Determination of in low-lactose products

Manali Aggrawal and Jeff Rohrer, Thermo Fisher Scientific, Sunnyvale CA 94085, USA

ABSTRACT RESULTS SAMPLE ANALYSIS RECOVERY CONCLUSIONS Purpose: To develop a High Performance Anion Exchange chromatography coupled with Pulsed Method accuracy was evaluated by measuring recoveries of sugar standards containing allolactose, Separation Figure 3 shows the separation of carbohydrates in lactose-free dairy samples S#1,S# 3, and S#5 . An HPAE-PAD method was successfully developed and validated for lactose analysis in 11 Amperometric Detection (HPAE-PAD) method for the determination of allolactose, lactose, and lactose, and lactulose spiked into lactose-free samples. Figure 5a, 5b shows the comparison of the along with a 1 ppm sugar standard mix. S#1 contains large amounts of , and small commercial lactose-free products including dairy as well as baked products. lactulose in lactose-free products. chromatogram of unspiked and spiked S#1 run under two eluent concentration; 14 mM and 23 mM To obtain a fast separation with baseline resolution of lactose from other structurally similar sugars, amounts of allolactose and lactose. S#3 is a 99% lactose-free low fat milk smoothie, it was diluted 40 Methods: We used the recently introduced Thermo Scientific™ Dionex™ CarboPac™ PA210-4μm KOH. As shown here, lactose is very well separated from similar using both eluent the method was evaluated at different isocratic eluent conditions ranging from 12–25 mM KOH. times to bring the lactose concentration into the calibration range. S#5 (aged cheese) contains a very . The Dionex CarboPac PA210-4µm column allows the separation of lactose from structurally similar column to separate lactose from other structurally similar milk sugars. conditions. However, the lactulose peak appears misshapen with much of the peak area eluting later Initially 14 mM KOH was chosen for the analysis as the peaks were well separated. Later during the small amount of lactose. In ripened hard cheese, the sugars (e.g., lactose, galactose ,and glucose) saccharides allolactose, lactulose, and epilactose in less than 9 min with overall cycle time of 30 Results: In this work, 11 commercial lactose-free products including both dairy products and baked than the standard using 14 mM KOH and symmetrical using 23 mM KOH. Similarly, other samples analysis of samples we observed that there were unknown peaks eluting at the retention time of are naturally metabolized by microflora. Thus, aged cheese contains a very low concentration of min. goods, were tested and analyzed for their lactose content using HPAE-PAD. In addition, structurally exhibited the same behavior using both 14 and 23 mM KOH and spiking experiments confirmed that lactulose and thus lactulose could not be distinguished from other unknown saccharides in samples lactose. similar sugars, allolactose and lactulose, were determined and quantitated in these samples. evaluated. Using 23 mM KOH, the peaks are sharper and lactulose was well separated from none of the samples tested here contained lactulose. . The method showed good precision and accuracy with a recovery range of 80–110% in the interfering peaks. Figure 1 shows the chromatographic profile (run @ 23 mM KOH) of a 5 ppm sugar Figure 3. Chromatograms of dairy samples along with an11 sugars standard mix samples tested. standard mix. Peak Name Peak Name Peak Name Figure 5a. Chromatogram of S#1 and Figure 5b. Chromatogram of S#1 and 295 INTRODUCTION 1 Fucose 5 Fructose 9 Lactulose spiked S#1 with 14mM KOH eluent spiked S#1 with 23mM KOH eluent Figure 1. Chromatogram of a 5 ppm sugar standard mix Lactose is the major found in milk and is metabolized to glucose and galactose by the 2 Arabinose 6 Sucrose 10. Epilactose 93.4 REFERENCES 100.3 enzyme lactase. Lactose-intolerant individuals have a lactase deficiency, therefore, lactose is not 3 Galactose 7 Allolactose 11. Raffinose 1 Peak No. Name 1 Peak No. Name 1 Allolactose 1 Allolactose completely metabolized, thus these individuals have difficulty digesting milk products, resulting in Peak Name Peak Name Peak Name nC 4 Glucose 8 Lactose 1. Seki, N.; Saito, H. Lactose as a source for lactulose and other functional lactose derivatives, Int. 140 2 Lactose 2 Lactose uncomfortable intestinal symptoms such as diarrhea and bloating. The demand for low-lactose dairy 1 Fucose 5 Fructose 9 Lactulose Dairy J. 2012, 22, 110–115. 1 2 3 Lactulose 3 Lactulose products is increasing and more lactose-free products are now available. These commercially 2 Arabinose 6 Sucrose 10. Epilactose Aged cheese (#5) nC 4 3 Galactose 7 Allolactose 11. Raffinose Kefir milk smoothie (#3) 2. Fransen, C.; Vliegenthart, J et al. Α-D-Glcp-(1↔1)-β-D-Galp-containing oligosaccharides, novel available lactose-free products are produced by enzymatically breaking down lactose into glucose 100 nC 5 Half n half milk (#1) 2 products from lactose by the action of β-galactosidase, Carbohydr. Res. 1998, 314, 101–114. and galactose. However, the enzymatic hydrolysis process is not 100% efficient and the resulting 3 4 Glucose 8 Lactose 2 1 2 3 4 5 7 8 9 products contain varying amounts of residual lactose. A concentration of <0.01% is generally nC 6 1011 1 ppm sugar std 3 3 3. Thermo Fisher Scientific. Application Note AN 248, Determination of Lactose in Lactose-Free Milk accepted as “lactose free”. Milk products undergo heat treatment procedures to eliminate microbes 6 7 8 18 1ppm std 0.0 2.0 4.0 6.0 8.0 10.0 12.0 1ppm std Products by High-Performance Anion-Exchange Chromatography with Pulsed Amperometric that can cause food spoilage. During this process, lactose and other saccharides present are min 30.0 9 10 11 30.0 Detection. converted thermally, enzymatically (β-galactosidase), or by bacterial fermentation (lactic acid 23.4 6.0 8.0 10.0 12.0 23.1 1 40 Figure 4. Chromatograms of dairy samples along with an 11 sugars standard mix min 6.0 8.0 10.0 12.0 bacteria) to many different derivatives. Allolactose, lactulose, and epilactose are some of the min 4. Van Scheppingen, W.; van Hilten, P.; Vijverberg, P.; Duchateau, A. Selective and sensitive common derivatives that are formed during these processes. Allolactose is a lactose isomer formed 20 determination of lactose in low-lactose dairy products with HPAEC-PAD, In J. Chromatogr. B. by β-galactosidase during the transgalactosylation reaction.2 Lactulose and epilactose are potentially 0.0 2.0 4.0 6.0 8.0 10.0 12.0 295 Peak Name Peak Name Peak Name Figure 6. Chromatogram of unspiked and spiked Lactose-free milk chocolate min 2017, 1060, 395-399. prebiotic isomers of the milk sugar lactose. The increased market demand for lactose-free products 1 Fucose 5 Fructose 9 Lactulose 2 Arabinose 6 Sucrose 10. Epilactose has created a need for a fast, reliable, and sensitive method to analyze these products. HPAE-PAD is 120.0 one of the most powerful techniques for carbohydrate determinations. It is a well-established The retention time for lactose is ~ 7.25 min. Epilactose and lactulose are nicely separated with 23 3 Galactose 7 Allolactose 11. Raffinose Peak No. Name 3,4 mM KOH despite being chemically similar and theoretically hard to separate. Not all 11 carbohydrates 4 Glucose 8 Lactose 1 technique for lactose and lactulose determinations. In this work, we developed an HPAE-PAD 1 Allolactose ACKNOWLEDGEMENTS nC method using the recently introduced Dionex CarboPac PA210-4μm column to separate lactose from are completely resolved, but the goal of this separation was to resolve allolactose, lactose, lactulose, 2 Lactose other structurally similar milk sugars. and epilactose. Shortbread cookie(#9) 3 Lactulose We would like to thank Dr. Detlef Jensen, EU Sales Support Expert for Ion Chromatography and nC 100 2 Environmental Market, Thermo Fisher Scientific, Dreieich, Germany for providing the lactose free Calibration and Quantification Sandwitch cookie (#7) 3 4 5 cookie and chocolate samples. 1 2 6 7 8 9 10 11 Calibration curves for these 3 sugars are shown in Figure 2. The coefficient of determination (r2) is 1 ppm sugar std 60.0 MATERIALS AND METHODS 3 greater than 0.999 for all sugars. Over the course of the analysis, the relative standard deviations of 18 0.0 2.0 4.0 min 6.0 8.0 10.0 12.0 Sample Preparation the retention times of all 3 peaks (n = 6) ranged from 0.03-0.15%. 32.1 TRADEMARKS/LICENSING Step 1 0.0 2.0 4.0 min 6.0 8.0 10.0 12.0 Weigh 1 g of sample in a 100 mL volumetric flask and add 10 mL DI water to the Table 1 summarizes the calibration data for calibration curves obtained by injecting calibration Figure 4 is a chromatogram of S#7 and S#9 along with a 1 ppm sugar standard mix. S#7 shows a © 2018 Thermo Fisher Scientific Inc. All rights reserved. All trademarks are the property of Thermo sample.[For cheese and cookie samples, additional heating at 65 ºC for 15 min is required] standards between 0.25–20 mg/L. very broad allolactose peak. It could indicate that there is another sugar eluting very close to Step 2 Add 200 μL Carrez I solution and 200 μL Carrez II solution to the mixture, shaking after allolactose. S#9 is a lactose-free and sugar-free cookie sample, thus only very small peaks were Figure 6 shows the chromatograms of unspiked and spiked S#10 (lactose-free milk chocolate Fisher Scientific and its subsidiaries. This information is not intended to encourage use of these products in any manner that might infringe the intellectual property rights of others. each addition. Bring the volume to 100 mL with DI water. The samples are treated with Figure 2. Calibration curves of allolactose, lactose and lactulose detected for glucose, fructose, and sucrose and almost no peaks for allolactose, lactose, and samples). The recovery percentages were calculated using the formula shown below: Carrez I and Carrez II solutions to remove fats, proteins and other redox compounds that lactulose. However, epilactose and raffinose peaks were detected. None of the samples had 퐶 푠푝푖푘푒푑 푠푎푚푝푙푒 − 퐶 푢푛푠푝푖푘푒푑 푠푎푚푝푙푒 25.0 20.0 Lactulose can interfere with analysis. Lactose 15.0 lactulose. Table 2 lists the concentration of lactose in all lactose free products tested in this study. 푅푒푐표푣푒푟푦 % = ∗ 100 Allolactose 퐶 푎푛푎푙푦푡푒 푎푑푑푒푑 Step 3 Centrifuge a portion of this sample at 3000 RPM. Aspirate the supernatant and filter through 20.0 15.0 12.5 a 0.2 μm syringe filter. Table 2. Concentrations of allolactose and lactose in lactose-free samples 10.0 Table 3. Recovery of allolactose, lactose, and lactulose in all samples Step 4 Prepare a Thermo Scientific™ Dionex™ OnGuard™ IIA, 2.5 cc cartridge by flushing it with 15.0 Allolactose Lactose 10.0 15 mL DI water at a flow rate of less than 2 mL/min, then discard the effluent. Load 8 mL of 7.5 Amount RSD Amount RSD Allolactose Lactose Lactulose

10.0 S# Sample Label Area (nC*min) Area sample, discard the first 6 mL into a waste container, and collect the next 2 mL for analysis. (nC*min) Area 5.0 g/100g (%) g/100g (%) Sample# Found added Recovery Found added Recovery Found added Recovery 5.0 (nC*min) Area 1 Half &half milk 100%lactose free 0.058 0.88 0.009 0.86 5.0 2.5 (mg/L) (mg/L) % (mg/L) (mg/L) % (mg/L) (mg/L) % The Dionex OnGuard IIA cartridge treatment removes the anionic contaminants and neutralizes the 2 Fat free Milk 100%lactose free 0.114 0.13 0.029 0.04 1 5.78 0.99 85.4 0.88 1.10 97.5 0.00 0.98 80.7 0.0 0.0 0.0 sample matrix, thereby minimizing baseline disturbances from a highly acidic sample. 0.0 10.0 20.0 25.0 0.0 10.0 20.0 25.0 0.0 10.0 20.0 25.0 3 Milk smoothie Up to 99%lactose free 0.06 0.47 3.16 0.32 2 11.41 0.98 96.7 2.89 1.02 106 0.00 1.00 87.4 Concentration (mg/L) Concentration (mg/L) Concentration (mg/L) 4 Fat free Yogurt Lactose free 0.006 0.48 0.002 0.00 3 0.15 1.00 98.6 7.89 1.00 100 0.00 1.01 86.0 Method 5 Aged Cheese None 0.008 0.39 0.006 0.84 4 0.56 1.02 95.5 0.18 0.99 87.0 0.00 0.97 80.3 5 Not calculated 0.65 1.00 91.3 0.00 1.00 81.3 LOD and LOQ were calculated by first measuring the peak-to-peak noise in a representative one- 6 Butter lactose free 0.003 0.30 0.001 0.17 System Thermo Scientific™ Dionex™ ICS-5000+ HPIC™ System 6 0.32 0.99 97.9 0.08 0.99 94.5 0.00 0.98 93.0 minute segment of the baseline where no peaks elute, but close to the peak of interest. The signal 7 Sandwich cookie lactose free; <0.1g/100g - - 0.009 0.41 Column Dionex CarboPac PA210-4µm + guard 7 Not calculated 0.93 1.10 88.1 0.00 0.98 94.7 was determined from the average peak height of three injections of 0.25 mg/L standard. The LOD 8 Butter cookie lactose free; <0.1g/100g 0.018 0.05 0.007 0.03 8 1.78 1.12 92.5 0.70 0.98 91.6 0.00 1.00 80.8 Eluent 0 – 10 min: 23 mM KOH ; 10-15 min: 100 mM KOH ; 15-30 min: 23 mM KOH and LOQ were determined by 3x and 10x S/N, respectively, of a 0.25 mg/L standard. The estimated 9 Shortbread cookie Lactose free and sugar free - - - - LODs and LOQs for the three sugars are summarized in Table 1. 9 0.00 1.00 85.7 0.00 1.11 107 0.00 1.01 88.5 Temp. 30 ºC 10 Milk Chocolate lactose free; <0.1g/100g 0.083 0.65 0.046 0.72 10 0.00 1.11 110 0.29 0.98 97.0 0.00 1.02 81.6 Table 1. Results for calibration, LOD and LOQ for allolactose, lactose and lactulose 11 Caramel cookie Lactose free and gluten free - - 0.003 0.12 11 8.34 2.01 90.1 4.57 1.99 101 0.00 2.10 91.4 Flow rate 0.8 mL/min

Inj Volume 10 µL Range Calibration 2 LOD LOQ • Lactose concentration was <0.01g/100g in all products tested except S#2 and S#10, each with Carbohydrate r Recoveries of all three sugars for all samples were 80 to 110% (Table 3), with the exception of (mg/L) type (mg/L) (mg/L) less than 0.1g/100g. Detection PAD (Au) Disposable Waveform A (TN21) allolactose in aged cheese and sandwich cookie. This is thought to be due to the poor peak shape for Allolactose 0.25-20 Quad, WithOffset 0.99999 0.009 0.029 • S#3 has about 3 g/100g (~3%) of lactose in it. While the label says it’s up to 99% lactose free. allolactose in these two samples. The poor peak shape is due to the co-elution of an unknown Data Analysis Lactose 0.25-20 Linear, WithOffset 0.99981 0.010 0.034 saccharide or other electrochemically active compound, which interferes with the spike recovery experiment and thus recovery was not calculated for allolactose in these two samples. The recovery The Thermo Scientific™ Chromeleon™ Chromatography Data System (CDS) was used for all data • Lactose-free butter has the lowest amount of allolactose and lactose. Lactulose 0.25-20 Quad, WithOffset 0.99993 0.017 0.056 percentages for lactose in all the sample matrices are in the range of 87–107%. acquisition and processing. • S#9 (lactose-free and sugar-free cookie) does not have any detected lactose or allolactose

PO72674-EN 0318S

Thermo Fisher Scientific • 5781 Van Allen Way • Carlsbad, CA 92008 • thermofisher.com