally considered to be the major photosynthetic product in all red algae, except for the order Ceramiales, where it is replaced by digeneaside (Kremer and Kirst 1982). Moreover, the red algal order Bangiales contains iso¯oridoside as well as ¯oridoside (Karsten et al. 1993). However, other sugars, polyols such as glucitol and galactitol, organic acids, amino acids and derivatives of amino acids, are reported to be present in considerable concentrations in some red algae (Kremer and Vog1 1975; Kremer and Kirst 1982). When solid or semi-solid samples are analysed using normal solution nuclear magnetic resonance (NMR) spectroscopy, the resulting spectral information is severely limited by the increased linewidths, due to dipolar interactions, chemical shift anisotropy and dierences in magnetic susceptibility of the sample. The signal broadening caused by dipolar couplings and chemical shift anisotropy is due to the restricted molecular motion in solid or semi-solid samples. However, when a semi-solid sample is spun at a rate in the lowkHz range, at an angle to the external magnetic ®eld [the magic angle (Q) 54.7°], the eects on linewidths of dipolar couplings and chemical shift anisotropy, which both are scaled by 3cos2Q ) 1, are drastically decreased. Furthermore, signal broadening due to magnetic susceptibility eects is also reduced. This results in markedly reduced resonance linewidths and thus signi®cantly improved spectral quality. These modi®cations are the basis for the technique called high-resolution magic angle spinning (HR-MAS) NMR spectroscopy (Fitch et al. 1994). When HR-MAS NMR spectroscopy is used to analyse intact plant material, it is possible to record both one-dimensional and two-dimensional NMR spectra of low-molecular-mass compounds in solution inside the studied organism, as well as spectra of polymers. The technique is very sensitive, only requiring a few minutes to record 1H NMR spectra of compounds present at medium to low microgram levels in the plant sample.
A. Broberg et al.: HR-MAS NMR of major metabolites in a red alga
High-resolution magic angle spinning NMR spectroscopy has recently been described as an analytical tool in the study of some metabolites in animal tissue (Cheng et al. 1996, 1997; Moka et al. 1997). Included in these studies are both one- and two-dimensional HRMAS NMR experiments. In the study of plant material, one-dimensional 1H NMR experiments have been employed (Ni and Eads 1993a,b), but 13C NMR experiments have also been used in studies of conifer seeds (Sayer and Preston 1996), as well as of yeast cell wall glucans (Krainer et al. 1994). The scope of this paper is to illustrate the usefulness of HR-MAS NMR as an analytical tool, both in the study of the content of low-molecular-weight compounds in plants, as well as in the study of metabolism in plants. To exemplify the promising abilities of HRMAS NMR, some metabolites in the red alga Gracilariopsis lemaneiformis are identi®ed in situ, using HRMAS NMR, and some metabolic events, in the same red alga, following a change in cultivation conditions, are monitored using the same technique. Materials and methods Algal material. The red alga Gracilariopsis lemaneiformis [(Bory) Dawson, Acleto, et Foldvik] was cultivated in the laboratory as previously described (Lignell et al. 1987; Yu and PederseÂn 1990). To yield complementary information, i.e. to decrease the concentration of ¯oridoside in the alga and hence simplify the corresponding NMR spectra, some algal thalli were kept at normal salinity (3.3%) in darkness for prolonged periods (2±3 weeks), before the HR-MAS NMR analysis. The red alga Ceramium rubrum (Ceramiales) was collected on the west coast of Sweden in October 1997. Cultivation of G. lemaneiformis at dierent salinities. Gracilariopsis lemaneiformis was transferred from normal cultivation conditions (3.3% salinity) to 5.0% salinity in the dark. After 4 d the salinity was lowered to 1.0% and the alga was kept in the dark for another 2 d. The alga was analysed with HR-MAS NMR after 4 d at 5.0% salinity and at the end of the experiment. Analysis by NMR. Prior to analysis with HR-MAS NMR, the algal samples were submerged in 2H2O for 2 min. The algal samples (typically 1±2 mg dry weight), were inserted in the magic angle spinning rotor (zirconia, 4 mm outer diameter, spherical sample volume; Bruker, Karlsruhe, Germany) and 2H2O was carefully added, without the introduction of air bubbles. The rotor was subsequently sealed with the rotor spacer, sealing screw and ®nally the rotor cap. During the experiments the samples were spun at 5±15 kHz, at the magic angle (54.7 °). Sodium trimethylsilylpropionate (dH 0.00) was added to the surrounding 2H2O as a reference. All the HR-MAS NMR spectra were recorded at 600 MHz (Bruker DRX-600, HR-MAS SB BL4 probe), at ambient temperature, using a Carr-Purcell-Meiboom-Gill pulse sequence (Meiboom and Gill 1958) [90° ) (s ) 180°) s)n ) acquisition] for one-dimensional T2 ®ltered HR-MAS 1H-NMR experiments (s 300 ls, n 500) and standard pulse sequences for 1H-1H correlation spectroscopy (COSY), 1H-1H total correlation spectroscopy (TOCSY, mixing time 110 ms), and 1H-13C heteronuclear multiple quantum coherence (HMQC) experiments. The twodimensional experiments were recorded with pre-saturation of the HDO-signal. All pulse programs were supplied by Bruker. Following the HR-MAS NMR analysis, the algal samples were lyophilised and weighed.
301 To simplify the interpretations of the recorded HR-MAS NMR spectra, comparisons were made with pure samples of ¯oridoside and iso¯oridoside. Moreover, digeneaside, isethionic acid (2hydroxyethanesulfonic acid) and the amino acids citrulline and taurine (2-aminoethanesulfonic acid), were partially isolated from aqueous extracts of G. lemaneiformis to record complementary 600MHz NMR spectra [one-dimensional 1H, 1H-1H COSY, 1H-1H TOCSY, 1H-13C HMQC, 1H-13C heteronuclear multiple bond correlation (HMBC), 1H-1H nuclear Overhauser enhancement spectroscopy (NOESY)] to verify the results from HR-MAS NMR. Digeneaside was also partially isolated from aqueous extracts of C. rubrum for comparison. Fractionation of algal components. Fractionation of the components in G. lemaneiformis was achieved by grinding the alga in liquid N2, followed by extraction with distilled water [3 ml (gFW))1]. The cell debris was removed by centrifugation and the supernatant lyophilised. Freeze-dried extract (50 mg) was dissolved in distilled water (1 ml) and loaded onto a column of Bio-Gel P-2 (2.6 ´ 90 cm; Bio-Rad, Hercules, Calif., USA), eluted with water at 0.5 ml min)1. The chromatography was monitored using a refractive-index detector, and the interesting fractions were analysed by NMR spectroscopy (one-dimensional 1H, recorded on a Bruker DRX-400 spectrometer). After the NMR analysis, some fractions were rechromatographed on the same column to improve the puri®cation. The extract from C. rubrum was prepared as above. The algal extract (20 ml from 6.5 g FW alga) was loaded onto an anionexchange resin column (Dowex 1, OH), 1 ´ 10 cm; Sigma, St. Louis, Mo., USA). Following washing with distilled water, the column was eluted with 15 ml 0.1 M HCL (aq). The collected eluent was neutralised with 0.1 M NaOH (aq), lyophilised and analysed by NMR spectroscopy (1H) at 400 MHz. Analysis by GC-MS. Analysis by GC-MS [HP 5890/5970 (HewlettPackard, Palo Alto, Calif., USA), carrier gas He, DB-5 fused-silica column (0.25 mm ´ 30 m; J&W Scienti®c, Folsom, Calif.)] was performed after silylation of lyophilised algal extracts, using pyridine:hexamethyldisilazane:chlorotrimethylsilane (9:3:1, by vol.), at ambient temperature for 2 h. The temperature program employed was 120±260 °C at 5 °C min)1, after a 5-min hold at 120 °C; injector temperature 240 °C, and GC-MS interface temperature 260 °C. Analysis by fast atom bombardment mass spectrometry (FABMS). High-resolution FABMS analysis was performed on a Jeol SX/ SX102A spectrometer (Jeol, Japan) in the negative mode, with glycerol as matrix, using matrix peaks as references.
Results and discussion Content of low-molecular-weight compounds in G. lemaneiformis. The red alga G. lemaneiformis was shown by HR-MAS NMR spectroscopy (Fig. 1) to contain ¯oridoside as the major heteroside, when cultivated at 3.3% salinity under normal conditions. The 1H NMR signals of pure ¯oridoside in 2H2O solution matched the corresponding signals in the algal sample very well. This ®nding is consistent with the theory that ¯oridoside is the major photosynthetic and reserve product in all orders of the Rhodophyta except the Ceramiales (Kremer and Vogl 1975; Kremer and Kirst 1982). Surprisingly, the heterosides iso¯oridoside (approximately 1:1 mixture of D- and L-forms) and digeneaside were also found in the same alga (Fig. 1). The signals of the anomeric protons of the D- and L-forms of iso¯oridoside overlap, resulting in a signal resembling a triplet
302
A. Broberg et al.: HR-MAS NMR of major metabolites in a red alga
Fig. 1. HR-MAS 1H NMR spectrum of Gracilariopsis lemaneiformis cultivated under normal conditions. The spectrum was recorded by 256 scans on 1.5 mg DW of algal sample at 15 kHz. The depicted acids are mainly are mainly present as salts at physiological pH
(Fig. 1). Exactly the same pattern was achieved when authentic iso¯oridoside (1:1 mixture of D- and L-forms) was analysed with normal high-resolution NMR. By incubating G. lemaneiformis at 3.3% salinity in the dark for 2±3 weeks, the ¯oridoside content decreased drastically, which simpli®ed the identi®cation of iso¯oridoside and digeneaside in the alga by a one-dimensional T2 ®ltered HR-MAS 1H NMR experiment (Fig. 2), and by 1H-1H COSY (Fig. 3), 1H-1H TOCSY (Fig. 4) and 1 H-13C HMQC (Fig. 5) HR-MAS NMR experiments. In the 1H-1H COSY spectrum (Fig. 3) some of the observed 1H-1H correlations for digeneaside are exempli®ed, whereas in the 1H-1H TOCSY spectrum (Fig. 4) the correlations to H-1 of iso¯oridoside (D/L-mixture) are demonstrated. The pattern of the iso¯oridoside cross-peaks in the TOCSY spectrum perfectly matched the pattern obtained when authentic iso¯oridoside was analysed with a TOCSY experiment using a normal high-resolution NMR probe head. The 1H-13C HMQC spectrum (Fig. 5) shows the 1H-13C correlations for digeneaside. The 13C chemical shifts obtained from this experiment match data from the literature very well (Karsten et al. 1994). The corresponding 1H chemical shifts, however, show some deviation from the literature data, possible due to dierences in pH. When NMR analysis was performed on digeneaside isolated from extracts of G. lemaneiformis, the data were consistent with an a-mannopyranosyl-(1 ® 2)-glyceric acid. Moreover, these data were identical with the corresponding HR-MAS NMR data. Furthermore, these NMR data
perfectly matched the 1H NMR data of digeneaside isolated form C. rubrum. The presence of ¯oridoside and iso¯oridoside was also shown by GC-MS analysis of silylated algal extracts from the same alga. The retention times and fragmentation patterns of ¯oridoside and iso¯oridoside were found to be identical to those of authentic samples. Moreover, the GC-MS analysis also yielded fragmentation data consistent with the presence of digeneaside. The detection of iso¯oridoside and digeneaside in G. lemaneiformis is surprising since iso¯oridoside is normally found in the red algal order Bangiales (Karsten et al. 1993), whereas digeneaside is found in the order Ceramiales (Kremer and Kirst 1982), and to our knowledge, these heterosides have never been reported in any other algal orders. Digeneaside has even been reported to be a chemotaxonomic marker of the red algal order Ceramiales (Kremer and Kirst 1982), but this has recently been questioned by several authors (Barrow et al. 1995; Karsten et al. 1995). Another major component in G. lemaneiformis was found to be 2-hydroxyethanesulfonic acid (isethionic acid, present as a salt at physiological pH) (Figs. 1±3). This was established by comparison of recorded HRMAS NMR data with NMR data from isolated isethionic acid and with literature data (1H and 13C chemical shits; Barrow et al. 1993; Holst et al. 1994) and by high-resolution FABMS in the negative mode. The 1H-1H COSY spectrum (Fig. 3) shows the correlations between the two methylene groups of isethionic acid.
A. Broberg et al.: HR-MAS NMR of major metabolites in a red alga
303
Fig. 2. HR-MAS 1H NMR spectrum of Gracilariopsis lemaneiformis after 2 weeks in the dark. The spectrum was recorded by 512 scans on 1.6 mg DW of algal sample at 15 kHz. The depicted acids are mainly present as salts at physiological pH
Fig. 3. HR-MAS 1H-1H COSY spectrum of Gracilariopsis lemaneiformis after 2 weeks in the dark. Some 1H-1H correlations for digeneaside (D), taurine (T) and isethionic acid (I) are exempli®ed. The spectrum was recorded on 1.6 mg DW of algal sample at 15 kHz. The depicted acids are mainly present as salts at physiological pH
304
A. Broberg et al.: HR-MAS NMR of major metabolites in a red alga
Fig. 4. HR-MAS 1H-1H TOCSY spectrum of Gracilariopsis lemaneiformis after 2 weeks in the dark. The 1H-1H correlations to H-1 of iso¯oridoside (IF, D/L-mixture) are indicated in the spectrum. The spectrum was recorded on 1.6 mg DW of algal sample at 15 kHz (mixing time 110 ms)
High-resolution FABMS analysis resulted in the molecular mass 124.9937, consistent with the formula C2H5O4S (0.002% from the calculated theoretical value 124.9909) for the studied negative ion of isethionic acid. Isethionic acid has recently been reported to exist in the red algal orders Ceramiales (Barrow et al. 1993) and Gigartinales (Holst et al. 1994). Within the Gigartinales, the presence of isethionic acid has not been shown in the Gracilariaceae (Holst et al. 1994), the family to which G. lemaneiformis belongs. The HR-MAS NMR spectra also showed that the amino acid taurine was present in G. lemaneiformis (Fig. 1±3). The 1H-1H COSY spectrum shows the correlations between the two methylene groups of taurine. The identity of taurine was con®rmed by comparing 1H NMR spectra of taurine, isolated from algal extracts, with spectra of authentic taurine. Moreover, when pure taurine was added to the isolated taurine, the 1H NMR signals showed perfect overlap. Another major metabolite of G. lemaneiformis proved to be the amino acid citrulline (Fig. 1±2). The amino acid was identi®ed by comparing HR-MAS NMR spectra recorded on intact algae with normal NMR spectra of samples puri®ed by size-exclusion chromatography and with NMR spectra of pure citrulline. The observation of citrulline and taurine in G. lemaneiformis is expected since high concentrations of these amino acids are often found when red algae are cultivated under conditions of rich nitrogen supply (Lignell and PederseÂn 1987). When the red alga Gracil-
aria sordida (Gigartinales) was cultivated with a rich nitrogen supply at 13 °C, the concentration of citrulline and taurine was found to be 670 lg and 130 lg N (g FW))1, respectively (Lignell and PederseÂn 1987). The simple fractionation of the extracts of G. lemaneiformis using size-exclusion chromatography yielded essentially pure digeneaside and isethionic acid. Floridoside and iso¯oridoside (D- and L-forms) were isolated as a mixture, whereas citrulline was partially eluted together with taurine. Anion-exchange chromatography on the extract from C. rubrum yielded mainly digeneaside. Eects of changing cultivation conditions. When G. lemaneiformis was kept in the dark at 3.3% salinity for two weeks, the concentration of ¯oridoside decreased drastically (Fig. 1,2). Also the concentration of iso¯oridoside decreased relative to digeneaside during this period of darkness (Fig. 1,2). When G. lemaneiformis was transferred from normal cultivation conditions to 5.0% salinity in the dark and analysed after 4 d, the concentrations of ¯oridoside and iso¯oridoside appeared to have increased relative to digeneaside (Fig. 6). When the salinity subsequently was lowered to 1.0%, the content of ¯oridoside decreased drastically, and also the content of iso¯oridoside decreased relative to digeneaside (Fig. 6). In a similar cultivation experiment on G. sordida performed by Ekman and co-workers (Ekman et al. 1991), the ¯oridoside content was found to have increased after
A. Broberg et al.: HR-MAS NMR of major metabolites in a red alga
305
Fig. 5. HR-MAS 1H-13C HMQC spectrum of Gracilariopsis lemaneiformis after 3 weeks in the dark. The 1H-13C correlations of digeneaside (D) are indicated in the spectrum. The spectrum was recorded on 1.5 mg DW of algal sample at 5 kHz. Digeneaside is present as a salt at physiological pH
3 d at 7.0% salinity in the dark. After an additional 5 d at 1.0% salinity in the dark, the ¯oridoside concentration was found to be drastically lower than the original concentration. The results are expected since ¯oridoside is regarded as an important osmotically active compound in many red algal species, except for the Ceramiales (Kirst and Bisson 1979; Ben-Amotz and Avron 1983) and the concentration of ¯oridoside in the alga is thus expected to parallel the osmolarity of the surrounding medium. The roles of iso¯oridoside and digeneaside in G. lemaneiformis are unknown, but since these compounds have been proposed to have osmoregulating functions in the algal orders Bangiales (Kauss 1967) and Ceramiales (Ben-Amotz and Avron 1983), respectively, similar roles could be possible also in G. lemaneiformis. Also the concentration of taurine appears to have increased when the salinity increased from 3.3% to 5.0% (Fig. 6). Dark treatment and upand down-shock of salinity in darkness are two known methods for reducing the content of both starch and ¯oridoside in red algae (Ekman et al. 1991). The upand down-shock experiments on G. lemaneiformis (Fig. 6) were not intended to determine the roles of digeneaside and iso¯oridoside in the studied alga, but merely to demonstrate how HR-MAS NMR can be used to follow physiological responses to changing environmental parameters, and thus we did not replicate
the experiment suciently to obtain statistically reliable results. Care must be taken when analysing these results, since the outcome depends on which compound in the compared spectra is chosen as reference. In this example, the signal of isethionic acid has been used to standardise the three spectra shown in Fig. 6. To obtain more reliable data, the spectra must be standardised relative to a reference compound added to the sample in the magic angle spinning rotor. Analytical and general considerations. Each HR-MAS NMR analysis (Figs. 1±6) of the studied red alga, was performed on 1.5±1.9 mg DW of alga. The ¯oridoside concentration in red algae (except the order Ceramiales) has been reported to be 1±23% of the algal dry weight (Kirst 1980). However, in the majority of the species the concentrations are in the range 2±4% of the algal dry weight (Kirst 1980). The HR-MAS NMR spectrum shown in Fig. 1 was recorded on 1.5 mg DW of alga. This corresponds roughly to 30±60 lg ¯oridoside, demonstrating the sensitivity of the method. These data were collected by 256 scans, taking approximately 15 min. To gain a good signal-to-noise ratio, and thus NMR signals that are easily interpreted, this duration of NMR run is often required, but is ultimately dependent upon the concentration of the analytes of interest.
306
A. Broberg et al.: HR-MAS NMR of major metabolites in a red alga
Fig. 6. HR-MAS 1H NMR spectra of Gracilariopsis lemaneiformis after cultivation at dierent salinities. A, normal cultivation (3.3% salinity). B after 4 d at 5.0% salinity in the dark. C, after 4 d at 5.0% salinity, followed by 2 d at 1.0% salinity in the dark. Spectra A±C were recorded at 10 kHz by 256 scans on 1.9, 1.5 and 1.6 mg DW of algal sample, respectively. Digeneaside and isethionic acid are present as salts at physiological pH
During the experimental time, metabolic event might change the actual concentration of the studied metabolites. However, experimental errors like this can be circumvented by accumulating data for a shorter time and comparing the results from longer and shorter NMR experiments. The two-dimensional experiments shown in Figs. 3± 5, required approximately 5 h (COSY and TOCSY) and 15 h (HMQC), and were recorded on 1.6 mg (COSY and TOCSY) and 1.5 mg DW (HMQC) of algal material. These long experimental times are necessary to obtain good signal-to-noise ratios but are, of course, dependent of the concentrations of the metabolites. However, the possible changes in composition of the analysed samples during the experiments might not be a problem, since these experiments are mostly run to assist in the interpretation of the corresponding 1H NMR spectrum, which does not take a long time at all. Moreover, by using an HR-MAS NMR probehead with gradients, the required experimental times can be reduced for many types of two-dimensional experiments. Conclusions. In our view, HR-MAS NMR can serve as a powerful tool in the analysis of the contents of many dierent kinds of organisms. In principle, it is possible to perform a complete structure analysis using this technique. However, due to overlapping signals, low con-
centrations of the compounds of interest etc., this might prove dicult. More important might be the possibility of obtaining an idea of the composition of the organism of study, before the start of the fractionation procedure. During the puri®cation it is very helpful to know the original composition as a reference to decide what is interesting and what is not. This technique should also prove very useful in many studies of metabolism. When the studied compounds are already known and described, it is often possible to follow the increase and decrease in concentrations of these compounds using HR-MAS NMR, even if the concentrations are very low relative to other compounds in the organism, as long as some signals of interest are separated from the dominating signals. If the interesting signals overlap in the one-dimensional experiment, it might also be possible to follow metabolic events using a two-dimensional technique, e.g. 1H-13C HMQC, with very well resolved cross-peaks. However, the long experimental times for two-dimensional experiments might limit the usefulness of these experiments. This study was supported by grants from the Swedish Research Council for Engineering Sciences, the Swedish Natural Science Research Council, the Swedish Council of Forestry and Agricultural Research and from Carl Tryggers Foundation. Mr Suresh Gohil is acknowledged for performing the FABMS analysis and Mrs Astrid Camitz for performing the cultivation experiments on G. lemaneiformis.
A. Broberg et al.: HR-MAS NMR of major metabolites in a red alga
References Barrow KD, Karsten U, King RJ (1993) Isethionic acid from the marine red alga Ceramium ¯accidum. Phytochemistry 34: 1429± 1430 Barrow KD, Karsten U, King RJ, West JA (1995) Floridoside in the genus Laurencia (Rhodomelaceae, Ceramiales) ± a chemosystematic study. Phycologia 34: 279±283 Ben-Amotz A, Avron M (1983) Accumulation of metabolites by halotolerant algae and its industrial potential. Annu Rev Microbiol 37: 95±119 Cheng LL, Lean CL, Bogdanova A, Wright SC, Jr, Ackerman JL, Brady TJ, Garrido L (1996) Enhanced resolution of proton NMR spectra of malignant lymph nodes using magic-angle spinning. Magn Res Med 36: 653±658 Cheng LL, Ma MJ, Becerra L, Ptak T, Tracey I, Lackner A, GonzaÂlez RG (1997) Quantitative neuropathology by high resolution magic angle spinning proton magnetic resonance spectroscopy. Proc Natl Acad Sci USA (medical sciences) 94: 6408±6413 Ekman P, Yu S, PederseÂn M (1991) Eects of altered salinity, darkness and algal nutrient status on ¯oridoside and starch content, a-galactosidase activity and agar yield of cultivated Gracilaria sordida. Br Phycol J 26: 123±131 Fitch WL, Detre G, Holmes CP (1994) High-resolution 1H NMR in solid-phase organic synthesis. J Org Chem 59: 7955±7956 Holst PB, Nielsen Se, Anthoni U, Bisht KS, Christophersen C, Gupta S, Parmar VS, Nielsen PH, Sahoo DB, Singh A (1994) Isethionate in certain red algae. J Appl Phycol 6: 443±446 Karsten U, Barrow KD, King RJ (1993) Floridoside, L-iso¯oridoside, D-iso¯oridoside in the red alga Porphyra columbina. Plant Physiol 103: 485±491 Karsten U, Barrow KD, Mostaert AS, King RJ, West JA (1994) 13 C- and 1H-NMR studies on digeneaside in the red alga Caloglossa leprieurii. A re-evaluation of its osmotic signi®cance. Plant Physiol Biochem 32: 669±676 Karsten U, Bock C, West JA (1995) 13C-NMR spectroscopy as a tool to study organic osmolytes in the mangrove red algal genera Bostrychia and Stictosiphonia (Ceramiales). Phycol Res 43: 241±247 Kauss H (1967) Iso¯oridoside und Osmoregulation bei Ochromonas malhamensis. Z P¯anzenphysiol 56: 453±465
307 Kirst GO (1980) Low MW carbohydrates and ions in Rhodophyceae: quantitative measurement of ¯oridoside and digeneaside. Phytochemistry 19: 1107±1110 Kirst GO, Bisson MA (1979) Regulation of turgor pressure in marine algae: ions and low-molecular-weight organic compounds. Aust J Plant Physiol 6: 539±556 Krainer E, Stark RE, Naider F, Alagramam K, Becker JM (1994) Direct observation of cell wall glucans in whole cells of Saccharomyces cerevisiae by magic-angle spinning !3C-NMR. Biopolymers 34: 1627±1635 Kremer BP, Kirst GO (1982) Biosynthesis of photosynthates and taxonomy of algae. Z Naturforschung 37 C: 761±771 Kremer BP, Vogl R (1975) Zur chemotaxonomischen Bedeutung des [14C]- Markierungsmusters bei Rhodophyceen. Phytochemistry 14: 1309±1314 Lingnell AÊ, Ekman P, PederseÂn M (1987) Cultivation technique for marine seaweeds allowing controlled and optimized conditions in the laboratory and on a pilotscale. Bot Mar 30: 417± 424 Lignell AÊ, PederseÂn M (1987) Nitrogen metabolism in Gracilaria secundata Harv. Hydrobiologia 151/152: 431±441 Meiboom S, Gill D (1958) Modi®ed spin-echo method for measuring nuclear relaxation times. Rev Sci Instrum 29: 688± 691 Moka D, Vorreuther R, Schicha H, Spraul M, Humpfer E, Lipinski M, Foxall PJD, Nicholson JK, Lindon JC (1997) Magic angle spinning proton nuclear magnetic resonance spectroscopic analysis of intact kidney tissue samples. Anal Commun 34: 107±109 Ni QW, Eads TM (1993a) Liquid-phase composition of intact fruit tissue measured by high-resolution proton NMR. J Agric Food Chem 41: 1026±1034 Ni QX, Eads TM (1993b) Analysis by proton NMR of changes in liquid-phase and solid-phase components during ripening of banana. J Agric food Chem 41: 1035±1040 Sayer BG, Preston CM (1996) A carbon-13 magic angle spinning nuclear magnetic resonance study of the germination of conifer seeds. Seed Sci Technol 24: 321±329 Yu S, PederseÂn M (1990) The a-galactosidase of Gracilaria tenuistipitata and G. sordida (Gracilariales, Rhodophyta). Phycologia 29: 454±460