Produced for lactose-intolerant consumers, lactose-free dairy products are manufactured by enzymatic hydrolysis, which can still leave residual levels of lactose in the product. For a product to be labelled “lactose-free”  it should be tested by an appropriately sensitive method for the absence of lactose. This article describes the use of a high performance anion exchange chromatography method with pulsed amperometric detection and the results obtained in the analysis of lactose-free and normal dairy samples.
by Pranathi Perati, Brian De Borba and Jeffrey Rohrer


Lactose, the major disaccharide found in milk products, is catabolised into glucose and galactose in a reaction catalysed by the enzyme lactase. Lactose-intolerant individuals are lactase-deficient, which results in  incomplete digestion of lactose and gives rise to abdominal discomfort. While this condition does not pose a serious risk to health, the global and ethnic prevalence of lactose intolerance has created a large market for lactose-free products, which are manufactured  by breaking down lactose into glucose and galactose by enzymatic hydrolysis. However, the resulting milk products can still contain varying amounts of residual lactose. This has created the need for simple, reliable, accurate and, most importantly, sensitive analytical methods to quantify lactose in dairy products marketed as lactose-free.
Milk undergoes structural and chemical changes when it is heat-treated, but the extent of the change depends on the time and temperature of the heating. Lactulose is a disaccharide containing galactose and fructose and is not naturally found in raw milk, but is formed from the isomerisation of lactose during the heat treatment of milk. The level of lactulose is used as an indicator of heat damage to milk in the dairy industry.

Currently, the  available analytical methods for the determination of lactose are time-consuming. Additionally, they do not differentiate between different carbohydrates and lack the sensitivity to determine lactose in products marketed as lactose-free. The study presented in this paper describes a sensitive and accurate method to determine lactose and lactulose in dairy products, including supposedly lactose-free products, using high-performance anion-exchange chromatography with pulsed amperometric detection (HPAE-PAD) which is a widely used technique for the determination of monosaccharides, disaccharides, oligosaccharides, smaller polysaccharides, sialic acids and other sugar acids. Being a direct detection technique, HPAE-PAD eliminates errors associated with analyte derivatisation [1]. The use of a CarboPac PA20 anion-exchange column in combination with PAD provides high-resolution separations of small and larger carbohydrates with sensitive detection. The method described in this study was used to determine low concentrations of lactose present in commercially available Gouda, Havarti, cottage cheeses and low-fat yogurt that are marketed as being lactose-free.
 
Experimental
Full, detailed descriptions of the experimental conditions and results of the work described in this paper are available from the authors. Basically the experimental set-up involved the use of a Dionex ICS-3000 chromatography system consisting of a Dual or Single Pump, electrochemical detector, AS Autosampler, and Chromeleon Chromatography Data System (CDS) software. Preparation of all samples and standards included treatment with Carrezz reagent and clean-up using OnGuard IIA cartridges, which remove anionic contaminants and neutralize acidic samples.

Results and discussion Chromatography and interference Studies
Our past experience indicated that if lactulose was present in the lactose-free samples, it would elute very close to lactose, and therefore we needed to be able to separate the two reproducibly. To optimize the separation of lactose and lactulose in the presence of expected sample carbohydrates, a mixed carbohydrate standard was prepared. We developed an optimized gradient to separate this mixed standard, in which the retention times of galactose, glucose, sucrose, lactose and lactulose were 9.63, 10.65, 13.79, 22.98, and 24.36 min, respectively. All the carbohydrates were well resolved from each other, including lactose and lactulose.

Column-to-column reproducibility
Due to the close elution of lactose and lactulose, the separation was investigated on three columns from three different lots to determine method ruggedness. Figure 1 shows that all three columns showed good separation of lactose and lactulose, demonstrating that the separation could be achieved with different lots of resin.
 
Short-term reproducibility
Table 1 shows intraday reproducibility measured from 30 consecutive injections of a mixed carbohydrate standard containing 30 mg/L galactose, glucose and sucrose, and 3 mg/L each of lactose and lactulose. The method exhibited good short term reproducibility based on the intraday retention time RSDs that ranged from 0.12 for lactulose to 0.16 for sucrose. The peak area RSDs ranged from 1.03 for galactose to 5.07 for lactulose.
 
Determination of linearity for lactose and lactulose
Calibration standards were prepared in deionized water. Table 2 summarizes the calibration curve data obtained by injecting standards between 0.25-100 mg/L of lactose and 0.5 – 100 mg/L of lactulose. The calibration curves for both compounds were linear with correlation coefficients (r2) of 0.9966 and 0.9942 for lactose and lactulose, respectively.

Method Detection Limits (MDLs) for lactose and lactulose
The Method Detection Limit (MDL) is defined as the minimum concentration of an analyte that can be identified, measured, and reported at the 99% confidence level that the analyte concentration is greater than zero. It is basically a measure of the precision of preparing and analyzing low-level samples according to the method. The MDLs for lactose and lactulose were determined by making seven injections of a low-level solution fortified with lactose and lactulose at 3-5 times the estimated MDL.  The calculated MDLs for lactose and lactulose in DI water obtained by this method are 0.12 mg/L and 0.23 mg/L for lactose and lactulose, respectively. Table 2 summarizes the data for this determination.

Sample analysis
The optimized separation method was applied to different dairy products, namely  lactose-free Gouda, Havarti, and cottage cheeses as well as low-fat yogurt. Figure 2a shows overlaid chromatograms of fortified and unfortified Gouda cheese. Trace 1 shows a separation of unfortified cheese with no lactose detected. Trace 2 shows the separation of the same Gouda cheese sample fortified with 10 mg/L each of lactose and lactulose. This chromatogram shows that lactose and lactulose are well separated both from each other and also from potential matrix-related interferents. It can be seen that the unfortified sample did not overload the column and cause lactose to be undetected. Figure 2b shows overlaid chromatograms of fortified and unfortified Havarti cheese. Trace 1 shows a separation of unfortified Havarti cheese with no lactose detected. Trace 2 shows the separation of a Havarti cheese sample fortified with 10 mg/L each of lactose and lactulose. The chromatography is similar to Figure 2a.
 
Figure 3a shows the separation of carbohydrates in low fat yogurt. The prepared yogurt samples were diluted 1:10 to prevent overloading with lactose. The diluted yogurt sample showed some galactose and glucose, and 33.5 mg/L or 0.0035% of lactose. Figure 3b shows the separation of carbohydrates in lactose-free cottage cheese. The chromatogram shows that lactose-free low fat cottage cheese contains high concentrations of galactose and glucose and 2.7 mg/L or 0.00027% of lactose.
 
Lactose-free milk and whole milk were also analysed and showed lactose concentrations of 0.6 mg/L or 0.00006% and 560 mg/L or 0.056%, respectively. A duplicate of each of the samples was fortified with known amounts of lactose and lactulose prior to sample preparation. Recoveries were calculated after analysis of both native and spiked samples. Recoveries of lactose and lactulose for all samples were 86-102% [Table 3].

Conclusion
The results of the work described in this paper show that high performance anion exchange chromatography with pulsed amperometric detection (HPAE-PAD) accurately determines lactose at low concentrations in lactose-free products without error and without the labor associated with analyte derivatization. The method resolves lactose from all other carbohydrates commonly found in these samples including galactose, glucose, sucrose and lactulose.

Acknowledgements
We would like to acknowledge G. Larcher of the Dionex Europe Applications Laboratory, H. Kruth from Dionex GmbH, and S. Wende, C. Wiedemann (PhD) and E. Kitzelmann (PhD) from Landwirtschaftliches Zentrum Baden-Württemberg (LAZBW) - Milchwirtschaft Wangen  for the initial method development work on this project.

References
1. Eluent Preparation for High-Performance Anion-Exchange Chromatography with Pulsed Amperometric Detection; Technical Note 71, LPN 1932Dionex Corporation, Sunnyvale, CA.
2. Determination of Lactose in Lactose-Free Milk Products by High–Performance Anion Exchange Chromatography with Pulsed Amperometric Detection (HPAE-PAD), Application Note 248, LPN 2507 Dionex Corporation, Sunnyvale, CA.

CarboPac, Chromeleon and OnGuard are registered trademarks of Dionex Corporation.

The authors
Pranathi Perati*, Brian De Borba and
Jeffrey Rohrer
Dionex Corporation
Sunnyvale, California, USA
* Corresponding author:.
E-mail: pranathi.perati@dionex.com
Fax: +1-408-732-2470