Food Sci. Techno/, Int., 2 (2), 82-83, 1996

Technical Paper

Quantitative Determination of Erythritol from Various Natural by HPLC

Tatsujl SHINDOUI and Hiroaki ISHIZUKA2,*

1 Division of Chemical and Biologica/ A nalysis, Nikken Chemicals Co., Ltd., 1-146 Kitabukuro-machi Omlya, Saitama 330, Ja pa n 2Development Department, Nikken Chemicals Co., Ltd., Ginza East Bldg. 16-14, Ginza 7-chome, Chuo-ku, Tokyo 104. Japan

Received March l, 1995

Erythritol in various natural cheeses was quantitatively determined using high-performance liquid chromato- graphy (HPLC). The peak detected at the position of erythritol by HPLC was identified by gas chromatography- mass spectrometry (GC-MS). Erythritol was detected in samples ripened with fungi and contents ranged from 0.01-2.96 mg/g, and surface-ripened cheeses with white fungi especially contained more erythritol compared to blue-vein cheeses with blue fungi inside. No erythritol was found in cheese samples ripened with bacteria and without ripeness. In the case of surface-ripened cheeses, much more amounts (1.99-6.86 mg/g) of erythritol was found in their surface part compared to the inner part (0-0.33 mglg). These results seem to suggest that erythritol is being produced by conventional microorganisms, Penicillium sp., used in ripeness.

Keywords: erythritol, polyol, cheese, HPLC, sweetener

Erythritol, known as meso-erythritol, is a tetrose sugar blocks were removed and analyzed separately. The remaining alcohol. Its sweetness Is approximately 70-80% of sucrose and blocks were then analyzed as the inner part. is being considered as a new non-caloric (Noda & Oku, 1992) and non-cariogenlc (Kawanabe et al. 1 992) sweetener. At Results and Discussion present, erythritol is commercially produced in Japan from Twenty cheese sarnples, analyzed for their erythritol con- yeast (Ishizuka et al, 1989) and its annual production tents, were classified into 5 groups according to the micro- reached 2,000 tons in 1994. It is widely used in table sugars, organisms used for their ripening processes. The results of the beverages and confectioneries. HPLC analysis are shown in Table I . The peak detected at Erythritol is known to widely exist in nature as has been the same retention time as erythritol by HPLC was further reported in our previous paper (Shindou et al, 1989). confirmed by its chemical identity using GC-MS. It is clearly However, the content or existence of erythritol in natural shown that cheese samples ripened with fungi contained cheeses has not been known until now. In this paper, the erythritol. No erythritol was found in cheese samples ripened content and existence of erythritol in natural cheeses commer- with bacteria or had not ripened. These results suggest that cially obtained in Japan has been studied. erythritol found in cheese samples had been produced by fungi during the ripening process. Actually no erythritol was Materials and Methods found in milk, the starting material of the cheese. The The imported natural cheeses and domestic cheeses were erythritol content of the surface-ripened cheeses with white purchased from a department store in Tokyo and Omiya-city, fungi varied from 0.56 to 2.96 mg/g, the average content of I l Japan. As an erythritol standard, a sample prepared by samples being 1.41 mg/g. On the contrary, the erythritol recrystallization from commercially obtained erythritol (Nik- content of 3 blue-veln cheeses ripened with blue fungi fell in ken Chemicals Co., Ltd.) was used. the lower range of 0.01 to 0.55 mg/g. Cambozola cheese was A 20 g portion was accurately weighed from cheese generally classified as blue-vein cheese, but its erythritol samples, homogenized in a solution composed of 30 ml of content (0.55 mg/g) is relatively close to those of the surface- dlstilled water and 0.5 ml of 1,4:3,6-dianhydro-D-glucitol ripened cheese. This is because that during the preparation of solution (50 mg/ml) and centrifuged at 8,000Xg for 10 min. Cambozola cheese, white fungi were used in addition to blue The obtained supernatant was successively treated with 2 g of fungi. In surface-ripened cheese, the white mycelium of activated charcoal and 2-3 g of mixed resin (MB-3, Japan Penicillium camemberti grows on the surface of cheese Organo Co.). The obtained solution was passed through a sprayed with spores but does not penetrate the interior. In the membrane-filter (MILLEX-HV, 0.45 /Im, 25 mm i.d., Japan blue-vein cheese, the mycelium of Penicillium roqueforti Millipore, Ltd.) and 50 pl of the filtrate was analyzed using grows along the spiked perforations of the inner portion, as HPLC. The HPLC and GC-MS analyses of erythritol were the blue spores are included in the starter. These facts carried out as described in our previous paper (Shindou et al, suggested that Penicillium sp, produced a greater amount of l 989). In some cases, a 5 mm thick surface layer of the cheese erythritol under more aerobic conditions and/or P. camem- berti produced a greater amount of erythritol than P. * To whom correspondence should be addressed. roqueforti . Quantitative Determination of Erythritol from Various Natural Cheeses by HPLC 83

Table 1. Erythntol contents m natural cheeses Cheese samples Type of cheese Country produced Erythritol (mg/g) a) Ripened with white fungi Baraka surface-ripened 2,18 (6.1) Brle surface-ri pened France 1.79 (6.0) de Meaux surface-ripened France 0.56 (5.2) surface-ri pened Denmark 1.46 (4.8) Camembert Hokkaido surface-ripened Japan l.13 (4.9) Camembert Koiwai surface-ripened Japan l.04 (5.7) Caprice de Dieux surface-ri pened France 0.97 (6.7) Gaperon d' surface-ri pened France l. 18 (6.8) Isigny Pyrograve surface-ripened France 0.85 (5.9) Isigny sainte m6re suTface-ri pened France l.39 (6.2) R6v6rend surface-ri pened France 2,96 (7,4) b) Ripened with blue fungi Blue cheese Bl ue-vein Denmark 0.01 (4.6) Gorgonzola dolcebelde Blue-vein Italy 0.27 (8.2) Papillon selection Blue-veln France 0.03 (4.0) c) Ripened wlth blue fungi and whlte fungi Cambozola Blue-vein Germany 0.55 (5.5) d) Ripened wlth bacteria Gouda I M Hard Netherlands ND Petit Livarot Washed France ND Pyramid Cendree C h~vre France ND e) Without ripeness Cream cheese Cream Australia ND Mascarpone Fresh Italy ND ( ): Coefficient value %, n=3; ND: Not detected <0.01 (mglg).

Table 2. Erythntol contents In surface and Inner part of surface-ripened equal amounts (0.94-0.96 mg/g for surface part, 0.82- 1.10 cheeses. mglg for inner part) of erythritol were found in the 2 parts of Erythritol (mglg) 2 Japanese cheeses. There seems to be a significant difference Cheese samples Surface part Inner part between the French and Japanese cheeses with respect to the Baraka 4.13 (8.3) ND Penicillium strains used for ripening or in their manufacturing Isigny Pyrograve I .99 (4.7) 0.33 (5.6) Isigny sainte m6re 6.04 (7.7) O. 18 (6, I ) processes. Detailed kinetics of erythritol production during the R6v6rend 6.86 (9. l) ND Camembert Hokkaldo 0.97 (5.4) l.lO (5.1) ripening process and its contribution to the taste of cheese Camembert Koiwal 0.94 (6.2) 0.82 (5.9) remain to be studied. ( ): Coefficlent value %, n-3; ND: Not detected <0.01 (mglg). References Ishizuka, H., Wako, K., Kasumi, T, and Sasaki T. (1989). Breeding of a mutant of A ureobasidium sp. with high erythritol production. J. Ferment. Bioeng., 68, 3 l0-3 14. In the next experiment, Iocalization of erythritol within the Kawanabe, J., Hirasawa, M., Takeuchi, T., Oda, T. and lkeda, T. surface-ripened cheeses with white fungi was exarnined. The ( 1 992). Noncariogenicity of erythritol as a substrate. Caries Res., 26, results of 6 samples are shown in Table 2. In the case of 4 358-362. French cheeses, greater amounts ( I .99-6.86 mg/g) of eryth- Noda, K. and Oku, T. ( 1992). Metabolism and disposition of erythritol after oral admlnistration to rats. J. Nutr., 122, 1266-1272. ritol were found in their surface parts compared to the inner Pfyffer, G.E. and Rest, D.M. ( 1980). The polyol pattern of some fungi part (0-0.33 mglg). Pfyffer & Rest (1980) reported that not hitherto investigated for sugar alcohols. Exp. Mycol., 4, 160- 1 70. erythritol was found in the mycelia of 3 Penicillium strains. Shindou, T., Sasaki, Y., Miki, H., Eguchi, T., Hagiwara, K. and These results and facts strongly support the previous sugges- lchikawa, T. (1989). Identification of erythritol by HPLC and tion that erythritol found in cheese was produced by GC-MS and quantitative measurement in pulps ofvarious fruits. J. Agric. Food Chem., 37, 1474-1476. Penicillium used for ripening. On the contrary, essentially