Patterns of Free Amino Acids in German Convenience Food Products: Marked Mismatch Between Label Information and Composition

ORIGINAL ARTICLE Patterns of free amino acids in German convenience food products: marked mismatch between label information and composition

M Hermanussen1, U Gonder2, C Jakobs3, D Stegemann4 and G Hoffmann4

1University of Kiel Germany, Aschauhof, Altenhof, Germany; 2Nutritionist, Hu¨nstetten, Germany; 3Metabolic Unit, Department of Clinical Chemistry, VU University, Medical Center, Amsterdam, The Netherlands and 4Department of General Pediatrics, Children’s University Clinic, Heidelberg, Germany

Background/Objectives: Free amino acids affect food palatability. As information on amino acids in frequently purchased pre- packaged food is virtually absent, we analyzed free amino acid patterns of 17 frequently purchased ready-to-serve convenience food products, and compared them with the information obtained from the respective food labels. Subjects/Methods: Quantitative amino acid analysis was performed using ion-exchange chromatography. g-Aminobutyric acid (GABA) concentrations were verified using a stable isotope dilution gas chromatography/mass spectrometry (GC-MS) method. The patterns of free amino acids were compared with information obtained from food labels. Results: An obvious mismatch between free amino acid patterns and food label information was detected. Even on considering that tomatoes and cereal proteins are naturally rich in glutamate, the concentrations of free glutamate outranged the natural concentration of this amino acid in several products, and strongly suggested artificial enrichment. Free glutamate was found to be elevated even in dishes that explicitly state ‘no glutamate added’. Arginine was markedly elevated in lentils. Free cysteine was generally low, possibly reflecting thermal destruction of this amino acid during food processing. The meat and brain-specific dipeptide carnosine (CARN) was present in most meat-containing products. Some products did not contain detectable amounts of CARN in spite of meat content being claimed on the food labels. We detected GABA at concentrations that contribute significantly to the taste sensation. Conclusion: This investigation highlights a marked mismatch between food label information and food composition. European Journal of Clinical Nutrition (2010) 64, 88–98; doi:10.1038/ejcn.2009.116; published online 23 September 2009

Keywords: amino acids; convenience food; GABA; glutamate; monosodium glutamate

Introduction humans (Finlayson et al., 2007), with food palatability being the strongest predictor of food intake when availability is not Louis-Sylvestre et al. (1984) showed in rats that the daily limited (Mattes, 2008). Recent innovations in food techno- presentation of a new choice of palatable food led to a logy using fermentation, extraction, encapsulation, fat sustained increase in nutrient intake and to overweight, and replacement and many other techniques, leading to new concluded that variety and high palatability are per se food ingredients, reduction or removal of undesirable sufficient factors to overcome weight-regulatory mechan- food components, addition of specific functional ingredi- isms. A complex reward system with distinct ‘liking’ ents, modification of food composition and masking of (hedonic/affective) and ‘wanting’ (incentive salience/moti- undesirable flavors (Hsieh and Ofori, 2007), further vation) components mediates food choice, controls appetite improved palatability. Consumption of convenience food and helps regulate the overall nutrient consumption also in (percentage of total food intake) increases with age from approximately 3% in 3- to 8-year-old children to 7% in 14- to Correspondence: Professor M Hermanussen, University of Kiel Germany, 18-year-old boys and to 5% in 14- to 18-year-old girls. Aschauhof 3, Altenhof 24340, Germany. Currently, German adolescents consume more than 700 E-mail: [email protected] Received 24 April 2009; revised 30 July 2009; accepted 5 August 2009; chemically designed convenience food creations with, on published online 23 September 2009 average, 14 ingredients and a high number of flavorings and Free amino acids in German convenience food products M Hermanussen et al 89 food additives (Alexy et al., 2008). Similar data were observed Table 1 Top-30 list of most frequently purchased ready-to-serve by the German Market Research Institute (Gesellschaft fu¨r German convenience food products Konsumforschung), Nuremberg, Germany: the total super- Maggi ravioli pikanta market selling subdivides into 45% non-alcoholic and Maggi ravioli tomato saucea alcoholic beverages, 22% fresh articles (vegetables, fruit, Schnell&Lecker TK chicken fricassee a meat and meat products), 13% milk and dairy products, 3% Lidl pizza salami a deep frozen products and 2% sweets, with the remaining Primana peas Gut Frielingshof TK chicken breast fillet percentage consisting of canned goods, dressings, oil and Miracoli spaghetti (4 persons)a so on. (German Market Research Institute, personal Wagner pizza piccolinia communication, August 2008). Pastini spaghetti tomato Apart from its effect on palatability, food processing affects Dr Oetker oven fresh pizza pepperoni salami Lentils ‘chef’a free amino acid contents. However, as information on amino Meica curry king sausagea acids in frequently purchased pre-packaged food is virtually Dr Oetker pizza pizza ristorante salame absent, we analyzed the free amino acid patterns of 17 Knorr gourmet forest mushrooms frequently purchased ready-to-serve convenience food pro- Miracoli spaghetti (2 persons) Primana lentilsa ducts and compared them with information obtained from Gut Frielingshof TK chicken cordon bleua the respective food labels. Wagner stone oven pizza The national food law requires labeling, and declaring Gut Frielingshof TK chicken nuggets pan macronutrient and energy contents and the relative con- Dr Oetker oven fresh pizza speziale Primana spaghetti tomato saucea centrations of major ingredients of pre-packaged food. Yet, Wagner stone oven pizza salami the exact composition of these products usually remains Knorr gourmet onions obscure. We contacted 15 market leaders before this study Maggi master class onions a requesting additional information on their products, but all Erasco chicken noodles Cucina lasagne Bolognese requests remained unanswered. Similar experience has been Dr Oetker pizza ristorante speziale reported by other researchers (Kersting, personal commu- Landgut TK chicken fricassee nication, 2007). We therefore decided to simply use food Dulano lasagne label information and to derive estimates of all major com- Primana chicken fricassee ponents by inferring from gross macronutrient content, and Five additional frequently purchased convenience food dishes. the ranking order of the listed ingredients. We inferred protein Bassermann ravioli; Maggi 5-minute terrine (potatoes with roasted onions); and protein-bound amino acid compositions, and related Erasco rice pot with dumplings; Maggi Magic Asia fried noodles with them to the free amino acid patterns analyzed in this study. vegetables; and Lidl masters herbal baguette. aDishes that underwent analysis. Five additional products were recommended by the German Market Research Institute (Gesellschaft fu¨r Konsumforschung) Nuremburg, Germany. Materials and methods

The top-30 list of most frequently purchased German ready- Table 2 Percentage of macronutrient content of frequently purchased to-serve convenience food products in 2006 was provided by convenience food products and percentage of water content of ready-to- the German Market Research Institute (Gesellschaft fu¨r serve dish Konsumforschung), Nuremburg, Germany. Top hits were Protein Carbo Fat %Water Maggi ravioli pikant and Maggi ravioli tomato sauce; the a following items consisted of eight types of pizza: seven Maggi ravioli pikant 17 50 33 84 a preparations of chicken meat (Cordon Bleu, soups, fricassee Maggi ravioli tomato 17 65 18 82 Lidl pizza salamia 14 50 36 53 and so on), four types of spaghetti, two lasagnes, two lentil Primana peasa 25 42 33 88 products, three ready-to-eat soups, one pot of peas and one Miracoli spaghettia 13 64 23 69 a curry sausage (Table 1). Of these, we chose a representative Wagner pizza piccolini 17 43 41 54 a sample of 12 dishes for analysis, and for a variety of reasons Lentils ‘chef’ 24 47 28 79 Meica curry sausagea 16 20 64 64 we added another five frequently purchased dishes that were Primana lentilsa 25 58 17 79 not listed among the top-30. All products contained multiple Chicken cordonbleua 44 36 20 65 a additives stimulating sweet, sour, salt and umami taste and Primana spaghetti 18 72 10 70 a were considered appropriate for ‘full lunch’. The dishes are Erasco chicken noodles 24 38 39 90 Bassermann ravioli 11 73 15 85 usually consumed as ‘single-pot meals’ and supposed to Maggi 5-min-terrine 13 80 5 79 satisfy the consumer without additional appetizer or side Erasco rice pot 14 42 44 86 dishes. None of the products was particularly spicy. Flavoring Maggi noodles ‘asia’ 11 60 29 66 with monosodium glutamate was explicitly mentioned in Lidl herbal baguette 8 53 38 39 some dishes. Most products were high in protein, but aProducts belonging to the top-30 list of frequently purchased convenience otherwise roughly in accordance with current national food food.

European Journal of Clinical Nutrition Free amino acids in German convenience food products M Hermanussen et al 90 recommendations. Fat contributed to o30% of total energy protein (5.5%), carbohydrate (12.9%) and fat (1.7%). Six in 10/17 dishes (Table 2). gram pork meat contains 1.2 g protein and 0.5 g fat For analysis, all dishes were prepared as described by the (Hartmann et al., 2005). Lentils are characterized by a manufacturer, that is, warmed, boiled, oven-baked or fried. protein–fat–carbohydrate ratio of 31:2:67 (Hartmann et al., After homogenizing, amino acids were classically quantified 2005). For example, 14% lentils nearly complete the labeled by ion-exchange chromatography with an amino acid quota for protein, yet at the expense of sufficient carbo- analyzer (Biotronic LC 3000, Biotronic). Samples were hydrate. As carrots, potatoes and onions contribute only deproteinized by mixing 200 ml of sample with 50 mlofa little to total macronutrient content, the amounts of starch sulfosalicylic acid solution (200 g/l). After centrifugation at and wheat flour had to be assumed to be near 2% each, to 500 g for 10 min, the supernatant was mixed with sample yield the labeled quota of total carbohydrate. Inferring is not dilution buffer (1:1). A 200 nmol/ml aqueous calibrator a trivial task, but we used a simple algorithm that took care solution containing 37 physiological amino acids (Laborservice of the labeled composition of total macronutrient content Onken Laborservice Onken GmbH, 63584 Gru¨ndau, and the ranking order of ingredients. We inferred that Germany) was used for external quantification. In all, 50-ml Primana lentils with lean pork were composed of water at samples were injected and the free amino acids were eluted approximately 40±2% 4 dried lentils (14±0.7%) 4 pork from a cation exchange resin using buffers with varying (exactly 6%) 4 carrots (4±0.4%) 4 potatoes (4±0.4%) 4 ionic strength and pH (Biotronic). They are detected with onions (4±0.4%) 4 tomato concentrate (3±0.3%) 4 ninhydrin (after column derivatization) and monitored at modified starch (2±0.2%) 4 wheat flour (2±0.2%) 4 plant 570 nm and at 440 nm (for proline and hydroxyproline). oil (0.9%). We similarly inferred the composition of all other The g-aminobutyric acid (GABA) concentrations were convenience products, and evaluated the composition for controlled in a second reference laboratory using a stable each product individually. Inferring was considered success- isotope dilution GC-MS method described by Kok et al. ful when the algorithm provided unequivocal results within

(1993): 0.5 g of the food sample was solved in 10 ml of H2O. limits of ±5% for each essential component, and ±10% for From this mixture, 10–500 ml was taken and 0.2 nmol of each remaining small component. In two dishes, the isotopically labeled internal standard (2H2-GABA) was algorithm failed. Regardless of how Erasco chicken noodles added. To this, 800 mlof1M phosphate buffer, pH 11.5, with fresh vegetables and garden herbs were composed, the and 50 ml methylchloroformate were added to form the resultant admixture of essential components, noodles, methylformate derivative of GABA. HCl, 150 ml and 6 M, was vegetables, chicken meat (5%) and chicken fat (2%), never continuously added and the sample extracted with 4 ml of fitted the protein–fat ratio declared on the food label, even ethylacetate. The organic layer was evaporated and the within the limits of ±10%. However, we kept the chicken residue derivatized with 10 ml triethylamine and 100 mlof7% dish in the sample as the inaccuracy of the estimate did not pentafluorobenzylbromide in acetonitrile. The formed greatly affect the overall protein composition. In addition, methylformate-pentafluorobenzyl-derivative of GABA was Miracoli spaghetti could not satisfyingly be inferred, but extracted with hexane and was analyzed using single ion monitoring gas-chromatography electron-capture negative- ion mass spectrometry. Table 3 Major protein donors (percentage of all protein in the The GC separation was achieved on a CPSil 88 column respective dishes) of frequently purchased convenience food products (25 m Â 0.32 mm (internal diameter)). The column oven was Cereals/ Meat Tomato Cheese Legumes held at 150 1C for 1 min and then programmed to 240 1Cat noodles conc. the rate of 200 1C/min. a The MS responses of m/z À160 for GABA and m/z À162 for Maggi ravioli pikant 30 44 26 a 2H2-GABA were quantified. Maggi ravioli tomato 41 36 23 Lidl pizza salamia 34 28 3 28 Primana peasa 45 51 Miracoli spaghettia 77 12 10 a Inferring the protein composition of convenience food products Wagner pizza piccolini 33 19 2 44 a According to national food law, macronutrient content has Lentils ‘chef’ 41 2 54 Meica curry sausagea 94 6 to be listed and major ingredients have to be ranked in Primana lentilsa 922 69 declining order. The absolute amounts of major ingredients Chicken cordonbleua 777 11 a can usually be inferred from the above information as Primana spaghetti 77 11 12 a exemplified. Erasco chicken noodles 39 57 Bassermann ravioli 50 16 25 The manufacturer of canned Primana lentils with lean pork Maggi 5-min-terrine 62 12 13 declares: water, lentils, pork (6%) (consisting of pork, salt Erasco rice pot 9 57 7 spices, dextrose, lactose and smoke), carrots, potatoes, Maggi noodles ‘Asia’ 58 1 Lidl herbal baguette 79 onions, tomato concentrate, modified starch, wheat flour, plant oil, salt, leeks, vinegar, aroma, guar flour, sugar, aProducts belonging to the top-30 list of frequently purchased convenience caramel, spices and ascorbic acid. Macronutrient content: food.

European Journal of Clinical Nutrition Free amino acids in German convenience food products M Hermanussen et al 91 Table 4 Free amino acids (mg/100 g) in frequently purchased German convenience food

GLU ASP ARG CARN ASN ALA GABA PHE SER LYS LEU

Maggi ravioli pikanta 75 31 5 20 11 25 6 5 4 3 Maggi ravioli tomatoa 208 28 12 7 17 12 22 5 5 4 3 Lidl pizza salamia 162 42 19 22 22 23 28 14 11 17 18 Primana peasa 251 6 58 19 25 15 9 5 6 10 6 Miracoli spaghettia 205 63 24 31 23 48 12 12 12 10 Wagner pizza piccolinia 78 36 9 39 23 17 23 12 9 16 13 Lentils ‘chef’a 29 14 37 21 25 9 5 4 5 6 4 Meica curry sausagea 53 31 6 56 15 23 26 9 7 6 6 Primana lentilsa 27 14 38 18 22 10 4 3 7 5 3 Chicken cordonbleua 45 16 9 123 6 25 2 13 15 27 23 Primana spaghettia 125 65 23 1 46 21 39 17 9 12 9 Erasco chicken noodlesa 373 9 4 6 10 4 3 2 3 3 Bassermann ravioli 98 31 10 18 9 21 6 5 5 5 Maggi 5-min-terrine 106 30 16 5 19 24 21 10 8 8 14 Erasco rice pot 254 3 10 12 10 3 1 2 2 1 Maggi noodles ‘asia’ 691 13 24 o1 22 8 10 11 9 16 Lidl herbal baguette 11 11 12 4 5 3 1 1 3 1

THR GLN VAL ILEU GLY HIS TYR TRP MET ORN CYS Maggi ravioli pikanta 51 3 2 14 2 1 o1 Maggi ravioli tomatoa 41 3 2 23 2 3 1 1o1 Lidl pizza salamia 10 6 11 8 8 6 4 5 5 3 o1 Primana peasa 51 7 4 63 4 2 3 1 Miracoli spaghettia 1137 6575 52o1 o1 Wagner pizza piccolinia 812107 6105 8 4 4o1 Lentils ‘chef’a 57 5 3 43 3 1 2 1 2 Meica curry sausagea 65 5 4 64 3o12 Primana lentilsa 5104 3 4 4 3 1 2 1 1 Chicken cordonbleua 13 10 15 12 13 7 10 3 7 3 1 Primana spaghettia 1146 6486 832o1 Erasco chicken noodlesa 31 3 2 41 1 2 1o11 Bassermann ravioli 5 2 3 2 2 3 2 2 1 o1 o1 Maggi 5-min-terrine 8 6 11 8 5 4 5 3 4 1 Erasco rice pot 2 1 1 1 1 1 1 o1 o1 Maggi noodles ‘asia’ 9 8 13 11 7 4 3 7 3 o11 Lidl herbal baguette 1 4 3 1 2 1 1 1

Abbreviations: ALA, alanine; ARG, arginine; ASP, aspartate; ASN, asparagine; CARN, carnosine; CYS, cysteine; GABA, g-aminobutyric acid; GLN, glutamine; GLU, glutamate; GLY, glicine; HIS, histidine; ILEU, isoleucine; LEU, leucine; LYS, lysine; MET, methionine; ORN, ornithine; PHE, phenylalanine; SER, serine; THR, threonine; TYR, tyrosine; TRP, tryptophan; VAL, valine. aMark products belonging to the top-30 list of frequently purchased convenience food. remained in the sample as absolute amounts of the essential the dishes. Some dishes are particularly rich in free components, spaghetti, tomato and cheese, were noted on glutamate. On average, free glutamate contributes to 40% the label. Knowing the composition of the dishes enabled (s.d. 24%) of all free amino acids, but we found considerable inferring the respective protein-bound amino acid contents. variation among the dishes. In Erasco chicken noodles, Erasco rice pot and Maggi noodles ‘Asia’, this amino acid contributes to more than 80% of all free amino acids. Results Figure 1a illustrates the mean and s.d. of the proportions for each amino acid. The large s.d. of carnosine (CARN) reflects Frequently purchased convenience food is characterized by that some of the dishes contain meat (that is, contain high protein content (Table 2). In the 12 products of the top- CARN), whereas others are vegetarian (no CARN). 30 list, on average 21% of nutritional energy was provided by The spectra of free amino acids do not match the spectra of protein. Major protein donors were meat, cheese, cereals and protein-bound amino acids. Table 5 shows the protein- tomatoes (Table 3). bound amino acid compositions of the 17 dishes (except for Table 4 shows the concentration of free amino acids in aspartate, CARN, GABA, glutamine and ornithine, which are the 17 dishes. The concentrations vary considerably, and the not listed in German nutritional tables (Hartmann et al., spectra, that is, the proportion of each amino acid relative 2005)). The concentrations vary less, and the proportion of to the other amino acids, differ considerably between each amino acid relative to the other amino acids also differs

European Journal of Clinical Nutrition Free amino acids in German convenience food products M Hermanussen et al 92 pattern of free amino acids a 70.00

60.00

50.00

40.00 SD mean 30.00

20.00 percent of all amino acids

10.00

0.00

GLU ASN ALA PHE SER LYS LEU THR VAL GLY HISTYR TRP CYS *ASP ARG *GLN ILEU MET*ORN *CARN *GABA

pattern of protein-bound amino acids b 70.00

60.00

50.00

40.00 SD mean 30.00 percent of all amino acids

20.00

10.00

0.00

GLU ASN ALA PHE SER LYS LEU THR VAL GLY HISTYR TRPMET CYS *ASP ARG *GLN ILEU *ORN *CARN *GABA Figure 1 (a) The mean and s.d. of the proportions of each amino acid, carnosine (CARN) and g-aminobutyric acid (GABA), in percentage of all 20 amino acids. Free glutamate contributes to 40% (s.d. 24%) of all free amino acids and shows considerable variation among the 17 convenience food dishes. (b) The mean and s.d. of the proportions for each protein-bound amino acid, derived from the protein composition of the ingredients of frequently purchased German convenience food. Asterisks indicate amino acids that are not listed in German nutritional tables (Hartmann et al., 2005). Protein-bound amino acids vary less than free amino acids

less between the dishes. Figure 1b illustrates the mean and Why are convenience food dishes so rich in glutamate? s.d. of the proportions for each protein-bound amino acid. Even considering that tomatoes (35%) and cereal proteins The figure shows that protein-bound amino acids vary less (ca. 30%) are naturally very rich in glutamate, an average than free amino acids. For example, protein-bound gluta- glutamate proportion of 40% outranges the natural mate accounts for only 24% of all protein-bound amino occurrence of this amino acid in convenience food dishes, acids, with a s.d. of only 6.5%. and strongly suggests artificial enrichment. The portion of

European Journal of Clinical Nutrition Free amino acids in German convenience food products M Hermanussen et al 93 Table 5 Protein-bound amino acids (mg/100 g) in the ingredients of frequently purchased German convenience food

GLU ARG ASN ALA PHE SER LYS LEU

Maggi ravioli pikanta 910 150 300 150 130 130 180 210 Maggi ravioli tomatoa 950 130 270 130 130 140 150 200 Lidl pizza salamia 2170 410 620 340 400 430 560 680 Primana peasa 990 520 630 310 260 270 470 430 Miracoli spaghettia 1360 140 230 130 190 200 120 290 Wagner pizza piccolinia 2440 400 630 350 470 490 590 820 Lentils ‘chef’a 900 370 760 270 260 280 470 400 Meica curry sausagea 1500 510 840 480 340 370 730 620 Primana lentilsa 880 350 760 230 250 270 420 380 Chicken cordonbleua 2990 1040 1500 880 740 750 1460 1370 Primana spaghettia 1350 140 230 130 190 200 120 290 Erasco chicken noodlesa 460 110 150 90 90 90 120 150 Bassermann ravioli 590 80 160 70 80 90 80 120 Maggi 5-min-terrine 860 120 200 110 130 140 110 200 Erasco rice pot 420 150 210 130 100 100 160 180 Maggi noodles ‘asia’ 1210 230 300 160 210 230 150 300 Lidl herbal baguette 1990 230 310 200 310 330 200 480

THR VAL ILEU GLY HIS TYR TRP MET CYS Maggi ravioli pikanta 110 140 130 120 80 90 30 60 50 Maggi ravioli tomatoa 110 130 120 120 80 90 30 50 50 Lidl pizza salamia 310 440 380 300 250 330 100 180 110 Primana peasa 270 310 290 290 170 200 60 110 80 Miracoli spaghettia 120 180 160 130 80 110 40 60 80 Wagner pizza piccolinia 370 540 460 290 240 390 120 200 110 Lentils ‘chef’a 230 270 230 240 180 180 50 100 150 Meica curry sausagea 380 440 400 440 300 300 90 210 100 Primana lentilsa 200 250 210 210 160 160 50 80 160 chicken cordonbleua 730 880 910 800 510 650 200 440 220 Primana spaghettia 120 180 160 130 80 110 40 60 80 Erasco chicken noodlesa 70 100 100 90 50 60 20 40 30 Bassermann ravioli 60 80 70 70 40 50 20 30 30 Maggi 5-min-terrine 90 130 120 110 60 80 30 50 50 Erasco rice pot 100 130 110 100 70 80 30 50 30 Maggi noodles ‘asia’ 140 190 180 170 90 130 50 60 70 Lidl herbal baguette 190 300 270 210 130 190 70 110 120

Abbreviations: ALA, alanine; ARG, arginine; ASN, asparagine; CYS, cysteine; GLU, glutamate; GLY, glicine; HIS, histidine; ILEU, isoleucine; LEU, leucine; LYS, lysine; MET, methionine; ORN, ornithine; PHE, phenylalanine; SER, serine; THR, threonine; TYR, tyrosine; TRP, tryptophan; VAL, valine. aMark products belonging to the top-30 list of frequently purchased convenience food. free glutamate within all free amino acids is high even in spectra of free amino acids in Lidl pizza salami and dishes that explicitly state ‘no glutamate added’ (for vegetarian Miracoli spaghetti with characteristic peaks of example, in Bassermann ravioli free glutamate contributes aspartate, glutamate and alanine found in several dishes. The to 37% of all free amino acids, that is, even exceeds the figure underlines striking resemblance although the pizza is natural glutamate content of tomato concentrate). Arginine supposed to contain meat (obvious by the small CARN peak), was elevated in lentils and in the herbal baguette. Only the whereas the vegetarian spaghetti do not. Figure 4 shows the free asparagine appeared to largely correspond to protein- spectra of free amino acids in Maggi ravioli tomato sauce bound asparagine (ratio free/bound amino acid close to 1); fortified with monosodium glutamate, and Bassermann the ratio of the other amino acids were generally below 1; ravioli claimed to have not been fortified with the flavoring free cysteine concentrations were lowest (Figure 2). The agent glutamate. Glutamate peaks are visible in both dipeptide CARN that is highly concentrated in muscle and spectra. We suspect that particularly tomatoes significantly brain tissue was present in most meat-containing products. contributed to the spectra of free amino acids. Bassermann ravioli, however, did not contain detectable amounts of CARN in spite of meat being claimed on the food labels (Table 4, Figure 4). Surprisingly, we detected the Discussion neurotransmitter GABA at concentrations between 0.05 and 0.002%. Control analysis in a second reference laboratory Recent food habits have grossly changed, with obvious verified these results (Table 6). Figure 3 exemplifies the preferences for variability and richness in protein and flavor.

European Journal of Clinical Nutrition Free amino acids in German convenience food products M Hermanussen et al 94 ratio free/bound amino acids 3.50

3.00

2.50

2.00 SD

ratio mean 1.50

1.00

0.50

0.00

GLU ASN ALA PHESER LYS LEUTHR VAL GLY HIS TYR TRP CYS *ASPARG *GLN ILEU MET*ORN *CARN *GABA Figure 2 The ratio of free versus protein-bound amino acids. A ratio of 1 indicates that the concentration of a designated free amino acid corresponds to its expected occurrence in the protein of the food products. Asterisks indicate missing values for protein-bound amino acids.

Table 6 GABA (mg/100 g) concentration in five convenience and one food products. In view of the increasing acceptance of fast food product (McDonald cheeseburger) convenience food (Alexy et al., 2008), we examined the Heidelberga Amsterdama compositions of 20 amino acids plus CARN and GABA in 17 frequently purchased ready-to-serve dishes, and compared Chicken cordon bleu 2.27 0.63 them with information provided by food labels. The present Lentils chef’ 4.48 4.80 investigation revealed a marked mismatch between food Lidl pizza salami 27.94 32.39 label information and food composition. Maggi ravioli tomato 21.74 25.90 Miracoli spaghetti 47.58 33.78 Variety and high palatability are per se sufficient to McDonald cheeseburger 11.60 10.37 overcome appetite regulation (Louis-Sylvestre et al., 1984), with food palatability being the strongest predictor of food Abbreviation: GABA, g-aminobutyric acid. intake when availability is not limited (Mattes 2008). aHeidelberg indicates results obtained by amino acid analysis using ion- exchange chromatography on an automated amino acid analyzer, Amsterdam Palatability mainly influences meal size, whereas variety indicates control measurements by stable isotope dilution GC-MS method also exerts an effect on short-term inter-meal intervals (Kok et al., 1993). (Rogers and Blundell 1984). Palatability can be significantly enhanced when combining the ligands of the five proto- typical tastes, particularly using combinations of sodium Modern convenience food is dominated by meat, wheat, chloride, glutamate and other amino acids at appropriate tomatoes and cheese. Meat is naturally rich in protein concentrations (Halpern 2000). Glutamate stimulates appe- containing approximately 16% of glutamate (Hartmann tite at the level of taste perception (Bellisle 2008) as it serves et al., 2005). Wheat and barley, although being major sources as the physiological ligand of the taste receptor umami, the of carbohydrates, contain approximately 10% protein that dominant taste of food containing L-glutamate, such as consists of more than 30% glutamate. Tomatoes and cheese chicken broth, meat extracts and ageing cheese. Umami is are naturally rich sources of free glutamate. Recent innova- responsible for the immediate sensory effect of monosodium tions in food technology using fermentation, extraction, glutamate on the palatability of food. Apart from glutamate, encapsulation, fat replacement and many other techniques, also aspartate (Stapleton et al., 1999), several peptides and leading to new food ingredients, reduction or removal of chicken protein hydrolyzates (Maehashi et al., 1999) produce undesirable food components, addition of specific func- strong umami taste. Monosodium glutamate was evaluated tional ingredients, modification of food composition and by the Joint FAO/WHO Expert Committee on Food Additives masking of undesirable flavors (Hsieh and Ofori 2007), in 1988, and was considered safe (Walker and Lupien 2000; have additionally modified many traditional dishes, and Beyreuther et al., 2007). Glutamate is still widely used as created a variety of newly designed and highly palatable flavouring agent in many restaurant kitchens (for example,

European Journal of Clinical Nutrition Free amino acids in German convenience food products M Hermanussen et al 95

Figure 3 The spectra of free amino acids in Lidl pizza salami (upper part) and Miracoli spaghetti (lower part). Miracoli spaghettis do not contain meat (no carnosine (CARN) peak). The spectra reveal strong resemblance with almost identical peaks of aspartate (ASP), glutamate (GLU), alanine (ALA) and g-aminobutyric acid (GABA). The NH4 peaks are uncertain because of technical reasons.

German Ministry of Food, Agriculture and Consumer Food labels provide information regarding ingredients that Protection, 2008). Glutamate is capable of stimulating do not match with the content of the final ready-to-serve appetite at central hypothalamic levels (Stanley et al., 1996; dish. The mismatch between food label information and Duva et al., 2002, 2005; Hermanussen et al., 2006). Recently, amino acid composition is particularly evident in the case of He et al. (2008) showed that also in East Asians monosodium glutamate. Possibly because of popular concerns regarding its glutamate intake seems to be positively related to body mass safety, a rising number of brands have recently claimed not index, quite in contrast to the popular opinion, considering to have added monosodium glutamate (German Market the Chinese being ‘immune’ to the appetite-stimulating Research Institute, personal communication, August 2008); effect of this amino acid. however, elevated levels of this amino acid are still found,

European Journal of Clinical Nutrition Free amino acids in German convenience food products M Hermanussen et al 96

Figure 4 The spectra of free amino acids in Bassermann ravioli claimed to have not been fortified with the flavoring agent glutamate (upper part) and Maggi ravioli tomato sauce fortified with monosodium glutamate (lower part).

even in dishes that explicitly state ‘no glutamate added’. In uncertain to what extent free amino acids have deliberately addition, the patterns of other free amino acids do not reflect been added to these dishes, or whether the patterns of these the protein composition of their ingredients. Arginine and amino acids in the final meals simply reflect food processing. free alanine were usually elevated; the levels of free Convenience food also contains GABA, but there is an tryptophan, being a precursor of serotonin and a co- amazing scarcity of scientific literature on GABA in human regulator of appetite, varied greatly with high levels in pizza food. GABA is naturally present in many plants (Carratu` and spaghetti, whereas the generally low levels of free et al., 2008). GABA concentrations may also depend on food cysteine possibly reflect the thermal destruction of this storage and processing. Makino et al. (2008) recently showed amino acid during food processing. At present, it remains that GABA levels in tomatoes can be increased by storing

European Journal of Clinical Nutrition Free amino acids in German convenience food products M Hermanussen et al 97 under adjusted aerobic atmospheres. Microbial fermentation Acknowledgements augments GABA in tea (Jeng et al., 2007) and in sour dough (Rizzello et al., 2008), and it has been suggested for appli- We thank Gesellschaft fu¨r Konsumforschung, Nuremburg, cation in dairy and other products (Park and Oh, 2007). Oral Germany. This study was supported by Deutsche Gesellschaft GABA shows effects on taste. Obata et al. (1997) first fu¨r Auxologie, and Norddeutsches Zentrum fu¨r Wachstums- suggested that a GABA-uptake system is present in the taste forschung. buds and that GABAergic neurotransmission is involved in the sensation of taste. Recently, Hisaki and co-workers showed by taste tests that GABA at concentrations near References 0.01% significantly contributes to taste sensation (Hisaki Alexy U, Sichert-Hellert W, Rode T, Kersting M (2008). Convenience et al., 2008 and personal communication, 2008). In contrast food in the diet of children and adolescents: consumption and to the scientific literature, non-scientific publications composition. Br J Nutr 99, 345–351. on GABA are numerous, with many patents stressing Bellisle F (2008). Experimental studies of food choices and palat- economic interests in this substance. A European Patent ability responses in European subjects exposed to the umami taste. Asia Pac J Clin Nutr 17, 376–379. Office quick search on GABA; (title or abstract) revealed 2347 Beyreuther K, Biesalski HK, Fernstrom JD, Grimm P, Hammes WP, patents in their worldwide database (http://ep.espacenet. Heinemann U et al. (2007). Consensus meeting: monosodium com/quickSearch?locale ¼ en_EP). For example, for produ- glutamate—an update. Eur J Clin Nutr 61, 304–313. Carratu` B, Boniglia C, Giammarioli S, Mosca M, Sanzini E (2008). cing an enzyme-processed cacao product having strong Free amino acids in botanicals and botanical preparations. J Food taste and mild flavor, the patent JP2007043931 suggests an Sci 73, C323–C328. enzyme-processed cacao product that enhances the con- Duva MA, Tomkins EM, Moranda LM, Kaplan R, Sukhaseum A, tent of total free amino acid by at least 1.5 times and the Bernardo JP et al. (2002). Regional differences in feeding and other behaviors elicited by N-methyl-D-aspartic acid in the rodent contents of glutamic acid, asparagine acid and arginine by at hypothalamus: a reverse microdialysis mapping study. Brain Res least 1.8 times. Patent WO2006114918 suggests a taste- 925, 141–147. improving agent comprising arginine as an active ingredient. Duva MA, Tomkins EM, Moranda LM, Kaplan R, Sukhaseum A, Patent JP2008048744 provides a new food material or food Stanley BG (2005). Origins of lateral hypothalamic afferents associated with N-methyl-d-aspartic acid-elicited eating studied using sprouted grains for improving taste such as sweet- using reverse microdialysis of NMDA and Fluorogold. Neurosci Res ness and deliciousness with alanine, and allowing to easily 52, 95–106. ingest GABA. Patent JP2007006853 provides a method for Finlayson G, King N, Blundell JE (2007). Is it possible to dissociate increasing the palatability of food and drink or chocolate by 0liking0 and 0wanting0 for foods in humans? A novel experimental procedure. Physiol Behav 90, 36–42. adding at least 0.05% of GABA. This is close to the German Ministry of Food, Agriculture and Consumer Protection concentration we found in the German convenience food (2008): http://www.verbraucherschutz.sachsen-anhalt.de/wirueberuns/ (Table 4). jahresberichte/jb04/jb2004-03-00.pdf(accessed 22 September 2008). The four examples clearly illustrate that economic Halpern BP (2000). Glutamate and the flavor of foods. JNutr130 considerations have started affecting the biochemical com- (4S Suppl), 910S–914S. Hartmann BM, Bell S, Vasquez-Caicedo AL, Go¨tz A, Erhardt J (2005) position of everyday food long before appropriate scientific et al. Bundeslebensmittelschlu¨ssel (German nutritional tables). research has provided sound evidence for its safety. As we do Version II 3 1 Bundesforschungsanstalt fu¨r Erna¨hrung und know that protein content (Hermanussen and Tresguerres, Lebensmittel (Federal Research Institute for Nutrition and Food). Karlsruhe (http://www.blsdb.de/uploads/media/BLS_Lizenz- 2007) and organic acids in food affect eating, satiety and modell_Bestellformular.pdf). body weight regulation, the market success of the new food, He K, Zhao L, Daviglus ML, Dyer AR, Van Horn L, Garside D, et al., paralleled by epidemic obesity, strongly underscores the INTERMAP Cooperative Research Group. (2008) Association of necessity of additional scientific research. We are concerned monosodium glutamate intake with overweight in Chinese adults: the INTERMAP Study. Obesity 6, 1875–1880. and particularly frightened that, at present, any chemi- Hermanussen M, Garcı´a AP, Sunder M, Voigt M, Salazar V, Tresguerres cally designed product claimed to be food can be manufac- JAF (2006). Obesity, voracity and short stature: the impact of tured everywhere and almost ubiquitously be distributed glutamate on the regulation of appetite. Eur J Clin Nutr 60, 25–31. without being properly labeled and controlled for product Hermanussen M, Tresguerres JAF (2007). Overweight, appetite control, and the role of glutamate and excess nutritional protein safety, and about its effect on body weight and on long-term during child development. Hum Ontogenet 1, 23–35. health. Hisaki K, Wada K, Shinohara K, Nakamura Y, Ueno H (2008). The Effect of GABA on Human Taste Sensations and the Influence of Food Components on the activity of glutamate decarboxylase, GABA Synthesizing Enzyme. International Symposium on Olfac- Conflict of interest tion and Taste: San Francisco, CA. (21–26 July, Abstract P346). Hsieh YH, Ofori JA (2007). Innovations in food technology for Theauthorsdeclarenoconflictofinterest.Noadditional health. Asia Pac J Clin Nutr 16 (Suppl 1), 65–73. financial support has been received apart from using the routine http://ep.espacenet.com/quickSearch?locale ¼ en_EP (accessed 22 September 2008). laboratory equipment of the Children’s University Clinic, Jeng KC, Chen CS, Fang YP, Hou RC, Chen YS (2007). Effect Heidelberg, and the Metabolic Unit, Department of Clinical of microbial fermentation on content of statin, GABA, and Chemistry, VU University, Medical Center, Amsterdam. polyphenols in Pu-Erh tea. J Agric Food Chem 55, 8787–8792.

European Journal of Clinical Nutrition Free amino acids in German convenience food products M Hermanussen et al 98 Kok RM, Howells DW, van den Heuvel CC, Guerand WS, Thompson Park KB, Oh SH (2007). Production of yogurt with enhanced levels of GN, Jakobs C (1993). Stable isotope dilution analysis of GABA in gamma-aminobutyric acid and valuable nutrients using lactic acid CSF using simple solvent extraction and electron-capture nega- bacteria and germinated soybean extract. Bioresour Technol 98, tive-ion mass fragmentography. J Inherit Metab Dis 16, 508–512. 1675–1679. Louis-Sylvestre J, Giachetti I, Le Magnen J (1984). Sensory versus Rizzello CG, Cassone A, Di Cagno R, Gobbetti M (2008). Synthesis of dietary factors in cafeteria-induced overweight. Physiol Behav 32, angiotensin I-converting enzyme (ACE)-inhibitory peptides and 901–905. gamma-aminobutyric acid (GABA) during sourdough fermentation Maehashi K, Matsuzaki M, Yamamoto Y, Udaka S (1999). Isolation of by selected lactic acid bacteria. J Agric Food Chem 56, 6936–6943. peptides from an enzymatic hydrolysate of food proteins and Rogers PJ, Blundell JE (1984). Meal patterns and food selection characterization of their taste properties. Biosci Biotechnol Biochem during the development of obesity in rats fed a cafeteria diet. 63, 555–559. Neurosci Biobehav Rev 8, 441–453. Makino Y, Soga N, Oshita S, Kawagoe Y, Tanaka A (2008). Stimulation Stanley BG, Willett VL, Donias HW, Dee MG, Duva MA (1996). of gamma-aminobutyric acid production in vine-ripe tomato Lateral hypothalamic NMDA receptors and glutamate as physio- (Lycopersicon esculentum mill fruits under modified atmos- logical mediators of eating and weight control. Am J Physiol 270 pheres. J Agric Food Chem 56, 7189–7193. (2 Pt 2), R443–R449. Mattes RD (2008). Food palatability, rheology, and meal patterning. Stapleton JR, Roper SD, Delay ER (1999). The taste of monosodium J Parenter Enteral Nutr 32, 572–574. glutamate (MSG), L-aspartic acid, and N-methyl-D-aspartate Obata H, Shimada K, Sakai N, Saito N (1997). GABAergic neuro- (NMDA) in rats: are NMDA receptors involved in MSG taste? transmission in rat taste buds: immunocytochemical study for Chem Senses 24, 449–457. GABA and GABA transporter subtypes. Brain Res Mol Brain Res 49, Walker R, Lupien JR (2000). The safety evaluation of monosodium 29–36. glutamate. J Nutr 130 (4S Suppl), 1049S–1052S.

European Journal of Clinical Nutrition