Determination of Retinol by Aqueous Reverse Phase Open Column System

colorimetric and fluorometric methods are most commonly used for vitamin A analysis. The colorimetric method has disadvantages which include rapidly fading blue color, sensitivity of reagent to moisture and cost of the reagent. The recognized fluorometric method, the AOAC procedure, employs column chromatography before the fluorometric determination to improve sensitivity. This \ method is time consuming and requires technical skill. The more recently developed HPLC method is rapid, reproducible and quantitative, but the instrument is expensive and requires skilled operators. It was found that packing the HPLC adsorbent (50 um C1s) in the open column and eluting it with isocratic, aqueous solvent system (methanol/water) as the mobile phase gave good separation of retinal from the extracted sample. Retinol values of milk and infant formula obtained by this method were very similar to that of AOAC and HPLC methods. Retinal correlation coefficient between RP-C1s and HPLC estimates and RP-c 18 with AOAC estimates were 0.988 and 0.999 respectively.


THESIS ABSTRACT
The AOAC method and other open column techniques which are used to determine retinol content, are time consuming, require technical skills, and do not always result in a complete separation of retinol from other fluorescent components. In some cases cis-trans isomers are formed due to the long term exposure of retinol to oxygen, light, adsorbents and solvents. The colorimetric method hes disadvantages which include a rapidly fading blue color, sensitivity of the reagent to moisture, and interference from carotenoids.
The more recently developed HPLC method is rapid, reproducible, nondestructive, quantitative and gives high resolution of closely related compounds. The instrumentation, however, is expensive to purchase and maintain, uses expensive solvents and is not available to nutritional scientists in many parts of the world. Therefore there is clearly a need for a method that has the simplicity of the AOAC method coupled with the accuracy and speed of the HPLC method for the determination of retinol in food products and serum.
It was found that packing the HPLC adsorbent (50 um C1a) in an open column and eluting it with an isocratic, aqueous solvent system consisting of methanol and water as the mobile phase gave good separation of retinol from the extracted sample.
The method was evaluated by measuring the retinal concentration in milk, infant formula, margarine, egg yolk and liver and was compared with both AOAC and HPLC methods.
The method was used to measure the retinal concentration in serum as compared to the HPLC method.
The retinal values of milk, infant formula, egg yolk, margarine and liver obtained by RP-C18 open column were \ very similar to that of AOAC and HPLC methods, and there was no significant differences among these meth .ods when compared over a set of samples. A correlation coefficient between RP-c 1 a and HPLC estimates and RP-C1a and AOAC estimates were 0.993 and 0.999, respectively, indicating that a highly significant correlation exists between these methods for the determination of retinal from the different samples. The study showed that the RP-C18 method is comparable with both the HPLC and the AOAC methods for retinal analysis from these food products.
The recovery study for retinal from the analyzed samples was found to be 97%, 98% and 96% for RP-C 1 s, HPLC and AOAC respectively, indicating that a good recovery was obtained by this method.
Also the study showed that the retinal values of serum obtained by RP-c 18 method were very similar to that of HPLC and indicated no significant differences among the two methods when compared over a different set of samples.
A correlation coefficient between the RP-C1a and HPLC estimates was 0.963 for retinal analysis, indicating a highly significant correlation between these two methods.
Also the study showed that the RP-C1a method gave  Table 1. Retinol content of milk and infant formula determined by reverse phase INTRODUCTION Vitamin A is found in animal products such as milk.
It is generally added to milk that is partially or wholly defatted and to infant formulas. It is added in specified amounts to these products usually as retinyl palmitate.

'
It is time consuming and in some cases cis-trans isomers are formed due to the long-term exposure of vitamin A to oxygen, light, adsorbents and solvents , Thompson and Maxwel 1 1977).
The recent applications of HPLC to the analysis of biological materials has provided the nutritional biochemist with highly useful techniques with a wide variety of applications. The advantages of HPLC over other methods includes its speed, high resolution of closely related compounds and a good recovery of vitamin A in foods (Van DeWeerdof et al. 1973, Frolik andOlson, 1984). The disadvantages of HPLC are that it is expensive to purchase and maintain, and is not available to nutritional scientists in many parts of the world.
Shu-Whei Tsai (1986) reported that packing the HPLC adsorbent (C 18 ) in an open column and eluting the sample 4 with an isocratic, obtained with less than 1.5% variation in results found between these two methods indicating that they are comparable with each other.
The comparison between RP-c 1 8 and AOAC method also presented in Table 3. The actual values ranged from 0.984 to 1.034 with an average value of 1.008 indicating that the two methods are equal since the value is close to one.
The standard deviation of 0.017 and a coefficient of variation of 1.7% were obtained with less than 2.6% variation in results found between these two methods, indicating that they are comparable with each other.   '' \ l ' I ., ..    . These chromatographic methods may or may not result in complete resolution of closely related compounds, and in some cases tis-trans isomers are formed due to the long exposure to solvents, adsorbents, light and oxygen .
In 1977 Thompson and Maxwell introduced a reverse Phase HPLC system for the determination of vitamin A from saponified margarine. Thompson et al (1980)  Vitamin A is found in animal products such as liver ' of lamb and chicken as mixed esters of vitamin A mainly as retinyl palmitate and in egg yolk as retinol with lesser amounts of retinyl esters and retinaldehyde (Parrish, 1977  occasional mixing .

II. Egg Yolk and liver
One gram of sample was ground with anhydrous sodium sulfate to a dry powder in a mortar. The lipids were extracted from the dried material with ethyl ether until no fluorecence appeared to the extract. The extract was 29 filtered to remove insoluble material, and then evaporated to dryness in a round bottom flask using a rotary evaporator ). Sufficient methanol was added to the residue to dissolve it and then 60% (W/V) aqueous KOH was added, 1 ml to every 10 ml methanolic solution. The alkaline mixture was shaken and immersed in a 7ooc water bath for 30 min. with occasional mixing (Bauernfeind, 1972).

'
Extraction: The saponified sample was cooled and transferred to 500 ml separatory funnel. Ethyl ether (50 ml
In spectrophotometric method, it is assumed that nothing except vitamin A is absorbed in the 325 to 330 nm region. Various lipids, sterols and vitamin E and D absorb in the same general region of the spectrum as vitamin A (Parrish, 1977). The fluorometric method does not always result in a complete separation of vitamin A from other fluorescent components, time consuming and in some cases cis-trans isomers are formed due to the long term exposure of vYtamin A to oxygen, light, absorbents and solvents . The colorimetric method has disadvantages which include rapidly fading blue color, sensitivity of the reagent to moisture, interference from carotenoids, time consuming and cost of the reagent (Parrish, 1977).
The recent application of high pressure liquid chromatography (HPLC) to analyses of biological material has provided the nutritional biochemist with highly useful 50 I I I techniques with a wide variety of applications.
The advantages of HPLC over other methods include its speed, high resolution of closely related compounds, nondestructive conditions, applicability to very small samples and selective quantitation of retinal in blood based upon retinol's high intrinsic UV absorption   .
The disadvantages of HPLC are that it is expensive to purchase an~ maintain, and is not available to nutritional scientists in many parts of the world.

Separation of Serum
The serum was separated by centrifugation of blood at 3000 RPM for 15 minutes in a cold room. Serum was transferred to a test tube, the head space was filled with nitrogen gas and the tube was tightly sealed. The serum was stored at -20°c until analysis.

Extraction of serum
The extraction procedure used was similar to that outlined by  and Woollard G.A. and Woollard D.C. (1984) with a modification.  times according to the procedure described in Arroyava et al. (1982). The concentration of the purified retinal was measured by a spectrophotometer set at 325 nm. The extinction coefficient for all trans retinal in petroleum ether is 183Q (Barua et al., 1973). The purified retinal was dissolved in chromatographic solvent system (ACCN:DCM:MEOH 70:20:10) .
Peak areas were measured by spectrophotometer set at 325 nm for retinol. A standard curve was prepared from peak area versus a known concentration of injected retinal.

Statistics
The results are expressed as mean.        (Williams and Caliendo, 1984).
Vitamin A is found in animal products such as liver, milk, butter and cheese as mixed esters of vitamin A and in egg yolk as retinal with lesser amounts of retinaldehyde and retinyl esters (Parrish, 1977).
Vitamin A is generally added to milk that is partially or wholly defatted, margarine and infant formulas. It is added in specified amounts to these products in the form of retinyl esters mainly as retinyl palmitate. Retinyl palmitate is often the main source of 69 vitamin A activity added to these foods, but sometimes the pigment 13carotene accounts for substantial fraction or all of the vitamin A activity of margarine .
Vitamin A is found in plasma in two major forms, retinal and its ester. Retinal is about 90% or more under most conditions, the retinyl ester is about 10 to 17 percent (Sommer, et al. 1977).
Since the discovery of the importance of vitamin A, ' nutritionists have been interested in the vitamin A content of foods and diets, and vitamin A has been determined in many natural and some processed foods containing both vitamin A and provitamin A. Vitamin A determination on foods has assumed greater importance because of the growing interest in nutrient requirements, dietary standards, nutrient labeling and warranties among consumers, processors, and regulatory agencies (Parrish, 1977).
In foods, the concentration of vitamin A is small in comparison to other components that may interfere with analysis or the vitamin A is held so that it cannot be extracted without pretreatment. Saponification, which is an alkaline digestion, is generally used in the vitamin A determination for foods to free the vitamin from the stabilizing matrix, from the lipids in which it might be dissolved, or from substances in the food that might interfere with the extraction.
In this process, the esterified vitamin A is converted to the alcoholic form, retinal, without destroying the vitamin (Parrish, 1977).
Several methods for vitamin A determination in blood and food products have been published, and all these methods depend on the physical and chemical properties of retinal which include the following: it is absorbed at 325 nm, it fluoresces at 480 nm, and it forms colored products with antimony trichloride and trifluoroacetic ' acid. The most common methods are spectrophotometric, fluorometric, colorimetric, and chromatographic methods (Parrish, 1977 which vitamin A reacts with antimony trichloride in chloroform yielding a blue color is by far the most widely used colorimetric method to determine vitamin A in foods (Carr and Price, 1926). This method has disadvantages which include rapidly fading blue color, sensitivity of the reagent to moisture, interference from carotenoids, and cost of the reagent (Parrish, 1977). A new colorimetric procedure was developed by Neeld and Pearson (1963) based on the blue color produced by trifluoroacetic acid in th~ presence of vitamin A. The major advantage of the trifluoracetic acid method is that the blue color is more stable and less susceptible to interference by traces of moisture.
The fluorometric method for vitamin A determination is based on the molecular fluorescence properties of the retinol structure (Thompson et al., , 1978. The presence of fluorescent components other than retinol such as phytofluene and lipids in the extracted food or blood samples raise the problem of nonspecificity. To overcome this problem, Garry et al. (1970) and  introduced a chromatographic step before fluorometry to remove the interfering substances. Today, the method for determining vitamin A in margarine recognized by the AOAC calls for fluoremetric determination following alumina column chromatography (AOAC, 1980).

72
The chromatographic methods are not complete analytical methods. They merely separate extracts of sample into components, and spectrophotometric, colorimetric or fluorometric analysis are required to measure the components separated. Both adsorption and partition chromatography are used to remove substances interfering in determination of vitamin A. Other methods include high-pressure liquid chromatography (HPLC) and gas-liquid chromatography (Parrish, 1977).
In col'umn chromatography, a number of adsorbents have been used for separating vitamin A from other fat soluble substances. Neutral alumina was one of the earliest techniques and most commonly used method to separate retinal from its esters . This method of adsorption chromatography utilizes alumina that has been deactivated, usually with 5 percent water, and most often employed diethyl ether, benzene or acetone in hexane as the eluting solvent.
Recoveries vary from 85 to 96 percent for retinal and retinyl esters (Ross and Zilversmit, 1977). Thin-layer and paper (partition) chromatography are generally not adaptable to quantitative determination of vitamin A in foods. They are used for qualitative estimations, checks on homogeneity, forms of vitamin A in extracts and presence of other vitamins and lipids (Parrish, 1977).
The most common adsorbents are silica gel and alumina. 73 The usual conditions under which thin-layer and paper separation are carried out promote oxidations and loss of vitamin A (Sajak et al., 1979). These chromatographic methods may or may not result in complete resolution of closely related compounds, require several hours and in some cases cis-trans isomers are formed due to the long exposure to solvents, adsorbents, light and oxygen. The recoveries are usually in the range of 70 to 85 percent  ' Gas-liquid chromatography has a limited use in the determination of vitamin A because of the instability of retinal to heat. Retinal and retinyl esters were reported to be rapidly dehydrated to anhydroretinol on several gas chromatographic columns even at lower temperatures as low as 15o 0 c. (Frolik and Olson, 1984 (Frolik and Olson, 1984). The disadvantages of HPLC are that it is expensive to purchase and maintain, is not available to nutritional scientists in many parts of the world, and it requires skilled operators.
Shu-Whei Tsai (1986) reported that packing the HPLC adsorbent (C 18 ) in the open column and eluting it with the isocratic, nonaqueous solvent system (Acetonitrile/methanol/ chloroform) as the mobile phase gave good separation of o(and 8 -carotene from 8cryptoxanthin and lycopene, ' and the results were the same as that of AOAC method for carrots, and very similar to that of the HPLC method for spinach and peaches.