Alkaline Phosphatase, Acid Phosphatase, L actoperoxidase, and
L ipoprotein L ipase Activities in Industrial E we’s Milk and Cheese
F. Cha ´varri, A. Santisteban, M. Virto,* and M. de Renobales*
Departamento de Bioquı ´ mica y Biologı ´ a Molecular, Facultad de Farmacia, Universidad del Paı ´ s
Vasco/Euskal Herriko Unibertsitatea, Apartado 450, E-01080 Vitoria-Gasteiz, Spain
Alkaline phosphatase activity in raw, industrial ewe’s milk increased steadily >2-fold between
J anuary [1.7 units (U)/mL] and J une (3.75 U/mL), whereas acid phosphatase increased 4-fold in
J anuary and February (17 mU/mL) and then remained constant until the end of lactation. By
contrast, lipoprotein lipase exhibited a downward trend and lactoperoxidase decreased 2-fold during
lactation. When assayed at cheese-ripening temperatures, acid phosphatase retained 16% of its
activity at 37 °C, whereas lactoperoxidase retained between 30 and 45% of its activity at 20 °C.
The rate of hydrolysis of model triacylglycerols by lipoprotein lipase was highest for tricaprylin.
Although alkaline phosphatase in raw milk cheeses was variable from 1 to180 days of ripening, no
apparent reactivation was observed. The activity of acid phosphatase increased 2-fold during the
180 days of ripening in the cheeses made in summer, whereas in winter and spring much smaller
increases were observed. Both raw milk cheeses made in summer and all pasteurized milk cheeses
had very low levels of lactoperoxidase throughout ripening.
K eywords: Alkalinephosphatase; acid phosphatase; lactoperoxidase; lipoprotein lipase; ewe’s milk;
ewe’s milk cheese; ovine cheese; lactation period; cheese ripening.
The quality of sheep’s milk cheeses tends to be more
variablethan that of cow’s milk cheeses duetoa number
of factors such as the characteristics of the milk from
different breeds of sheep, stage of lactation when cheese
is made, or use of various renneting agents (commercial
calf rennet, artisanal and commercial lamb rennet
pastes), among others. Sheep’s milk has been charac-
terized with respect to fat and protein content (Anifan-
takis, 1986; Gonza ´lez-Llano and Ramos, 1989; Muir et
al., 1993) with occasional references tothelevels of some
enzymic activities. Although some 60 enzyme activities
havebeen identified in milk, their significancetocheese
quality has not been well studied (Fox et al., 1996).
Alkaline phosphatase, acid phosphatase, lactoperoxi-
dase, and lipoprotein lipase are four enzyme activities
of technological significance in the dairy industry. The
alkaline phosphatase test has been widely used for a
long time(Aschaffenburg and Mullen, 1949) as a control
for the efficiency of pasteurization. A positive result
with this test is the accepted parameter tocharacterize
a raw milk product, whereas it is taken as a sign of
improper pasteurization (or contamination with raw
milk) in a pasteurized milk product. However, positive
alkalinephosphataseresults havebeen reported in blue-
veined cheeses (Rosenthal et al., 1996) and in fresh,
unripened, Hispanic-style cheeses (Pratt-Lowe et al.,
1987) made from pasteurized milk. In both cases this
enzyme activity was of microbial origin. Its behavior
in raw milk cheeses has not been well characterized,
including how long it could be detected. For regulatory
purposes, thedetermination of its level of activity during
cheese maturation could provide useful information.
Thus, it was of interest to follow this activity in both
raw and pasteurized milk cheeses of long ripening
Acid phosphatase has a greater thermal stability and
it is most active at pH values typical of cheese ripening.
Its activity could be of major importance in the cheese-
ripening process as it is very active against phospho-
protein substrates such as the casein of milk (Andrews,
1991).Phosphopeptides have been reported to be
resistant to proteolytic attack, although not bitter
(Dulley and Kitchen, 1972). Thus, a high acid phos-
phatase activity during ripening could result in exces-
sive proteolysis and flavor defects.
The lactoperoxidase system (lactoperoxidase/thiocy-
anate/hydrogen peroxide) is a natural antimicrobial
system present in milks from many species (Perraudin,
1991). The enzyme lactoperoxidase catalyzes the oxida-
tion of thiocyanate by hydrogen peroxide, the antimi-
crobial effect being due to intermediate reaction prod-
ucts (Pruitt and Reiter, 1985). Thiocyanate is widely
distributed in animal tissues, with levels in bovine milk
ranging between 1 and 10 ppm (Wood, 1975). The third
component of the system could be generated in milk by
leucocytes and by lactic acid bacteria. The literature
refers primarily to the bovine, porcine, caprine, guinea
pig, and human enzymes (Wolfson and Sumner, 1993;
Zapicoet al., 1990), but littleinformation has been found
about the level of this activity in sheep’s milk as well
as in any kind of cheese.
The natural lipolytic enzyme in milk is lipoprotein
lipase. Its level of activity is usually low because of the
difficult access of the enzyme to its substrates in milk.
Unlike in milk, in which lipolytic activity causes rancid-
ity (Deeth, 1993), the presence of this lipoprotein lipase
or other lipases in dairy products (such as cheeses)
would be interesting as their reaction products, free
* Address correspondence to either author (telephone +34-
945-183097; fax +34-945-130756; e-mail firstname.lastname@example.org).
2926J. Agric. Food Chem . 1998, 46, 2926−2932
S0021-8561(97)00968-0 CCC: $15.00©1998 Am erican Chem ical Society
Published on Web 07/18/1998
fatty acids and their derivatives, could impart specific
tastes and flavor to the processed product (Kim and
Lindsay, 1990; Chilliard, 1982).
Idiazabal cheese is a raw ewe’s milk cheese, typical
of the Basque Country region of Spain. Most of the
production is commercialized within Spain between 90
and 180 days of ripening, with very few cheeses being
sold after that time. Although the Denomination of
Origin does not allow the use of pasteurized milk, trial
batches were made to study its characteristics for
potential export. The work presented herein is part of
a larger study tocharacterize the main biochemical and
microbiological aspects of the ripening of Idiazabal
cheese, relating them to sensorial parameters in the
finished product. Knowledge of how these enzymic
activities vary along the cheese-making period, both in
milk and during ripening of cheeses, may allow the
cheesemaker to introduce appropriate modifications in
the fabrication process.
MATERIALS AND METHODS
Milk and Cheese Samples. Bulk ewe’s milk was from
several commercial flocks of latxa sheep used in the Basque
Country region of Spain tomanufactureIdiazabal cheese. Milk
from an individual flock was a gift of Dr. L. Oregi (Basque
Government Department of Agriculture, Experimental Station,
Arkaute, Alava, Spain). The lactation period for this flock
extended from early March until J une. Milk samples were
taken after the milk from all ewes had been pooled and were
kept on ice. Enzyme activities were assayed on the day milk
The lactation period for an individual flock extends ≈4
months between November and J uly, depending on the
altitude at which the flock is located, with flocks on lower
elevations starting lactation earlier. Industrial cheese-making
(using mixed milk from different flocks) extends from J anuary
until the end of J une. Thus, industrial milk in J anuary
contains a large percentage of milk from flocks early in their
lactation stage, while industrial milk in J une corresponds to
the late lactation period of most flocks. In April milk could
be a mixture (of unknown composition) of milk from late, mid,
and early lactation period flocks. Milk had been refrigerated
for up to48 h in commercial farmhouses and pasteurized upon
receipt (when needed) in a local cheese factory (Queserı ´ as
Araia, S.A., Araia, Spain) prior tocheese-making. Pasteurized
milk samples were immediately stored on ice until enzymic
activities were assayed. Cheeses were made according to the
traditional method for the industrial production of Idiazabal
cheese approved by its Denomination of Origin (Basque
Government, 1986), using commercial calf rennet (Ch. Hansen,
Madrid, Spain), in 200 L vats. Cheeses (approximately 1 kg
and 15 cm diameter) were made in three consecutive weeks
in J anuary, April, and J une (total of nine fabrications), as
indicated in figure legends. Three different types of cheeses
were made for each fabrication: (C) control cheese made from
raw milk with no starter added; (R) cheese made from raw
milk with a starter cultureadded (90% Lactococcus lactis sbsp.
diacetylactis and 10% Lactobacillus lactis sbsp. lactis, isolated
from Idiazabal cheese, Pe ´rez-Elortondo, unpublished results);
(P) cheese made from pasteurized milk with the same starter
culture as for cheese R. For each fabrication and type of
cheese, samples from two different cheeses were taken for
analysis after 1, 90, and 180 days of ripening. Cheese extracts
were prepared immediately, and enzyme assays were carried
out on the same day. Other cheese samples were wrapped in
plastic film and aluminum foil and frozen at -80 °C to
determine enzyme stability.
E nzymic Assays. Enzyme activities were determined in
quadruplicate both in milk samples and in cheese extracts.
Milk was used either whole or defatted by centrifuging it at
5000g for 20 min. Cheese extracts were prepared as follows:
5.0 g of cheese (taken uniformly along the radius of the piece)
were homogenized in a Potter-type homogenizer on ice with
25.0 mL of 0.12 M Tris-HCl, pH 8.0, buffer and centrifuged at
5000g for 20 min at 4 °C. After the fat layer was removed,
the aqueous layer was filtered through Whatman No. 1 filter
paper and the filtrate was used for lactoperoxidase and
alkalineand acid phosphatasedeterminations. For lipoprotein
and esterase determinations, the filtrate was concentrated
with a PM10 membrane in an Amicon ultrafiltration unit on
Alkaline phosphatase was determined as described by
McComb et al. (1979) in 0.9 M 2-amino-2-methyl-1-propanol
(AMP, Sigma Chemical, Madrid, Spain), pH 10.45, with
p-nitrophenyl phosphate(pNPP, Sigma Chemical) as substrate
(assay concentration was 15.8 mM), at 35 °C. Total volume
was 2.0 mL, and reaction rates were determined at 400 nm.
The molar absorptivity of p-nitrophenol at 400 nm was taken
as 19 000 M-1cm-1. TheAMP buffer system was used instead
of the carbonate buffer used by other workers (Kitchen et al.,
1970) to ascertain that if any reactivation of the alkaline
phosphatase occurred in pasteurized cheeses during cheese
maturation, it would easily be detected. Values for alkaline
phosphatase activity obtained in the presence of AMP buffer
were ∼2-fold higher than those obtained in the traditional
carbonate buffer, due to the activating effect of AMP for
phosphatases (Tietz, 1979). No Mg2+was added to assay
mixtures because no increase in activity was observed in its
Acid phosphatase was determined essentially as described
by Larsen and Parada (1988). Milk (80 µL) or cheese extract
(0.3 mL) was added toan appropriate volume of 0.1 M sodium
acetate buffer, pH 4.4 (total assay volume was 1.0 mL),
containing pNPP (final concentration in theassay mixturewas
7.5 mM). Mixtures were incubated at 37 °C for 1 h, and
reactions werestopped with 1.0 mL of 12% (w/v) trichloroacetic
acid. After the precipitated proteins had been separated by
centrifugation at 16000g for 10 min, 0.7 mL of the clear
supernatant was mixed with 1.5 mL of 2 M NaOH. Absor-
bance was read at 405 nm.
Lactoperoxidase was determined as described by Pruitt et
al. (1990) in 0.1 M phosphate buffer, pH 6.0, and at 20 °C.
The assay concentrations of 2,2′-azinobis(3-ethylbenzthiazo-
line-6-sulfonic acid) (ABTS) and H2O2were 3.88 mM and 0.1
mM, respectively, 2.5 mL being the total assay volume.
Reaction rates were determined at 415 nm, and the molar
absorptivity coefficient of ABTS was taken as 32 400 M-1cm-1.
Lipoprotein lipase was adapted from the method of Egelrud
and Olivecrona (1972). The substrate emulsion was prepared
as follows: 0.86 g of tricaprylin was sonicated with a solution
of 1.3% (by weight) bovine serum albumin in 17.8 mL of 0.37
M Tris-HCl buffer, pH 8.6, containing 0.66 M NaCl and 20 µL
of heparin (1000 international units/mL). Assay mixtures (2.0
mL total volume) containing 0.4 mL of decomplemented
human serum and 0.8 mL of substrate emulsion in 0.37 M
Tris-HCl buffer, pH 8.6, were incubated at 37 °C for 1 h (milk)
or 4 h (cheese extracts). Reactions were stopped with 5.0 mL
of hexane/2-propanol/10 N sulfuric acid (1:4:0.1 by volume).
Free fatty acids were extracted with 3.0 mL of water and 3.0
mL of hexane and titrated with 0.005 M methanolic KOH,
using phenolphthaleine as indicator. When ewe’s milk fat
[extracted according to the modified method of Folch, as
described by Hamilton et al. (1992)] was used as substrate,
the free fatty acids produced were extracted and analyzed by
gas-liquid chromatography as described (Cha ´varri et al.,
All enzyme units are reported in international units (U), 1
U being the amount of enzyme that catalyzes the production
of 1 µmol of product/min under the described conditions.
Activity in milk is given in units per milliliter and in cheese
as units per kilogram of cheese. All determinations were done
in quadruplicate, and the results presented are the mean (
Thiocyanate concentration in milk samples was determined
as described by Gaya et al. (1991).
Statistical Analysis. The BMPD statistical package (Dix-
on, 1983) was used for statistical treatment of the results.
Enzym e Activities in Ewe’s Milk Cheese RipeningJ. Agric. Food Chem ., Vol. 46, No. 8, 1998 2927
Analysis of variance was employed to test for statistically
significant differences (p < 0.05) in the levels of the various
enzymeactivities over thelactation period. Bonferroni’s paired
t test was used to test for statistically significant differences
(p < 0.05) between values for any two of the three different
times in the lactation period, winter, spring, and summer.
RESULTS AND DISCUSSION
E nzyme Activities in Industrial Milk. Alkaline
phosphatase activity in industrial ewe’s milk was
observed to increase >2-fold between J anuary (early
lactation, average of 1.75 U/mL) and J une (late lacta-
tion, average of 3.75 U/mL) (Figure 1A), in agreement
with the 2-fold variation in activity reported for bulk
bovinemilk collected during different seasons (Andrews,
1991). Differences among the three lactation stages
were statistically significant (p < 0.05). When the milk
from a single flock was analyzed, a similar trend was
observed, although the actual values were lower (Table
1), in contrast with the results reported by Anifantakis
and Rosakis (1983) for three Greek flocks in which the
activity was maximal in mid-lactation, with lower
values in both early and late lactation. The level of this
enzyme in ewe’s milk (when determined in the more
frequently used carbonate buffer) was comparable to
that reported by Anifantakis and Rosakis (1983) for
ewe’s milk but >2-fold higher than that reported for
bovine milk (Kitchen et al., 1970) and ∼5-fold higher
than that reported for goat’s milk (Williams, 1986). The
optimum pH was 10.5 (Figure 2), and the optimum
temperature was 40 °C. As expected, pasteurization
completely inactivated this enzymic activity. The re-
sidual activity observed for one of the J une samples
could be due to improper pasteurization, because total
viable counts for this sample were reduced by 25% upon
pasteurization, whereas in the rest of the samples an
average 34% reduction was observed (de Vega, 1996).
Between 70 and 80% of the alkaline phosphatase
activity of whole milk was found in skim milk. When
cheese was made, only 40-45% of alkaline phosphatase
was lost in the whey.
Acid phosphataseactivity (Figure1B) increased 4-fold
in J anuary and then remained constant (17 mU/mL)
until the end of J une.This value is ≈1 order of
F igure 1. Enzyme activities in ewe’s crude (b) and pasturized (O) milk throughout lactation period: (A) alkaline phosphatase;
(B) acid phosphatase; (C) lactoperoxidase; (D) lipoprotein lipase. Values reported represent the mean of four determinations with
the error bars indicating the standard deviation. When no error bars appear, the magnitude of the standar deviation is smaller
than the symbol used.
T able 1.E nzyme Activities in R aw E we’s Milk from a Single F locka
date alkaline phosphatase, U/mL
1.819 ( 0.135a
2.257 ( 0.086b
2.653 ( 0.204c
acid phosphatase, mU/mL
29.29 ( 0.45a
25.99 ( 0.10b
21.61 ( 0.05c
9.854 ( 0.127a
10.565 ( 0.340b
8.653 ( 0.382c
lipoprotein lipase, U/mL
0.160 ( 0.014a
0.243 ( 0.021b
0.117 ( 0.003a
J une 9
aEnzyme activities were determined as described under Materials and Methods. Pooled milk samples from all ewes were taken on the
dates indicated after the morning milking. Values reported represent the mean ( standard deviation of four replicates. For each enzyme
activity different letters indicate statistically significant (p < 0.05) differences.
F igure 2. Variation of enzymeactivity in industrial, raw milk
with pH: alkaline phosphatase ([); acid phosphatase (9);
lactoperoxidase (0); lipoprotein lipase (b). Assay conditions
(except pH) were as described under Materials and Methods.
2928 J. Agric. Food Chem ., Vol. 46, No. 8, 1998 Cha ´ varri et al.
magnitude higher than that reported for bovine milk
(Kitchen et al., 1970), although no indication is given
in this publication as to variations between early and
late lactation. The activity of acid phosphatase in the
milk from a single flock was somewhat higher than that
observed in bulk milk (Table 1) and it decreased toward
the end of the lactation period, the differences between
periods being statistically significant (p < 0.05). This
activity had a broad pH optimum between 4.5 and 5.5
(Figure 2). When the assay temperature was lowered
to10 °C (approximate ripening temperature for Idiaza-
bal cheese is between 8 and 12 °C), the enzyme retained
16% of its activity at 37 °C. Both the pH profile and
the fact that it was active at low temperatures indicate
that it can be active under cheese-ripening conditions.
Pasteurization did not inactivate it to any significant
extent, as expected. The activity found in skim milk
was between 65 and 80% that of whole milk. Between
30 and 34% of the milk acid phosphatase was found in
the whey after renneting.
Lactoperoxidase activity in industrial milk (Figure
1C) was variable during the winter and spring months,
with an average value of 4 U/mL, decreasing by an
average factor of 2 by the end of J une (p < 0.05). This
value is ≈3-fold higher than the average value of 1.4
U/mL reported for bovine milk (Stephens et al., 1979).
By contrast, lactoperoxidase activity in the milk of a
singleflock gavea mean valueof 9.7 U/mL, 2-fold higher
than the mean value reported by Medina et al. (1989)
for Manchega sheep’s milk at mid-lactation, which could
indicate differences due to the breed of sheep. The
lactoperoxidase pH optimum was 5.5 (Figure 2). When
the enzyme was assayed at cheese-ripening tempera-
tures, it retained between 30 and 45% of the activity
found at 20 °C, suggesting that the enzyme could also
be active during cheese ripening. Over 90% of the
lactoperoxidase activity of whole milk was found in the
skim milk. After renneting, 65% of this enzymeactivity
appeared in the whey. Pasteurization reduced the level
of activity by an average 10-30% in winter and spring,
with one sample exhibiting a 70% reduction in activity.
Pasteurization did not inactivate any of the J une
samples, which initially had very low activity. Thiocy-
anate content in milk depends on the composition of the
pastures, with cows on natural pastures containing
clover giving milk with up to 15 ppm of thiocyanate
(Pruitt and Reiter, 1985). The concentration of thiocy-
anate (Table 2) was very high at the end of J anuary,
when most of the sheep would be receiving dry fodder.
Thelowest values wereobserved in spring and summer.
Considering that the levels of lactoperoxidase were not
limiting and that approximately 15 ppm of thiocyanate
and 8.5 ppm of hydrogen peroxide would be adequate
for the lactoperoxidase system to be activated (Bjo ¨rck
et al., 1979), addition of these compounds could help
improvethemicrobiological quality of refrigerated milk.
Lipoprotein lipase activity (Figure 1D) was routinely
assayed with tricaprylin as model substrate.This
activity, although variable, exhibited a slight downward
trend as lactation progressed, with the differences
between J anuary and J unebeing statistically significant
(p < 0.05). When model substrates were used, the
hydrolysis of tricaprylin proceeded at the highest rate,
with the rates for tricaprin, trilaurin, and olive oil being
comparable (Figure 3A). Results with tripalmitin were
highly variable (not shown) due todifficulties in obtain-
ing a stable emulsion at the assay temperature of 37
°C. When ewe’s milk fat was used as substrate, caprylic
and capric acids appeared as products at an apparently
lower rate than palmitic acid (typical results are shown
in Figure 3B). However, taking intoconsideration that
palmitic acid represented 24% of the total fatty acids
in this ewe’s milk fat and caprylic and capric acids
represented 3 and 9%, respectively (de Renobales et al.,
unpublished observations), these results did not con-
tradict those obtained with model substrates.
results were consistent with those reported for purified
bovine lipoprotein lipase (Deckelbaum et al., 1990),
which showed that this enzyme exhibited a higher rate
of hydrolysis toward triacylglycerols containing medium-
chain fatty acids than toward those containing long-
chain fatty acids. Results for oleic acid (which repre-
sented 24% of the total fatty acids in this ewe’s milk
fat) were highly variable. Lipoprotein lipase activity in
milk from a singleflock presented slight variations with
an increase toward mid-lactation (Table 1). Pasteuriza-
tion caused an average 73-95% inactivation of this
enzyme. The low percent inactivation (45%) observed
in the second J une sample was most likely due to
improper pasteurization of this batch, as mentioned
E nzyme Activities during Cheese R ipening. The
amount of alkalinephosphatasedetermined in raw milk
cheeses with nostarter culture added showed a general
trend to increase from J anuary to J une (Figure 4A),
most likely due tothe increase reported above for ewe’s
milk. Within each fabrication, and as ripening pro-
gressed, the levels of alkaline phosphatase were vari-
able, in some cases decreasing by different amounts
from 1 to 180 days, whereas in others, after a decrease
at 90 days, levels roseagain at 180 days. This apparent
“reactivation” observed in some cases was most likely
due to sample variation because of the following rea-
sons: (1) none of the cheeses made with pasteurized
milk showed any alkaline phosphatase activity except
those made with improperly pasteurized milk (Figure
1A, second J une fabrication), in which a low and
constant level was observed throughout ripening (data
not shown); (2) different cheeses (rather than samples
of the same cheese) were analyzed at different ripening
times. Therefore, it is reasonable to conclude that no
microbial alkaline phosphatase was produced in these
cheeses, as has been reported in somecases (Pratt-Lowe
et al., 1987; Rosenthal et al., 1996) or that reactivation
of themilk enzymeoccurred (Lyster and Aschaffenburg,
1962; Linden, 1979). In cheeses made with raw milk
and starter culture added (group R) the activity levels
and the variability within each fabrication were not
different from those reported for the control fabrication
in Figure 4A. When the activity was measured at
various distances from the outer rind to the center of
the cheese, a slight decrease in activity was observed
in thecenter, but it was not statistically significant. This
was most likely due to the small cheese size and to the
fact that the temperature is never >35 °C during the
T able 2.
Milk throughout L actation Perioda
T hiocyanate Concentration in Bulk R aw E we’s
J an 31 18.26 ( 2.64 April 11 9.18 ( 0.15 J une 13 1.64 ( 0.22
Feb 75.70 ( 0.22 April 18 2.12 ( 0.05 J une 20 2.08 ( 0.18
Feb 144.13 ( 0.16 April 25 2.12 ( 0.08 J une 27 1.55 ( 0.16
[SCN], ppmdate [SCN], ppmdate [SCN], ppm
aMilk samples were taken on dates specified. Determinations
were made as described under Materials and Methods. Values are
the mean ( standard deviations of four replicates.
Enzym e Activities in Ewe’s Milk Cheese Ripening J. Agric. Food Chem ., Vol. 46, No. 8, 1998 2929
manufacturing process, thus avoiding a partial thermal
inactivation of the enzyme. By contrast, in Grana-
Padano and Parmigiano-Reggiano cheeses (28 cm di-
ameter), which are cooked at 53 °C, a decreasing
gradient of alkaline phosphatase activity from the outer
rind tothecenter of thepiecehas been found (Pellegrino
et al., 1995; Resmini and Pellegrino, 1996). Alkaline
phosphatase levels were unaltered in cheese samples
kept at -80 °C for as long as 2 years.
The levels of acid phosphatase measured on the first
day of ripening in raw milk cheeses were comparable
regardless of thetimeof theyear (Figure4B). However,
cheeses made in J une exhibited a 2-fold increase in acid
phosphatase activity between days 1 and 180, whereas
those made in winter and spring showed more modest
increases. Cheeses made with raw milk and starter
culture added (group R) gave essentially the same
results as the control group reported in Figure 4B.
Cheeses madewith pasteurized milk exhibited a similar
pattern, except that the level of acid phosphatase
activity in cheeses made in J une was 17% lower than
that reported for raw milk cheeses in Figure 4B. An
increase in acid phosphatase activity during ripening
could result in a higher degree of hydrolysis of the
phosphopeptides (Andrews, 1991), which would be
rendered more susceptible toproteolytic attack (Dulley
and Kitchen, 1972). This is in agreement with the
sensorial analysis of the raw milk cheeses, which
correlated flavor defects in cheeses made in J une with
excessive proteolysis, whereas pasteurized cheeses with
a starter culture added made in J une had less proteoly-
sis and significantly (p < 0.05) better sensorial scores
(Mendı ´ a, 1998). The possibility that this increase in
acid phosphataseis duetocertain microbial populations
is currently being investigated. Acid phosphatase activ-
ity remained constant in cheese samples kept at -80
°C during 2 years.
The levels of lactoperoxidase activity (Figure 4C) in
day 1 cheeses closely reflected the levels found in the
milk with which cheeses were made (Figure 1C). The
activity of lactoperoxidase in cheeses made in winter
and spring with raw milk was ≈1 order of magnitude
higher than that in cheeses made in summer. Group R
cheeses made with raw milk and added starter culture
gave somewhat higher values for lactoperoxidase activ-
ity than those reported in Figure 4C, although the
differences were not significant. However, in group P
cheeses made with pasteurized milk the level of lacto-
peroxidase activity was an average 50% lower than in
raw milk cheeses throughout lactation, in contrast with
the relative thermal stability of this enzyme observed
in pasteurized milk (Figure 1C). This could be due toa
gradual loss of activity of the heat-treated enzyme
during cheese-making when the enzyme is at temper-
atures between 20 and 38 °C for over 1.5 h. The fact
that lactoperoxidaseactivity decreased by 50% in cheese
F igure 3. Lipoprotein lipase activity in industrial, raw milk with model substrates (A) and ewe’s milk fat (B): tricaprilyn (2);
tricaprin (b); trilaurin (]); olive oil (0).
2930 J. Agric. Food Chem ., Vol. 46, No. 8, 1998 Cha ´ varri et al.
samples kept at -80 °C after 5 months also indicates
that this enzyme is not as stable as acid phosphatase
despite its behavior during pasteurization.
Several attempts were made to determine both lipo-
protein lipase and esterase activities in cheese extracts
at various times during ripening. Results were incon-
clusive due primarily to the very low levels of each
activity found, which were difficult to reproduce in a
systematic manner. Lipoprotein lipase activity tended
tobe associated with the fat layer in variable amounts.
Results presented in this paper indicatethat thelevel
of alkaline and acid phosphatases, lactoperoxidase, and
lipoprotein lipase in raw ewe’s milk changed signifi-
cantly (p < 0.05) between early and late lactation.
Lactoperoxidase levels were sufficiently high in winter
and spring toallow the activation of the lactoperoxidase
system as a means of improving the bacteriological
quality of the milk. This system could potentially be
activated in raw milk cheeses madein winter and spring
toimprovetheir hygenic quality. Duetothemuch lower
levels of lactoperoxidase found in pasteurized milk
cheeses, or in raw milk cheeses made in summer, the
activation of the lactoperoxidase system in these cases
could bequestionable. Theincreasein acid phosphatase
levels in cheeses made in J une needs to be further
investigated toidentify its origin in view of theapparent
correlation with increased proteolysis and flavor defects.
The observed increase in alkaline phosphatase levels
in raw milk at theend of lactation did not causepositive
responses to the alkaline phosphatase test in pasteur-
ized cheese samples over the 180 day ripening period,
except for a particular fabrication that was not properly
pasteurized. Thus, pasteurization would inactivatethis
enzyme in ewe’s milk, as expected, despite the higher
levels found at the end of lactation.
We thank M. A. Marquı ´ nez for excellent technical
assisstance, cheesemaster A. Pe ´rez de Albe ´niz for mak-
ing the cheeses, and Queserı ´ as Araia (Araia, Alava,
Spain) for the use of their pilot cheese manufacturing
F igure 4. Enzyme activities during ripening of raw ewe’s milk cheeses manufactured at the specified times in the lactation
period: (A) alkaline phosphatase; (B) acid phosphatase; (C) lactoperoxidase. The bars correspond to, from left to right in each
grouping, 1 day of ripening, 90 days of ripening, and 180 days of ripening.
Enzym e Activities in Ewe’s Milk Cheese Ripening J. Agric. Food Chem ., Vol. 46, No. 8, 1998 2931
plant and ripening chambers. F.C. acknowledges a
fellowship from the Dpt. of Industry, Agriculture and
Fisheries of the Basque Government.
Andrews, A. T. Phosphatases. In Food Enzimology; Fox, P. F.,
Ed.; Elsevier Applied Science: Essex, U.K., 1991; Vol. I, pp
Anifantakis, E. M. Comparison of the physico-chemical proper-
ties of ewe’s and cow’s milk. Int. Dairy Fed. Bull. 1986, 202,
Anifantakis, E. M.; Rosakis, P. S. Alkaline phosphatase
activity of sheep’s milk and some factors affecting it. Egypt.
J . Dairy Sci. 1983, 11, 173-182.
Aschaffenburg, R.; Mullen, J . E. C. A rapid and simple
phosphatase test for milk. J . Dairy Res. 1949, 16, 58-67.
Basque Government (Department of Agriculture and Fisher-
ies). Idiazabal cheese. Denomination of Origin Control
Board. Bol. Of. Pais Vasco 1986, 252, 6357-6373.
Bjo ¨rck, L.; Claesson, O.; Schulthess, W. The lactoperoxidase/
thiocyanate/hydrogen peroxide system as a temporary pre-
servative for raw milk in developing countries. Milchwis-
senschaft 1979, 34, 726-729.
Cha ´varri, F.; Virto, M.; Martı ´ n, C.; Na ´jera, A. I.; Santisteban,
A.; Barro ´n, L. J . R.; de Renobales, M. Determination of free
fatty acids in cheese: comparison of twoanalytical methods.
J . Dairy Res. 1997, 64, 445-452.
Chilliard, Y. Variations physiologiques des activite ´s lipasiques
et de la lipolyse spontane ´e dans les laits de vache, de che `vre
et de femme: revue bibliographic. Lait 1982, 62, 126-154.
Deckelbaum, R. J .; Hamilton, J . A.; Moser, A.; Bengtsson-
Olivecrona, G.; Butbul, E.; Carpentier, Y. A.; Gutman, A.;
Olivecrona, T. Medium-chain versus long-chain triacylglyc-
erol emulsion hydrolysis by lipoprotein lipase and hepatic
lipase: implications for the mechanisms of lipase action.
Biochemistry 1990, 29, 1136-1142.
Deeth, H. C. Lipaseactivity and its effect on milk quality. Aust.
J . Dairy Technol. 1993, 48, 96-98.
deVega, M. C. Contribucio ´n a la caracterizacio ´n microbiolo ´gica
de la leche de oveja para la produccio ´n de Queso Idiazabal.
Master’s Thesis, Universidad del Paı ´ s Vasco/Euskal Herriko
Dixon, W. J . BMPD Statistical Software Manual; University
of California Press: Los Angeles, 1983.
Dulley, J . R.; Kitchen, B. J . Phosphopeptides and bitter
peptides produced by rennet. Aust. J . Dairy Technol. 1972,
Egelrud, R.; Olivecrona, T. The purification of a lipoprotein
lipase from bovine skim milk. J . Biol. Chem. 1972, 247,
Fox, P. F.; O’Connor, T. P.; McSweeney, P. L. H.; Guinee, T.
P.; O’Brien, N. M. Cheese: physical, biochemical and nutri-
tional aspects. Adv. Food Nutr. Res. 1996, 39, 163-328.
Gaya, P.; Medina, M.; Nu ´n ˜ez, M. Effect of the lactoperoxidase
system on Listeria monocytogenes behavior in raw milk at
refrigeration temperatures. Appl. Environ. Microbiol. 1991,
Gonza ´lez-Llano, D.; Ramos, M. Composicio ´n quı ´ mica de la
leche de oveja. Ovis 1989, 2, 9-19.
Hamilton, S.; Hamilton, R. J .; Sewell, P. A. Extraction of lipids
and derivative formation. In Lipid Analysis: A Practical
Approach; Hamilton, R. J ., Hamilton, S., Eds.; Oxford
University Press: Oxford, U.K., 1992; Chapter 2.
Kim, H. A. J .; Lindsay, R. C. Method for the quantitative
analysis of volatile free and total branched fatty acids in
cheese and milk fat. J . Dairy Sci. 1990, 73, 1988-1999.
Kitchen, B. J .; Taylor, G. C.; White, I. C. Milk enzymesstheir
distribution and activity. J . Dairy Res. 1970, 37, 279-288.
Larsen, R. F.; Parada, J . L. Acid phosphatases in somecheeses
and starters. Sci. Aliment. 1988, 8, 285-294.
Linden, G. Biochemical study of some aspects of milk alkaline
phosphatasereactivation. Milchwissenschaft 1979, 34, 329-
Lyster, R. L. J .; Aschaffenburg, R. The reactivation of milk
alkaline phoshatase after heat treatment. J . Dairy Res.
1962, 29, 21-35.
McComb, R. B.; Bowers, G. N. Appendix: Procedures. In
AlkalinePhosphatase; McComb, R. B., Bowers, G. N., Posen,
S., Eds.; Plenum Press: London, 1979; pp 903-907.
Medina, M.; Gaya, P.; Nu ´n ˜ez, M. The lactoperoxidase system
in ewe’s milk: levels of lactoperoxidase and thiocyanate.
Lett. Appl. Microbiol. 1989, 8, 147-149.
Mendı ´ a, C. Cambios proteolı ´ ticos y organole ´pticos del Queso
Idiazabal elaboradoen diferentes e ´pocas del an ˜o: Influencia
de la pasteurizacio ´n de la leche y del cultivoiniciador. Ph.D.
Dissertation, Universidad Pu ´blica de Navarra, 1998.
Muir, D. D.; Horne, D. S.; Law, D. J . R.; Steele, N. Ovine milk.
1. Seasonal changes in composition of milk from a commer-
cial Scottish flock. Milchwissenschaft 1993, 48, 363-366.
Pellegrino, L.; Resmini, P.; Fantuzzi, U. Valutazione ed
interpretazione della attivita ` della fosfatasi alcalina nel
controllo dei formaggi Grana Padano e Parmigiano Reg-
giano. Ind. Latte 1995, 31, 17-30.
Perraudin, J . P. Lactoperoxidase. A natural preservative.
Dairy Ind. Int. 1991, 56, 21-24.
Pratt-Lowe, E. L.; Richardson T.; Geiger R. M. Inactivation of
microbial alkaline phosphatase in Hispanic-style cheeses
with theScharer rapid phosphatasetest. J . Dairy Sci. 1987,
Pruitt, K. M.; Reiter, B. Biochemistry of peroxidase system:
antimicrobial effects. In TheLactoperoxidaseSystem: Chem-
istry and Biological Significance; Pruitt, K. M., Tenovuo, J .
O., Eds.; Dekker: New York, 1985; Chapter 8.
Pruitt, K. M.; Kaman, D. N.; Miller, K.; Mansson-Ramentulla,
B.; Ramentulla, F. Quantitative, standardized assays for
determining the concentrations of bovine lactoperoxidase,
human salivary peroxidase and human myeloperoxidase.
Anal. Biochem. 1990, 191, 278-286.
Resmini, P.; Pellegrino, L. Evaluation and interpretation of
alkaline phosphatase activity in hard cheeses made with
raw milk. Food Aust. 1996, 48, 452-454.
Rosenthal, I.; Bernstein, S.; Rosen, B. Alkaline phosphatase
activity in Penicillium roqueforti and blue-veined cheeses.
J . Dairy Sci. 1996, 79, 16-19.
Stephens, S.; Harkness, R. A.; Cockle, S. M. Lactoperoxidase
activity in guinea pig milk and saliva: correlation in milk
of lactoperoxidase activity with bactericidal activity against
Escherichia coli. Br. J . Exp. Pathol. 1979, 60, 252-258.
Tietz, N. W. Alkaline phosphatase: general comments. In
Clinical Enzymology; Griffiths, J . C., Ed.; Masson Publish-
ing: New York, 1979.
Williams, D. J . A modification of the Aschaffenburg and
Mullen alkaline phosphatase test suitable for goat’s milk.
Aust. J . Dairy Technol. 1986, March, 28-30.
Wolfson, L. M.; Sumner, S. S. Antibacterial activity of the
lactoperoxidase system: a review. J . Food Prot. 1993, 56,
Wood, J . L. Biochemistry. In Chemistry and Biochemistry of
Thiocyanic Acid and Its Derivatives; Newman, A. A., Ed.;
Academic Press: London, 1975; pp 156-221.
Zapico, P.; Gaya, P.; Nu ´n ˜ez, M.; Medina, M. Lactoperoxidase
and thiocyanate contents of goat’s milk during lactation.
Lett. Appl. Microbiol. 1990, 11, 90-92.
Received for review November 12, 1997. Revised manuscript
received May 11, 1998. Accepted May 11, 1998. This work
was supported by grants from theEuropean Union (AIR-CT93-
1298) and the Spanish Ministry of Education and Science
2932 J. Agric. Food Chem ., Vol. 46, No. 8, 1998Cha ´ varri et al.