ABSTRACT

Vitamin B₁₂ contents in the edible cyanobacterium Nostoc flagelliforme, also known as hair vegetable, were assayed using a microbiological method. We detected high vitamin B₁₂ contents in samples of naturally grown cells (109.2 ± 18.5 μg/100g dry weight) and cultured cells (120.2 ± 53.6 μg/100g dry weight). However, commercially available hair vegetable samples, which comprised fake substitutes and Nostoc, had variable contents (4.8 - 101.6 μg/100g dry weight) because concomitant fake items contain very low vitamin B₁₂ contents. To evaluate whether natural and cultured N. flagelliforme samples contained vitamin B₁₂ or pseudovitamin B₁₂, corrinoid compounds were purified and identified as pseudovitamin B₁₂ (approximately 72%) and vitamin B₁₂ (approximately 28%) using silica gel 60 TLC bioautography and LC/MS. The results suggested that N. flagelliforme contains substantial amounts of pseudovitamin B₁₂, which is inactive in humans.

1. Introduction

Nostoc flagelliforme is an edible cyanobacterium, which grows naturally in some semidesert regions of China and Mongolia. When dried, the cyanobacterium resembles black hair and hence the name hair vegetable (“Facai” in Chinese), which is one of the most expensive ingredients in Chinese cuisine [1]. At present, fake items and mixtures of pure N. flagelliforme with fake substitutes (approximately 90%) are flooding the market [1].

N. flagelliforme contains many nutrients [2], including a novel acidic polysaccharide (nostoflan) that has potent antiviral activity [3]. Takenaka et al., [4] demonstrated the oral acute and subacute safety of dried N. flagelliforme in rats. Therefore, N. flagelliforme is also suitable for pharmaceutical use. Several studies [5,6] have reported that most of the corrinoids found in certain edible cyanobacteria may not be bioavailable in mammals. Watanabe et al. [7] also demonstrated that pseudovitamin B₁₂ (adeninylcyanocobamide or pseudo B₁₂; Figure 1),

which is inactive in humans, is the predominant corrinoid in the edible cyanobacteria used as a health food by humans. N. flagelliforme, hair vegetable, is already used as health food, but there is no information on about B₁₂ contents in pure N. flagelliforme and commercially available hair vegetable, or whether the corrinoids are authentic B₁₂ or inactive corrinoids.

In the present study, we characterized corrinoid compounds from N. flagelliforme sources, including naturally grown samples, cultured samples, and commercially available hair vegetable samples.

2. Materials and Methods

2.1. Materials

Authentic B₁₂ was obtained from Sigma (St Louis, Missouri, USA). Silica gel 60 thin-layer chromatography (TLC) aluminum sheets were obtained from Merck (Darmstadt, Germany). All other reagents were highgrade commercially available reagents. N. flagelliforme Born. et Flah. was harvested from Alxa, Inner Mongolia, China, during the summer of 1996. After washing in water, the cyanobacterium was dried in sun and used for the analyses. It was also aseptically cultured in a Nostoc-N liquid medium (K₂HPO₄, 40 mg/L; MgSO₄∙7H₂O, 70 mg/L; Na₂SiO₃∙7H₂O 60 mg/L; CaCl₂∙2H₂O 36 mg/L; FeSO₄∙7H₂O, 4.8 mg/L, EDTA 2Na, 1 mg/L; H₃BO₃, 2.86 mg/L; MnCl₂∙4H₂O, 1.8 mg/L; ZnSO₄∙7H₂O, 222 μg/L; Na₂MoO₄∙2H₂O, 390 μg/L; CuSO₄∙5H₂O, 80 μg/L; and Co(NO₃)₂∙6H₂O, 50 μg/L; at pH 7.5) at 20 - 25˚C with aeration under illumination (40 μmol/m²/s). Commercially available hair vegetable samples were purchased from the markets in Japan.

2.2. Samples of N. flagelliforme

Samples A-E were naturally grown samples, F-J were cultured samples, and K-N were commercially hair vegetable samples.

2.3. Extraction and Assay of Corrinoids from N. flagelliforme Samples

Dried N. flagelliforme samples (five different lots of naturally grown and cultured samples and four commercially available hair vegetable samples) were used for the assays. First, 0.5 g of each sample was suspended in 40 mL of distilled water and homogenized with an ultrasonic disruptor UD-200 (Tomy, Tokyo, Japan). Total corrinoids were extracted after boiling at pH 4.8 in the presence of 4.0 × 10⁻⁴% KCN and determined using the Lactobacillus delbrueckii ATCC 7830 microbiological assay method, according to the method described in the Standard Tables of Food Composition in Japan. L. delbrueckii ATCC 7830 can utilize deoxyribosides, deoxyribonucleotides (known as alkali resistant factor), and B₁₂. Thus, accurate B₁₂ contents were calculated by subtracting the results for alkali resistant factor from those of total B₁₂ [8].

2.4. Bioautography of Corrinoid Compounds Using Vitamin B₁₂-Dependent Escherichia coli 215

Bioautography of corrinoid compounds was performed as previously described [9]. B₁₂ extracts (20 mL) prepared as mentioned above were partially purified and concentrated using a Sep-Pak Plus^® C18 cartridge (Waters Corp., Milfora, USA), which was washed with 5 mL of 75% (v/v) ethanol and equilibrated with 5 mL of distilled water. The C18 cartridge was washed with 5 mL of distilled water, and B₁₂ compounds were eluted using 2 mL of 75% (v/v) ethanol. The eluate was evaporated in a centrifugal concentrator (Integrated SpeedVac^® System ISS110; Savant Instruments Inv., NY, USA). The residual fraction was dissolved in 5.0 mL of distilled water. Concentrated B₁₂ extracts (1 μL) and authentic and pseudo B₁₂ (each 50 μg/L) were spotted onto the silica gel 60 TLC sheet and developed in the dark using 2-propanol/ NH₄OH (28%)/water (7:1:2 v/v) at room temperature (25˚C). After drying the TLC sheet, it was overlaid with agar containing basal medium and precultured E. coli 215, and incubated at 37˚C for 20 h. The gel plate was then sprayed with methanol solution containing 2,3,5- triphenyltetrazolium salt, and B₁₂ compounds were visualized as red, indicating E. coli growth.

2.5. Liquid Chromatography-Electrospray Ionization/Multistage Mass Spectrometry (LC/ESI-MS/MS) Analysis

Each extract (40 mL) was partially purified and concentrated using a Sep-Pak^® Plus C18 cartridge (Waters Corp) as described above. The eluate was evaporated in a centrifugal concentrator (Integrated Speed VacR System ISS110), and the residual fraction was dissolved in 5.0 mL of distilled water. The purified extract was loaded onto an immunoaffinity column [EASI-EXTRACT^® Vitamin B₁₂ Immunoaffinity Column (P80) R-Biopharm AG, Darmstadt, Germany], and the corrinoids were purified according to the manufacturer’s recommended protocol. Nostoc corrinoids, authentic pseudo B₁₂, and B₁₂ were dissolved in 0.1% (v/v) acetic acid and filtered using a Nanosep MF centrifuge device (0.4 μm, Pall Corp., Tokyo, JAPAN) to separate small particles. We analyzed an aliquot (2 μL) of the filtrate using a LCMS-IT-TOF coupled with an Ultra-Fast LC system (Shimadzu, Kyoto, JAPAN). Each purified corrinoid was injected into an Inert Sustain column (3 μm, 2.0 × 100 mm, GL Science, Tokyo, JAPAN) and equilibrated with 85% solvent A [0.1% (v/v) acetic acid)] and 15% solvent B (100% methanol) at 40˚C. Corrinoid compounds were eluted using a linear gradient of methanol (15% solvent B for 0 - 5 min, increasing the concentration from 15% to 90% solvent B for 5 - 11 min, and decreasing the concentration from 90% to 15% solvent B for 11 - 15 min). The flow rate was 0.2 mL/min. ESI conditions were determined by injecting authentic pseudo B₁₂ or B₁₂ into the MS detector to determine the optimum parameters for detecting the parent B₁₂ compound and daughter ions. ESI-MS was operated in the positive ion mode. Argon was used as the collision gas. Pseudo B₁₂ (m/z 672.777) and B₁₂ (m/z 678.292) as [M+2H]²⁺ were confirmed by comparing the observed molecular ions and the retention times.

2.6. Analytical High Performance Liquid Chromatography (HPLC)

Each immunoaffinity-purified B₁₂ fraction (10 μL) was analyzed with a reversed-phase HPLC column (WakosilII 5C18RS, 4.6 × 150 mm; 5 μm particle size; Wako Pure Chemical Industries, Osaka, Japan). Corrinoids were isocratically eluted with 20% (v/v) methanol solution containing 1% (v/v) acetic acid at 40˚C and monitored by measuring the absorbance at 361 nm. The flow rate was 1 mL/min. Retention times of authentic B₁₂ and pseudo B₁₂ were 8.6 min and 10.7 min, respectively. The relative content ratio of B₁₂ and pseudo B₁₂ in various N. flagelliforme samples was calculated on the basis of peak areas with identical retention times of B₁₂ and pseudo B₁₂.

2.7. Evaluation of True and Fake N. flagelliforme

Because fake materials generally contain starch [2], commercially available hair vegetable samples were tested using the iodine-starch reaction. The dried samples (0.1 g) were added to 10 mL of distilled water and boiled for 30 min. The treated samples were cooled to a room temperature and centrifuged at 10,000 × g for 10 min at 25˚C. Each supernatant solution (0.2 mL) was added to 1.4 mL of distilled water and treated with 0.4 mL of 25% Lugol solution (MP Biomedicals, LLC, Ohio, USA). The solution was allowed to stand for 30 min, and absorbance was measured at 600 nm. Microscopic analysis was performed using a BH-2 type microscope (Olympus Corp., Tokyo, Japan) with a digital camera QV-200 (Casio Computer Co. Ltd, Tokyo, Japan), as previously described [2].

3. Results and Discussion

3.1. Vitamin B₁₂ Contents

B₁₂ contents were analyzed in various sources of N. flagelliforme i.e., naturally grown samples, cultured samples, and commercially available hair vegetable samples, using the L. delbrueckii ATCC 7830 microbiological assay method (Table 1). High B₁₂ contents were detected in naturally grown cells (109.2 ± 18.5 μg/100g dry weight) and cultured cells (120.2 ± 53.6 μg/100g dry weight). However, commercially available hair vegetable samples had very variable and lower B₁₂ contents [45.1 ± 40.6 (range, 4.8 - 101.6) μg/100g dry weight]. B₁₂ contents of natural and cultured cells were similar to those of other edible cyanobacteria, i.e., Spirulina sp. (127.2 - 244.3 μg/100g dry weight) [10], Suizenji-nori (Aphanothece sacrum, 143.8 μg/100g dry weight) [11], and Ishikurage (Nostoc commune, 98.8 μg/100g dry weight) [12].

3.2. E. coli 215 Bioautography Analysis

Corrinoids found in all Nostoc samples were analyzed using the E. coli 215 bioautogram after separation by silica gel 60 TLC (Figure 2). Corrinoids found in all Nostoc samples and the commercially available hair vegetable sample K were separated to yield two spots, the Rf values of which were identical to those of authentic pseudo B₁₂ and B₁₂, respectively. No or faint spots were obtained with commercially available hair vegetable samples L-N because of their lower B₁₂ contents.

3.3. LC/ESI-MS/MS Analysis

N. flagelliforme extracts were purified using a B₁₂ immunoaffinity column and analyzed by LC/ESI-MS/MS (Figure 3). Authentic B₁₂ and pseudo B₁₂ were eluted as

peaks with retention times of 7.55 and 7.42 min, respectively. Mass spectrum of authentic B₁₂ indicated that a doubly-charged ion with an m/z of 678.2897 [M+2H]²⁺ was prominent (Figures 3(a) and (b)). The exact mass calculated from its formula (C₆₃H₈₈CoN₁₄O₁₄P) was 1354.5674, and the isotope distribution data showed that B₁₂ was the major divalent ion under the LC/ESI-MS conditions. For authentic pseudo B₁₂ with an exact mass of 1343.5375 (C₅₉H₈₃CoN₁₇O₁₄P), a doubly-charged ion with an m/z of 672.77861 [M+2H]²⁺ was prominent (Figures 3(d) and (e)). The MS/MS spectra of B₁₂ and pseudo B₁₂ indicated that the dominant ions at m/z 359.0982 and m/z 348.0684, respectively, were attributable to the nucleotide moiety of each corrinoid compound (Figures 3(c) and (f)). Nostoc corrinoids purified from the naturally grown sample E were eluted to yield several total ion peaks, indicating the presence of impurities (Figure 4(a)). Ion peaks at m/z 672.77 and m/z 678.29 for pseudo B₁₂ and B₁₂, respectively, were also detected, and their retention times were identical to those of authentic pseudo B₁₂ and B₁₂. The mass spectra at the retention times of 7.42 and 7.55 min showed that both pseudo B₁₂ and B₁₂ divalent ions were formed at m/z 672.7735 (Figure 4(b)) and m/z 678.2888 (Figure 4(d)), respectively. There spective MS/MS spectra of each compound were identical to those of authentic pseudo B₁₂ (Figure 4(c)) and B₁₂ (Figure 4(e)). Similar results were obtained with other naturally grown and cultured cell samples (data not shown).

3.4. Relative Content Ratios of B₁₂ and Pseudo B₁₂ in Various Nostoc Samples

Table 2 summarizes the relative contents ratios of B₁₂ and pseudo B₁₂ in various Nostoc samples. The ratios of B₁₂ (approximately 28%) and pseudo B₁₂ (approximately 72%) are shown for naturally grown (A-E) and cultured (G and H) samples and for commercially available hair vegetable samples (K-M). Pseudo B₁₂ was the predominant corrinoid in cultured samples F and J. In contrast, sample I contained approximately 76% of B₁₂. The variable ratios of B₁₂ and pseudo B₁₂ in the cultured samples

may have been due to differences in the culture conditions, but we have no detailed information on the key factor that affected B₁₂ and pseudo B₁₂ ratios. These results indicate that most Nostoc samples and commercially hair vegetable samples contained pseudo B₁₂ (major) and B₁₂ (minor).

3.5. Evaluation of True and Fake N. flagelliforme

Because hair vegetable is one of the most expensive ingredients in Chinese cuisine, certain fake items represent a large proportion of the commercially available hair vegetable [2]. No color change was observed in all cultured samples, whereas all commercially available hair vegetable samples (K-N) exhibited significant staining by the iodine-starch method (optical densities of 0.29, 0.65, 0.34, and 1.19, respectively, at 600 nm). As shown in Figure 5, microscopic analysis indicated that although naturally grown and cultured N. flagelliforme possessed a bead-like morphology, no such morphology was found in the commercially available hair vegetable sample N (fake item only). The remaining hair vegetable samples K-M contained both Nostoc and fake substitutes. These microscopic data coincided with the results of iodine-starch reaction. Our results indicated that because naturally grown N. flagelliforme contain substantial amounts of

pseudo B₁₂, which is inactive in humans [7] and because the fake items have very low B₁₂ contents, commercially available hair vegetable is not suitable for use of B₁₂ source, regardless of the presence of the fake items.

Cyanobacteria have the ability to synthesize pseudo B₁₂ [13], which functions as a coenzyme of methionine synthase to catalyze the synthesis of methionine from homocysteine and N⁵-methyltetrahydrofolate [14]. In the present study, the cultured Nostoc sample I predominantly contained B₁₂ but not pseudo B₁₂ (Table 2), suggesting that N. flagelliforme may synthesize both B₁₂ and pseudo B₁₂ de novo. Further biochemical and genetic studies are required to elucidate the detailed physiological functions of each corrinoid in this terrestrial cyanobacterium.

REFERENCES

NOTES

^*Corresponding author.