246

Journalof FoodProtection, Vol. 64, No. 2, 2001, Pages 246– 251 Copyright q,International Association forFood Protection

RoquefortineC Occurrence inBlue Cheese

CARLO FINOLI, 1 ANGELA VECCHIO, 2* ANTONIETTA GALLI, 2 AND IVAN DRAGONI 3

1Dipartimentodi Ingegneria e TecnologieAgro-Forestali, Universita `degliStudi di Palermo, viale delle Scienze 10,90128 Palermo, Italy; 2Dipartimentodi Scienze eTecnologieAlimentari e Microbiologiche,Universita `degliStudi di Milano, via Celoria 2, 20133 Milano, Italy; and 3Istitutodi Ispezione degli Alimenti di Origine Animale, Universita `degliStudi di Milano, via Celoria 6, 20133 Milano, Italy

MS00-186:Received 9June2000/ Accepted 6September2000

ABSTRACT Downloaded from http://meridian.allenpress.com/jfp/article-pdf/64/2/246/1671860/0362-028x-64_2_246.pdf by guest on 27 September 2021 Severalstrains of areused for the production of mold-ripenedcheeses, and some of themare able to produce .The aims of the research were the determination of roquefortine C andPR toxinin domestic and imported blue cheeses,the identiŽ cation of thepenicillia used as starter,and the investigation of their capacity for producing toxins in culture media.Roquefortine C wasalways found in the cheeses at levels ranging from 0.05 to 1.47 mg/ kg,whereas the PR toxinwas neverfound. The identiŽ cation of thefungal strains present in thedomestic cheeses included Penicilliumglabrum, , and Penicilliumcyclopium inthe Gorgonzola ‘ ‘dolce’’ and Penicilliumroqueforti inthe Gorgonzola ‘ ‘naturale’’; inone case, the presence of Penicilliumcrustosum wasobserved. The strains isolated from the foreign cheeses belonged to P.roqueforti. Thestrains were able to producebetween 0.18 and 8.44 mg/ literof roquefortinein yeastextract sucrose medium andbetween 0.06 and 3.08 mg/ literand less than 0.05 mg/ literwhen inoculated in milk at 20 8Cfor14 days and 4 8C for 24 days,respectively. Linear relations between production of roquefortinein culture media and cheeses did not emerge. PR toxin rangedfrom less than 0.05 to 60.30 mg/ literin yeast extract sucrose medium and was produced in milk at 20 8C from only onestrain. The low levels and the relatively low toxicity of roquefortine make the consumption of blue cheese safe for the consumer.

Mostpenicillia reside in soil or onvegetables, and their aspatulin, penicillic acid, mycophenolic acid, isofumigo- presencein foods usually is connected with an accidental clavinA andB, roquefortineC, andPR toxin (11). The contaminationduring production; in some cases, such as presentresearch has been limited to the last two toxins be- fermentedsalami and ripened cheese, they are added to causethey are the mycotoxins more easily produced from developspeciŽ c organolepticcharacteristics (7). Some thecultures used as starters for theproduction of Gorgon- cheesesare manufactured with strains of Penicilliumro- zola.Roquefortine has neurotoxic properties (34); PR toxin queforti, i.e.,the blue-veined cheeses such as Gorgonzola, inhibitsRNA andprotein synthesis (18, 19) andis carci- Roquefort,Bleu de Bresse, Cabrales, Danablu, and Edel- nogenicfor rats (22) andmutagenic for Salmonella Typhi- pilzka¨se,and strains of Penicilliumcamemberti are used for murium (33). whitesurface mold cheeses such as Camembert, Brie, and Fungalstrains used as starters in the dairy industry Coulommiers.Once these molds were naturallyfound in were ableto produce roquefortine in vitro (10, 26). The themilk and cheese, whereas today spore suspensions are presenceof roquefortinewas foundalso in cheeses (15, 26, directlysprayed on the surface of cheeses or added to the 28, 32, 35).PRtoxinwas isolatedŽ rst inthe extracts of milkor curd. The most important biochemical characteris- culturemedium of P.roqueforti by Wei et al. (36) in 1973. ticsare the proteolytic activity, which plays a fundamental Itwas neverfound in blue cheeses (9,23, 26, 27) , because roleduring blue cheese ripening, and the lipolytic activity, itreacts with nitrogenous compounds, converting itself to whichinduces triglyceride hydrolysis and subsequent b- PR-imine,which has a lowertoxicity (1).Siemensand Za- ketoacidand methylketone formation. wistowski (30),ina studyon derivatives of the PR toxin P.camemberti and P.roqueforti were usedwithout incheeses, reported PR-imine at levels of 19 to 42 mg/kg particularproblems in cheesemaking until Gibel et al. (14) in50 ofthe60 samplesanalyzed. The authors demonstrated publisheda studyin whichthey assumed that such cultures thatin vitro, in the presence of bovine serum, the conver- couldrepresent a riskfor consumerhealth because of the sionof thisderivative in themore dangerous original parent formationof toxic metabolites. The authors suggested that was possible.PR toxinwas foundin culture media up to P.camemberti extractsseemed to be able to induce carci- 200mg/ liter (9). nogeniceffects in mice. P.roqueforti, atpresentreclassiŽ ed Gorgonzolabelongs to the group of the blue cheeses afterpolymerase chain reaction identiŽ cation into three spe- havingin common the speciŽ c characteristicof the for- cies named P.roqueforti,Penicillium carneum, and Peni- mationof bluegreenish veins in their interior ,producedby cilliumpaneum (5) ,isable to produce several toxins, such mold of the P.roqueforti type,with tonality of moreor less intensecolor (12,13, 24, 25) .Itis a cheeseof ‘ ‘denomin- *Authorfor correspondence. Tel: 0039022367181; Fax: 0039022361576; azionedi origine’’ (6),originatingfrom Gorgonzola,a town E-mail: [email protected]. nearMilan, and its production is governed by appropriate J.FoodProt., Vol. 64, No. 2 ROQUEFORTINE COCCURRENCE INBLUECHEESE 247 regulationsestablished by theConsortium for theProtection TABLE 1. Mycotoxinrecovery percentages and standard devia- ofGorgonzola Cheese. Only the dairies authorized by the tions consortiumcan market Gorgonzola cheese with the brand Added name ‘‘g,’’whichguarantees its authenticity. Depending on amount Cheese YES medium Milk themanufacturing processes, two varieties of Gorgonzola (mg/kg) (%) (%) (%) canbe obtained:a traditionaltype, ‘ ‘Gorgonzolanaturale,’ ’ Roquefortine25 97 6 5.2 61 6 5.7 53 6 4.2 witha consistingand homogenous texture of the paste, 200 101 6 4.3 65 6 5.1 62 6 4.8 sometimesslightly chalky, a strongtaste, with molds of a shininggreen color ,verydeveloped and uniformly distrib- PR toxin 500 45 6 4.7 65 6 2.3 48 6 3.7 utedin the paste; and a creamytype, ‘ ‘Gorgonzoladolce,’ ’ 1000 46 6 3.4 69 6 1.8 50 6 2.6 witha soft,smeared texture of the paste, sometimes with smalleyelike holes, a sweettaste, with green-colored molds littledeveloped and uniformly distributed in thepaste. The matography(TLC) andin high-performance liquid chromatogra- 7 Downloaded from http://meridian.allenpress.com/jfp/article-pdf/64/2/246/1671860/0362-028x-64_2_246.pdf by guest on 27 September 2021 productionof Gorgonzolacheese is ofabout4.2 3 10 kg/ phy(HPLC) eluentfor roquefortine (3) andin methanol for PR yearand 28% is exported (data from theConsortium for toxinfor HPLC analysis.Working standard solutions of roque- theProtection of Gorgonzola Cheese). fortineand PR toxinwere prepared from crystalline toxins (Sigma Theobjectives of this study were toinvestigate the ChemicalCo., St. Louis, Mo.). The purity of the standards was ability of Penicillium speciesto produce mycotoxins in blue determinedin absolute ethanol by UV analysisusing a spectro- cheeses,to isolateand identify the molds, and to determine photometerUV/ VISLambda15 Perkin Elmer (Perkin Elmer ifthesemolds produce mycotoxins in variousculture media Corp.,Norwalk, Conn.) and compared with reported spectra (8). incubatedin different conditions. BeforeHPLC analysis,all samples were analyzed by TLC. MATERIALS AND METHODS TLC analysis. HPTLC Kieselgel60 F 254 andFertigplatten Kieselgel60 (Merck) (10 by 10 cm), activated at 110 8C for 2 h, TwentyItalian Gorgonzola cheeses, supplied by the Consor- wereused for TLC ofroquefortine and PR toxin,respectively. tiumfor the Protection of Gorgonzola cheese with technical in- Thedevelopment was monodimensional using, in sequence, n- formation,and 10 foreign blue cheeses, purchased in various Eu- hexane,diethyl ether, and chloroform– methanol– ammonium hy- ropeancountries, have been analyzed. droxide(9:1:0.1) for roquefortine and n-hexane, diethyl ether, and toluene–ethyl acetate– acetic acid (6:4:1) for PR toxin.After de- Isolationand classiŽcation of penicillia. Themolds were velopment,plates were air dried. The roquefortine plates were isolatedafter extracting a portionof thegreen zone of the cheese observedunder UV light(UVLS-58 mineral light multiband 254 witha sterileneedle. The mycelium was transferred to a maltagar to366 nm, San Gabriel, Calif.) at 254 nm, and the toxin appeared plateand incubated at 20 8Cfor7 daysaccording to Pitt and Hock- likea darkspot. After spraying plates with 1% CeSO 4 in 6 N ing (21).Thepurity of the strain isolated was tested with a mi- sulfuricacid, roquefortine formed an orange spot (26). For the croscope,and the strain was kept on malt agar slant at 4 8C. The detectionof PR toxin,plates after exposure to UV lightat 254 identiŽcation and classiŽ cation of molds were carried out using nmfor about 1 minwere observed at 366 nm; the dark blue thefollowing media: malt extract agar ,25%glycerol nitrate agar , uorescentspot changed to green  uorescence (4). andCzapek yeast extract agar according to Pitt and Hocking (21). HPLC analysis. Roquefortineanalysis was carried out ac- Mycotoxinproduction. Thecapability of thespecies isolat- cordingto Auerbach et al. (3) andPR toxinanalysis according to edto produce mycotoxins was tested using yeast extract sucrose Siemensand Zawistowski (29).HPLC determinationswere done (YES) (yeastextract, 20 g/ liter;sucrose, 150 g/ liter)broth incu- usinga PerkinElmer Series 4 LiquidChromatograph, equipped bated at 208Cfor14 days. Some strains were also tested in re- with a 20-mlloopand connected to a PerkinElmer LC 75spec- constituted(10%) sterile skim milk, sterilized at 0.5 atm for 30 trophotometricdetector .Thesystem was linked to a personalin- min,and incubated at 20 and 4 8Cfor14 and 24 days,respectively. tegratorPerkin Elmer Nelson 1020. The chromatographic column Thesetemperatures were chosen because P.roqueforti can grow was a 5-mmSupelcosilLC-18 N 8 58230(150 by 4.6mm) supplied atmesophilic to psychrotrophic temperatures, which commonly witha SupelguardCartridge LC-18 N 8 59564(20 by 4.6 mm) areused in blue cheese manufacture. One hundred milliliters of guardcolumn (Supelco Inc., Bellefonte, Pa.). Each extract was YES orsterileskim milk, placed in a500-ml ask,was inoculated injectedin triplicate. All solvents were HPLC grade. withfungal spores to obtain a Žnalconcentration of 10 4 spores/ ml.The spore concentration was conŽ rmed by microscope by Mycotoxinrecovery. Culturemedia and cheeses were spiked meansof direct counting chamber (31). withknown amounts of mycotoxin to determine the method perfor- mance.The analyses were carried out in triplicate, and mean recovery Mycotoxinextraction and puriŽcation. Theextraction of percentagesand standard deviations are reported in T able1. roquefortineand PR toxinfrom the culture media was carried out accordingto Schochet al. (26);theextraction of roquefortinefrom Inuence of technological parameters on the roquefortine 50g ofcheese samples was carried out according to Ware et al. presence. Therelations between presence of roquefortinein con- (35) andthat of PR toxinfrom 50 g ofcheesesamples according sortiumcheeses and some technological parameters, such as salt toSiemens and Zawistowski (29).Theresidue of PR toxinwas concentration,length of ripening, and interval between one spik- subsequentlycleaned up by a LiChrolutSPE Si 2024silica gel ingand an other of thecheese to produce aeration channels, were column(Merck, Darmstadt, Germany) eluted with 5 mlof examined.T oestimatethe salt concentration, a coefŽcient of dif- mixturesof hexaneand ethyl acetate according to Engel and Pro- fusion of Na1 equal to 1 3 1028 cm2/s was used (2, 20), and it kopek (9).Allresidues were reduced to dryness under a gentle wasassumed that the salt was added all at once. The values, nitrogenstream and dissolved in chloroform for thin layer chro- calculatedusing the graphical resolution of Fick equation (20), 248 FINOLIET AL. J.FoodProt., Vol. 64, No. 2

TABLE 2. Roquefortineoccurrence in blue cheeses Roque- Sample fortine no. Type Origin (mg/kg)

1 Gorgonzola‘ ‘dolce’’ Italy 0.13 2 Gorgonzola‘ ‘dolce’’ Italy 0.08 3 Gorgonzola‘ ‘dolce’’ Italy 0.17 4 Gorgonzola‘ ‘dolce’’ Italy 0.31 5 Gorgonzola‘ ‘dolce’’ Italy 0.13 6 Gorgonzola‘ ‘dolce’’ Italy 0.16 7 Gorgonzola‘ ‘dolce’’ Italy 0.42 FIGURE 1. Roquefortinedistribution in blue cheese. 8 Gorgonzola‘ ‘dolce’’ Italy 0.15 9 Gorgonzola‘ ‘dolce’’ Italy 0.40 10 Gorgonzola‘ ‘dolce’’ Italy 0.21 Gorgonzola‘ ‘naturale’’ andin the foreign blue cheeses, Downloaded from http://meridian.allenpress.com/jfp/article-pdf/64/2/246/1671860/0362-028x-64_2_246.pdf by guest on 27 September 2021 11 Gorgonzola‘ ‘dolce’’ Italy 0.17 whereasit is presentonly in afew Gorgonzola‘ ‘dolce.’’ In 12 Gorgonzola‘ ‘dolce’’ Italy 0.13 this type, Penicilliumglabrum isinstead predominant, and 13 Gorgonzola‘ ‘dolce’’ Italy 0.48 Penicilliumcrustosum 14 Gorgonzola‘ ‘dolce’’ Italy 0.15 in one case ispresent. A poolof 15 Gorgonzola‘ ‘dolce’’ Italy 0.48 penicilliacontaining P.roqueforti,P. glabrum , and Peni- 16 Gorgonzola‘ ‘dolce’’ Italy 0.07 cilliumcyclopium was presentin two samples of Gorgon- 17 Gorgonzola‘ ‘naturale’’ Italy 1.43 zola‘ ‘dolce.’’ Themolds inoculated in culture media 18 Gorgonzola‘ ‘naturale’’ Italy 0.68 showeda veryabundant growth in YES anda veryabun- 19 Gorgonzola‘ ‘naturale’’ Italy 1.44 dantor abundant growth at 20 8Canda scantygrowth at 20 Gorgonzola‘ ‘naturale’’ Italy 0.15 48Cinmilk. In YES, P. glabrum, isolatedfrom domestic 21 Danablu Denmark 0.97 Gorgonzola,produced roquefortine levels ranging from 22 Danablu Denmark 1.33 0.18to 4.54mg/ literand PR toxinlevels between 0.05 and 23 Danablu Iceland 1.47 40.10mg/ liter; P.roqueforti producedroquefortine at levels 24 Roquefort France 0.05 between0.31 and 8.44 mg/ literand the PR toxinbetween 25 Roquefort Greece 0.69 26 Roquefort Greece 0.52 0.24and 60.30 mg/ liter;one strain pool produced 2.88 mg/ 27 Edelpilzka¨se Germany 0.30 literof roquefortine and 1.92 mg/ literof PR toxin,and the 28 Edelpilzka¨se Austria 0.24 otherpool produced 3.06 mg/ literof roquefortineand 6.13 29 Stilton GreatBritain 0.21 mg/literof PR toxin. P.crustosum producedonly roque- 30 Stilton Ireland 0.65 fortineat a levelof 1.36 mg/ liter.Twelvestrains (Ž ve P. roqueforti, Ž ve P. glabrum, twostrain pool) inoculated in milkproduced roquefortine at values ranging from 0.06to arereferred to the surface at 0.8 times the radius and 0.8 times 3.08mg/ literat 20 8Candless than 0.05 at 4 8C.Regarding thehalf height of the cheese. PRtoxin,no strain was ableto produce it at4 8C, and only onestrain produced the toxin at a levelof 0.28 mg/ literat RESULTS 208C (Table 4). Theresults reported concerning mycotoxin determi- Therelation between the roquefortine occurrence in the nation,not corrected for recovery,are the average of two cheeseand the production in YESbyallthe strains isolated, experiments. except for P.crustosum, isshown in Figure 2 andthat Roquefortinewas foundin all samples with a maxi- betweenthe toxin presence in thecheese and the production mumvalue of 1.44 mg/ kgin the Gorgonzola from thecon- inthe milk in Figure 3. Meaningfullinear relations between sortiumand 1.47 mg/ kgin the other blue cheese (T able2 productionin culture media and cheeses did not emerge. andFig. 1), whereas the PR toxinwas neverfound. Regardingthe existence of eventual relations between Themorphological characteristics of the colonies iso- presenceof roquefortine in consortium cheeses and some latedare described in T able3, and the list of the strains technologicalparameters, a correlationbetween the amount isolatedis shown in T able4. P.roqueforti ispresent in all ofNaCl in the cheese and the production of toxin by P.

TABLE 3. Morphologicalcharacteristics of the colonies of Penicillium isolateda

P.roqueforti P. glabrum P.crustosum

Diameter,mm 40–70 40–50 25–40 Color Whitegray with center olive green Darkgreen with white edge Greenwith white edge Texture Velvet Velvet Velvet ProŽ le Flat Flat/furrowed Flat Back color Greenbrownish Yelloworange greenish Pale,yellow brown a MoldidentiŽ cation based on Pitt and Hocking (21). J.FoodProt., Vol. 64, No. 2 ROQUEFORTINE COCCURRENCE INBLUECHEESE 249

TABLE 4. Productionof roquefortine and PR toxinin YES and inmilk incubated at 20 8C for 14 days YES Milk (mg/liter) (mg/liter)

Sample Isolated Roque- Roque- no. strains fortinePR toxin fortinePR toxin

1 P. glabrum 1.75 1.17 —b — 2 P. glabrum 0.80 0.05 — — 3 P. glabrum 0.18 9.31 0.06 ,0.05 4 P. glabrum 3.29 1.54 1.01 ,0.05 5 P. glabrum 2.72 2.68 1.05 ,0.05 6 P. glabrum 4.54 2.42 0.09 ,0.05

7 P. glabrum 0.28 40.10 0.56 ,0.05 Downloaded from http://meridian.allenpress.com/jfp/article-pdf/64/2/246/1671860/0362-028x-64_2_246.pdf by guest on 27 September 2021 8 P. glabrum 0.71 9.09 — — 9 P. glabrum 1.82 3.03 — — 10 P. glabrum 0.27 4.20 — — FIGURE 3. Relationbetween roquefortine production in cheese 11 P. glabrum 0.40 7.50 — — versusmilk. 12 P.roqueforti 1.97 0.24 — — 13 P.roqueforti 0.31 1.54 — — 14 Poola 2.88 1.92 0.22 ,0.05 roqueforti canbe observed (Fig. 4). The salt presence in 15 Pool 3.06 6.13 0.13 ,0.05 factfavors, for valuesof 1 to3%, mold growth and prob- 16 P.crustosum 1.36 ,0.05 — — ablyalso toxin production. Also, the length of ripening,in 17 P.roqueforti 2.15 4.30 0.21 ,0.05 the case of P.roqueforti, seemedlinked to an increase of 18 P.roqueforti 2.54 3.39 0.23 0.28 thetoxin production (Fig. 5). Although a clearrelation be- 19 P.roqueforti 0.40 3.74 — — tweenthe interval of the production of aeration channels, P.roqueforti 20 0.59 2.47 — — whichranged from 10to 20 days, and the toxin presence 21 P.roqueforti 0.31 8.33 — — incheeses was notevidenced, it was observedthat shorter 22 P.roqueforti 2.44 16.20 3.08 ,0.05 23 P.roqueforti 8.44 4.05 2.74 ,0.05 intervalscorresponded with a greaterproduction of toxin 24 P.roqueforti 4.28 3.33 1.13 ,0.05 by P.roqueforti. Concerning P. glabrum,theproduction of 25 P.roqueforti 0.65 15.59 — — thetoxin does not seem related to considered parameters. P.roqueforti 26 3.08 12.40 — — DISCUSSION 27 P.roqueforti 1.58 8.40 — — 28 P.roqueforti 2.54 11.62 — — Theability of somestrains of thegenus Penicillium (P. 29 P.roqueforti 0.86 60.30 — — roqueforti,P. glabrum,P. cyclopium )usedin themanufac- 30 P.roqueforti 0.67 13.31 — — turingof blue cheeses to produce mycotoxins has been a Poolincludes P.roqueforti,P. glabrum, and P.cyclopium. shown.The presence of penicillia different from P. roque- b Strainsnot inoculated in milk. forti isnot strange in Gorgonzola; in the production of this cheese,the use of various cultures of selected strains of penicilliais allowed. The use of penicilliamixtures confers tothe cheese its characteristic veins and texture, typical of

FIGURE 2. Relationbetween roquefortine production in cheese FIGURE 4. Relationbetween roquefortine production in cheese versusYES medium. by P.roqueforti versusNaCl concentration. 250 FINOLIET AL. J.FoodProt., Vol. 64, No. 2

ACKNOWLEDGMENTS

Thiswork was supportedby a grantfrom the Ministero dell’Universita `edellaRicerca ScientiŽca eTecnologica(MURST 60%). Theauthors thank Dr. A. Bareggifor his skillful aid in the analyses.

REFERENCES

1.Arnold, D. L.,P .M.Scott,P .E.McGuire,H. Harwig,and E. A. Neva. 1978.Acute toxicitystudies on roquefortine and PR toxin, metabolitesof Penicilliumroqueforti, inthe mouse. Food Cosmet. Toxicol.16:369 –371. 2.Atkins, P .W.1982.Chimica Žsica. Zanichelli,Bologna, Italy. 3.Auerbach, H., E.Oldenburg,and F .Weissbach 1998.Incidence of Penicilliumroqueforti androquefortine C insilages. J. Sci.Food Agric.76:565– 572. Downloaded from http://meridian.allenpress.com/jfp/article-pdf/64/2/246/1671860/0362-028x-64_2_246.pdf by guest on 27 September 2021 4.Betina, V .1985.Thin-layer chromatography of mycotoxins.J. Chro- matogr.334:211–276. FIGURE 5. Relationbetween roquefortine production in cheese 5.Boysen, M., P.Skouboe,J. Frisvad,and L. Rossen.1996. Reclassi- by P.roqueforti versusripening time. Žcationof the Penicilliumroqueforti groupinto three species onthe basis ofmolecular geneticand biochemical proŽ les. Microbiology 142:541–549. 6.Decreto Presidentedella Repubblica n. 1269 del 30 Ottobre 1955. thesingle producer .Thecultures containing P. glabrum, Riconoscimentodelle denominazioni circa imetodidi lavorazione, whichcan become predominant, seem to be more suitable caratteristiche merceologichee zonedi produzione dei formaggi. Gazzetta UfŽciale dellaRepubblica Italiana n.295, 22 December for theproduction of Gorgonzola ‘ ‘dolce’’; infact, the do- 1995. mesticmarket demand is for acreamypaste cheese with 7.Dragoni, I., C. Cantoni,A. Papa,and L. Vallone.1997. Muffe ali- nottoo evident or colored veins. mentie micotossicosi,2nd ed. Citta `StudiEditore, Milano. Roquefortinewas alwayspresent in cheeses, whereas 8.Engel, G. 1979.UV -densitometrischeAnalyse von Roquefortin und PrT oxin.J. Chromatogr.170:288– 291. thePR toxinwas neverfound. Regarding the roquefortine 9.Engel, G., andM. Prokopek.1979. Kein Nachweis von Penicillium producedfrom P.roqueforti, itsconcentration in thecheese roqueforti-Toxinin Ka ¨se. Milchwissenschaft34:272– 274. wouldseem in uenced by the salt added and the ripening 10.Engel, G., and M. Teuber.1978. Simple aid for the identiŽ cation of time.The concentrations of roquefortine found are to be Penicilliumroqueforti Thom:growth in acetic acid.Eur .J.Appl. Microbiol.Biotechnol. 6:107– 111. considerednormal for thistype of productand are in agree- 11.Engel, G., andM. Teuber.1989. T oxinmetabolites from fungal mentwith other authors (15,26, 28, 35) .Thelow levels, cheese starter cultures( Penicilliumcamemberti and Penicilliumro- therelatively low toxicity, and the noncancerogenity of this queforti),p.163– 192. In H.P.VanEgmond (ed.), Mycotoxins in toxinrender blue cheese not hazardous for theconsumer dairyproducts. Elsevier Applied Science, London. 12.Ghitti, C., B.BianchiSalvadori, and C. Rottigni.1996. Il formaggio (11, 32). Gorgonzola.Sci. Tecn. Latt. cas. 47:49–57,115– 124, 195– 216, 417– Toreducethe presence of secondary metabolites in 440. blueand white surface mold cheeses, some authors suggest 13.Ghitti, C., B.BianchiSalvadori, and C. Rottigni.1997. Il formaggio theuse of strains with low productive ability (11, 17). We Gorgonzola.Sci. Tecn. Latt. cas. 48:85–93, 375– 381, 438– 444, 503– have,therefore, correlated the presence of the toxin in the 508. 14.Gibel, W .,K.Wegner,andG. P.Wildner.1971.Experimental inves- cheeseand culture media. The strains isolated, usually be- tigationson thecarcinogenicity of Penicilliumcamemberti var. can- longing to P.roqueforti and P. glabrum, were capableof didum. Arch.Geschwulstforschung 38:1– 6. producingroquefortine both in YES andmilk at 20 8C, and 15.Lo ´pez-Diaz, T.M.,C.Roma´n-Blanco,M. T.Garc ´õ a-Arias, M.L. theŽ rst mediumwas moresuitable for thetoxin production; Garcõ´a-Fernandez,and M. L.Garc õ´a-Lo´pez.1996. Mycotoxins in twoSpanish cheese varieties.Inter .J.FoodMicrobiol. 30:391– 395. however,theycan be considered low producers. Relations 16.Lund, F .,O.Filtenborg,and J. CFrisvad.1995. Associated mycof- betweenconcentration of roquefortinein culture media and loraof cheese. FoodMicrobiol. 12:173– 180. inthe cheese were notshown. All strains, except for P. 17.Moss, M. O.1992.Secondary metabolism andfood intoxication- crustosum, producedPR toxinin YES brothwith values up moulds.J. Appl.Bacteriol. Symp. Suppl. 73:80s– 88s. 18.Moule, Y .,M.Jemmali, andN. Darracq. 1978.Inhibition of protein to60 mg/ liter,whereasmilk was alesssuitable medium. synthesisby PR toxin,a mycotoxinfrom Penicilliumroqueforti. Anaspect that should not be neglectedin cheese man- FEBSLett. 88:341– 344. ufacturingis the control of the cultures used as starters, 19.Moule, Y .,M.Jemmali, andN. Rousseau.1976. Mechanism ofthe becausesome Penicillium speciesproduce toxic metabo- inhibitionof transcription by PR toxin,a mycotoxinfrom Penicillium roqueforti. Chem.Biol. Interact. 14:207– 216. lites.This is the case of the occurrence, probably due to 20.Peri, C., andB. Zanoni1994. Manuale di T ecnologieAlimentari I, accidentalpollution, of P.crustosum ableto produce pen- Cusl,Milano. itrem A (16).Inconclusion, the search of toxicmetabolites 21.Pitt, J. I.,andA. D.Hocking.1997. Fungi and food spoilage, 2nd infoods and the selection of fungal strains with reduced ed.Blackie Academic &Professional,London. 22.Polonelli, L., L.Lauriola,and G. Morace.1982. Preliminary studies abilityto produce these substances is justiŽ ed by the risk onthe carcinogenic effects of Penicilliumroqueforti (PRtoxin)on oftheir possible exposure for longperiods even though to rats. Mycopathologia78:125– 127. lowdosages. 23.Polonelli, L., G.Morace, F.Delle Monache,and R. A.Samson.1978. J.FoodProt., Vol. 64, No. 2 ROQUEFORTINE COCCURRENCE INBLUECHEESE 251

Studieson the PR toxinof Penicilliumroqueforti. Mycopathologia 31.Splittstoesser ,D.F.1992.Direct microscopiccount, p. 97– 104. In 66:99–104. C.Vanderzantand D. F.Splittstoesser(ed.), Compendium of meth- 24.Salvadori del Prato, O. 1999.Il re deglierborinati: il Gorgonzola. odsfor the microbiological examination of foods, 3rd ed. American Latte 24(5):70–78. PublicHealth Association,W ashington,D.C. 25.Salvadori del Prato, O. 1999.Il re deglierborinati: il Gorgonzola. 32.Teuber ,M.,andG. Engel.1983. Low risk of mycotoxinproduction Latte 24(6):34–46. incheese. Microbiol.Aliments Nutr.1:193–197. 26.Schoch, U., J.Lu¨thy,and C. Schlatter.1984.Mykotoxine von P. 33.Ueno, Y .,K.Kubota,T .Ito,and Y .Nakamura. 1978.Mutagenicity roqueforti and P.camemberti inKa ¨se, I:vorkommenchemisch iden- ofcarcinogenicmycotoxins in Salmonellatyphimurium. Cancer Res. tiŽzierter Mykotoxine.Milchwissenschaft 39:76– 80. 38:536–542. 27.Scott, P .M.,andS. R.Kanhere.1979. Instability of PRtoxinin blue 34.Wageneger ,R.E.,N.D.Davis,and U. L.Diener.1980. cheese. J.Assoc. Off.Anal. Chem. 62:141– 147. androquefortine production by Penicilliumcommune. Appl. Envi- 28.Scott, P .M.,andB. P.C.Kennedy.1976. Analysis of blue cheese ron.Microbiol. 39:882– 887. forroquefortine and other alkaloids from Penicilliumroqueforti. J. Agric.Food Chem. 24:865– 868. 35.Ware, G.M.,C.W.Thorpe,and A. E.Pohland.1980. Determination 29.Siemens, K., and J. Zawistowski.1992. Determination of Penicillium ofroquefortine in blue cheese andblue cheese dressingby high roqueforti toxinby reversed-phase high-performanceliquid chro- pressure liquidchromatography with ultraviolet and electrochemical matography.J. Chromatogr.609:205– 211. detectors.J. Assoc. Off.Anal. Chem. 63:637– 641. Downloaded from http://meridian.allenpress.com/jfp/article-pdf/64/2/246/1671860/0362-028x-64_2_246.pdf by guest on 27 September 2021 30.Siemens, K., andJ. Zawistowski.1993. Occurrence ofPR imine,a 36.Wei, R. D.,P.E.Still,E. B.Smalley,H. K.Schnoes,and F .M. metaboliteof Penicilliumroqueforti, inblue cheese. J.FoodProt. Strong.1973. Isolation and partial characterization ofa mycotoxin 56:317–325. from Penicilliumroqueforti. Appl.Microbiol. 25:111– 114.