Open Access
Dairy Sci. Technol.
Volume 90, Number 5, September–October 2010
Page(s) 509 - 520
Published online 19 January 2010

© INRA, EDP Sciences, 2010


Vitamin B12 plays many important roles in normal metabolism [6]. Subnormal vitamin B12 levels have been linked to reversible dementia [30], while vitamin B12 deficiency results in pernicious anaemia and is also manifested by neurological changes which could culminate in brain damage [6, 16]. Deficiency had also been associated with cerebellar ataxia [26], triggering seizures in epileptics [29] and reversible involuntary movements in adults [8]. More recently, low folate and vitamin B12 levels have been linked to psychiatric diseases, such as bipolar disorder [28]. Furthermore, not only were low vitamin B12 levels recorded in HIV-infected patients [14], but cobalamins demonstrated anti-retroviral activity against HIV-1 in vitro [23, 32], and the possibility that adequate vitamin B12 may slow the onset of AIDS in HIV-positive individuals was also reported [2].

Since chemical synthesis is extremely complex, only biosynthetic processes are used for the commercial production of vitamin B12 [27]. The microorganisms most widely used in these fermentations include Propionibacterium spp. [13, 27], Pseudomonas denitrificans [21], methanogens (Methanosarcina barkeri, Methanobacterium spp.) and Bacillus spp. [24, 34]. However, since some Propionibacterium species were granted GRAS (generally recognised as safe) status by the United States Food and Drug Administration [27], as well as European QPS (qualified presumption of safety) status [4], they are the preferred organisms for vitamin B12 production [27]. Although Propionibacterium freudenreichii retains most of the synthesised vitamin intracellularly [35], the vitamin is readily released in acidic solutions (pH < 5.0), including the human stomach, thus ensuring bioavailability when ingested [31].

Since vitamin B12 deficiencies are prevalent among the poor in Southern Africa [10, 18], there is a need to produce fortified functional foods to combat these deficiencies. However, since activity may be lost during processing and storage, quantification of vitamin B12 is needed to ensure quality control, to satisfy government regulatory requirements, including nutrition labelling [3, 12].

Vitamin B12 assay techniques include microbiological assays, radioisotope dilution assay, electrothermal atomic absorption spectrometry, capillary electrophoresis, chemiluminescence techniques, biomolecular interaction analysis (BIA) and a combination of fluorescence resonance energy transfer (FRET) and flow-injection analysis (FIA) [1, 6, 12, 22, 33]. The microbiological assay is still the official method, but is very laborious [1]. High-performance liquid chromatography (HPLC) has been employed by a number of workers to assay vitamin B12 in multivitamin and mineral tablets. These matrices are less complex than foodstuffs and, therefore, extraction and resolution of the vitamin is simple in comparison [20, 22, 25]. More recently, reversed-phase HPLC methods with immunoaffinity extraction and UV detection [15], and coulometric electrochemical detection [19] were used to assay vitamin B12 in petfood and various infant formulae [15] and fortified fruit juice and various seafoods [19]. However, these food matrices did not include complex dairy products. Furthermore, none of the above-mentioned studies included extraction and analysis of vitamin B12 from propionibacteria (PAB). Hence, the aim of this study was to develop and optimise both an extraction protocol and an HPLC method that will afford a more rapid vitamin B12 analysis of milk and fermented dairy products, as well as of synthetic media cultured with P. freudenreichii strains.


2.1. Chemicals, reagents and other materials

Vitamin B12 (cyanocobalamin) (Fluka) standard was used. Methanol (BDS) and acetonitrile (BDS) were of HPLC grade. All other chemicals were of Analar grade. Reverse osmosis Milli-Q water (18 MΩ), purified with a Milli-Q water purification system (Millipore, Microsep, Bellville, South Africa), was used for all solutions and dilutions.

The KCN-acetate buffer consisted of (g·L−1) 20 g KCN (Saarchem) and 544.32 g sodium acetate (CH3COONa) (Saarchem). Prior to dilution to the required concentration, the pH was adjusted to 4.5 using glacial acetic acid (BDH Chemicals) [19]. The medium (B12 medium) used to enhance vitamin B12 synthesis consisted of (g·L−1): polypeptone (BBL) 12.5; casamino acids (Difco) 11.0; yeast extract (Difco) 2.5; NaH2PO4·2H2O (Merck) 2.25; MgCl2·6H2O (Merck) 0.4; K3P04 (Merck) 1.76; CoCl2·6H2O (Merck) 0.018; FeSO4·7H2O (Merck) 0.01; d-calcium-pantothenate (Fluka) 0.004; (+)-Biotin (Fluka) 0.0003; Tween 80 (Saarchem) 1.0 mL; glucose (BDH Chemicals) 40; and 5,6-dimethylbenzimidazole (DBI) (Merck) 0.07. The pH of the medium was adjusted to 7.0 with 40 g·L−1 NaOH prior to autoclaving at 121 °C for 15 min [35]. Yeast extract sodium lactate (YEL) broth contained in g·L−1: yeast extract (Oxoid) 5.0; sodium lactate syrup (500 g·L−1) (Saarchem) 20.0; peptone (Biolab) 2.0; KH2PO4 (Merck) 10.0; and Tween 80 (Saarchem) 1.0 mL [7].

Eight strains of P. freudenreichii were used. Strains J8, J9, J10, J15 and J16 were obtained from the culture collection of the Department of Food Science, University of Stellenbosch, strain J17 (ATCC 13673) from the American Type Culture Collection and strains J18 and J19 from the All Russian Collection of Microorganisms, Pushino IBFM, Russian Academy of Sciences. The inoculum volume was 200 μL·100 mL−1 broth, with a cell count of 1 × 107 cfu·mL−1 (variation of 5% at final count). The test organism used in the microbiological assay, Lactobacillus delbrueckii ssp. lactis (ATCC 7830), was obtained from the South African Bureau of Standards (SABS).

UHT skim milk, 10 g·L−1 fat content, UHT full cream milk, 33 g·L−1 fat content and “Amasi”, 50 g·L−1 fat content, a fermented milk product generally consumed in Southern Africa, were purchased from a local supermarket. Kefir grains were obtained from the Department of Food Science, University of Stellenbosch. Reactivation of the Kefir grains was achieved by adding 30 mL milk to 10 g grains and incubating at 25 °C for 12 h, with agitation. The milk phase was drained using a sterilised metal strainer. The grains were then inoculated into 100 mL UHT full cream milk and incubated at 25 °C for 12 h, after which the grains were separated from the Kefir beverage. Reactivated grains were also co-inoculated with strain J15 to obtain Kefir beverage samples fortified with vitamin B12.

2.2. Standard solutions

The stock solution was prepared bimonthly by weighing out 4.0 mg cyanocobalamin and diluting it to 200 mL with KCN-acetate buffer in amber glassware and covering the container with aluminium foil. This solution, stored at 4 °C, was used to make dilutions as required for both the HPLC and microbiological assays and was completely stable for a minimum of five months. Since cobalamin powders are very hygroscopic [6] and sensitive to light [17], prior to preparing a fresh stock solution, the powder was dried to constant mass in a desiccator containing fresh silica gel in the dark.

2.3. Linearity, repeatability, recovery and validity

2.3.1. Linearity

The inherent sensitivity of the HPLC method was determined by analysing a series of vitamin B12 standard solutions, increasing linearly in concentration from 0.005 to 15 μg·mL−1 sample. The range of concentrations was chosen based on vitamin B12 levels (in μg·mL−1 sample) reported for dairy products (0.003–0.012 [11]) and synthesised by PAB (0.02 [5]; 0.14 [9]; and 12.50 [35]).

Seven samples per concentration were analysed (n = 7) and the resultant computed vitamin B12 concentration was plotted against the actual (i.e. expected) vitamin B12 concentration. A linear regression line, calculated by the method of least squares, was indicated. In addition, the correlation coefficient (R 2) and the regression coefficient (R) were calculated.

2.3.2. Repeatability

An homogeneous sample was divided into 14 aliquots, and the vitamin B12 concentration was determined in two series of seven aliquots on two successive days.

2.3.3. Recovery

A sample or placebo, i.e. a sample that initially contained no cyanocobalamin, was spiked with a known quantity of cyanocobalamin. After the normal extraction procedure, the added amount of cyanocobalamin was measured to determine the percentage recovery.

2.3.4. Validity

Using a selection of samples, the validity of the HPLC results was established by performing the microbiological assay on aliquots of the same extract. All assays were performed in triplicate.

2.4. Sample extraction

The ratio of KCN-acetate extraction buffer (pH 4.5) to broth sample was 10:4 (v/v). The samples were protected from light as described. Autoclaving at 121 °C for 25 min was followed by rapid cooling and centrifugation at 15 000× g for 10 min at 4 °C. This was followed by (1) filtration through a 0.22-μm syringe filter (Cameo, Osmonics, Sigma Aldrich) into amber sample vials with solid caps (HPLC method) and (2) suitable dilutions, followed by inoculation (microbiological assay). HPLC samples that were not analysed immediately were stored at 4 °C in the dark.

The presence of cyanide in the extraction buffer is critically important to ensure quantitative conversion of all forms of vitamin B12 to cyanocobalamin, the most stable form of the vitamin, while the heat treatment liberates all protein-bound cobalamin [6]. Hence, this extraction method was used for both the HPLC and microbiological assays.

2.5. Sample purification and concentration

Since the vitamin B12 content of some dairy products is as low as 0.003 μg·mL−1 [11], sample purification and concentration was investigated to improve the sensitivity of the analysis. Solid phase extraction (SPE) columns were used to both purify and concentrate the sample. Purification involved application of 9–12 mL sample extract to a preconditioned SPE column (Chromabond SB/3 mL/500 mg, Macherey-Nagel, Düren, Germany) and collecting the eluate, i.e. the purified sample extract, for HPLC analysis. When sample concentration followed purification, immediately after collecting the purified sample extract, any vitamin B12 remaining on the SPE column was eluted with one column volume 400 mL·L−1 methanol (HPLC grade, BDS). This second eluate was collected separately. A preconditioned Chromabond C18ec/6 mL/1000 mg column (Macherey-Nagel) was then used to concentrate the sample. To achieve this, the first eluate (i.e. the purified sample extract) was applied to the column. The vitamin B12, which remains as a visible pink to red band on the column, was then eluted using the second eluate (collected separately during the purification step), followed by 1 mL 900 mL·L−1 methanol. After flash evaporation of the methanol in a rotary evaporator at 40 °C, the concentrate was diluted with Milli-Q water to 3 mL in a volumetric flask. Syringe filtration (0.22-μm Cameo, Osmonics, Sigma Aldrich) into amber glass vials preceded the HPLC analyses.

2.6. HPLC

The HPLC separations were performed using a Hewlett-Packard 1100 HPLC system, equipped with a vacuum degassing unit, a quaternary pump and a Rheodyne 7125 injection valve fitted with a 20-μL sample loop. A 50-μL syringe was employed to ensure improved accuracy with respect to delivery of the desired 20-μL sample volume. A variable wavelength detector set at 361 nm was used. The HP Chemstation software (Hewlett-Packard, Waldbron, Germany) integrated, recorded and calculated the peak areas and was used to store data.

A range of chromatographic conditions was employed in an attempt to optimise peak resolution and response (peak area). These included the use of different analytical columns, adjusting the flow rate and increasing and decreasing the slope of the solvent gradient. Optimum conditions were found when the analytical column used was a 5- μm BDS-C18 column (250 × 2 mm internal diameter, i.d.) produced by Hewlett-Packard (Waldbron, Germany). A guard cartridge (ThermoQuest, Needham, Massachusetts) (4 × 4 mm i.d.) packed with 5 μm Hypersil BDS-C18 packing material was installed ahead of the analytical column. The guard cartridge was consistently replaced after 150–200 injections. The mobile phase was a 15:85 to 50:50 (v/v) methanol-water linear gradient over 25 min with a flow rate of 0.35 mL·min−1. The retention time of cyanocobalamin was 12.46 ± 0.48 min under these conditions.

Utilising the calibration function of the HP Chemstation software to construe the calibration curve, quantification of the vitamin B12 levels of samples was performed by the external standard method. Hence, a standard solution diluted to a known concentration on the day was analysed in duplicate and the software programmed to use the resultant peak area to calculate the concentration of vitamin B12 in the samples. Regeneration of the column between injections entailed post-run flushing with 100% acetonitrile (5 min), followed by 100% water (5 min), and then recycled to initial conditions (5 min).

2.7. Microbiological assay

The microbiological assay method of Angyal [1], with modifications, was employed. The test organism, L. delbrueckii ssp. lactis ATCC 7830, was routinely cultured in MRS broth or MRS agar (Biolab). One day prior to the assay, the culture was inoculated into Bacto Lactobacilli broth AOAC (Difco) liquid culture medium and incubated at 37 °C. On the day of the assay, the cultures were centrifuged (10 000× g, 10 min, 4 °C), the cells were washed and resuspended in a 10 mL sterile physiological saline solution (p.s.s.) under aseptic conditions. To obtain the inoculum, this cell suspension was diluted with sterile p.s.s. to 60 %T (read against sterile saline set at 100 %T).

The requisite standard solutions (with cyanocobalamin concentrations of 0.01, 0.014 and 0.02 ng·mL−1) were prepared by diluting the stock solution with the KCN-acetate buffer and not with 25% alcohol as suggested by Angyal [1].

As mentioned earlier, since the KCN buffer ensures superior vitamin B12 stability in the sample extract [6], it was preferred to the metabisulphite buffer described by Angyal [1]. Furthermore, since replicate samples were assayed using the HPLC, the extraction procedure was as described earlier.

After preparation and sterilisation of the standard and assay tubes, a Distriman repeater pipette equipped with a gamma-sterilised tip (Gilson, Villiers-le-Bel, France) was used to deposit one drop of inoculum directly onto the surface of the tube contents. Check tubes (replicates of the primary standard tubes) were prepared as prescribed.

The tubes were then incubated at 37 °C for up to 72 h. Two check tubes were removed on a 24 h basis and titrated against 4 g·L−1 NaOH. After a maximum of 72 h, or when the check tube with the highest concentration of vitamin B12 primary standard required 8–12 mL NaOH, the tubes were refrigerated at 4 °C to retard further acid production. The relative growth in all the standard and assay tubes was then determined by titrating against 4 g·L−1 NaOH to a set end-point of pH 6.8, using a Methrohm 702 SM Titrino automatic titrator (Swiss Lab Technologies, Standton, South Africa).

Averages of the replicate values were used to plot the standard curves and these, in turn, were employed to reference the amount of vitamin B12 in the assay tubes. The vitamin B12 content of the original sample was calculated taking all dilution factors into account.

2.8. Data analyses

All statistical analyses of data were performed using SPSS 16.0 for Windows®. The relative standard deviation (RSD) was calculated for the recovery data as a measure of inter-assay variability.


3.1. Linearity and sensitivity

The correlation coefficient (R 2) and the regression coefficient (R) were 0.9992 and 0.9996, respectively. Confidence of linearity, signified by R > 0.998, is closely related to the rate of recovery, i.e. the accuracy of the analysis [19]. The linear appearance of the plot depicted in Figure 1, combined with the fact that the correlation coefficients approached unity, indicated that the measured vitamin B12 concentration was linear over the actual vitamin B12 concentration range from 0.005 to 15 μg·mL−1 sample. The HPLC method could, therefore, be adjudged sufficiently sensitive to be used as an analytical tool.

thumbnail Figure 1.

Linearity of vitamin B12 analyses. Error bars indicate the SD.

The limit of quantification (LOQ), i.e. the lowest vitamin B12 concentration where reproducible results were obtained, was 0.005 μg·mL−1 sample. This compares favourably to the LOQ of 0.01 μg·mL−1 sample reported by Heudi et al. [15]. However, since it excludes the lowest vitamin B12 concentration reported for dairy products, viz. 0.003 μg·mL−1 [11], the sensitivity of the method was improved by employing sample concentration successfully (Fig. 3).

3.2. Repeatability, recovery and validity

When a standard sample with expected vitamin B12 concentration of 2.63 μg·mL−1 was divided into 14 aliquots and analysed repeatedly on different days, reproducible results were obtained. Mean vitamin B12 concentration, 2.62 ± 0.02 μg·mL−1 (day 1) and 2.61 ± 0.02 μg·mL−1 (day 2), did not differ significantly (P = 0.73) and was close to the expected value of 2.63 μg·mL−1. Good repeatability (intermediate precision) was, therefore, demonstrated.

The recovery of vitamin B12 from the samples and placebo (Tab. I) ranged from 98.6% to 103.2% (average values), with the RSD ranging from 2.68% to 6.79%. It was, therefore, reasonable to conclude that the extraction regime and the analytical procedure effected complete dissolution of vitamin B12 in the matrices analysed in this study.

Table I.

Recovery results for HPLC determination of vitamin B12 in different samples, with vitamin B12 added in various known concentrations.

Statistical analysis of the results of the microbiological assays vs. the HPLC analyses of the same samples indicated that no significant difference existed (P > 0.05) (Tab. II). The HPLC method is, therefore, a valid alternative to the microbiological assay.

Table II.

Comparison of vitamin B12 results from HPLC and microbiological assays (n = 3). Values are reported as mean ± SD.

3.3. Sample extraction and HPLC assays

Sample chromatograms depict the results for various samples (Fig. 2). The excellent resolution of the B12 peak, together with the absence of baseline noise, illustrated the excellent selectivity of the HPLC method. The results also reinforced the fact that the extraction and analytical procedures produced acceptable results when applied to these food products and to synthetic media fermented by P. freudenreichii spp.

thumbnail Figure 2.

Sample chromatograms of the vitamin B12 standard (A, 0.005 μg·mL−1) and (B, 15 μg·mL−1), vitamin B12 extracted from B12 medium fermented by strain J15 (C, 3.61 μg·mL−1), fortified skim milk (D, 1.28 μg·mL−1), fortified Amasi (E, 1.28 μg·mL−1) and fortified Kefir (F, 1.31 μg·mL−1). Retention time (t R) next to the peak marks the B12 peak.

3.4. Sample purification and concentration

When comparing the vitamin B12 levels before (Figs. 3A and 3C) and after (Figs. 3B and 3D) sample concentration using SPE, it is clear that this technique resulted in a substantial increase in the apparent concentration of vitamin B12 in a broth sample cultured with strain J15 (from 0.77 to 3.77 in μg·mL−1, i.e. an almost five-fold increase, Figs. 3A and 3B), as well as in a fortified Kefir sample (from 0.34 to 3.36 μg·mL−1, i.e. an almost 10-fold increase, Figs. 3C and 3D).

thumbnail Figure 3.

Chromatograms depicting sample purification (A and C), followed by sample concentration (B and D) using SPE techniques. Samples were extracted from B12 medium cultured with strain J15 (A and B) and fortified Kefir (C and D). Vitamin B12 concentration was measured (in μg·mL−1): 0.77 (A) and 0.34 (C) before and 3.77 (B) and 3.36 (D) after sample concentration. Retention time (t R) marks each B12 peak.

The significance of this enhanced response is that it increases the sensitivity of the analysis. In other words, while the LOQ was established at 0.005 μg·mL−1, with a five-fold increase in sensitivity, as demonstrated for the J15 sample, the LOQ will be decreased to 0.001 μg·mL−1. Hence, using SPE concentration makes this method suitable for the analysis of dairy products with vitamin B12 levels as low as 0.003 μg·mL−1.

Sample purification resulted in cleaner chromatograms. This is illustrated by the absence of any peaks other than the vitamin B12 peak (the KCN buffer and solvent peaks are not shown since they elute at t R < 6 min) (Figs. 3A3D). Furthermore, the chromatograms depicted in Figures 2C and 3A were both obtained when analysing sample extracts of strain J15 cultured in B12 medium. It is clear that the compounds eluting immediately after vitamin B12 (Fig. 2C) are absent after purification using SPE (Fig. 3A). Hence, this is further confirmation that effective purification was achieved.


The extraction protocol was shown to effect complete recovery of vitamin B12 from the sample matrices. The HPLC method was found to be selective, linear (R > 0.998), accurate (recovery from 98.6% to 103.2%) and sensitive (LOQ = 0.005 μg·mL−1). It was also shown that the sensitivity could be improved using sample concentration. The method, therefore, is a rapid alternative to the more time consuming microbiological assay.


  1. Angyal G., Methods for the microbiological analysis of selected nutrients, AOAC International, Gaithersburg, Maryland, USA, 1996. [Google Scholar]
  2. Anonymous, Vitamin B12 may slow AIDS, Tufts Univ. Health Nutr. Lett. 15 (1998) 2. [Google Scholar]
  3. Anonymous, Regulations relating to the fortification of certain foodstuffs, G.N.R. 2003/2003, Foodstuffs, cosmetics and disinfectants act and regulations. Act no. 54 of 1972, Lex Patria Publishers, Johannesburg, South Africa, 2003. [Google Scholar]
  4. Anonymous, Scientific opinion of the panel on biological hazards on a request from EFSA on the maintenance of the QPS list of microorganisms intentionally added to food or feed, EFSA J. 923 (2008) 1–48. [Google Scholar]
  5. Babuchowski A., Laniewska-Moroz L., Warminska-Radyko I., Propionibacteria in fermented vegetables, Lait 79 (1999) 113–124. [CrossRef] [EDP Sciences] [Google Scholar]
  6. Ball G.F.M., Vitamins in Foods, Analysis, Bioavailability and Stability, CRC Press, Boca Raton, FL, USA, 2006. [Google Scholar]
  7. Britz T.J., Riedel K.-H.J., Propionibacterium species diversity in Leerdammer cheese, Int. J. Food Microbiol. 22 (1994) 257–267. [CrossRef] [PubMed] [Google Scholar]
  8. Celik M., Barkut I.K., Öncel C., Forta H., Involuntary movements associated with vitamin B12 deficiency, Parkinsonism Relat. Disord. 10 (2003) 55–57. [CrossRef] [PubMed] [Google Scholar]
  9. Cerná J., Hrabová H., Biologic enrichment of fermented milk beverages with vitamin B12 and folic acid, Milchwissenchaft 32 (1977) 274–277. [Google Scholar]
  10. Dannhauser A., Joubert G., Nel M., Nutritional status of preschool children in the Bloemfontein district, South Afr. J. Food Sci. Nutr. 8 (1996) 14–22. [Google Scholar]
  11. De Man J.M., Principles of food chemistry, 3rd edn., Aspen Publishers Inc., Gaithersburg, Maryland, USA, 1999. [Google Scholar]
  12. Gao Y., Fuo F., Gokavi S., Chow A., Sheng Q., Guo M., Quantification of water-soluble vitamins in milk-based infant formulae using biosensor-based assays, Food Chem. 110(2008) 769–776. [CrossRef] [Google Scholar]
  13. Gardner N., Champagne C.P., Production of Propionibacterium shermanii biomass and vitamin B12 on spent media, J. Appl. Microbiol. 99 (2005) 1365–2672. [Google Scholar]
  14. Hepburn M.J., Dyal K., Runser L.A., Barfield R.L., Hepburn L.M., Fraser S.L., Low serum vitamin B12 levels in an outpatient HIV-infected population, Int. J. Sexually Transmitted Dis. & AIDS 15 (2004) 127–133. [Google Scholar]
  15. Heudi O., Kilinç T., Fontannaz P., Marley E., Determination of vitamin B12 in food products and in premixes by reversed-phase high performance liquid chromatography and immunoaffinity extraction, J. Chromatogr. A 1173 (2006) 63–68. [CrossRef] [Google Scholar]
  16. Hoffbrand V., Provan D., Macrocytic anaemias, Br. Med. J. 314 (1997) 430–433. [PubMed] [Google Scholar]
  17. Kelly R.J., Detection of methyl cobalamin and adenosyl cobalamin using a nitroxide radical trap, Anal. Commun. 35 (1998) 257–259. [CrossRef] [Google Scholar]
  18. Labadarios D., The iron status of South African children 6–71 months of age: results of the national survey, 1994, South Afr. J. Food Sci. Nutr. 8 (1996) 10. [Google Scholar]
  19. Lebiedzińska A., Marszałł M.L., Kuta J., Szefer P., Reversed-phase high-performance liquid chromatography method with coulometric electrochemical and ultraviolet detection for the quantification of vitamins B1 (thiamine), B6 (pyridoxamine, pyridoxal and pyridoxine) and B12 in animal and plant foods, J. Chromatogr. A 1173 (2007) 71–80. [CrossRef] [PubMed] [Google Scholar]
  20. Li H.B., Chen F., Jiang Y., Determination of vitamin B12 in multivitamin tablets and fermentation medium by high-performance liquid chromatography with fluorescence detection, J. Chromatogr. A 891 (2000) 243–247. [CrossRef] [PubMed] [Google Scholar]
  21. Li K.-T., Liu D.-H., Li Y.-L., Chu J., Wang Y.-H., Zhuang Y.-P., Zhang S.-L., Improved large-scale production of vitamin B12 by Pseudomonas denitrificans with betaine feeding, Bioresour. Technol. 99 (2008) 8516–8520. [Google Scholar]
  22. Markopoulou C.K., Kagkadis K.A., Koundourellis J.E., An optimized method for the simultaneous determination of vitamins B1, B6, B12, in multivitamin tables by high performance liquid chromatography, J. Pharm. Biomed. Anal. 30 (2002) 1403–1410. [CrossRef] [PubMed] [Google Scholar]
  23. Mathé G., The non-enumerable described retrovirus integrase inhibitors are not a lure, as evidenced by ten years of clinical experience, Biomed. Pharmacother. 54 (2000) 3–6. [CrossRef] [PubMed] [Google Scholar]
  24. Mazumder T.P., Nishio N., Fukuzaki S., Nagai S., Production of extracellular vitamin B12 compounds from methanol by Methanosarcina barkeri, Appl. Microbiol. Biotechnol. 26 (1987) 511–516. [CrossRef] [Google Scholar]
  25. Moreno P., Salvadó V., Determination of eight water- and fat-soluble vitamins in multivitamin pharmaceutical formulations by high-performance liquid chromatography, J. Chromatogr. A 870 (2000) 207–215. [CrossRef] [PubMed] [Google Scholar]
  26. Morita S., Miwa H., Kihira T., Kondo T., Cerebellar ataxia and leukoencephalopathy associated with cobalamin deficiency, J. Neurol. Sci. 216 (2003) 183–184. [CrossRef] [PubMed] [Google Scholar]
  27. Murooka Y., Piao Y., Kiatpapan P., Yamashita M., Production of tetrapyrrole compounds and vitamin B12 using genetically engineering of Propionibacterium freudenreichii. An overview, Lait 85 (2005) 9–22. [CrossRef] [EDP Sciences] [Google Scholar]
  28. Ozbek Z., Kucukali C.I., Ozkok E., Orhan N., Aydin M., Kilic G., Sazci A., Kara I., Effect of the methylenetetrahydrofolate reductase gene polymorphisms on homocysteine, folate and vitamin B12 in patients with bipolar disorder and relatives, Prog. Neuropsychopharmacol. Biol. Psychiatry 32 (2008) 1331–1337. [CrossRef] [PubMed] [Google Scholar]
  29. Thiel R.J., Fowkes S.W., Down syndrome and epilepsy: a nutritional connection, Med. Hypotheses 62 (2004) 35–44. [CrossRef] [PubMed] [Google Scholar]
  30. Tucker K.L., Mahnken B., Wilson P.W.F., Jacques P., Selhub J., Folic acid fortification of the food supply. Potential risks and benefits for the elderly population, J. Am. Med. Assoc. 276 (1996) 1879–1885. [CrossRef] [PubMed] [Google Scholar]
  31. Vorobjeva L.I., Propionibacteria, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1999. [Google Scholar]
  32. Weinberg J.B., Sauls D.L., Misukonis M.A., Shugars D.C., Inhibition of productive Human Immunodeficiency Virus-1 infection by cobalamins, Blood 86 (1995) 1281–1287. [PubMed] [Google Scholar]
  33. Xu H., Li Y., Liu C., Wu Q., Zhao Y., Lu L., Tang H., Fluorescence resonance energy transfer between acridine orange and rhodamine 6G and its analytical application for B12 with flow-injection laser-induced fluorescence detection, Talanta 77 (2008) 176–181. [CrossRef] [PubMed] [Google Scholar]
  34. Yang Y., Zhang Z., Lu J., Maekawa T., Continuous methane fermentation and the production of vitamin B12 in a fixed-bed reactor packed with loofah, Bioresour. Technol. 92 (2004) 285–290. [CrossRef] [PubMed] [Google Scholar]
  35. Ye K., Shijo M., Miyano K., Shimizu K., Metabolic pathway of Propionibacterium growing with oxygen: enzymes, 13C NMR analysis, and its application for vitamin B12 production with periodic fermentation, Biotechnol. Prog. 15 (1999) 201–207. [CrossRef] [PubMed] [Google Scholar]

All Tables

Table I.

Recovery results for HPLC determination of vitamin B12 in different samples, with vitamin B12 added in various known concentrations.

Table II.

Comparison of vitamin B12 results from HPLC and microbiological assays (n = 3). Values are reported as mean ± SD.

All Figures

thumbnail Figure 1.

Linearity of vitamin B12 analyses. Error bars indicate the SD.

In the text
thumbnail Figure 2.

Sample chromatograms of the vitamin B12 standard (A, 0.005 μg·mL−1) and (B, 15 μg·mL−1), vitamin B12 extracted from B12 medium fermented by strain J15 (C, 3.61 μg·mL−1), fortified skim milk (D, 1.28 μg·mL−1), fortified Amasi (E, 1.28 μg·mL−1) and fortified Kefir (F, 1.31 μg·mL−1). Retention time (t R) next to the peak marks the B12 peak.

In the text
thumbnail Figure 3.

Chromatograms depicting sample purification (A and C), followed by sample concentration (B and D) using SPE techniques. Samples were extracted from B12 medium cultured with strain J15 (A and B) and fortified Kefir (C and D). Vitamin B12 concentration was measured (in μg·mL−1): 0.77 (A) and 0.34 (C) before and 3.77 (B) and 3.36 (D) after sample concentration. Retention time (t R) marks each B12 peak.

In the text