HAL Id: hal-00929606https://hal.archives-ouvertes.fr/hal-00929606

Submitted on 1 Jan 1998

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

Evaluation of proteolysis in Parmesan cheese usingelectrophoresis and HPLC

Helmut Karl Mayer, Christa Rockenbauer, Hannes Mlcak

To cite this version:Helmut Karl Mayer, Christa Rockenbauer, Hannes Mlcak. Evaluation of proteolysis in Parmesancheese using electrophoresis and HPLC. Le Lait, INRA Editions, 1998, 78 (4), pp.425-438. �hal-00929606�

Lait (1998) 78, 425-438© InralElsevier, Paris

425

Original article

Evaluation of proteolysis in Parmesan cheese usingelectrophoresis and HPLC*

Helmut Karl Mayer**, Christa Rockenbauer, Hannes Mlcak

Department of Dairy Research & Bacteriology, University of Agriculture, Gregor Mendel-StrafJe33, A-//80 Vienna, Austria

Abstract - The proteolytic changes occurring during the ripening of Parmesan cheese were studiedusing urea-polyacrylamide gel electrophoresis (urea-PAGE) of caseins, HPLC analysis offree aminoacids (FAA) and Kjeldahl determination of soluble nitrogen fractions. An electrophoretic ripening indexfor the evaluation of proteolysis in Parmesan cheese was established. The separation of caseins byalkaline PAGE (12 % T, 2.6 % C, pH 8.9, 5 M urea) was followed by the densitometric analysis ofthe y- and ~-casein fractions. The relationship between the resulting coefficients (y-Cn/~-Cn) and theage of reference samples of Original ltalian Grana Padano (6-22 months) was linear up to 15 months,allowing an evaluation of the extent of proteolysis and therefore a deduction of the age of the Parmesansamples analysed. Threshold levels (y-Cn/~-Cn) were proposed for the verification of the required ageof Parmesan cheese. The coefficients (y-Cn/~-Cn) as weil as the 13-casein content of two addition alseries of references and of 117 commercial Parmesan sampI es are presented. Commercial Parmesansamples retailed as a loaf or as prepacked slices were generally found to fulfill the requirementsconcerning endoproteolytic changes during ripening. However, man y grated Parmesan samplestaken from retail outlets in Austria showed poor quality, which was probably due to the adulterationwith products with low proteolysis (e.g., cheese rind, very young cheese). HPLC analysis ofFAA wasalso used to characterize the ripening process. Although FAA content ofreference samples of GranaPadano showed a very high degree of variability, a distinct increase of FAA content could be obser-ved during the ripening period. However, two series of reference samples, which had completelydifferent electrophoretic casein patterns, could not be distinguished by HPLC analysis of the totalamount of FAA. As an experiment, two loaves of Parmesan cheese were removed from the ripe-ning room at an age of 4 and 2 months and were subsequently stored at 6 "C for 18 months. Duringthe storage period, endoproteolytic changes were slowed down and breakdown of C:XS1- and ~-caseinas weil as the accumulation of degradation products proceeded at a reduced rate, and could bedetected using urea-PAGE casein patterns, but not by HPLC analysis of FAA content. In conse-quence of these findings, no accurate evaluation of the age of commercial Parmesan samples was pos-sible by means of the total amount of FAA. Since Kjeldahl determination of WSN and HPLC analysisof FAA content give insufficient information about the distinct endoproteolytic changes, which werefound to be typical for Parmesan cheese, urea-PAGE of the casein fraction has to be do ne to enable

* Oral communication at the IDF symposium 'Ripening and quality of cheeses', Besançon, France,February 26-28, 1996.

** Correspondence and reprints. E-mail: mayer@mail.boku.ac.at

426 H.K. Mayer et al.

the evaluation of the extent of proteolysis and therefore the deduction of the age of commercial Par-mesan samples analysed. © InraJElsevier, Paris

Parmesan cheese / proteolysis / electrophoresis / free amino acid / soluble nitrogen

Résumé - Évaluation de la protéolyse du fromage parmesan par électrophorèse et HPLC. Laprotéolyse du parmesan a été suivie au cours de l'affinage par électrophorèse (urée-PAGE) descaséines, analyse par HPLC des acides aminés libres et détermination par Kjeldahl des fractionsazotées solubles. L'analyse des profils protéiques obtenus par électrophorèse a permis d'établir un indexd'affinage grâce auquel l'âge des échantillons de fromages du commerce peut être estimé. L'utilisationde cet index a en outre permis de montrer que beaucoup de parmesan vendu sous forme de râpé étaitde mauvaise qualité. L'analyse des acides aminés libres ne permet pas en revanche de déterminer cor-rectement l'âge du fromage. © InraJElsevier, Paris

fromage parmesan / protéolyse / électrophorèse / acide aminé libre / azote soluble

1. INTRODUCTION

Parmesan cheese has considerable com-mercial importance, therefore severalmethods have been reported to control qua-lity and authenticity of Parmesan. One ofthe prerequisites established by productionstandards for Parmigiano-Reggiano, GranaPadano and other Italian hard cheese varie-ties labelled as Original Italian ParmesanChee se is the minimum age of 12 months[l, 12]. In most cases, attention has beenfocused on the parameters that quantify andcharacterize proteolysis occurring duringcheese ripening [9].

The degradation of the casein fractionsproceeds rapidly during the first 8-10 monthsof ripening, slowing down considerablyduring further maturation of Parmesancheese. This has been shown by followingthe variations in soluble nitrogen with respectto total nitrogen over time [6, 7, 29].

The changes in insoluble proteins havebeen studied using electrophoretic tech-niques [9, 22, 24]. It has been shown that~-casein is hydrolysed very rapidly duringthe first months of ripening, disappearingafter approximately one year [l, 6]. Con se-quently, the Y2-and Y3-caseins continue toincrease, and Yl-casein tends to decreaseconcomitantly, which confirms the impor-

tant role of plasmin in maturation [15,23].Due to manufacture technology, this indi-genous milk proteinase remains active andcontributes to proteolysis during the ripe-ning of Parmesan cheese [14]. Using fastatom bombardment-mass spectrometry, itwas also found that the majority of oligo-peptides arises from ~-casein [2-4].

In Parmesan cheese, the total amount offree amino acids (FAA) rapidly reacheshigher levels than in other cheeses [7,17-19]. The relative percentages of certainFAA were used to establish useful chemo-metric models for evaluating authenticityofParmigiano-Reggiano [25, 26]. Based onthe relative amounts of individu al FAA, atwo-dimensional chemometric model wasestablished for distinguishing traditionalGrana Padano from similar cheeses,although great differences were observedamong the patterns of samples with the sameripening period but produced by differentdairies [27].

The purpose of the present study was: i)to study proteolytic changes occurringduring the ripening of reference samples ofParmesan chee se using urea-PAGE ofcaseins, HPLC analysis of FAA and Kjel-dahl determination of soluble nitrogen; ii)to establish an electrophoretic ripening indexfor the evaluation of the extent of proteoly-

Proteolysis in Parmesan cheese

sis and the deduction of the age of Parmesancheese, and to analyse commercial Parmesansamples with respect to this proteolysisindex; and iii) to determine proteolysis intwo loaves of Parmesan, which were remo-ved from the ripening room at an age of 4 or2 months and were subsequently stored at6 "C for 18 months, and to study the effectof this storage period on the endoproteoly-tic changes, which were found to be typi-cal for Parmesan chee se ripened accordingto the traditional technology.

2. MATERIALS AND METHODS

2.1. Cheese samples

Reference samples (n = 24) of Original ltalianGrana Padano of different ages (6-22 months)were supplied by commercial dairy plants in ltaly(Series 1). Reference samples (n = 10) of GranaPadano (10-20 months), which had been certifiedby the Consorzio Grana Padano, were obtainedfrom Prof. P. Resmini (Università degli Studi diMilano, 1-20133 Milan) (Series II), additionalreference samples (n = 13, 1-18 months) wereobtained directly from the Consorzio GranaPadano (1-20121 Milan) (Series III).

Commercial Parmesan samples (n = 49 as aloaf or prepacked slices and n = 68 as grated pro-ducts) were taken from retail outlets in Austriaand Italy.

Parmesan cheese was produced in an Aus-trian cheese factory according to the traditionaltechnology. The cheeses had a large cylindricalshape with a diameter of 40 cm and a height of20 cm. At first, the manufacturing protocol forParmesan cheese was followed exactly, but thentwo cheeses were removed from the ripeningroom at an age of 4 (Loaf A) and 2 (Loaf B)months. Cheese loaves were subsequently sto-red at lowered temperature (6 "C) for 18 months,so that their final age was 22 (Loaf A) and20 months (Loaf B). At regular intervals, sampleswere taken using a cheese borer and kept frozenfor further analyses.

2.2. Preparation of cheese proteins

After removal of the rind, cheese sampi eswere grated, dried at 45 "C, defatted by extractionwith petrolether in a Soxhlet apparatus and dried

427

at 30 oc. The resulting powder was dissolved in10 rnmol-L -i Tris-glycine buffer, pH 8.3, contai-ning 6 rnol-L"! urea and 3 % (v/v) 2-mercapto-ethanol to give a final concentration of 1 % (w/v).

2.3. Urea-polyacrylamide gelelectrophoresis (urea-PAGE)

Urea-PAGE analysis of caseins was perfor-med by using a dual cooled vertical slab gel elec-trophoresis unit SE 600 (Hoefer Scientific Ins-truments, San Francisco, CA, USA). Separationgel (12 % T; 2.6 % C; 380 mmol-L'" Tris-citratebuffer, pH 8.9; 5 mol-L -1 urea), stacking gel(5 % T; 7.5 % C; 120 mol-L:' Tris-citrate buffer,pH 7.2; 5 mol-L"! urea) and electrode buffer(10 mol-L"" Tris-glycine buffer, pH 8.3) wereprepared according to the procedure of Mayer[22], which was modified from Maurer [20].Electrophoresis was carried out at 15 "C at aconstant voltage of 200 V for 30 min and 400 Vfor 4 h. Gels were simultaneously fixed and stai-ned with 0.15 % (w/v) Coomassie brilliant blueG-250 in water/methanol/acetic acid (60/401lO:v/v/v), destaining was done in water/metha-nollacetic acid (70/3017:v/v/v) [13]. Ali electro-phoresis chemicals were of plusonet" quality(Pharmacia LKB Biotechnology, Uppsala, Swe-den); other chemicals were of analytical grade(Merck, Darmstadt, Germany).

2.4. Densitometry

Densitometric evaluation of colour photos(Polaroid SX-70 instant film) ofurea-PAGE gelswas performed at 619 nm using Shimadzu(Kyoto, lapan) CS-91O densitometer and C-R1Aintegrator. Quantification was based on the mea-surement of peak areas of y-casein (calculatedas the sum ofY2-' Y3-and Yj-caseins [10]) and of~-casein. Reference samples of Grana Padano ofdifferent ages were used for calibration curves.Coefficients (y-casein/~-casein) were plottedagainst the age of cheese (months) and linearrelationships were calculated for Series 1and III.Determination of the ~-casein content (g/100 gcheese solids non-fat) was performed by using~-casein (Sigma Chemical Co., St Louis, MO,USA) as an extemal standard.

2.5. HPLC analysis of free amino acids(FAA)

ln Erlenmeyer flasks, 3 g of grated chee sewere suspended in 27 g 100 mmol-L:! citrate-

428 H.K. Mayer et al.

HCI, pH 2.2 buffer. The mixture was homoge-nized with an Ultra Turrax (8 000 rpm for 2 min),moderately stirred for 20 min and filtered througha Schleicher & Schuell 595 1/2 folded filter.Five g 3 % (w/v) 5-sulfosalicylic acid were addedto 1 g cheese filtrate and after 20 min of stirring,the suspension was filtered through a Schleicher& Schuell 595 1/2 folded fil ter. One g filtratewas diluted (1:50) with 50 mmol-L"! borate-HCl,pH 9.0 buffer and 20 J.lLsampi es were used forHPLC analysis of free amino acids. Ali chemicalswere of analytical grade (Merck).

Free ami no acid analysis was carried out by ahigh-performance liquid chromatography(HPLC) method according to the WatersAccQ.Tag® precolumn derivatization procedureusing the Waters AccQ.Tag® chemistry package[30]. RP-HPLC was performed on a WatersChromatography System (Waters Corp., Mill-ford, MA, USA) using a model 600 E multisol-vent delivery system, an U6K inje ct or, guardcolumn and an AccQ.Tag® amino acid anal y-sis column (Nova-Pak! C 18, 4 um), Columneluates were monitored at 395 nm (excitation at250 nm) using a Shimadzu (Tokyo, Japan) RF-535 fluorescence detector interfaced with PCrunning Waters Maxima 820 software for auto-matie quantitation and documentation.

To prepare eluent A from Waters AccQ.Tag®Eluent A concentrate, 200 mL of the concentratewere added to 2 L Milli-Q'" water (Millipore,Bedford, MA, USA). Eluent B was HPLC-gradeacetonitrile (Labscan Ltd, Dublin, Ireland), eluentC was Milli-Q'" water (Millipore). Ali eluentsand derivatized sampi es were filtered throughMillipore filters (HA-aqueous, 0.45 um, FH-organic, 0.5 um and GY, 0.22 um), Sam pieswere analysed in duplicate, injection volume was5 J.lL.Elution was carried out at 37 "C followingthe instructions for single-pump gradient deli-very systems [30]. The Waters amino acid hydro-lysate standard solution, containing œ-amino-n-butyric acid as an internai standard, was used forcalibration in the range of 2.5 to 40 pmol.Concentrations of individual amino acids weredisplayed as pmol, which were subsequentlyconverted to the total amount of amino acids cal-culated as g FAA per 100 g of cheese protein.

2.6. Kjeldahl determination of solublenitrogen fractions

The following fractionation procedure wasused only for Loaf A and Loaf Band was per-

formed according to the instructions reportedpreviously [7, 8, 28] with modifications.

The total nitrogen (TN) content of the cheesesamples was determined by the Kjeldahl methodafter mineralization [II].

Grated cheese (20 g) was suspended in 180 gdeionized water at 50 oc. The mixture was homo-genized with an Ultra Turrax apparatus(9 000 rpm for 2 min), moderately stirred for 1 hat 50 "C and subsequently homogenized asabove. After centrifugation (10 000 g for 30 minat 5 OC) the suspension was filtered through aSchleicher & Schuell 595 112 folded filter.Cheese filtrate was finally filtered through a Mil-lipore filter (GS, 0.22 um), nitrogen soluble inwater (WSN) was determined in duplicate usingthe Kjeldahl method.

Twenty mL of 24 % (w/v) trichloracetic acid(TC A) solution was added to 20 mL of cheesefiltrate. The suspension was held at room tem-perature overnight and then filtered through aSchleicher & Schuell 602 112folded filter. Thenitrogen content of the TCA-soluble fraction(TCA-N) was determined by the Kjeldahlmethod.

Twenty mL of 10 % (w/v) phosphotungsticacid (PT A) solution was added to 20 mL ofcheese filtrate and held at room temperature over-night. The nitrogen content of the PTA-solublefraction (PT A-N) was determined after filtrationusing the Kjeldahl method. Ali chemicals were ofanalytical grade (Merck).

Nitrogen fractions soluble in water (WSN),12 % TCA (TCA-N) and 5 % PT A (PT A-N)were calculated as relative content of the totalnitrogen (% ofTN).

3.RESULTS

3.1. Urea-PAGE casein patternsand densitometry

Figure 1 shows the eleetrophoretie pat-terns obtained by urea-PAGE analysis ofthe reference samples of Grana Padano ofSeries 1. As a result of the action of plas-min, which seems to be one of the mostactive endopeptidases in Grana Padano,~-easein is hydrolysed rapidly during thefirst year of ripening and the y-caseins areaccumulated as its major degradation pro-duets. These peptides are present even in

Proteolysis in Parmesan cheese

4 years old Parmigiano-Reggiano cheese(figure l , right lane). Since the true compo-sition of the y-casein fraction depends onthe genetic variants of ~-casein occurringin the bulk milk used for cheesemaking [10],this fraction was summarized as 'y-casein'for simplification. <Xs1-caseinis the principaltarget for proteolysis during the early stagesofripening. Consequently, <Xs1-I-casein,asits primary degradation product, continues to

429

increase until the 12th month of ripening,and is subsequently hydrolysed to yield<Xs1-II-caseinand other peptides. As the dif-ferentiation and identification of these degra-dation products was not an object of thisstudy, this fraction was termed '<Xs1-II-casein' [16].

Referring to the breakdown of ~-casein,the ripening characteristics of Parmesancheese becomes even more evident when

o 6 8 12 15 18 213ft 48*Origin- -

1!It_"-<1s1-Cn

- <1s1-I-Cn}<1s1-II-Cn

(+)

Figure 1. Urea-PAGE analysis of reference samples of Grana Padano (Series 1) and of two additio-nal samples of Parmigiano-Reggiano (indicated by an asterisk). Numbers refer to the age of cheese(months).Figure 1. Électrophorégramme urée-PAGE des échantillons de référence de Grana Padano (série 1)etde deux échantillons de Parmigiano-Reggiano (indiqués par un astérisque). Les numéros corres-pondent à l'âge du fromage (en mois).

o B-Cn 6

B-Co

....

12 18'\'-Cn

B-en

.;co.;co.;co..co

rJV\.IU~l/V

..co.;co.;co.;co

Figure 2. Densitometric curves of selected reference samples of Grana Padano (Series 1). Numbersrefer to the age of chee se (months).Figure 2. Courbes densitométriques de quelques échantillons de référence de Grana Padano (série 1).Les numéros correspondent à l'âge du fromage (en mois).

430 H.K. Mayer et al.

comparing the densitometric curves of selec-ted reference samples of Grana Padano(Series 1) of different ages (figure 2). Theripening characteristics found in samples ofSeries 1were in good agreement with thoseobserved in reference samples of GranaPadano of Series II. [3-casein is hydrolysedrapidly and intense bands are accumulated inthe y-casein region. However, the proteoly-sis of cxs1-casein occurred more slowly andseemed to be more varied (results notshown).

Reference samples of Grana Padano ofSeries III showed quite surprising electro-phoretic patterns of caseins, which weresimilar to those of very young Grana Padanochee se. Neither the breakdown of [3-casein,nor a distinct degradation of cxs1-casein,

l'-en '-.1(1

B-en

··c

~œ

•.sIc;:)

•(S:tC;;:I•C:;:,

li •'~II{

n,Ccl-l.n

Figure 3. Urea-PAGE analysis and densitometriccurve of an 18-month-old reference sample ofGrana Padano (series III).Figure 3. Électrophorégramme urée-PAGE etcourbe densitométrique d'un échantillon de réfé-rence de Grana Padano âgé de 18 mois (sérieIII).

could be observed during the ripening per-iod of 18 months (figure 3). Based on theelectrophoretic study of caseins, in any casethe degree of proteolysis is not related tothe labelled age of these samples, which hadbeen certified by the Consorzio GranaPadano. This mysterious discrepancy getseven more evident when comparing the den-sitometric curves of selected referencesamples of Series 1 (figure 2) and Series III(figure 3). In striking contrast to referencesamples of Series 1, samples of Series IIIdo not show any noticeable breakdown of [3-casein during the ripening period of18 months.

Figure 4 shows the electrophoretic caseinpatterns of Loaf A and Loaf B. At first, inparticular Loaf A shows an even higherextent of casein breakdown in comparison tothat observed in reference samples of Series1. However, during further storage at low-ered temperature, endoproteolytic changesare slowed down. Breakdown of CXS1- and[3-casein as weIl as accumulation of degra-dation products proceeds at reduced rate.At the end of the storage period (18 monthsat 6 "C), Loaf A shows an electrophoreticpattern similar to that of a 12-month-oldreference sample of Grana Padano of Series1, whereas Loaf B does not reach this stan-dard level of casein breakdown.

3.2. Electrophoretic proteolysis index

As a general principle, a ratio of a degra-dation product to its primary compound wasconsidered to be suitable for providing asignificant proteolysis index [22]. Sincehydrolysis of [3-casein and concomitantaccumulation ofy-caseins seems to be typi-cal for Parmesan cheese, coefficients (y-casein/ê-casein) were plotted against theage of chee se (figure 5). The relationshipbetween these coefficients and the age ofreference samples of Series 1 was linear upto 15 months, allowing an evaluation of theextent of proteolysis and therefore a deduc-tion of the age of Parmesan samples analy-

Proteolysis in Parmesan cheese

4 5 6 7 9 11 13 14 15 16 18 22

Origin AJ.

} 'Y-Cn

- B-Cn

Ijllta'''~M-Cts1-Cn

- lXs1-I-Cn} lXs1-II-Cn

2 3 4 5 7 9 11 12 13 14 16 20 .. BOngln

- B-en

- Cts1-Cn

- lXs1-I-Cn} lXs1-II-Cn

i (+)

Figure 4. Urea-PAGE analysis of Parmesansamples. A, Loaf A; B, Loaf B. Numbers refer tothe age of cheese (months).Figure 4. Électrophorégramme urée-PAGEd'échantillons de Parmesan. A, meule A; B,meule B. Les numéros correspondent à l'âge dufromage (en mois).

sed. Based on this calibration a standardlevel (y-Cn/~-Cn) of 1.512 (cr= 0.106) wasfound for 12-month-old Grana Padanocheese. Taking into account a 2 cr confi-dence level we propose to use a thresholdlevel of 1.3 (y-Cn/~-Cn) to verify the requi-red age of 12 months for all Parmesan-typecheeses retailed as a loaf or prepacked slices,since ~-casein breakdown is even moreextensive in Parmigiano-Reggiano in com-parison to Grana Padano [1]. Although coef-ficient y-Cn/~-Cn shows a high degree ofvariability, most samples of Series II sho-

431

wed an even higher extent of proteolysis incomparison to samples of Series 1 and allsamples except one had a coefficient ~ 2.0(one 10-month-old sample had a coefficientof 1.6). This means that all samples of SeriesII more than fulfilled the requirements ofthe proposed threshold level of 1.3 (y-Cn/~-Cn). Samples of Series III do not show anyappreciable increase referring to this pro-teolysis index, due to the reasons describedabove.

At first, in particular Loaf A shows evenhigher coefficients (y-Cn/~-Cn) in compa-rison to that observed in the referencesamples of Series 1. However, during fur-ther storage at lowered temperature, coef-ficients increase very slowly. Because of itsvery high initial coefficient, Loaf A exceedsthe proposed threshold level of 1.3 at an ageof 14 months, although the standard levelof 1.5 is not reached before an age of18 months. Loaf B does not exceed theselevels during the storage period.

As the minimum age for grated GranaPadano was established to be 9 months [12],the corresponding standard level for gratedParmesan cheese is 1.0 (y-Cn/~-Cn). Sincecommercial grated Parmesan samples maycontain ::;;20% cheese rind [12, 26], whichhas a very low extent of proteolysis [22,24],a threshold level of 0.8 (y-Cn/~-Cn) is pro-posed for all grated Parmesan products. Thisvalue corresponds to the proteolysis index ofan approximately 6-month-old loaf of GranaPadano (Series 1).

Equivalent results were obtained with the~-casein content of the reference samples(figure 6). ~-casein decreases rapidly insamples of Series 1 and II. The relationshipbetween ~-casein content and the age ofreference samples of Series 1 was linear inthe range analysed. In most cases, samplesof Series II showed an even more extensivebreakdown of ~-casein. In striking contrastto that, even very old samples of Series IIIhad a higher ~-casein content th an everyother reference sample analysed in thisstudy. Thus the question arises whether these

432 H.K. Mayer et al.

Coefficient (gamma-caseins/B-casein)8, ;

3

......... Series 17

0 Series Il

6 -1t- 'Series III'

-'i'- Losf A5 -- -A- Loaf B

4---.--._1

Series 1:y • 0.118x + 0.096r • 0.995s ·0.106

.! ·.·..·._.·M'~M.~,_. ~"' .,

o ~~q::a:~~:=;=~~=1o 2 4 6 8 ro ~ M ~ ffi W ~

Age of cheese (months)

B-casein (g/100g cheese solids non-fat)40 1-;-....,......:=;--+....::.,;.;..:~~~=i=:::::c~

35

30

25

20

15 --- Seriea 10 Serieall10

--'JIlt- 'Serlea III'

5 -'V- Loaf A-e-. Loa! B

Series l,Y • -1.006x + 27.65

. r • 0.990a • 0.804

o 0O-+--+--+-r----;---;--,--,---;.--r----.--.J

o 2 4 6 8 10 12 14 16 18 20 22Age of cheese (months)

chee se sampI es were in fact Grana Padanocheese produced following the traditionaltechnology of manufacture. There is no rea-sonable explanation for the phenomenonthat very little endoproteinase activity and,in particular, nearly no p-casein breakdownoccurs in Grana Padano chee se samples of

Figure 5. Comparison of threeseries of reference samples ofGrana Padano as weil as ofLoaf A and B referring to thecoefficient (y·Cn/p-Cn). Linearregression up to 15 months ofripening is shown for Series l.Figure 5. Comparaison destrois séries d'échantillons deréférence de Grana Padanoainsi que des meules A et Bpour le coefficient (y-Cn/p-Cn).La régression linéaire jusqu'à15 mois est montrée pour lasérie l.

Figure 6. Comparison of threeseries of reference samples ofGrana Padano as weil as ofLoaf A and B referring to theB-casein content. Linear regres-sion is shown for Series 1.Figure 6. Comparaison destrois séries d'échantillons deréférence de Grana Padano etdes meules A et B pour lateneur en caséine p. La régres-sion linéaire est montrée pourla série I.

Series III, unless these chee ses were notripened using traditional ripening condi-tions, but were stored at lowered temperatureor ripened following a new technology.

At first, in particular, Loaf A has a verylow p-casein content in comparison to thatobserved in reference samples of Series 1.

Proteolysis in Parmesan chee se

However, during further storage at 6 "C,~-casein content decreases at a reduced rate.ln consequence, ~-casein content of loaves(in particular, LoafB) is relatively high at anadvanced age in comparison to that ofreference sampI es of Series 1. Loaf Aexceeds the standard level of approxima-tely 15 g ~-casein per 100 g cheese non-fatsolids, which corresponds to a 12-month-old reference sample of Series 1, only at anage of 18 months, whereas Loaf B does notreach this level during the storage period.

3.3. HPLC analysis of free amino acids

Figure 7 shows the total amount of freeamino acids (FAA) plotted against the age ofParmesan cheese. Although FAA contentof reference samples of Series 1and II showsa very high degree of variability, a distinctincrease of FAA content can be observedduring the ripening period. Referring to theanalysis ofFAA, reference samples of SeriesIII show a very high extent of proteolysis,which is in striking contrast to the resultsobtained by urea-PAGE analysis of caseins.

Free amino acids (9/100g cheese proteln)35 ; ..,;.

20

15

10 0 Series Il

t<. 'Series III'

5 -9-- Loaf A

-A- Loaf B

o -l-r---r--;--r-~~~~~-.Jo 2 4 6 8 W ~ M ~ re ~ ~

Age of cheese (months)

433

If we consider the fact that reference samplesof Series 1 and III have completely diffe-rent electrophoretic patterns (figures J, 2and 3), it is quite surprising that in almost ailcases, samples of Series III show a higherFAA content. In consequence of these fin-dings, reference samples of Series 1and IIIcannot be distinguished using HPLC ana-Iysis of total FAA, although there is a highlysignificant difference in the coefficient(y-Cn/~-Cn) (figure 5) and the ~-caseincontent (figure 6).

FAA content of Loaf A and B shows asteady increase during the storage at 6 oc.Taking into account the fact that Loaf A andB had a very limited ripening period of only4 and 2 months, respectively, it is quite sur-prising that at an advanced age, their FAAcontents stay within the variability of thoseof reference samples of Series 1 and II.

3.4. Soluble nitrogen fractions

Figure 8 shows the nitrogen fractionssoluble in water (WSN), 12 % TCA (TCA-N) and 5 % PTA (PTA-N) calculated as

Figure 7. Comparison of threeseries of reference samples ofGrana Padano as weil as ofLoaf A and B referring to theFAA content.Figure 7. Comparaison destrois séries d'échantillons deréférence de Grana Padano,ainsi que des meules A et Bpour la teneur en acides ami-nés libres.

434 H.K. Mayer et al.

Soluble nltrogen fractions (as % of TN)33 l, 1 ! 1; .

30.. -e- WSN/.lN (%)

-e- lCA-soluble NITN (%)

27·· -9- PlA-soluble N/lN (%)

121.··.+---.:-- --1'·'-·",··:Ji'-'!"-'-'-'O',-==~~···_·

91-" : : ,._•....;.. ·····+····i······,···· .. ; ,

6

3h---+---i--';==t=::::;=:!.i-...;--i-----~o 2 4 6 8 ~ ~ W m ~ w ~

Age of cheese (months)

Soluble nltrogen fractions (as % of TN)33 ; , l" 1 . •

30 -e- WSNITN (%)

-e- rCA-soluble N/lN ('lO)

27 -- -9- PlA-soluble N/lN ('lO)

24

21

15

12

9

6

3-l--i--+----T-~~=+=-i"---i--r-+-..Jo 2 4 6 8 ro ~ w ~ re ~ ~

Age of cheese (montns)

Figure 8. Changes of soluble nitrogen fractions as relative content of the total nitrogen (TN) duringthe storage period of Loaf A and B at 6 oc.Figure 8. Évolution des fractions solubles de la matière azotée exprimée par rapport à la teneurtotale en matière azotée (TN) au cours du stockage à 6 "C des meules A et B.

relative content of the total nitrogen (% ofTN). Soluble nitrogen fractions of Loaf Aand B show a steady increase during the sto-rage period at 6 oc. Because of its high ini-tiallevel of WSN, Loaf A reaches a veryhigh final WSN content of 30 %. Loaf Bstarts at a relatively low WSN level, butreaches 26 % WSN at the end of the sto-rage period.

3.5. Commercial Parmesan sampi es

Commercial Parmesan samples labelledas Grana Padano, Parmigiano-Reggiano orOriginalltalian Parmesan cheese were takenfrom retail outlets in Austria and Italy.Figure 9 shows the electrophoretic caseinpatterns of selected commercial Parmesansamples retailed as a loaf or as grated pro-ducts. Most of the samples, retailed as a loafor as prepacked slices, show electrophoreticpatterns that are in good accordance with

those obtained by the reference samples ofSeries 1and Il, having weak ~-casein bandsand very intense bands in the y-caseinregion. However, man y of the grated pro-ducts show quite surprising electrophoreto-grams, which are similar to those of veryyoung Grana Padano cheese or wholereference casein. Using urea-PAGE analy-sis, intense bands of as 1- and ~-casein, butno degradation products could be detected inmany of these commercial sarnples of gratedParmesan.

Figure 10 illustrates the relative distri-butions of the samples referring to their coef-ficients (y-Cn/~-Cn) as well as their ~-caseincontents. Less than 10 % of the samples(retailed as a loaf or as prepacked slices)have coefficients (y-Cn/~-Cn) lower than1.3, which was proposed as the thresholdlevel for verification of the required age of12 months. Concomitantly, only approxi-mately 10 % of the samples have a ~-casein

Proteolysis in Parmesan cheese

content higher than 15 g per 100 g chee senon-fat solids. This standard level corres-ponds to the degree of p-casein degradationof a 12-month-old reference sample ofGrana Padano (Series 1).

In striking contrast to these findings, morethan 60 % of the commercial grated Parme-san samples have coefficients (y-Cn/p-Cn)lower than 0.8, which was proposed as thethreshold level for verification of the requi-red age of 9 months. Concomitantly,approximately 60 % of the samples havea p-casein content higher than 20 g per 100 gcheese non-fat solids, which corresponds tothe extent of p-casein breakdown of a 8-month-old reference sample of GranaPadano (Series 1).

4. DISCUSSION

Electrophoretic patterns obtained by urea-PAGE analysis of the reference samples ofGrana Padano of Series 1and II were in good

1 2 3 4 5 6 1 7Origin-...._ ........ ,.

=-

435

accordance with those reported in other stu-dies [1,6, 15,22,23]. cxs1-casein is the prin-cipal target for proteolysis during the earlystages of ripening, which was also obser-ved in other cheese varieties [21]. As a resultof the action of plasmin, which seems to beone of the most active endopeptidases inGrana Padano, p-casein is hydrolysedrapidly during the ripening period and they-caseins are accumulated as its major degra-dation products. As the sum of ail y-caseinswas used for the calculation of the coeffi-cient (y-Cn/p-Cn), we did not differentiatebetween the distinct y-caseins (Y1-' Y2- andY3-caseins deriving from the different gene-tic variants of p-casein), which can only beseparated by isoelectric focusing. As wecould find a residual p-casein band, even in22-month-old Grana Padano, we doubtedthis fraction to be really p-casein, because,using two-dimensional electrophoresis, thisband proved to contain several degradationproducts of p-casein, which have the sameelectrophoretic mobility in one-dimensio-nal urea-PAGE [1].

8 9 10 11 12 1~ 14

~

... .. }'Y-Cn

. - B-Cn

- Œs1-Cn

- Œs1-I-Cn

} Œs1-II-Cn

(+)Figure 9. Urea-PAGE analysis of selected commercial Parmesan samples. Lanes 1-6, retailed as aloaf; 7-14, as grated product.Figure 9. Électrophorégramme urée-PAGE de quelques échantillons de Parmesan du commerce.Lignes 1-6, vendu à la coupe; 7-14 vendu en râpé.

__ -~~""!f"...,.,.

436 H.K. Mayer et al.

Frequency (%)

40Parmesan samples:

35 / E:J loat _ grated

30

U1U2U3U4U5U6U7M6U9~Coefficient (gamma-caseins/B-casein)

Frequency (%)

24Parmesan samples:ŒJ loat _ grated

21__ • ~ ••• ~_ •• __ • __ 0"." _. '_._. __ . v_. .

:~~:~.. _---~

18

15

12

9

6

3

o2488w~wrere~~~UU~~M~~

B-casein (g/100g chee se soUds non-fat)

Figure 10. Relative distributions of commercial Parmesan samples retailed as a loaf or prepacked slices(n = 49) or as grated product (n = 68) referring to the coefficient (y-Cn/~-Cn) and ~-casein content.Figure 10. Distribution relative des échantillons de parmesan du commerce vendu à la coupe, en pré-emballé (n = 49), ou en râpé (n = 48), selon le coefficient (y-Cn/~-Cn) et la teneur en caséine ~.

Threshold levels ('Y-Cn/~-Cn) were pro-posed for the verification of the requiredage of Parmesan cheese retailed as a loaf oras grated product. The use of these thresholdlevels as a standard to assess the endopro-teolytic changes of commercial samplescould contribute to improve the quality ofParmesan cheese. In comparison to GranaPadano samples of Series 1and II, very sirni-lar electrophoretograms were also reportedfor a Parmesan-type chee se manufacturedin Austria [22].

Commercial Parmesan samples retailedas a loaf or as prepacked slices were gene-rally found to fulfill the requirementsconcerning endoproteolytic changes duringripening. However, many grated Parmesansamples taken from retail outlets in Austriashowed poor quality, which was probablydue to the adulteration with products withlow proteolysis (e.g., cheese rind, veryyoung cheese).

Reference samples of Grana Padano ofSeries III showed quite surprising electro-phoretic patterns of caseins, which weresirnilar to those of very young Grana Padanocheese. Neither the breakdown of ~-casein,nor a distinct degradation of (Xsl-casein,could be observed during the ripening periodof 18 months. Based on the electrophoreticstudy of caseins, in ail cases the degree ofproteolysis was not related to the labelledage of these samples, which had been certi-fied by the Consorzio Grana Padano. How-ever, using HPLC analysis ofFAA, samplesof Series III showed a very high extent ofproteolysis, which was in striking contrast tothe results obtained by urea-PAGE analy-sis of caseins.

As an experiment, two loaves of Parme-san cheese were removed from the ripeningroom at an age of 4 and 2 months and weresubsequently stored at 6 "C for 18 months.During the storage period, endoproteolyticchanges were slowed down and breakdown

Proteolysis in Parmesan chee se

of (XSl- and ~-casein as weil as accumula-tion of degradation products proceeded atreduced rate, which cou Id be detected usingurea-PAGE casein patterns, but not byHPLC analysis ofFAA content. In compa-rison to the results reported in other studies[5, 27], these loaves could not be distingui-shed from Grana Padano chee se producedaccording to the traditional technology usingHPLC analysis of their FAA content. At anage of 12 months, Loaf A and B have a FAAcontent of approximately 20 and 16 g per100 g of cheese protein, respectively, whichis also within the variability of the FAAcontent of l2-month-old reference samplesof Series 1 and II.

At an age of 12 and 22 months, Loaf Areaches a WSN content (as % of TN) ofapproximately 25 % and 30 %, respectively,which is also within the variability repor-ted for reference samples of Grana Padanoof the same age [5, 6].

ln conclusion, since Kjeldahl determina-tion of WSN as well as HPLC analysis ofFAA content give insufficient informationon the distinct endoproteolytic changes,which were found to be typical for Parmesancheese, urea-P AGE of the casein fractionhas to be done to enable the evaluation ofthe extent of proteolysis and therefore adeduction of the age of commercial Parme-san samples.

ACKNOWLEDGEMENTS

Thanks are due to Prof. H. Klostermeyer andDr I. Krause (FML Weihenstephan, Germany)for their advice and for providing Iiterature, toProf. P. Resmini (Università degli Studi diMilano, Italy) for supplying reference samplesof Grana Padano and a1so to Prof. F. Addeo (Uni-versità degli Studi di Napoli Frederico II, Italy)for his advice conceming electrophoresis. Wealso thank the Austrian cheese factory for sup-plying two loaves of Parmesan cheese.

437

REFERENCES

[1] Addeo F., Moio L., Stingo C., Typical charac-teristics of proteolysis in Pannigiano-Reggianocheese, in: Consorzio Parmigiano-Reggiano(Ed.), Triennial research on the composition andthe peculiar characteristics of Parmigiano-Reg-giano cheese, Reggio Emilia, Italy, 1988, pp.21-40.

[2] Addeo F., Chianese L., Salzano A., Sacchi R.,Cappuccio U., Ferranti P., Malorni A., Charac-terization of the 12% trichloracetic acid-inso-lubie oligopeptides of Parmigiano-Reggianocheese, J. Dairy Res. 59 (1992) 401-41 1.

[3] Addeo F., Chianese L., Sacchi R., Musso S.S.,Ferranti P., Malomi A., Characterization of theoligopeptides of Parmigiano-Reggiano cheesesoluble in 120 g trichloracetic acidIL, J. DairyRes. 61 (1994) 365-374.

[4] Addeo F., Garro G., Intorcia N., Pellegrino L.,Resmini P., Chianese L., Gel electrophoresisand immunoblotting for the detection of caseinproteolysis in cheese, J. Dairy Res. 62 (1995)297-309.

[5] Battistotti B., Corradini C; Italian cheese, in:Fox P.F. (Ed.), Cheese: chemistry, physics andmicrobiology, vol. 2, Chapman & Hall, Lon-don, 1993,pp.221-243.

[6] Bertozzi L., Panari G., Studies on proteolysisin Parmigiano-Reggiano cheese, Int. Dairy J. 3(1993) 4-6.

[7] Bütikofer U., Rüegg M., Ardô Y., Determinationof nitrogen fractions in cheese: evaluation of acollaborative study, Lebensm. Wiss. Technol.26 (1993) 271-275.

[8] Christensen T.M.I.E., Bech A.M., Werner H.,Methods for crude fractionation (extraction andprecipitation) of nitrogen components in cheese,FIUIDF Bulletin 261 (1991) 4-9.

[9] Creamer L.K., Electrophoresis of cheese,FIUIDF Bulletin 261 (1991) 14-28.

[10] Eigel W.N., Butler J.E., Ernstrom C.A., FarrelH.M., Harwalkar V.R., Jenness R., WhitneyR.McL., Nomenclature of proteins of cowsmilk: fifth revision, J. Dairy Sei. 67 (1984)1599-1631.

[11] FIL/lOF, Standard 25. Protein content of pro-cessed chee se products, FIL/IDF Secretariat,Brussels, Belgium, 1964.

[12] FoiBy H., Wann ist geriebener Parmesan 'ori-ginal italienischer Parmesan'? Ôsterr. Milch-wirtsch II (1994) 15-19.

[13] FoiBy H., Winterer H., Genetische Variantender Milchproteine bei ôsterreichischern Feck-vieh, Ôsterr, Milchwirtsch 30 (1975) 1-14.

[14] Fox P.F., Proteinases, in: Fox P.F. (Ed.), Advan-ced dairy chemistry, vol. 1, Elsevier, London,1992, pp. 310-321.

438 H.K. Mayer et al.

[15] Fox P.F., Cheese: an overview, in: Fox P.F.(Ed.),Cheese: ehemistry, physies and miero-biology, vol. l , Chapman & Hall, London, 1993,pp. 1-36.

[16] Grappin R., Rank T.C., OIson N.F., Seeondaryproteolysis of eheese during ripening: a review,J. Dairy Sei. 68 (1985) 531-540.

[17] Krause 1., Resmini P., Klostermeyer H., Beur-teilung geriebener Hartkâseproben des Handelsmit der Bezeiehnung 'Parmesan', Wissenseh.Jahresber FML Weihenstephan 32 (1989)113-114.

[18] Krause 1., Neekermann H., Henle T., Beurtei-lung von Parmesan-Kâse, Wissenseh. Jahres-ber FML Weihenstephan 33 (1990) 109-111.

[19] Krause 1., Neekermann H., Henle T., Kloster-meyer H., Zur Qualitâtsbeurteilung von Parme-san mittels Analyse der freien Arninosâuren undbiogenen Aminen, Lebensmittelchemie 47(1993)73.

[20] Maurer H.R., Disk-Elektrophorese, Walter deGruyter, Berlin, 1971.

[21] Mayer H., Methoden zur Kontrolle desEiweiBabbaues wâhrend der Kâsereifung, Milch-wirtseh Ber. 92 (1987) 156-163.

[22] Mayer H., The diffieulties inherent in esti-mating the age of hard eheese by PAGE, in:ICODRL Experts Meeting at NIZO (Ed.), Newinsights in eheese ripening, Ede, The Nether-lands, 1990, p. 16.

[23] Mayer H.K., Eleetrophoretie ripening index forthe evaluation of proteolysis and the deduetionof the age of Parmesan chee se, Z. Lebensm.Unters Forsch. 202 (1996) 465-470.

[24] Mayer H., FoiBy H., Elektrophoretisehe Rei-fungskontrolle bei Parmesan- Kâse, Lebensmit-telchemie 48 (1994) 132-133.

[25] Resmini P., Pellegrino L., Il eontrollo di qua-lita' deI formaggio grattugiato, Sei. Tee. Lau.-Cas. 37 (1986) 421-427.

[26] Resrnini P., Pellegrino L., Pazzaglia c., Hogen-boom J.A., Free ami no aeids for the qualitycontrol of Parmigiano-Reggiano eheese and par-tieularly for the grated produets, Sei. Tee. LaU.-Cas. 36 (1985) 557-592.

[27] Resrnini P., Hogenboom J.A., Pazzaglia c., Pel-legrino L., Free ami no aeids for the analytiealeharaeterization of Grana Padano eheese, Sei.Tee. Lan.-Cas. 44 (1993) 7-19.

[28] Rohm H., Benedikt J., Jaros D., Comparison oftwo precipitation methods for determination ofphosphotungstie aeid soluble nitrogen in cheese,Lebensm. Wiss. Teehnol. 27 (1994) 392-393.

[29] Seianealepore V., Longone V., Rapid methodsfor measuring the degree of proteolysis as eheeseripening index, Milchwissensehaft 43 (1988)357-359.

[30] Waters, Instruction Manual. Waters AeeQ Tag®ehemistry package, Waters Corp, Milford, MA,USA,1993.