VITAMIN ANALYSIS

Definition and Importance

Vitamins are defined as relatively low-molecular weight compounds which humans, and for
that matter, any living organism that depends on organic matter as a source of nutrients,
require small quantities for normal metabolism. With few exceptions, humans cannot
synthesize most vitamins and therefore need to obtain them from food and supplements.
Insufficient levels of vitamins result in deficiency diseases [e.g., scurvy and pellagra, which
are due to the lack of ascorbic acid (vitamin C) and niacin, respectively].

Importance of Analysis

Vitamin analysis of food and other biological samples has played a critical role in
determining animal and human nutritional requirements. Furthermore, accurate food
composition information is required to determine dietary intakes to assess diet adequacy and
improve human nutrition worldwide. From the consumer and industry points of view, reliable
assay methods are required to ensure accuracy of food labeling. This chapter provides an
overview of techniques for analysis of the vitamin content of food and some of the problems
associated with these techniques. Please note that the sections below on bioassay,
microbiological, and chemical methods are not comprehensive, but rather just give examples
of each type of analysis.

METHODS

Vitamin assays can be classified as follows:

1. Bioassays involving humans and animals.

2. Microbiological assays making use of protozoan organisms, bacteria, and yeast.

3. Physicochemical assays that include spectrophotometric, fluorometric, chromatographic,

enzymatic, immunological, and radiometric methods.

In terms of ease of performance, but not necessarily with regard to accuracy and precision,
the three systems follow the reverse order. It is for this reason that bioassays, on a routine
basis at least, are limited in their use to those instances in which no satisfactory alternative
method is available.

Extraction Methods

With the exception of some biological feeding studies, B vitamin assays in most instances
involve the extraction of a vitamin from its biological matrix prior to analysis. This generally
includes one or several of the following treatments: heat, acid, alkali, solvents, and
enzymes. In general, extraction procedures are specific for each vitamin and designed to
stabilize the vitamin. In some instances, some procedures are applicable to the combined
extraction of more than one vitamin, for example, for thiamine and riboflavin as well as some
of the fat-soluble vitamins (1, 2, 13). Typical extraction procedures are as follows:

1
• Ascorbic acid: Cold extraction with metaphosphoric acid/acetic acid.

• Vitamin B1 and B2: Boiling or autoclaving in acid plus enzyme treatment.

• Niacin: Autoclaving in acid (noncereal products) or alkali (cereal products).

• Folate: Enzyme extraction with α-amylase, protease and γ-glutamyl hydrolase(conjugase)

• Vitamins A, E, or D: Organic solvent extraction, saponification, and re-extraction with

organic solvents. For unstable vitamins such as these, antioxidants are routinely added to
inhibit oxidation.

Analysis of fat-soluble vitamins may require saponification, generally either overnight at

room temperature or by refluxing at 70◦C. In the latter case, an air-cooled reflux vessel as
depicted in Fig. 11-1 provides excellent control of conditions conducive to oxidation.

The B-group vitamins

The B-group vitamins have traditionally been determined by MBA but many applications of
HPLC have appeared in the literature over the last twenty years. MBAs have many
disadvantages in that they require dedicated laboratory facilities, microbiologically trained
staff, and are lengthy and prone to known and unknown interferences. Despite these
drawbacks, MBA often provides the only procedure capable of determining the amount of a
particular vitamin in a wide range of foodstuffs with reasonable accuracy. HPLC offers the
potential advantages of selectivity, so that we can determine individual vitamins of interest,
as opposed to a total growth response from MBA, and sometimes it provides rapid results,
but it does suffer from disadvantages related to sensitivity, extraction of the vitamins from
foodstuffs and often the need for pre-HPLC removal of interferences. Neither MBA nor
HPLC offer the analyst the solution to all problems associated with the determination of the
B-group vitamins, they both have advantages and disadvantages. For many studies they

2
should be regarded as complimentary techniques, the results from both providing the analyst
with essential information.

Microbiological assays

Microbiological assays for the determination of vitamins are dependent upon the specific
growth requirements of selected microorganisms, usually lactic acid bacteria. The first MBA
for riboflavin was performed by Snell in 1939 using lactic acid bacteria because their growth
requirements had been studied by dairy bacteriologists, and as agroup their requirements
were complex but specific. Lactobacilli have the additional advantage that their growth can
be easily followed by turbidometric or optical density measurement or by titration of the
lactic acid produced during growth. Their choice as assay organisms is amply justified as
they are still used today for the determination of B-group vitamins. The basis of a MBA is
addition of a dilution series of the sample extract to a basal medium which contains all the
growth requirements for the test organism except the vitamin to be determined; the mixture is
then inoculated with the test organism and incubated. The test organism will grow in
proportion to the vitamin content of the sample extract and quantitation is achieved by
inclusion of a range of vitamin standards in the MBA. The growth of the test organism is
measured as mentioned above.

The basic procedure for a MBA is the same for all the vitamins and can be broken down into
a series of stages as follows:

1. Preparation of media for maintaining stock cultures of the test organisms.

2. Preparation of a basal medium deficient in the vitamin to be determined.

3. Preparation of the inoculum medium and inoculum culture.

4. Extraction of the vitamin from the samples.

5. Setting up the assay.

6. Sterilisation of the assay tubes and media.

7. Inoculation of assay tubes with the test organism.

8. Incubation (18-24 h).

9. Measurement of the growth response of the test organism and calculation of results.

Stages 1-3 require media which are capable of supporting the growth of the test organism to
be used and for lactic acid bacteria they must contain amino acids, vitamins, purine and
pyrimidine bases, fermentable carbohydrate, mineral salts and buffers. These media are
available commercially which eliminates the possible variation in composition experienced
when media are prepared from basic ingredients in the laboratory. Test organisms must be
obtained from a recognised national culture collection and stock cultures have traditionally
been maintained as agar stabs. Preparation of the basal assay medium is conveniently

3
performed by rehydrating commercially available dried media according to suppliers
instructions. The preparation of the inoculum has shown variation over time, and to some
extent is dependent upon the method to be used at the end ofthe incubation period to measure
bacterial growth. One classic procedure involves taking a stab-culture from the stock culture
into a sterile lactobacilli broth and incubating overnight before required. The broth is
centrifuged, the supernatant discarded, the bacteria washed several times with sterile saline
and then suspended in sterile saline and used for inoculation of the assay tubes. This
technique is not very satisfactory for 18-24 h turbidometric assays (as opposed to 72 h
titrimetric assays) as the bacteria are in the lag phase of growth and need time to pass through
the acceleration phase and into the desired exponential phase which is required for the assay.
A technique used by Bell (1984) and still employed at the LGC involves growth overnight in
lactobacilli broth and the following morning one drop of the broth is subcultured into the
appropriate assay medium which contains a controlled amount of the vitamin to be
determined. After six hours growth the inoculum is diluted with sterile assay medium and
used for inoculation of the assay. This ensures that the test organism is approaching or in the
exponential phase of growth at the start of the assay. Another way of preparing inocula is the
use of glycerol cryo protected lactobacilli (Wilson and Home, 1982) which offers the
advantage of standardised inocula strength and is becoming increasingly popular. A large
volume of overnight inoculum is prepared and incubated overnight, glycerol added, and then
stored in 2 ml vials at -70°C. A thawed vial is then used for each assay.

4
Extraction of the Vitamin from the Test Material (Stage 4)
The vitamins are extracted from the food matrix in a form that can be utilized by the
particular assay organism being used. This generally involves autoclaving the food sample in
the presence of acid or, for acid labile vitamins, digesting the sample with suitable enzymes.
After precipitating the proteins at their isoelectric point (c. pH 4), the pH of the extract is
adjusted to that of the basal medium (typically pH 6.8). This step is necessary to ensure that
the pH of the medium is not altered by the addition of different amounts of the extract. The
extract is then diluted to bring the concentration of the vitamin to be assayed within the range
of the standard curve. Hopefully, the dilution factor is sufficiently high to dilute out any
interfering substances that would cause drift and invalidate the assay. The minimum dilutions
of foods necessary to avoid the inhibitory effects of food preservatives and neutralization
salts. Finally, the extracts are filtered to remove the precipitated protein and lipoidal material,
and to obtain a clear solution for assay.

5
Setting up the assay (stage 5) is almost the same for all vitamins. Test tubes are most
commonly used in racks of perhaps 90 tubes. 0-5 ml of sample extracts and standards are
pipetted into the tubes, the volume made up to 5 ml as necessary and 5 ml of assay medium
added to each tube. An alternative procedure (Bell, 1984) is to add 0-400 I.d of sample
extracts and standards to assay tubes followed by the addition of 10 ml of assay medium.
This has the advantage that the pH of sample extracts is less critical and single strength assay
media can be used, and smaller volumes are handled, making semi-automation of the system
easier. After addition of the media, tubes are capped and sterilised by autoclaving.

Inoculation (stage 7) involves adding the same amount of inoculum to each assay tube (e.g. I
drop or 100 Jll) then the whole rack of tubes is placed in a water bath at the required
temperature for incubation. Air incubators are not suitable because of the temperature
differential experienced between assay tubes. After incubation for 18 to 24 h bacterial growth
in standards and sample assay tubes should be visible and the most convenient way to
measure growth is by the use of a nephelometer or by measurement of optical density. It is
essential that the measurement of growth is determined after carefully defined times if results
are to be valid, and these times must be determined within the laboratory using a defined
assay protocol. Growth measurements from standards are plotted against vitamin
concentration and results for samples obtained by interpolation from this calibration line.
Setting up the assay and measurement of the growth response can be labour intensive
operations but these operations can be semi-automated.

6
QUANTIFICATION

At the end of the incubation period, the cells are uniformly suspended by shaking the tubes,
and time is allowed for the air bubbles to disperse before measurement. The turbidities of all
tubes are measured in a nephelometer using a neutral filter, colorimeter with a filter in the
region of 640 nm or spectrophotometer at 540-660-nm wavelength. The turbidity may be
expressed as an extinction, as a transmittance (in % T), as a difference 100 - T (in %) or
simply as a galvanometer reading. The arithmetic means of the replicates are calculated, and
the means for the standard solutions are plotted on semi logarithmic graph paper with the
turbidity values as ordinates (linear scale) and concentrations in ng/ml as abscissae
(logarithmic scale). The calibration curve is drawn through these points (Figure 7.2). The
vitamin contents of the sample tubes are read off from the calibration curve and the values for
the original samples are calculated from simple dilution factors. Values for the vitamin
content of a given sample calculated from at least three dilutions should check within the
limits of error of the assay, which is usually considered to be ±10-15%; that is, they should
not differ by more than 15% from their common mean. If this condition is not fulfilled, the
determination must be repeated.

7
The reliability of a determination can be assessed by testing for the presence or absence of
drift. This simply entails plotting on the calibration curve the mean turbidity values for the
sample dilutions against the corresponding dilution factors. In the example given in Figure
7.2, three dilutions, 1 : 200, 1: 500 and 1 : 1000, give mean percentage turbidities of 43%,
25% and 14%, respectively. Joining the points together gives a check curve which is roughly
parallel to the calibration curve, signifying the absence of drift. Drift is manifested by a check
curve that deviates widely from the calibration curve, either increasing or decreasing with
concentration of the sample as shown in Figure 7.3. The occurrence of drift in assay values is
evidence for the presence of interfering materials in the test solution presented for assay.
Snell (1948) recognized three general causes of drift: (1) a substance chemically unrelated to
the vitamin may stimulate or inhibit the response of the assay organism to suboptimal levels
of the vitamin (e.g. free fatty acids affect the response of L. casei to riboflavin and L.
plantarum to pantothenic acid); (2) substances chemically related to the vitamin may replace
the vitamin in the nutrition of the test organism, but the dose-response curve to the vitamin
and the related compound may be completely different (e.g. polyglutamyl folates with 4-7
glutamate residues cause positive drift in folate assays using L. casei); and (3) substances
physiologically related to the vitamin may also replace it for growth (e.g. D-alanine may

8
replace vitamin B6)' If a drift has been established, the determination is invalid and the assay
must be repeated. In this event, measures must be taken to remove the interfering substances
by adopting an effective extraction procedure. If these measures fail, the assay conditions
must be changed or a different assay organism employed. The microbiological assay
procedure is capable of being semi automated as described by Keagy (1986). Using a
microcomputer to control sample dilution, medium addition, turbidity determination and data
acquisition, an assay capacity of 600 tubes/day can be achieved, which is about twice that of
a comparative manual assay.

MBAs have been scaled down to microtitre plate assay formats and this development shows
much promise (Newman and Tsai, 1986). The AOAC manual of official methods (AOAC,

9
1990b) provides MBA procedures for vitamins B12, folic acid, niacin, pantothenic acid and
riboflavin. MBAs are simple in theory but often very difficult to perform. The laboratory
must be dedicated to the routine performance of MBAs; they cannot be used on an ad-hoc
basis in the way many chemical methods of analysis are used. Staff must be well trained in
the application of standard microbiological procedures and laboratory facilities and glassware
and equipment must be scrupulously clean. Cleaning regimes for glassware, etc. must be
committed to a defined protocol and all sources of reagents and water must be of defined
quality. Glassware for individual vitamin assays must be segregated. For very sensitive
assays, e.g. BI2 and folate, glassware will probably have to be subjected to a cleaning
regimen similar to the one used in this laboratory which involves boiling in a water bath
containing detergent (the type is important) for three hours prior to thorough washing in an
automated washing machine with deionised water rinses, followed by soaking in dilute acid,
further deionised water rinses and baking in an oven at 100°C overnight before use. If
cleaning or reagent quality is compromised, assays will fail; a result which is extremely
expensive in staff time.

Thiamin - vitamin BI

Introduction. VitaminB1exists in tissue as thiamin, thiamin monophosphate, thiamin

diphosphate (pyrophosphate, cocarboxylase), thiamin triphosphate and in protein bound
forms. The most common way to determine the vitamin BI content of food stuffs has been to
release and extract thiamin and its phosphate esters using acid hydrolysis followed by
enzymatic dephosphorylation of the esters and subsequent determination of thiamin. The
determination ofthiamin has traditionally been achieved by fluorimetry after oxidation of
thiamin to the fluorescent compound thiochrome, by using microbiological assay (MBA) and
more recently by high performance liquid chromatography (HPLC).

Assay organisms

Two Lactobacilli have been widely used as assay organisms for the determination of thiamin,
namely L. fermentum (ATCC No. 9338) and L. viridescens (ATCC No. 12706). Of the two,
the latter is preferred as it is less susceptible to inhibitory or stimulatory substances

Extraction

In order to ensure the complete utilization of total thiamin by L. viridescens, the extraction
procedure involves hot acid digestion, followed by enzymatic hydrolysis as a means of
liberating free thiamine from all bound forms. The enzyme treatment is omitted for the
analysis of grain products and milk, and for the determination of the added thiamin
hydrochloride in fortified foods. In a procedure recommended by Pearson (1967b), a suitable
weight of the finely ground or homogenized material is suspended or dissolved in at least 15
times its weight of 0.1 N HCl. The mixture is autoclaved for 15 min at 121°C or steamed in
an autoclave for 30 min. A lower autoclaving temperature of 108-109 °C is required for the
digestion of grain products, which contain mostly non phosphorylated thiamin. Distinctly
acid conditions (pH 1.0-1.5) must be maintained during the digestion. The mixture is then
cooled to room temperature and adjusted to pH 4.0-4.5 with 2.5 M sodium acetate solution. A

10
5-ml quantity of a freshly prepared 6% solution of Takadiastase (or other diastatic enzyme
preparation) in 2.5 M sodium acetate is added, and the mixture is incubated for 3 h at 45-50
°C or overnight at 37°C. If overnight incubation is used, the mixture must be protected from
bacterial action with a layer of toluene. After incubation, the extract is steamed in an
autoclave for 20 min to deactivate the enzyme, cooled, adjusted to pH 6.5-6.6, and diluted to
give an appropriate thiamin concentration for the assay. If the extract is cloudy at this point, it
may be filtered through Whatman No. 1 filter paper. A reagent blank is taken through the
same procedure.

MBA of thiamin. Lactobacillus fermenti was proposed for the determination of thiamin in
1944 but this organism was shown to respond to pentoses, reducing agents, fructose, maltose
and heat degradation products of glucose. Attempts to compensate for inhibitory and
stimulatory effects were not practicable and this organism was superseded by the use
ofLactobacillus viridescens (ATCC 12706,NCIMB 8965) for which commercial assay media
can be obtained. Thiamin is extracted from foodstuffs usually by autoclaving with acid
followed by treatment with an enzyme preparation and dilution to a suitable volume. The
assay range using L. viridescens is typically 0-25 ng per assay tube, and the assay should be
incubated at 30°C for 18-20 hours for turbidometric measurement. Thiamin assays take about
three days to produce results because of the overnight enzyme incubations during extraction

Riboflavin - vitamin B2

Introduction. Natural riboflavin occurs in foods as free riboflavin or as the protein bound
riboflavin-5'-phosphate (FMN, flavin mononucleotide) and flavin adenine dinucleotide
(FAD). Extraction ofthese bound forms of the vitamin is most commonly achieved by
hydrolysis with a dilute mineral acid (e.g. 0.1 M HCl); this stage in the extraction also
hydrolyses most FAD to FMN. Ifwe are to determine 'free' riboflavin, the FMN has to be
further hydrolysed enzymatically using a commercial enzyme preparation such as
Takadiastase or Clarase. As stated previously, these enzyme preparations contain phosphates
(as 'impurities') in addition to amylases, therefore, FMN can be dephosphorylated to yield
riboflavin; the amylase has the advantage of hydrolysing starch which aids sample digestion
when carbohydrate rich foods are being examined. Riboflavin can be determined
fluorimetrically and the AOAC prescribe a manual and an automated procedure for foodstuffs
as fmal actions (AOAC, 1990d). The suitability of this method has been questioned for many
years because it is time consuming and subject to many potential interferences.

Assay organisms

The organism traditionally used for determining riboflavin is Lactobacillus casei subsp.
rhamnosus (ATCC No. 7469). Lactic acid bacteria cannot utilize FAD, and the growth
response of L. casei, measured turbidimetrically, differs significantly between riboflavin and
FMN . As most of the vitamin B2 activity is present in food sources as FMN after acid
extraction, it would be more accurate to use FMN as the standard in the microbiological assay
instead of riboflavin.

11
Kornberg, Langdon and Cheldelin (1948) proposed the use of Enterococcus faecalis (ATCC
No. 10100) which, with a sensitivity to 0.1 ng riboflavin/ml, is 50 times more sensitive
compared with L. casei.

Extraction

The flavins are released from their intimate association with proteins by autoclaving the
sample at 121°C for 15 min in the presence of 0,1 N HCL For cooked wheaten products, such
as bread, the autoclaving time must be increased to 30 min . As a result of the acid digestion,
FAD, which cannot be utilized by lactic acid bacteria, is completely degraded to FMN and
riboflavin, and some of the FMN is also degraded to riboflavin. The autoclaving of food
samples with 0.1 N HCl hydrolyses the starch, but it also liberates sufficient amounts of free
fatty acids to cause an interference in the L. casei assay. The use of Takadiastase has been
suggested as a means of hydrolysing starch (Scott, Randall and Hessel, 1941), but Barton-
Wright and Booth (1943) reported that various commercial preparations of this enzyme
complex showed a marked stimulatory action over other hydrolysing agents (ptyalin, HCl and
H2S04), In any case, the use of enzymes tends to complicate the analysis, because most crude
enzyme preparations used for such purposes contain variable amounts of vitamins, which can
lead to high and unpredictable blank values.

The general extraction procedure used in L. casei assays for analysing foods of very low fat
content (such foods include many cereals) involves acid hydrolysis, followed by precipitation
of the denatured proteins at pH 4.5 (near their isoelectric point). The precipitated proteins,
together with the small amount of lipoidal material and any non-hydrolysed starch, are
removed by simple filtration through paper. Omission of the filtration step produced a
pronounced drift in riboflavin values for whole wheat flour, which was typical of cereal
products in general. Ether extraction could be used to remove· interfering lipids, but, for
samples of negligible fat content, the filtration step results in a valid assay (i.e. free from
drift), and is much simpler and quicker to accomplish. High-fat foodstuffs should be given a
preliminary extraction with petroleum ether to remove the bulk of the lipids before the acid
hydrolysis step. This initial extraction does not completely remove the stimulatory lipids, and
a further extraction with diethyl ether is necessary after filtering off the precipitated proteins.
Milk should be separated by centrifugation, and the serum shaken in a separating funnel with
diethyl ether. Petroleum ether is not a suitable fat solvent with milk, as the mixture tends to
form an emulsion. Working procedures for extracting the riboflavin from materials of low
and high fat content are outlined below. For the extraction of cereals and other dry materials
of negligible fat content, the method of Strong and Carpenter has been found to be
satisfactory. In this procedure a suitable weight (usually 5 g is sufficient) of the finely ground
sample is suspended in 50 ml of 0.1 N HCl, and then autoclaved for 15 min at 121°C. The
volume of 0.1 N HCl in millilitres must be equal to at least 10 times the dry weight of the
sample in grams. The hydrolysed extract must be protected from light as soon as it is
removed from the autoclave. After cooling, 2 ml of 2.5 M sodium acetate solution is added
and the pH is adjusted to 4.5 with 0.5 N NaOH using bromocresol green or bromophenol blue
as external indicator. A heavy precipitate of proteinaceous material forms at this point. The
volume is made up to 100 ml, and the extract is filtered (repeatedly, if necessary) through a

12
fluted Whatman No. 42 filter paper until a clear filtrate is obtained. A 50-ml aliquot of the
filtrate is adjusted to pH 6.8 with 0.5 N NaOH, using bromothymol blue as external indicator,
and diluted to 100 ml. It may be necessary to filter this final extract before assay to remove
any turbidity. A reagent blank is taken through the same procedure.

High-fat materials, such as wheat germ, maize, oats, soya bean, meat, cheese and mixed diets,
require a preliminary extraction with petroleum ether (b. p. 40--60 0c) before hydrolysis to
remove neutral fats. This involves extracting the dried, finely ground sample with petroleum
ether for 16-18 h in a Soxhlet apparatus. The defatted material is autoclaved, adjusted to pH
4.5, diluted and filtered as described in the above procedure for the extraction of cereals. A
50-ml aliquot of the filtrate is shaken with two or three 30-ml portions of diethyl ether in a
separating funnel. The combined ether layers are washed two or three times with water, and
the washings are added to the bulked aqueous layers. Finally, the pH of the extract is adjusted
to 6.8, and the extract is filtered, if necessary, and diluted to 100 ml (or other suitable
volume) for direct assay. The published extraction procedure using E. faecalis entailed the
addition of 20 ml of water and 3 ml of 1 N H2S04 to 1 g of the dry material to be assayed,
and autoclaving for 30 min. The pH was adjusted to 4.5-5.0, and the extract diluted to contain
c. 0.001-0.002 ~g riboflavin/ml.

MBA of Riboflavin. The riboflavin assay developed by Snell in 1939 used the organism
Lactobacillus casei and this organism is still used today and is recommended for use by the
AOAC (1990d), but it can be stimulated by starch and both inhibited and stimulated by fatty
acids and lipids. The assay range is typically 0-200 ng per assay tube. The test organism
Enterococcusfaecalis (ATCC 10100, NCIMB 7432) has the advantage of being more than ten
times more sensitive to riboflavin (assay range 0-10 ng per tube) and much less sensitive to
stimulation or inhibition. Riboflavin is extracted from foodstuffs usually by autoclaving with
dilute acid. E faecalis responds equally to FMN and riboflavin and so an enzymatic
hydrolysis during extraction is not required. Assays are usually incubated at 37°C for 22 to
24 h.

Niacin

Introduction. Niacin is the collective name for nicotinic acid and nicotinamide. Nicotinic
acid is readily converted to the physiologically active form, nicotinamide which functions as
a component of the coenzymes nicotinamide adenine dinucleotide (NAD) and nicotinamide
adenine dinucleotide phosphate (NADP). Nicotinic acid, NAD and NADP occur in almost all
foodstuffs and provide sources of the vitamin which are available to man, but the vitamin
does occur bound as nicotinyl esters to polysaccharides, peptides and glycopeptides which
are not metabolised by the human gut. If we are to determine 'available' niacin then care has
to be exercised over choice ofextraction conditions used. Acid hydrolysis will release niacin
from NAD and NADP but will not hydrolyse nicotinyl esters; alkali hydrolysis is required to
achieve this. Thus if we wish to determine 'available' niacin, acid hydrolysis should be used,
whereas alkali hydrolysis will yield 'total' niacin. Bound forms of niacin occur mainly in
cereal products. Pellagra (niacin deficiency disease) has been common in areas where maize
is a staple food because the niacin is present in a bound form and tryptophan is lacking in

13
maize protein (tryptophan is converted to nicotinic acid in the body), but in Mexico where
maize is usually eaten as tortillas Pellagra has not been a problem. This is because in the
preparation of tortillas, maize is mixed with lime which hydrolyses the nicotinyl esters of the
bound niacin and the vitamin becomes nutritionally available.

If sufficiently robust acid hydrolysis conditions are used for niacin extraction, nicotinamide is
hydrolysed to nicotinic acid and so one compound can be determined as opposed to two.

The procedural sequence for the microbiological analysis of niacin is outlined in Fig. 11-3.
Lactobacillus plantarum ATCCTM 8014 is the test organism. A stock culture needs to be
prepared and maintained by inoculating the freeze dried culture on Bacto Lactobacilli agar
followed by incubation at 37◦C for 24 h prior to sample and standard inoculation. A second
transfer may be advisable in the case of poor growth of the inoculum culture. In general,
growth is measured by turbidity. If lactobacilli are employed as the test organism, acidimetric
measurements can be used as well. The latter may be necessary if a clear sample extract
cannot be obtained prior to inoculation, and incubation (which is a prerequisite for
turbidimetry) cannot be obtained. In making a choice between the two methods of
measurement, one needs to bear in mind that a prolonged incubation period of 72 h is
required for acidimetry.

14
Folates

Folate is the general term including folic acid (pteroylglutamate, PteGln) and poly-γ-glutamyl
conjugates with the biological activity of folic acid. Folates present a diverse array of
compounds that vary by oxidation state of the pteridine ring structure, one-carbon moieties

15
carried by the specific folate, and the number of conjugated glutamate residues on the folate.
Folates are labile to oxidation, light, thermal losses, and leaching when foods are processed.
Because of the presence of multiple forms in food products and its instability, folate presents
a rather difficult analytical problem. To account for differences in biological availability of
synthetic folic acid used for food fortification and food folate, the Institute of Medicine Panel
on Folate, Other B Vitamins and Choline established the dietary folate equivalent (DFE)
value (5). Based on research showing that folic acid is 85% bioavailable whereas food folate
is only 50%, it can be stated that folic acid in fortified products is 85/50 or 1.7 times more
bioavailable than food folate. Therefore, the μg of DFEs provided equals the μg of food folate
plus (1.7 × μg folic acid). Calculation of the μg DFE for any food requires quantitation of
folic acid as a separate entity from food folate. Currently, quite sophisticated liquid
chromatography methods are necessary for accurate quantitation of folic acid and the
multiple forms of folates in foods. A collaborated microbiological procedure based on the
trienzyme extraction quantifies only total folate and cannot differentiate between added folic
acid and food folate.

Assay organisms

Three organisms, Lactobacillus casei subsp. rhamnosus (ATCC No. 7469), Enterococcus
hirae (ATCC No. 8043) and Pediococcus acidilactici (ATCC No. 8081), have been routinely
employed in folate assays because they respond specifically to certain metabolic forms of
folate, and therefore can be used to distinguish between the different forms present in the
assay material.

Principle Folate in the sample is extracted with a buffer at 100◦C (boiling water bath). The
extract is then digested with α-amylase and protease (i.e., to free macromolecularly bound
folates) and conjugase (i.e., to cleave poly-γ-glutamyl folates to PteGln3 or lower.) Growth
response of the assay microorganism is measured by percent transmittance. Transmittance
depends on folate concentration.

Critical Points Care must be exercised to protect labile folates from oxidation and
photochemical degradation. Reducing agents including ascorbic acid, β-mercaptoethanol, and
dithiothreitol are effective in preventing oxidation. Strict adherence to microbiological assay
techniques is necessary to assay folate with accuracy and precision.

Procedure Analysis of food folate by Lactobacillus casei (spp. rhamnosus) ATCCTM 7469
and a trienzyme extraction procedure (Fig. 11-4) is provided by AOAC International (2). The
analytical protocol has also been easily adapted using 96-well microtiter plates and a reader.

Calculations Results are calculated manually or from the regression line of the standard
curve responses using 4th degree polynomial plots and a computer program written to
conform to the AOAC microbiological analysis protocol. Software provided for microplate
readers is suitable for calculating results from analyses using 96-well microplates. Results are
reported as micrograms of vitamin per 100 g or per serving.

16
Vitamin B12

Vitamin B12 is a member of a group ofcompounds known as cobalamins and exists in

foodstuffs mainly in coenzyme forms which are wholly or partly bound to cellular protein
constituents. Good sources of the vitamin are offal meats followed by muscle meats, dairy
produce and fish. VitaminB12 is present in food stuffs at very low levels and MBAs are the
only way to estimate the B12 content of food satisfactorily.

Assay Organisms

The choice of test organism for the determination of vitamin B12 has been a topic of great
debate over the years. The two most commonly used are Ochromonas malhamensis and
Lactobacillus leichmannii. 0. malhamensis has a greater specificity for cobalamins than L.
leichmannii and is claimed to provide a truer estimate of B12 biological activity.

17
Extraction

The extraction procedure employed in the AOAC microbiological method for determining
vitamin B12 activity in vitamin preparations is also applicable to foods, having been found
satisfactory by inter laboratory collaborative analysis of a crude liver paste, condensed fish
solubles and a crude vitamin B12 fermentation product. The procedure entails homogenizing
the sample with a 0.1 M phosphate-citrate buffer at pH 4.5 containing freshly prepared
sodium metabisulphite (Na2S20S), and then autoclaving the mixture for 10 min at 121°C.
The metabisulphite converts the various cobalamins to the more stable sulphitocobalamin as
soon as they are released from their association with proteins. For the determination of
vitamin B12 activity in milk-based infant formula, protein is removed by filtration after
adjustment of the autoclaved extract to the point of maximum precipitation (c. pH 4.5).
Methods in which the sample is heated on a boiling water bath, rather than autoclaved, may
not completely extract all of the bound vitamin. Prior treatment with Takadiastase may be
employed for starchy samples that yield turbid extracts after filtration or centrifugation

For the microbiological determination of vitamin B12 in bovine milk using L. delbrueckii,
Gregory (1954) added 1 ml of 0.1 M sodium acetate buffer pH 4.6 and one drop of 1 % w/v
sodium cyanide to 1 ml of the test milk. The mixture was steamed for 30 min, and then
cooled, diluted, readjusted to pH 4.6 and filtered. An acetate buffer cannot be used in the
assay procedure published by the Analytical Methods Committee (1956), as the assay
organism, O. malhamensis, is inhibited by the presence of acetate in the medium. In this
procedure the test material is mixed with water and sodium cyanide, and the pH is adjusted to
between 4.6 and 5.0 with 1 N HCl. The mixture is allowed to stand for 30 min at room
temperature with occasional shaking, after which it is heated on a boiling water bath for 30
min. The cooled extract is diluted and then centrifuged. Starchy samples can be cleared by
treatment with Takadiastase.

[Numerous extraction procedures have been proposed but a common approach involves
addition of an acetate buffer to the sample followed by a small amount of sodium cyanide and
immersion of the mixture in a boiling water bath for thirty minutes. After cooling, the
mixture is diluted to a suitable volume and filtered. The cyanide converts the unstable
cobalamins into cyanocobalamins which are more stable and more closely match
cyanocobalamin which is used to construct the assay calibration line].

18