Analysis of Thiamin by Reverse Phase C18 Open Column Chromatography

A modification of the AOAC method for thiamin determination utilizing reversed phase (RP) c18 packing material was developed for analysis of milk, infant formula and breakfast cereal products. Thiamin was extracted from food samples and hydrolyzed following a similar Association of official Analytical Chemists (AOAC) (1984) procedure. The sample filtrate was purified by passing through the reversed phase c18 column. An isocratic mobile phase consisting of 3% KCl and methanol (70 30) was used to elute thiamin from the reversed phase c18 column. The sample eluate was oxidized to thiochrome and quantified on a fluorometer using similar AOAC procedure. The experimental technique was compared to the existing AOAC (1984) method, which employs a Bio-Rex 70 resin in the purification step of thiamin analysis. There was no significant difference in the thiamin content of the samples determined by the two methods. However, there was significant difference in the re~overy of thiamin, when taken through the method or added to food samples before extraction. The recoveries of the added thiamin were 97.97 ± 0.69% and 92.67 ± 1.25% for RP c18 and AOAC (1984), respectively, the coefficient of variation was 0.70% and 1.34% for RP c18 and AOAC (1984) methods respectively.

The RP c 18 method was found to be better than the AOAC (1984) method in terms of accuracy, precision, and reproducibility. And the RP c 18 proved to be faster than the AOAC ((1984)    form; it is also added to some foods as an essential nutrient. Thiamin is required for the normal health and growth of humans.
The establishment of food labelling, in· the last two decades , has resulted in a growing need for determination of micro-nutrients in labelled food products. This has created a need for methods that are fast, accurate, and applicable to different kinds of food matrices. Such methods should not require costly reagents or equipment, and should be suitable to both well and poorly equipped laboratories (Sebecic and Dragojevic, 1986 (Mauro and Wetzel, 1984 (Ayi, et al., 1985;Hilker and Clif ford, 1982;Ohta , et al., 1984) and combined with other water soluble vitamins (Ang and Moseley, 1980;Augustin, 1984;Fellman, et al., 1982;Fingals and Faulks,1984;Kamman, et al., -2-;Mauro and Wetzed, 1984;Skurray, 1981;Toma and Tabekhia, 1979;Wills, et al,. 1977;Wimalasiri and Wills, 1985 ). This technique has proved to have advantages over the official AOAC method in terms of sensitivity, speci ficity, accuracy and speed (Fellman, et al., 1982;Ohta, et al., 1984;Polsello and Rizzolo, 1986;), however the initial capital outlay and subsequent recurrent cost of maintaining the HPLC instrument are relatively high, and the equipment must be used by a skilled operator .
A comparison study (wills, et al., 1985) was made between HPLC, using a reverse phase c 18 column, and the AOAC (1980) methods for thiamin determination in different foods . The HPLC system gave thiamine values and recoveries which were 10% higher than those obtained by the AOAC procedure.  determined thiamin in potatoes by the HPLC system, equipped with a reversed phase c 18 column, which gave a recovery value of 95.2%, HPLC also gave higher thiamin values than those obtained by the AOAC fluorometric method. Wehling and Wetzel, (1984) reported similar thiamin levels in fortified cereal products analyzed with HPLC, using a reversed phase c 18 column, and AOAC techniques.
-3-In another study (Ayi, et al., 1985), the HPLC, with zorbax cN, 6µm column, and AOAC methods were utilized to determine vitamin B 1 content in infant formula products.
The results obtained by both methods were in agreement with those obtained by Wehling and Wetzel, (1984). HPLC recovery studies yielded values of 100 -102%.   which has been used in HPLC as th~ stationary phase by many investigators to separate and analyze thiamine in food matrices Ellefson, et al., 1981;Mauro and Wetzed, 1984;Wills, et al. , 1987;Wimalasiri and Wills, 1985). This material was packed into an open column and used in the purification -4-step of the AOAC procedure for thiamin analysis in milk, infant formula and breakfast cereal products.
The RP c 18 (50µ) packed into an open column has been used in this laboratory to separate carotene (Tsai et al,19 a9) and retinal (Al-Abdulaly and Simpson, 1989) from various foods. Furthermore it was used to separate · ribofl avin, a water soluble vitamin, from milk and milk products (Saibu, 1988

Methods:
All sample preparations were performed under subdued incandescent light.
Two methods for thiamin analysis were compared. Sample preparation, extraction, hydrolysis, oxidation to thiochrome and measurement on fluorometer were the same.
The only difference between the two methods was in the

Sample Extraction:
A modified AOAC method was used for extraction, enzyme hydrolysis, and oxidation of the samples.
-Extraction and hydrolysis of skim and whole milk: A 25 ml prepared sample were put in 250 ml Erlenmeyer flask. To the flask 50 ml 0.14 N HCl was added. The flask was covered with aluminum foil and -9-autoclaved at 121°C for 15 min. After cooling the extract to 50°c or lower, 5 ml of 2.5 N NaOAc were added to adjust the pH to 4.5 -5.0, which is appropriate for the enzymes and five ml of freshly prepared enzyme suspension were added. The solution consisted of two enzymes, 1-6% taka-diastase (phosphatase enzyme) to hydrolyze any phosphate esters of thiamin present to its free form. This step was necessary because thiochrome phosphates are not extracted by isobutyl alcohol (Hennessy and Cerecedo, 193 9;Rindi and deGiuseppe, 1961;, and 2-6% papain (proteolytic enzyme) to release thiamin bound to protein. The contents of the flask were mixed and incubated at 47°C for 3 hrs in a water bath, cooled to room temperature and transferred to a 100 ml volumetric flask. The extracts were then mixed thoroughly and filtered through No. 40 Whatman ash-free fi lter paper, which does not absorb thiamin, and the fi ltrate was stored in a refrigerator at 4°C.

-Extraction of infant formula (Enfamil and similac):
Eight ml of infant formula concentrate were used. The samples were extracted and hydrolyzed the same way as skim and whole milk, except that only one enzyme, Taka-diastase, was used in the hydrolysis step.
-10-Extraction of breakfast cereal (Kelloqq•s corn flakes): rnorder to facilitate the vitamin extraction the breakfast cereal was ground to a fine powder in a coffee beans grinder, to pass a 32 mesh screen. A 1 g sample was used. It was extracted and hydrolyzed as above, except that only one enzyme, Taka-diastase, was used in the hydro lysis step.
_ Extraction and hydrolysis of thiamin working standard: To a 250 ml Erlenmeyer flask, containing 45 ml 0.14 N HCl and 5 ml 2.5 N NaOAc, 5 ml thiamin intermediate standard (2 ug/ml) were added. It was extracted and hydrol yzed the same way as skim and whole milk, except that only Taka-diastase was used in the hydrolysis step.
Puri fication through the RP c 18 column: A glass chromatography column, fitted with a 50 ml reservoir at the top and a column (160 mm long * 11.5 ml od.) drawn into a capillary at bottom was used, It was fitted with a piece of glass wool, placed over the upper end of the capillary. The column was packed with RP c 18 to a height of 6 cm. The surface of the stationary phase was covered with another piece of glass wool. Nitrogen gas was used to control column flow rate, which was 1 ml / min. or less. Different mixtures of solvents were examined -11-to choose the right mobile phase to elute thiamin from the RP c 18 column. In another RP c 18 column, 5 ml thiamin working standard filtrate were run through the column in the manner described above.
The RP c 18 packed in the column was cleaned after elution with mixture of chloroform: methanol (60: 40), and then rinsed with a mixture of methanol water (70

Purification through the Bio-Rex 70 resin:
A glass column, fitted with 50 ml reservoir at top and a column (180 mm long * 9 mm od) drawn into a capillary at bottom, was fitted with a piece of glass wool placed over upper end of capillary. It was packed with Bio-Rex 70 -12-resin to a height of 10 cm. The flow rate was 1 ml/min, it was controlled by attaching polyethylene tube with metal clamps to the tip of capillary and adjusting the dial for flow rate.
A 10 ml sample filtrate was pipetted into the Bio-Rex 70 column. The filtrate was discarded, the column and reservoir were washed with three 5 ml portions of hot Ho (63°C), and the eluate was also discarded. Thiamin 2 was eluted from the resin by passing five 4.5 ml portions of hot (63°C) acid-25% KCl solution through the column.
The eluate was collected into a 25 ml volumetric flask, cooled and diluted to volume with acid-25 % KCl solution.
In another Bio-Rex 70 column 10 ml thiamin working standard filtrate were run through the resin as above.
The Bio-Rex 70 resin packed in the column was washed between runs with 1 N HCl, rinsed with deionized water, re-equiliberated by depacking th~ resin, and then placed in a beaker to be stirred with deionized water for 1 min before decanting. H 2 o washes were repeated until excess acid had been removed. The resin was repacked in the column {Guide to Ion Exchange, Bio Rad).
-13-conversion to thiochrome: To each of two 35 ml centrifuge tubes, 1.5 g NaCl and 5 ml of sample eluate were added. To the first centrifuge tube, 3 ml of the oxidizing reagent (1 % alkaline ferri cyanide solution) were added. The tube was swirled to ensure adequate mixing. Immediately, 11 ml isobutanol were added . The tube was stoppered and shaken vigorously for 90 sec. Thiamin is oxidized to thiochrome by potassium ferri cyanide in the presence of strong alkali Rosenberg, 1942). Thiochrome is soluble in isobutyl alcohol.
The second centrifuge tube (sample blank) was treated similarly, except that the oxidizing reagent was replaced with a 15 % NaOH solution.
To each of another 2 centrifuge tubes, 1.5 g NaCl and 5 ml thiamin working standard eluate were added. These tubes were treated in the same manner as mentioned above for the tubes containing the sam~le eluate. After the addition of isobutanol to all tubes, the tubes were shaken aga in for 2 min in a shaker box then centrifuged for 1 min.

Measurement of thiochrome:
Thiochrome fluorescence was measured on the The results of the RP c 18 and Bio-Rex 70 methods were expressed as mean ± SD and compared using the t-test. -17-

RESULTS AND DISCUSSION
The thiamin standard was first checked by UV/VIS absorption to confirm its purity. The resultant spectrum was found to be the same as documented in the literature.
The Taka-diastase and Papain were proved to be thiamin free . Taka-diastase was found effective in releasing phosphates from thiamin. The recoveries were 94% and 95% from thiamin monophosphate and thiamin pyrophosphate respectively. Therefore the Taka-diastase was considered suitable for the dephosphorylation step .
The sample eluted from the RP c 18 column was evaporated to near dryness to remove methanol present in the mobile phase ( 3% KCl: methanol 70: 30 ). Methanol was found to increase the fluorescent intensity of the sample, resulting in higher and inaccurate thiamin content calculation. Figure 1 shows the fluorescence spectrum of the Enfamil sample, eluted from the RP c 18 column, (a) after evaporation and (b) without evaporation. These spectrums clearly demonstrate the methanol effect on thiamin measurements (there was a 10% increase).
-18-    The recoveries of thiamin added to four replicate sampl es of each type of food analyzed, were determined.
The results, summarized in     (1986 ) found thiamin in breakfast cereal to be 171% of the value declared on the label. This was further supported by the findings of , They determined the thiamin content to be 141% of the declared leve l on the breakfast cereal label.
Thiamin has particular stability problems. Heat causes thiamin losses during the production process. So the manu facturers of the fortified products must take into account potential losses that may happen during the processing and storage of the product. Accordingly, declared levels can be lower than the amount present in the product (Martin et al, 1987).
Skurray (1981) Kamman et al, (1980) in agreement except for the corn flakes breakfast cereal.  , (1986). Liquid chromatography analysis of thiamin and its phosphates in food products using improlium as an internal standard. J.
Micronutr. Anal., 2, 189-99. Wehling, R., & Wetzel, D., (1984 condition known as beriberi. This is characterized by loss of appetite and weight (Lehninger, 1982). Polyneuritis involving degeneration of the pripheral nerves, coupled with high concentration of lactate and pyruvate in the blood, may develop in the latter stage of deficiency .
Thiamin pyrophosphokinase metabolizes thiamin to thiamin pyrophosphate (TPP) in animal cells (Kawasaki and Sanemori, 1985). TPP, the coenzyme form of the vitamin, plays an important role in glycolysis and the glycolytic pathway, the citric acid cycle and the pentose pathway (Lamden, 1972 (Lamden, 1972). In plant tissue thiamin is mostly found in the free form, while the most abundant form in animal tissue is the pyrophosphate.
Thiamin is found in many plants. Fruit and vegetables contain small amounts, while the outside coats of grains have high amount of the vitamin. In animals, thiamin is found in various organs ( heart, liver, kidney, and brain) (Rosenberg, 1942).
Foods like nuts, pork, yeast and cereal germs are particularly rich in thiamin   (Lamden, 1972).
Animal assays were the first methods developed to measure thiamin contents in food. Animals that have been utilized are the chick, the pigeon and the rat (Lamden, 1972). The most used animal assays were growth measurements and curative tests  . Animal assays are useful because they are specific for thiamin and are of important in determining all forms of thiamin that are physiologically available to animals. Also, because animal assays determine all biological forms of thiamin, there is no need for the extraction or pretreatment of the sample as required by other methods (Lamden, 1972).
The main disadvantages of these .assays are time required and high cost; The growth test takes 6 to 8 weeks, while the curative test is slightly faster. The use of large numbers of animals and specially prepared diet, coupled with the long feeding time needed, caused these tests to be expensive .
Different microbiological methods have been employed in thiamine determination. Fermentation and the growth of -51-or acid production by bacteria, yeasts, molds or fungi have been used . These methods are less expensive, faster, more sensitive, and give more reproducible results than animal methods. An microbiological assay takes from 4 hours to 3 days to complete. The main disadvantage of these methods is the effect of other materials such as the breakdown products of thiamin The microorganisms respond to these products in the same manner as thiamin, thus, these methods are not often reproducible .
Most of the colorimetric methods suggested for determination of thiamin depend on the reaction of the vitamin with a diazoitized reagent.
The most successful reagent was diazotized p-aminoacetophenon. It was introduced by Prebluda and Mccollum (1939). In alkaline solutions the reagent reacts with the thiazole portion of thiamin to form a purple-red color. The colored compound produced is insoluble in water. This reaction was developed into a satisfactory quantitative method by . They described a procedure for the extraction of the purple-red colored compound with exylene, and the purification and concentration of thiamin by means of zeolite adsorption.
It is the most widely used colorimetric method for thiamin determination.
-52-In colorimetric method larger amount of thiamin (20 -100 µg) is required in the test sample than for the thiochrome method. The determination although specific, is rather complex, tedious and time consuming (Mickelsen and Yamamoto, 1958).
In thiamin determination both colorimetric and thiochrome methods have similar initial steps. These include: extraction with dilute hydrochloric acid or sulfuric acid; hydrolysis of the phosphate esters of thiamin by phosphatase containing enzyme; adsorption on base or ion exchange column and elution of thiamin (Lamden, 1972).
Thiochrome method for thiamin determination had its beginning in the work of Peters in 1935, who discovered that the oxidation of thiamin resulted in its conversion to a strongly blue fluorescent material which was called thiochrome.  was the first to utilize this property of thiamin as the basis for a quantitative estimation of thiamin. Hennessy and Cercedo (1939) improved the method by introducing an enzyme for the hydrolysis of the phosphate esters of thiamin and used a base exchange zeolite to separate thiamin from impurities which interfere in the determination. Although their -53-procedure was studied and occasionally modified, it is still the basis for thiochrome method. Recently  ion exchange resin has replaced Decalso resin used in the purification step (Effelson et al, 1981).
The thiochrome method is the standard method of the Association of Official Analytical Chemists (AOAC). The manual thiochrome method will measure 1 to 20 µg of thiamin in the test sample (Lamden, 1972). The advantages of this method over the biological and colorimetric methods are: rapid analysis, good precision and a wide range of applications (Ellefson, 1981) Gas Chromatography (GC) was also applied for thiamin measurment. A nitrogen-phosphorus detector has been used to detect thiamin in meat, vegetables and cereals (Echols et al., 1983) and in milk (Echols et .al., 1985).
Advantages of the GC are simplicity of procedure, ease of standarization, and lack of cleanup. However, disadvantages of the of the GC procedure are the relatively large amount of sample needed, the high cost of the instrumentation and the long time required to run a GC analysis (Echols et al., 1983).
In recent years there have been an increase in the number of analytical methods utilizing high performance -54-liquid chromatography (HPLC) for thiamin determination in food. HPLC methods usually utilize the same acid extraction and enzyme hydrolysis used in the AOAC method, followed by protein precipitation with trichloroacetic acid when oxidation to thiochrome was adopted Ohta et al.,1984).
However, when the purpose was to measure non-derivatized thiamin, acid extraction followed by a clean-up and concentration step through disposable columns , decreased the analysis time. This extraction technique allowed the detection of thiamin phosphate esters as well as free thiamin by a UV detector Panijban et al., 1982;Vanderslice and Huang, 1986).
The HPLC method has many advantages over the manual AOAC thiochrome method such as direct analysis without derivatization, simultaneous determination of different vitamins in a single analysis, a reduced analysis time and good precision and accuracy (Polsello and Rizzolo, 1986).
However, the initial capital cost and subsequent recurrent cost of HPLC instrumentation are relatively high and a skilled operator is needed .
The RP c 18 (50µ) packed into an open column has been used in this laboratory to separate carotene (Tsai et al, 1989) and retinal (Al-Abdulaly and Simpson, 1989) from various foods. Furthermore, it was used to separate riboflavin, a water soluble vitamin, from milk and milk products (Saibu, 1988). -56-