The concentration levels of many vitamins in foods and even in vitamin-enriched foods can be extremely low and, therefore, often require highly sensitive tests. Dr. Jeanne Rader of the Science and Applied Technology division in the Office of Food Labeling at FDA in Washington says the sample matrix often determines which vitamin test procedure is most applicable. For example, HPLC is preferred when measuring folic acid in vitamin tablets, but microbiology-based tests are better for measuring the amount of folic acid in spinach. Determining the level of vitamins that occur naturally in foods is much more difficult than analyzing pharmaceutical products or fortified foods which contain vitamins added in free (unbound) forms of known structure and activity.
Naturally occurring vitamins often exist in numerous isomeric forms in food materials, and the biological activity of the various isomers can vary from 100% activity to 0% activity. Microbiology-based methods tend to measure all biologically active forms of the vitamin. HPLC can't determine which forms are biologically active. Furthermore, it is often impossible or extremely costly to purchase all the various isomeric forms of a vitamin to use as calibration standards for quantitating results. According to Rader, microbiological methods are useful for testing naturally occurring water-soluble vitamins, including biotin, folic acid, pantothenic acid, riboflavin, thiamine, pyridoxine, and cobalamin (vitamin B12).
The first step in the vitamin assay of a natural product, says Rader, is to treat the food sample with a variety of enzymes to release the vitamin which may be bound to other food chemicals. For example, proteases are used to release thiamin bound to food proteins. Once in the free state, the vitamin is ready to be analyzed by either HPLC or microbiology-based methods.
One laboratory that has considerable experience with microbiology-based vitamin tests is Corning's Hazleton Laboratories, Madison, WI. Approximately one-third of all the vitamin testing at Hazleton is based on microbiological methods, according to Doug Dillman, business division manager. "We now use different bacteria that have been genetically engineered so they can't synthesize from their media the vitamins they need to grow," he says.
"To test for all biologically active pantothenic acid isomers, for instance, you grow these genetically engineered microbes in a media free of all vitamin B12 isomers," explains Dillman. "Then you add an aqueous extract of the food material you want to test for vitamin B12 activity. The rate of growth of the microbes is proportional to the amount of vitamin B12 activity. By comparing the growth rate to media spiked with various concentrations of vitamin B12 isomers and comparing results to media spiked with only the aqueous food sample extract, it is possible to obtain quantitative estimates of vitamin B12 activity."
Although microbiological plate counting techniques were used to measure microbe growth in the past, these days labs use turbidity measurements to estimate microbe populations.
Turbidimetric techniques are much easier, faster, and usually more accurate.
"Standard calibration curves of turbidity (microbe growth) versus concentration of pantothenic acid don't show the greatest linearity," says Dillman. "They're somewhat sloppy. Therefore, we run six replicate analyses for each sample and average them for an estimate of vitamin B12 activity in the product."
Standard microbiological calibration curves for niacin, on the other hand, demonstrate excellent linearity and have extremely good linear correlation coefficients.
"The major advantage of micro tests is that all biologically active forms of the vitamin can be monitored," says Dillman. "Actually we are running a mini-bioavailability study each time we perform the analysis."
Vitamin B12, for instance, is actually a group of compounds known as cobalamins. These compounds are important components of several enzymes and aid in the metabolism of certain amino acids. Examples include cyanocobalamine, hydroxycobalamine, nitritocobalamine, the co-enzyme form of B12 (with an adenosine group), and methylcobalamine. In addition, dibencozide is a form of vitamin B12 used by some vitamin supplement companies. Since microorganisms don't respond to dibencozide, HPLC methods must be employed if this vitamin B12 isomer is to be measured.
This is a point of controversy with microbiological tests, says FDA's Rader. Biological activity for microorganisms and biological activity for mammals may be different. So even microbiological test results are estimates of vitamin activity.
"Micro has its place, and HPLC has its place," says Dillman. "HPLC is usually more expensive to run but is less time consuming. I'm a chemist and don't like to admit it, but micro methods definitely have a place in vitamin testing."
Testing water soluble vitamins in vitamin-supplemented foods like cereals is most accurate and most cost effective when testing is performed by HPLC rather than micro methods. The reason is that vitamins aren't bound up the same way as in natural products so they're easier to extract from the food matrix. Also, the vitamins are added in known isomeric forms, and obtaining appropriate standards is an easy matter. Under these conditions, HPLC methods are more accurate, cost effective and demonstrate greater precision than microbiological tests.
Winston Laboratories, Ridgefield Park, NJ, handles a large variety of routine, as well as non-routine, requests from food processors and has been exclusively in the food testing business continuously since 1920. It doesn't use micro testing at all for vitamin analysis.
"To get good results, you have to do it often on a regular basis," says Marvin Winston, president of Winston Labs. "It's both an art and a science to get micro testing to work. Tests are susceptible to contamination. Microorganisms can be finicky, and they don't always behave the same way time after time."
Winston favors HPLC instrumental methods. One change that's needed in vitamin analysis is that the CARR-Price method is still an official method, and it shouldn't be. "We don't perform that test for clients any longer," says Winston. "There's too much error involved for most food products. HPLC results are far more accurate."
Poor precision, questions regarding the true mammalian biological vitamin activity, long assay times and complicated procedures are all drawbacks of microbiology-based vitamin tests.
Sandy Bailey, group leader, food sciences, at Lancaster Laboratories, Lancaster, PA, says her lab is seeing a lot of requests for testing antioxidants like vitamin C and b-carotene, as well as vitamin E isomers including a, b-, g-, and d-tocopherol. The company is evaluating new HPLC columns and new extraction techniques for fat soluble vitamins prior to injection into an HPLC. One question that needs more study, according to Bailey, is whether or not saponification of the sample is necessary prior to extraction of the fat soluble vitamins.
Some labs are extracting fat soluble vitamins from foods using carbon dioxide in the form of a supercritical fluid. This technique, called supercritical fluid extraction (SFE), is rapid and generates little or no hazardous reagent and solvent waste.
"More and more labs will be using SFE in the future," says Dr. James Tanner from the office of Special Nutritionals, Center for Food Safety and Applied Nutrition, FDA. "It's generally more rapid than traditional solvent extraction and doesn't require costly disposal of chemical reagent wastes."
Another type of chromatography that may be more widely used for the analysis of water soluble vitamins in the near future is capillary electrophoresis. The technique, which is a hybrid of HPLC and electrophoresis, is capable of superior peak resolution and shorter analysis time compared to conventional HPLC methods.
Sample preparation methods prior to HPLC need further study. Inappropriate sample preparation techniques can transform a vitamin structure from a biologically active form (e.g., trans-beta-carotene) to an inactive isomeric form (cis-beta-carotene). The use of solid phase extraction has proven to be an efficient technique for simplifying sample clean-up prior to HPLC analysis.
Another problem is that different test methods measure different isomeric forms of vitamins, and not all forms are of equal biological activity. The old colorimetric (CARR-Price) method for measuring vitamin A activity in foods has significant shortcomings. For example, the cis-retinol isomer has the same test response as the all trans-retinol isomer. Because the cis isomer of vitamin A has less biological activity than the trans form, the CARR-Price method overestimates the vitamin A activity of foods that contain significant amounts of the cis isomer. Both methods have official approval even though results obtained from the two methods can be quite different.
Similarly, the approved spectrophotometric test procedure for measuring b-carotene, the primary contributor to vitamin A activity in plant materials, can overestimate vitamin A activity because the test cannot distinguish between the various types of carotenoids present.
There are dozens of carotenoid compounds that may be present in foods, and they have a wide range of biological activity. While HPLC methods often can detect all or most of the carotenoid compounds present, it is no simple matter to decipher which peak corresponds to which isomer. And even when this is possible, it is frequently impossible to quantitate the level of specific carotenoids because suitable pure standard compounds are unavailable. Usually chemists that employ HPLC techniques only consider trans-beta-carotene when calculating vitamin A activity. They sometimes include alpha-carotene, which has half the vitamin A activity of trans-beta-carotene, but almost always ignore cis-beta-carotene and the vitamin A activity contribution of other carotenoids.
Nonetheless, HPLC will undoubtedly prove to be an invaluable tool in understanding the role of carotenoids in human health and nutrition. For example, Harvard medical researchers recently discovered that lycopene, the carotenoid compound that gives tomatoes their red color, provides a protective effect against prostate cancer, a form of cancer that kills more than 40,000 men each year. Lycopene is twice as abundant in human blood as b-carotene, and it is the most abundant carotenoid in the prostate gland. Other carotenoid compounds may offer similar health benefits, and HPLC is a good tool for measuring these compounds in various food sources, as well as in human blood.
Kjeldahl vs. Combustion
The analysis of total protein in food products has traditionally been accomplished with the official Kjeldahl method. While the Kjeldahl technique is accurate, reliable and rugged, it does have some significant disadvantages. It is labor intensive; requires long analysis time; is relatively dangerous to perform; and generates significant waste products (acids, alkalis, copper, mercury, and/ or selenium).
The steps involved in Kjeldahl analysis include sample digestion in boiling sulfuric acid, neutralization with sodium hydroxide solution, distillation of the resulting ammonia gas into a trapping solution, titration with an acid solution, and finally determination of the amount of nitrogen and protein by calculation. The entire process may require over four hours and, excluding the digestion step, needs direct technician involvement. Handling flasks of boiling sulfuric acid and adding concentrated caustic solution to hot sulfuric acid make it one of the most dangerous tests performed in the food lab. Furthermore, Kjeldahl tests generate toxic waste, since mercury or selenium catalysts are used during the digestion step. The advent of the highly touted "automated" Kjeldahl instruments in recent years hasn't significantly improved any of these deficiencies.
More and more, combustion-based analyzers are replacing Kjeldahl systems in food laboratories. Actually, the Dumas combustion method for protein is not new; it was developed over 100 years ago - several years before the Kjeldahl technique. In the Dumas method, a sample is burned in an oxygen-rich atmosphere, the amount of nitrogen gas is measured, and the total protein present is calculated from the nitrogen content.
Early attempts at using the Dumas combustion method provided imprecise and inaccurate results, and the technique was soon abandoned. However, new high quality instruments have dramatically improved the accuracy and reliability of the Dumas test - and offer a significant advantage over the Kjeldahl method. With the combustion technique analysis time is only three to four minutes, compared to two to four hours for Kjeldahl; no hazardous waste is generated; cost per analysis ranges from 37 to 50 cents (for Kjeldahl it is approximately $1 per test); and incorporation of an autosampler is possible, allowing analysts to run over 100 samples unattended.
Leco Corp., St. Joseph, MI, the market leader in combustion nitrogen analyzers, markets an instrument called the FP-428, which provides protein results in approximately three minutes. Beverly, MA-based Fisons Instruments' model NA 2000 nitrogen and protein analyzer is another automated state-of-the-art combustion analyzer capable of high sample throughput. The major difference between instrument designs from various vendors is how total nitrogen created during the combustion step is measured.
Results from the combustion-based instruments show close agreement with Kjeldahl results for most sample types. In one study, 45 different meat and dairy food products were analyzed by the official AOAC International Kjeldahl method and by a combustion method. The samples covered a wide range of protein levels, from 3% to 90%. Nine different laboratories were used in the study, and all labs used the Leco FP-428 analyzer. Repeatability standard deviation ranged from 0.11 to 0.40 for Kjeldahl results and 0.12 to 0.41 for combustion results. Reproducibility standard deviations ranged from 0.20 to 0.49 for the Kjeldahl method and 0.18 to 0.46 for the combustion method.
Comparison of means, however, shows a trend for combustion values to be slightly higher than Kjeldahl - a difference that becomes statistically significant at high protein levels. Several non-protein nitrogen compounds might be recovered by combustion but lost in Kjeldahl digestion. One example involves nitrogen oxides, which often occur in cured meats in the form of sodium nitrite or sodium nitrate. A detailed comparison of the mean protein value for both techniques is shown in the accompanying table.
As with vitamin assays, the choice of which analytical method to use for protein assays is strongly dependent on the sample matrix. For fluid milk samples and many semi-fluid dairy products, for example, one of the newly developed Fourier transform infrared spectroscopy (FTIR) instruments should be considered. Near infrared analyzers appear to provide rapid and reliable results for cereals and grains. Laboratories that test a wide variety of different sample types should evaluate Kjeldahl or combustion instruments.
The level of protein in the food sample, the amount of moisture in the sample, possible interfering compounds, the homogeneity of the sample, the quantity of samples tested per day, and other factors must all be carefully considered when selecting the optimum test method for protein analysis.
The combustion technique for crude protein has received AOAC approval for cereal grains and oilseeds, meat/meat products, and animal feed. It is currently being reviewed for dairy products. In July 1994, the U.S. Dept. of Agriculture's Federal Grain Inspection Service (FGIS) replaced the Kjeldahl method with the combustion method as its reference method for determining protein content in wheat and soybeans. Kjeldahl's dominance as the favored test method for crude protein analysis may soon be over.
values =MEAN PROTEIN (%)
.png?width=850&auto=webp&quality=95&format=jpg&disable=upscale "Are New Methods Better?")
.png?width=800&auto=webp&quality=80&disable=upscale)