Importance of lactic acid bacteria in Asian fermented foods.

Lactic acid bacteria play important roles in various fermented foods in Asia. Besides being the main component in kimchi and other fermented foods, they are used to preserve edible food materials through fermentation of other raw-materials such as rice wine/beer, rice cakes, and fish by producing organic acids to control putrefactive microorganisms and pathogens. These bacteria also provide a selective environment favoring fermentative microorganisms and produce desirable flavors in various fermented foods. This paper discusses the role of lactic acid bacteria in various non-dairy fermented food products in Asia and their nutritional and physiological functions in the Asian diet.

Geographical and climatic dependencies of green tea (Camellia sinensis) metabolites: a (1)H NMR-based metabolomics study.

The effects of climatic conditions on green tea metabolites in three different growing areas of Jeju Island, South Korea, were investigated through global metabolite profiling by (1)H nuclear magnetic resonance (NMR) spectroscopy. Pattern recognition methods, such as principal component analysis (PCA) and orthogonal projection on latent structure-discriminant analysis (OPLS-DA), revealed clear discriminations of green teas from the three different growing areas. Variations of theanine, isoleucine, leucine, valine, alanine, threonine, glutamine, quinic acid, glucose, epicatechin (EC), epigallocatechin (EGC), epigallocatechin-3-gallate (EGCG), and caffeine levels were responsible for the discriminations. Green teas grown in an area with high temperature, long sun exposure time, and high rainfall had higher levels of theanine but lower levels of isoleucine, leucine, valine, alanine, EC, EGC, EGCG, and caffeine than those grown in areas with relatively low temperature, short sun exposure time, and low rainfall. These results indicate that high temperature, long sun exposure, and high preciptation stimulate theanine synthesis in green tea during the spring season. This study highlights how metabolomics coupled with multivariate statistical analysis can illuminate the metabolic characteristics of green tea associated with climatic variables, thereby allowing for the assessment of quality strategy in green tea production.

Metabolomic assessment of fermentative capability of soybean starter treated with high pressure.

Meju, a brick of dried fermented soybean naturally inoculated with microorganisms, is a starter used for producing traditional Korean fermented soybean products such as soybean paste (doenjang) and soy sauce (ganjang). In order to reduce aging time during production of soybean paste and soy sauce, high pressure (HP) treatment was applied to the meju starter at 500 MPa of pressure for 10 min at 15 degrees C. Fermentative behaviors of normal and HP-treated meju were assessed and compared through physicochemical and (1)H NMR-based metabolomic analysis. All mejues were incubated for 3 weeks at 30 degrees C. At 1 week of incubation, total bacterial population decreased mainly due to a reduction of water content by spontaneous evaporation during the incubation period. As the incubation time increased, glutamate, proline, betain, choline, and phosphocholine levels increased in both normal and HP-treated mejues, indicating that microorganisms in the mejues synthesize these metabolites to endure intracellular hyperosmotic stress induced by the reduction in water content. Through 3 weeks of incubation, the amino-type nitrogen contents and neutral protease activities in HP-treated meju were significantly higher (p < 0.05) than in normal meju, even though total bacterial content in HP-treated meju was 2 or 3 times lower. Moreover, marked increases in glycerol, acetate, tyrosine, and choline levels were observed in HP-treated meju compared to normal meju. In particular, higher levels of tyrosine in HP-treated meju were consistent with the increased neutral protease activities compared to normal meju, indicating an improvement in enzyme stability with HP treatment. These findings highlight a new or better understanding of the influence of the HP or physical treatments on fermentative products in food processing, such as those associated with soybean paste and soy sauce, regarding metabolic behaviors in fermentative starter induced by HP treatment.

Metabolomics reveals alterations in both primary and secondary metabolites by wine bacteria.

Lactic acid bacteria (LAB) were isolated from Korean Meoru (Vitis coigneties) wine and identified as Lactobacillus plantarum meoru0711 (KACC 91436C). The fermentative behavior and metabolic effects of L. plantarum during malolactic fermentation (MLF) were compared with those of the commercial Oenococcus oeni strain through 1H NMR- and GC-based metabolic profiling. Twenty-two primary metabolites of amino acids, carbohydrates, and organic acids, and 55 secondary metabolites of volatile compounds were identified in wines by 1H nuclear magnetic resonance (NMR) spectroscopy and gas chromatography (GC), respectively. Principal component analysis (PCA) revealed that malolactic (ML)-fermented and non-ML-fermented wines, and wines ML-fermented with O. oeni and L. plantarum were clearly differentiated. Both the primary and secondary metabolites were responsible for these differentiations. Compared to non-MLF wines, MLF wines were characterized by increased levels of primary metabolites such as lactic acid, phenylalanine, uracil, ornithine, alanine, threonine, leucine, isoleucine, and valine with decreased levels of monosaccharides, glycerol, malic, and citric acids. In addition, higher levels of secondary metabolites such as butanal, ethyl isobutylate, isobutanol, isoamyl acetate, 2-butanoate ethyl ester, isoamyl alcohol, ethyl hexanoate, glycine, acetic acid, and benzaldehyde characterized the MLF wine. Higher levels of primary metabolites such as tyrosine, monosaccharides, glycerol, alanine, 2,3-butanediol, valine, and leucine, and of secondary metabolites such as propyl acetate, isobutanol, isoamyl acetate, 1-butanol, ethyl hexanoate, prenyl alcohol, glycine, 2-hexen-1-ol, ethyl octanoate, acetic acid, benzaldehyde, and butyric, together with lower levels of lactic acid, were observed in the wines fermented by L. plantarum compared with those by O. oeni. This present study demonstrates that different genera of LAB affect both the primary and second metabolites in wine. Moreover, metabolomics with multivariate statistical analysis provide insight into wine fermentation.

Metabolomic insight into soy sauce through (1)H NMR spectroscopy.

Soy sauce, a well-known seasoning in Asia and throughout the world, consists of many metabolites that are produced during fermentation or aging and that have various health benefits. However, their comprehensive assessment has been limited due to targeted or instrumentally specific analysis. This paper presents for the first time a metabolic characterization of soy sauce, especially that aged up to 12 years, to obtain a global understanding of the metabolic variations through (1)H NMR spectroscopy coupled with multivariate pattern recognition techniques. Elevated amino acids and organic acids and the consumption of carbohydrate were associated with continuous involvement of microflora in aging for 12 years. In particular, continuous increases in the levels of betaine were found during aging for up to 12 years, demonstrating that microbial- or enzyme-related metabolites were also coupled with osmotolerant or halophilic bacteria present during aging. This work provides global insights into soy sauce through a (1)H NMR-based metabolomic approach that enhances the current understanding of the holistic metabolome and allows assessment of soy sauce quality.

Evidence of vintage effects on grape wines using 1H NMR-based metabolomic study.

The chemical composition of grape wines varies with grape variety, environmental factors of climate and soil, and bacterial strains, which can each affect the wine quality. Using (1)H NMR analysis coupled with multivariate statistical data sets, we investigated the effects of grape vintage on metabolic profiles of wine and the relationship between wine metabolites and meteorological data. Principal component analysis (PCA) showed a clear differentiation between Meoru wines that were vinified with the same yeast strain and Meoru grapes harvested from the same vineyard but with a different vintage. The metabolites contributing to the differentiation were identified as 2,3-butandiol, lactic acid, alanine, proline, gamma-aminobutyric acid (GABA), choline, and polyphenols, by complementary PCA loading plot. Markedly higher levels of proline, lactic acid and polyphenols were observed in the 2006 vintage wines compared to those of 2007 vintage, showing excellent agreement with the meteorological data that the sun-exposed time and rainfall in 2006 were approximately two times more and four times less, respectively, than those in 2007. These results revealed the important role of climate during ripening period in the chemical compositions of the grape. This study highlights the reliability of NMR-based metabolomic data by integration with meteorological data in characterizing wine or grape.

Metabolomic characterization of malolactic fermentation and fermentative behaviors of wine yeasts in grape wine.

Wine contains a number of metabolites that are produced during alcoholic and malolactic fermentations (MLF) or aging, which are important compounds for determining wine quality. This study investigated changes in metabolites in wines to characterize malolactic fermentation (MLF) and to assess fermentative behaviors of wine yeast strains using (1)H nuclear magnetic resonance (NMR) spectroscopy coupled with multivariate statistics. Principal component analysis (PCA) showed clear differentiation between non- and induced-malolactic fermented wines by wine lactic acid bacteria (LAB) and between wines fermented with various wine yeast strains. Metabolites such as glycerol, lactate, 2,3-butanediol, succinate, leucine, isoleucine, alanine, valine, proline, choline, gamma-aminobutyric acid (GABA), and polyphenols contributed to the differentiations. Decreased levels of malate and citrate along with increased levels of lactate were the metabolites most responsible for the differentiation of induced-MLF wines from non-MLF wines. In particular, high succinate levels provided evidence of an inhibitory effect of Saccharomyces bayanus against spontaneous MLF. Furthermore, dependence of metabolites on wine yeast strains was observed, demonstrating their different fermentative behaviors. This study demonstrates that wine fermentation by yeast and LAB can be characterized through global and multivariate statistical analysis of (1)H NMR spectral data.

Characterization of fermentative behaviors of lactic acid bacteria in grape wines through 1H NMR- and GC-based metabolic profiling.

The effects of five commercial Oenococcus oeni strains (MCW, Enoferm alpha, Wyeast, Vinibacti111, and Vinibacti222) on fermentative behaviors, and variations of metabolites in Meoru ( Vitis coigneties ) wines during malolactic fermentation (MLF) were investigated by metabolomic analysis of (1)H NMR and GC data sets. In the development of MLF with various O. oeni strains, the fastest conversions of malic acid to lactic acid occurred in wines fermented with Enoferm alpha and Vinibacti111 strains. Seventeen primary metabolites and 65 secondary metabolites of volatile compounds in the wines were identified by (1)H NMR spectroscopy and GC-MS, respectively. In pattern recognition models of principal component analysis (PCA) and orthogonal projection to latent structures discriminant analysis (OPLS-DA), significant differentiations between wines with O. oeni strains were identified by the secondary metabolites rather than by the primary metabolites, showing the effects of O. oeni strains only on the secondary metabolites. Twelve volatile compounds, 2-phenylethanol, isoamyl alcohol, 2-butanol, ethyl octanoate, ethyl hexanoate, hexadecanoic acid, diethyl succinate, butyl butyrate, octanoic acid, 9-hexadecanoic acid, isobutyric acid, and 2-ethyl-1-hexanol, contributed to the differentiation of wines according to O. oeni strain, including spontaneous MLF. This study demonstrates that O. oeni strains affect the secondary metabolites, which are easily identified through multivariate statistical analysis of GC-MS data set.

Metabolomic studies on geographical grapes and their wines using 1H NMR analysis coupled with multivariate statistics.

Environmental vineyard conditions can affect the chemical composition or metabolites of grapes and their wines. Grapes grown in three different regions of South Korea were collected and separated into pulp, skin, and seed. The grapes were also vinified after crushing. (1)H NMR spectroscopy with pattern recognition (PR) methods was used to investigate the metabolic differences in pulp, skin, seed, and wines from the different regions. Discriminatory compounds among the grapes were Na, Ca, K, malate, citrate, threonine, alanine, proline, and trigonelline according to PR methods of principal component analysis (PCA) or partial least-squares discriminant analysis (PLS-DA). Grapes grown in regions with high sun exposure and low rainfall showed higher levels of sugar, proline, Na, and Ca together with lower levels of malate, citrate, alanine, threonine, and trigonelline than those grown in regions with relatively low sun exposure and high rainfall. Environmental effects were also observed in the complementary wines. This study demonstrates that (1)H NMR-based metabolomics coupled with multivariate statistical data sets can be useful for determining grape and wine quality.

(1)H NMR-based metabolomic approach for understanding the fermentation behaviors of wine yeast strains.

(1)H NMR spectroscopy coupled with multivariate statistical analysis was used for the first time to investigate metabolic changes in musts during alcoholic fermentation and wines during aging. Three Saccharomyces cerevisiae yeast strains (RC-212, KIV-1116, and KUBY-501) were also evaluated for their impacts on the metabolic changes in must and wine. Pattern recognition (PR) methods, including PCA, PLS-DA, and OPLS-DA scores plots, showed clear differences for metabolites among musts or wines for each fermentation stage up to 6 months. Metabolites responsible for the differentiation were identified as valine, 2,3-butanediol (2,3-BD), pyruvate, succinate, proline, citrate, glycerol, malate, tartarate, glucose, N-methylnicotinic acid (NMNA), and polyphenol compounds. PCA scores plots showed continuous movements away from days 1 to 8 in all musts for all yeast strains, indicating continuous and active fermentation. During alcoholic fermentation, the highest levels of 2,3-BD, succinate, and glycerol were found in musts with the KIV-1116 strain, which showed the fastest fermentation or highest fermentative activity of the three strains, whereas the KUBY-501 strain showed the slowest fermentative activity. This study highlights the applicability of NMR-based metabolomics for monitoring wine fermentation and evaluating the fermentative characteristics of yeast strains.

1H nuclear magnetic resonance-based metabolomic characterization of wines by grape varieties and production areas.

(1)H NMR spectroscopy was used to investigate the metabolic differences in wines produced from different grape varieties and different regions. A significant separation among wines from Campbell Early, Cabernet Sauvignon, and Shiraz grapes was observed using principal component analysis (PCA) and partial least squares-discriminant analysis (PLS-DA). The metabolites contributing to the separation were assigned to be 2,3-butanediol, lactate, acetate, proline, succinate, malate, glycerol, tartarate, glucose, and phenolic compounds by PCA and PLS-DA loading plots. Wines produced from Cabernet Sauvignon grapes harvested in the continental areas of Australia, France, and California were also separated. PLS-DA loading plots revealed that the level of proline in Californian Cabernet Sauvignon wines was higher than that in Australian and French Cabernet Sauvignon, Australian Shiraz, and Korean Campbell Early wines, showing that the chemical composition of the grape berries varies with the variety and growing area. This study highlights the applicability of NMR-based metabolomics with multivariate statistical data sets in determining wine quality and product origin.

Tastes and structures of bitter peptide, asparagine-alanine-leucine-proline-glutamate, and its synthetic analogues.

Asn-Ala-Leu-Pro-Glu (NALPE) is a strong bitter peptide with a minimum response threshold (MRT) of 0.074 mM. To elucidate the relationship of spatial structure and bitterness on peptides, NALPE and its analogues, NALPW, NALPS, NALPL, NALPP, NALPD, and NALPR, were synthesized and sensorially evaluated. Structural analysis using computer simulation for each peptide revealed that the presence of a polar group and hydrophobic bitter amino acids, the composition of hydrophobic regions, the spatial orientation of the polar group and hydrophobic regions, and the proximity between polar groups and hydrophobic regions faced within the same plane space may be the major determinants for the taste type and intensity of peptide bitterness.

Identification of LDL-receptor transcription stimulating peptides from soybean hydrolysate in human hepatocytes.

Soybean protein and its hydrolysate have been reported to have cholesterol-lowering property, but the responsible components are still largely unknown. In previous study, we found that soybean protein hydrolysate (SPH) prepared with the protease from Bacillus amyloliquefaciens FSE-68, strongly stimulates transcription of low density lipoprotein receptor (LDL-R). To identify LDL-R transcription stimulating peptides in human hepatocytes, the SPH was fractionated with gel permeation chromatograpy and the active fraction was further separated by using reverse-phase chromatography. Several peptides in the most active fraction were identified by LC/MS and MS/MS analysis. LDL-R transcription stimulating peptides were synthesized on the basis of identified sequences, and their effect on LDL-R transcription was tested in vitro. Among the synthesized peptides, Phe-Val-Val-Asn-Ala-Thr-Ser-Asn (FVVNATSN) showed the strongest activity, and LDL-R transcription of hepatic cells was increased to 248.8% (compared to 100% of untreated control) by FVVNATSN at a concentration of 100 microM. This study provides direct evidence that peptides derived from soybean protein can influence LDL-R transcription in hepatocytes.

New approach for characterization of gelatin biopolymer films using proton behavior determined by low field 1H NMR spectrometry.

The behavior of protons in biopolymer films (BFs) formed with gelatin, water, and glycerol was investigated at various relative humidities (RHs) and concentrations of glycerol using a low field 1H NMR spectrometer. At a RH of approximately 0%, the distributed spin-spin relaxation times (T2) of protons in BFs showed two components: a rapidly relaxing proton with the shortest T2 derived from protons in the rigid backbone of the gelatin polymer such as CH1-, CH2-, and CH3-, and a slowly relaxing component with longer T2 from protons of the functional groups in amino acid residues in gelatin such as -OH, -COOH, and -NH3. These two components are referred to as nonexchangeable (T2N) and exchangeable protons (T2E), respectively, indicating the different mobility of the protons. The T2E increased as RH increased indicating the increase in relative mobility of protons due to the larger free volume in the BF matrix. Above a RH of 33%, the slowest relaxing component was found in all BFs and referred to as hydration-water protons (T2W) with the highest relative mobility of all protons in the films. It suggests that the free volume in BFs can be formed above a RH of 33% in the absence of glycerol. The behaviors of T2N, T2E, and T2W reveal the formation of free volume in the BF matrix associated with the presence of plasticizers (water and glycerol). The T2 behavior in BFs is consistent with the behavior of spin-lattice relaxation (T1). Our result is the first attempt to characterize using low field 1H NMR technology how all protons in a film matrix behave and to develop correlations between proton mobility and free volume in protein-based BFs plasticized with water and glycerol.

Cholesterol lowering mechanism of soybean protein hydrolysate.

Numerous attempts have been made to find the mechanism and component of the cholesterol lowering activity of soybean. In this study, it was proved that the peptides in soybean protein hydrolysate (SPH) made by certain proteases have a hypocholesterolemic effect. Among the mechanisms suggested, that is, blockage of bile acid and/or cholesterol absorption, inhibition of cholesterol synthesis, and stimulation of low-density lipoprotein receptor (LDL-R) transcription, SPH appeared to stimulate LDL-R transcription. When Hep T9A4 cells were incubated with soy protein hydrolysates by using the proteases from Bacillus amyloliquefaciens FSE-68, LDL-R transcription was strongly stimulated, but the other mechanisms were not affected. Among the six types of SPH, F1-15, hydrolyzed with the neutral protease to a degree of hydrolysis (DH) of 15%, showed the highest LDL-R transcription. The fractions of molecular weight of 200 and 3000 Da showed LDL-R transcription stimulating activity. The bioactivity is due to soybean peptides because the ethanol extract of soybean protein which contains isoflavones does not stimulate LDL-R transcription. In conclusion, dietary upregulation of LDL-R transcription by soybean may be consequent to an enhanced catabolism or a reduced synthesis of intracellular cholesterol. Therefore, we suggest that soy peptides can effectively stimulate LDL-R transcription in the human liver cell line and reduce blood cholesterol level.

Self-diffusion coefficient of water in tofu determined by pulsed field gradient nuclear magnetic resonance.

The self-diffusion coefficient of water in soybean protein dispersion and tofu was measured by pulsed field gradient (PFG) NMR. A soy protein isolate (SPI) dispersion (6 and 12%, w/w) in water, calcium cross-linked precipitate, and tofu were used for comparison. The self-diffusion coefficient of water (D) in the SPI dispersion, 2.23 x 10(-9) m2/s, was estimated lower than that of free water, 2.6 x 10(-9) m2/s at 25 degrees C, and decreased as the SPI concentration increased. It further decreased by the addition of calcium chloride, reflecting the obstruction effect induced by the precipitates in addition to the hydration and hydrodynamic interaction in the protein dispersion. The two water regions in tofu were interpreted by the two-site Kärger model: D1 and D2 of soft tofu were 2.26 (+/-0.11) x 10(-9) and 6.84 (+/-0.34) x 10(-11) m2/s, respectively. The relative amount of proton (water) was p1 = 0.98 and p2 = 0.02 at 100 ms of diffusion time. The self-diffusion coefficients of water decreased in pressed tofu, and their relative amounts of water changed to p1 = 0.93 and p2 = 0.07. It was suggested that D1 corresponded to obstructed water in the network structure and D2 corresponded to hydrated water on the surface layer of pores formed in the protein network of tofu. The pore sizes estimated from the diffusion length of obstructed water were 21.3 microm in soft tofu and 20.8 microm in pressed tofu. The removal of fat from pressed tofu led to a decrease in D2 from 6.26 (+/-0.31) x 10(-11) to 3.53 (+/-0.18) x 10(-11) m2/s, and the relative amount of hydrated water increased from 0.07 to 0.14, which indicated hydrophobic hydration.

Crystal structure of CD14 and its implications for lipopolysaccharide signaling.

Lipopolysaccharide, the endotoxin of Gram-negative bacteria, induces extensive immune responses that can lead to fatal septic shock syndrome. The core receptors recognizing lipopolysaccharide are CD14, TLR4, and MD-2. CD14 binds to lipopolysaccharide and presents it to the TLR4/MD-2 complex, which initiates intracellular signaling. In addition to lipopolysaccharide, CD14 is capable of recognizing a few other microbial and cellular products. Here, we present the first crystal structure of CD14 to 2.5 angstroms resolution. A large hydrophobic pocket was found on the NH2-terminal side of the horseshoe-like structure. Previously identified regions involved in lipopolysaccharide binding map to the rim and bottom of the pocket indicating that the pocket is the main component of the lipopolysaccharide-binding site. Mutations that interfere with lipopolysaccharide signaling but not with lipopolysaccharide binding are also clustered in a separate area near the pocket. Ligand diversity of CD14 could be explained by the generous size of the pocket, the considerable flexibility of the rim of the pocket, and the multiplicity of grooves available for ligand binding.

Effects of irradiated phytic acid on antioxidation and color stability in meat models.

Lipid oxidation and color stability of meats treated with irradiated phytic acid were investigated during storage for 2 weeks at 4 degrees C. The phytic acid in deionized distilled water (DDW) was degraded by irradiation at 10 and 20 kGy, and the irradiated phytic acid showed a strong antiradical activity. For measuring the antioxidant effects of irradiated phytic acid in food models, beef and pork were prepared with DDW (control), irradiated (10 and 20 kGy) or non-irradiated phytic acid, and ascorbic acid as a model system. Irradiated phytic acid significantly inhibited the lipid oxidation in meats compared to the control and ascorbic acid treated samples during storage (P < 0.05). The redness of the meats treated with phytic acid had a higher value than did the control and ascorbic acid treated samples, but a significant difference was not observed in the samples treated with phytic acid regardless of irradiation treatment. Irradiated phytic acid was also effective in inhibiting the loss of heme iron and metmyoglobin formation during storage. Results indicated that irradiation might be helpful for improving the antioxidant activity of phytic acid in meats.

Rheological properties of erythrocytes from male hypercholesterolemia.

Diet and general health status has close relation to the flow behavior of blood, which influences the circulation of the blood in the body. In this study, we have compared the rheological properties of erythrocyte, plasma and whole blood from high-cholesterol male subjects with healthy male subjects. Intravenous blood was taken from healthy males (n=10) and males with high cholesterol (n=14). Basic health profile, BMI, hematological count and lipid profile (total cholesterol, LDL, HDL and triglyceride) of the blood were determined. Viscosity and shear rate dependent flow behavior of the subjects blood were measured by cone and plate rheometer, and permeability of erythrocytes by pulsed field gradient NMR. Using the microchannel flow analyzer (MC-FAN), the microcirculation of erythrocyte and plasma were investigated. Our data showed a difference in viscosity and consistency index of the whole blood, and permeability (P<0.05) of erythrocytes between the two groups. Also, the time taken for the flow of erythrocyte and plasma through the MC-FAN was slower for the high-cholesterol group. Correlation study showed that consistency index of the blood is closely related to the level of LDL (P<0.05), and total cholesterol, HDL and LDL (P<0.01) highly correlated with the microcirculation of erythrocyte and plasma. A negative correlation (P<0.05) was found between total cholesterol, HDL and LDL, and permeability of erythrocytes. It is concluded that high level of cholesterol, LDL and HDL in vivo alter the morphology and flow behavior of blood cells that can subsequently increase the risk of impairing physical function and microcirculation.

Water self-diffusion in Chlorella sp. studied by pulse field gradient NMR.

The water self-diffusion behavior in chlorella water suspension was investigated by pulsed field gradient NMR technique. Three types of water was determined, which differs according to the self-diffusion coefficients; bulk water, extracellular and intracellular water. Intracellular and extracellular water self-diffusion were restricted, and the sizes of restriction regions were 3.4 microm and 17 microm, respectively. The water molecular exchange process between these three diffusion regions was investigated. The residence time and exchange rate constant for chlorella cells were obtained. The cell wall permeability determined from the rate constant as 3 x 10(-6) m/s agreed with the permeability 10(-6) m/s obtained from time dependence of intracellular water self-diffusion coefficient. The structural cluster model of chlorella cell is estimated to describe the extracellular water self-diffusion in chlorella water suspension.