View
3
Download
0
Category
Preview:
Citation preview
Results and Discussion
163
1. Comparison of different phenol- chloroform based methods for DNA
isolation from E.coli, S.aureus and S.typhi
In this part of the study, four different phenol chloroform based methods were
evaluated for DNA extraction from the three bacterial strains namely E. coli, S.
typhi and S. aureus.
Extraction of DNA was carried out by using 4 different phenol-chloroform based
methods. DNA was extracted and the concentration of DNA was checked by UV-
spectrophotometer.
Table 1, Table 2 and Table 3 shows the O.D. of extracted DNA from E. coli, S.
aureus and S. typhi respectively. Figure 1 shows the comparison O.D. of E. coli
DNA samples obtained by four different methods. It can be seen that method 4
gives the 260/280 ratio near to 1.8 for E. coli DNA, indicating that highest DNA
was recovered from E. coli bacterial cells using this method. Thus, out of the four
studied methods, method 4 is most efficient for E. coli DNA extraction.
Results and Discussion
164
Table 1. O.D. of extracted DNA by different methods from E. coli
Method O.D. at 260
nm
O.D. at 280
nm
260/280 ratio
Method 1 0.77 0.40 1.91
Method 2 0.68 0.404 1.68
Method 3 0.75 0.436 1.72
Method 4 0.83 0.453 1.83
Results are mean of three observations
Results and Discussion
165
0
0.5
1
1.5
2
2.5
Method 1 Method 2 Method 3 Method 4
Different methods for DNA extractions
O.D
.
O.D. at260 nm
O.D. at280 nm
260/280ratio
Figure 1. O.D. of E. coli DNA samples obtained by four different phenol-
chloroform based DNA isolation methods
Results and Discussion
166
Table 2 and Figure 2 show the comparison O.D. of S. aureus DNA samples
obtained by four different methods. It can be seen that method 2 gives the
260/280 ratio near to 1.8 for the S. aureus DNA, indicating that out of four studied
methods, method 2 is most efficient for S. aureus DNA extraction.
Table 3 and Figure 3 show the comparison O.D. of S. typhi DNA samples
obtained by four different methods. It can be seen that method 4 gives the
260/280 ratio near to 1.8 for the S. typhi DNA, indicating that out of four studied
methods, method 4 is the most efficient for S. typhi DNA extraction.
Results and Discussion
167
Table 2. O.D. of extracted DNA by different methods from S. aureus
Method O.D. at 260
nm
O.D. at 280
nm
260/280 ratio
Method 1 0.73 0.426 1.71
Method 2 0.78 0.421 1.85
Method 3 0.65 0.329 1.97
Method 4 0.63 0.370 1.70
Results and Discussion
168
0
0.5
1
1.5
2
2.5
Method 1 Method 2 Method 3 Method 4
different methods for DNA extraction
O.D
.
O.D. at 260 nm
O.D. at 280 nm
260/280 ratio
Figure 2. O.D. of S. aureus DNA samples obtained by four different phenol-
chloroform based DNA isolation methods
Results and Discussion
169
Table 3. O.D. of extracted DNA by different methods from S. typhi
Method O.D. at 260 nm O.D. at 280 nm 260/280 ratio
Method 1 0.72 0.444 1.62
Method 2 0.69 0.367 1.88
Method 3 0.62 0.352 1.76
Method 4 0.78 0.428 1.82
Results and Discussion
170
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
Method 1 Method 2 Method 3 Method 4
Different methods for DNA extraction
O.D
.
O.D. at 260 nm
O.D. at 280 nm
260/280 ratio
Figure 3. O.D. of S. typhi DNA samples obtained by four different phenol-
chloroform based DNA isolation methods
Results and Discussion
171
Extraction of genomic DNA in a reasonably intact and pure state forms the first
step in studies attempting to understand the molecular aspects of bacterial
pathogenesis, physiology, and epidemiology. Chemical methods for extraction of
genomic DNA rely primarily on the use of lysozyme in conjunction with other
lipolytic and proteolytic enzymes. However, many bacteria, such as
Staphylococcus aureus, are resistant to lysozyme (Lachica et al., 1971).
Increased recovery of genomic DNA, in particular from gram-positive bacteria
was obtained when the cells were pretreated with 70% ethanol (Kalia et al.,
1999). It is probable that 70 % ethanol exposure induces changes in the bacterial
cell wall and membrane, thereby increasing cellular porosity that could
accentuate subsequent bacterial lysis resulting in greater recovery of genomic
DNA.
When DNA is isolated from organisms, frequently there remains protein present
in the DNA solution; protein is tightly bound to DNA and complete removal of
protein is not always possible. To determine the concentration and purity of the
DNA solution, the absorbance of UV light is measured in a spectrophotometer.
Both protein and DNA absorb UV light, but they have different absorbance
curves. The peak of light absorption is at 260 nm for DNA and at 280 nm for
protein. When a spectrum of absorbance with varying wave length is run, both
curves slightly overlap in the area between, and including, 260 and 280 nm.
Thus, when a solution contains both protein and DNA, absorbance at 260 nm is
mainly due to the DNA present, but a little bit by the protein. At 280 nm, it is due
to the presence of protein. By dividing the two absorbance values, one can
calculate the purity of the DNA solution. If the solution relatively free of protein,
then one can take the absorbance at 260 nm as a measure for concentration of
DNA (Peng et al., 2007).
Results and Discussion
172
Of the 4 phenol chloroform based methods, method 4 was the most efficient
method for DNA extraction from E. coli. The same method also showed optimum
results when used for DNA isolation from S. typhi. However, for DNA extraction
from S. aureus, method 2 showed the best results.
Results and Discussion
173
2. Molecular and bacteriological examination of milk from
different milch animals with special reference to coliforms
The present study was planned to assess the quality of milk from different milch
animals to detect the presence of coliforms. For this, microbiological as well as
molecular methods (PCR) were used.
A total of twenty samples of milk were analysed. Raw milk of cow (5 samples),
buffalo (5 samples) and goat (5 samples) were collected from local vendors and
farmers of Vallabh Vidyanagar and Anand in the month of April, 2006. Amul
brand pasteurized milk (5 samples) was purchased from retailed outlets.
Cow milk samples were labeled as C1 to C5, buffalo milk samples from B1 to B5,
goat milk samples from G1 to G5 and pasteurized milk samples from P1 to P5.
The study was carried out in two stages. Milk samples from different milch
animals and pasteurized milk samples were analyzed during the first stage by
agar plate and sugar fermentation methods. In the second stage, confirmation of
bacteria was carried out by Polymerase Chain reaction.
Microbial analysis of milk samples
Total Plate count of milk samples
The results of total plate count are presented in Table No. 1. Total plate count in
milk samples ranged from 3.26 x 103 to 3.44 x 105 cfu/ml. Total plate count in
cow milk samples ranged from 0.83 x 104 to 3.44 x 105 cfu/ml and of buffalo milk
sample ranged from 1.07 x 104 to 9.31 x 104 cfu/ml. Among all milk samples,
goat milk showed highest plate count results ranging from 3.26 x 103 to 6.55 x
105. Plate count results in pasteurized milk samples ranged from 0.54 x 104 to
8.71 x 104 cfu/ml. Among all milk samples, pasteurized milk samples showed the
least number of cfu/ml.
Results and Discussion
174
VRBA count of milk samples
The results of VRBA count are presented in Table No. 2. VRBA count in milk
samples ranged from 1.53 x 101 to 8.73 x 102 cfu/ml.
VRBA count in cow milk samples ranged from 4.17 x 102 to 6.81 x 102 cfu/ml and
in buffalo milk samples, it ranged from 2.28 x 102 to 5.21 x 102 cfu/ml. Among all
milk samples, goat milk samples showed VRBA results ranging from 4.87 x 102
to 8.73 x 102 cfu/ml. VRBA results in pasteurized milk samples ranged from 1.53
x 101 to 1.31 x 102 cfu/ml.
Yeast and mold count of milk samples
The results of Yeast and mold count are presented in Table No. 3. Yeast and
mold count in all milk samples ranged from 1.02 x 101 to 2.23 x 102 cfu/ml. Yeast
and mold count in cow milk samples ranged from 1.04 x 102 to 2.23 x 102 cfu/ml
and in buffalo milk samples, it ranged from 1.41 x 102 to 2.12 x 102 cfu/ml. Goat
milk showed Yeast and mold count results ranging from 1.49 x 102 to 2.17 x 102
cfu/ml. Only one pasteurized milk sample showed the presence of Yeast and
molds, which was 1.02 x 101 cfu/ml. Highest count of yeast and mold was
observed in cow milk samples, while the least count of yeast and mold was
observed in pasteurized milk samples.
BGLB results of milk samples
BGLB (Brilliant Green Lactose Bile broth) results are presented in Table 4. Out
of a total of 20 milk samples, 9 samples showed positive results in BGLB broth.
LST results of milk samples
LST (Lauryl Sulphae broth) results are presented in Table 5. Out of total 20 milk
samples, 9 samples showed positive result in LST broth confirming the presence
of coliform group of bacteria. The same 9 samples were also positive for BGLB
test.
Results and Discussion
175
Table 1. Total Plate count of milk samples
Sample
No.
Total Plate
count
(cfu/ml)
Total Plate
Count
(log cfu/ml)
C1 4.51 x 104 4.451
C2 5.74 x 104 4.574
C3 3.44 x 105 5.344
C4 0.83 x 104 4.083
C5 6.58 x 104 4.658
B1 3.94 x 104 4.394
B2 4.12 x 104 4.412
B3 7.83 x 104 4.783
B4 1.07 x 104 4.107
B5 9.31 x 104 4.931
G1 2.74 x 105 5.274
G2 2.01 x 104 4.201
G3 6.55 x 105 5.655
G4 7.72 x 104 4.772
G5 3.26 x 103 3.326
P1 8.71 x 104 4.871
P2 0.54 x 104 4.054
P3 4.76 x 104 4.476
P4 2.64 x 104 4.264
P5 8.15 x 104 4.815
cfu = colony forming unit; C1-C5: Cow milk
B1-B5: Buffalo milk; G1-G5: Goat milk
P1-P5: Pasteurized milk
Results and Discussion
176
Table 2. VRBA count of milk samples
Sample
No.
VRBA
count
(cfu/ml)
VRBA
Count
(log cfu/ml)
C1 6.81 x 102 2.681
C2 4.81 x 102 2.481
C3 5.71 x 102 2.571
C4 6.70 x 102 2.670
C5 4.17 x 102 2.417
B1 2.28 x 102 2.228
B2 4.87 x 102 2.487
B3 5.21 x 102 2.521
B4 5.10 x 102 2.510
B5 2.82 x 102 2.282
G1 5.71 x 102 2.571
G2 8.73 x 102 2.873
G3 7.7 x 102 2.770
G4 5.76 x 102 2.576
G5 4.87 x 102 2.487
P1 2.21 x 101 1.221
P2 1.31 x 102 2.131
P3 1.71 x 101 1.171
P4 1.53 x 101 1.153
P5 1.68 x 101 1.168
cfu = colony forming unit; C1-C5: Cow milk
B1-B5: Buffalo milk; G1-G5: Goat milk
P1-P5: Pasteurized milk
Results and Discussion
177
Table 3. Yeast and mould count of milk samples
Sample
No.
Yeast and
mould
count
(cfu/ml)
Yeast and
mould
Count
(log cfu/ml)
C1 1.37 x 102 2.137
C2 2.23 x 102 2.223
C3 -- --
C4 1.04 x 102 2.104
C5 -- --
B1 1.41 x 102 2.141
B2 -- --
B3 -- --
B4 2.12 x 102 2.212
B5 -- --
G1 1.49 x 102 2.149
G2 -- --
G3 2.08 x 102 2.208
G4 2.17 x 102 2.217
G5 -- --
P1 -- --
P2 -- --
P3 -- --
P4 1.02 x 101 1.102
P5 -- --
cfu = colony forming unit; C1-C5: Cow milk
B1-B5: Buffalo milk; G1-G5: Goat milk
P1-P5: Pasteurized milk
Results and Discussion
178
Table 4. Gas production by milk samples in
Brilliant Green Lactose Bile Broth (BGLB)
Sample
No.
BGLB
reaction
C1 --
C2 --
C3 + ve
C4 --
C5 + ve
B1 + ve
B2 + ve
B3 --
B4 + ve
B5 + ve
G1 --
G2 + ve
G3 --
G4 + ve
G5 + ve
P1 --
P2 --
P3 --
P4 --
P5 --
-- = absent; + ve = Present
C1-C5: Cow milk; B1-B5: Buffalo milk
G1-G5: Goat milk; P1-P5: Pasteurized milk
Results and Discussion
179
Table 5. Gas production by milk samples in
LST (Lauryl tryptose broth)
Sample
No.
LST
reaction
C1 --
C2 --
C3 + ve
C4 --
C5 + ve
B1 + ve
B2 + ve
B3 --
B4 + ve
B5 + ve
G1 --
G2 + ve
G3 --
G4 + ve
G5 + ve
P1 --
P2 --
P3 --
P4 --
P5 --
-- = absent; + ve = Present
C1-C5: Cow milk; B1-B5: Buffalo milk
G1-G5: Goat milk; P1-P5: Pasteurized milk
Results and Discussion
180
ImViC test
Results from ImViC tests are presented in Table 6. Out of the total 20 milk
samples tested, 9 samples showed positive result for indole production. None of
the samples showed positive result in VP broth. 9 samples showed positive
results in Methyl Red test. None of the samples were found to be positive for
citrate production.
Indole positive bacteria such as Escherichia coli produce tryptophanase, an
enzyme that cleaves tryptophan, producing indole and other products. When
Kovac's reagent (p-dimethylaminobenzaldehyde) is added to a broth with indole
in it, a dark pink color is developed (Bacteriological Analytical Manual, 1998). The
methyl red (MR) and Voges-Proskauer (VP) tests were read from a single
inoculated tube of MR-VP broth. After 24-48 hours of incubation the MR-VP broth
was split into two tubes. One tube was used for the MR test; the other was used
for the VP test. MR-VP media contains glucose and peptone. All enteric bacteria
oxidize glucose for energy; however the end products vary depending on
bacterial enzymes. Both the MR and VP tests were used to determine what end
products result when the test organism degrades glucose. E. coli is one of the
bacteria that produce acids, causing the pH to drop below 4.4. When the pH
indicator methyl red is added to this acidic broth it will be cherry red, a positive
MR test (B.A.M., 1998).
In the present study, the pasteurized milk samples showed the presence of
bacteria. Gruetzmacher and Bradley (1999) stated that factors that limit the shelf
life of refrigerated pasteurized milk and the microbial quality of raw milk are time
and temperature of pasteurization, presence and activity of post pasteurization
contaminants, types and activity of pasteurization resistant microorganisms and
the storage temperature of milk after pasteurization.
Results and Discussion
181
Table 6. ImViC test of milk samples
Sample
No.
Indole
production
MR-VP Methyl Red Citrate
production
C1 - ve - ve - ve - ve
C2 - ve - ve - ve - ve
C3 + ve - ve + ve - ve
C4 - ve - ve - ve - ve
C5 + ve - ve + ve - ve
B1 + ve - ve + ve - ve
B2 + ve - ve + ve - ve
B3 - ve - ve - ve - ve
B4 + ve - ve + ve - ve
B5 + ve - ve + ve - ve
G1 - ve - ve - ve - ve
G2 + ve - ve + ve - ve
G3 - ve - ve - ve - ve
G4 + ve - ve + ve - ve
G5 + ve - ve + ve - ve
P1 - ve - ve - ve - ve
P2 - ve - ve - ve - ve
P3 - ve - ve - ve - ve
P4 - ve - ve - ve - ve
P5 - ve - ve - ve - ve
+ ve = Positive, - ve = Negative
Results and Discussion
182
PCR Analysis
PCR analysis of milk samples is expressed in Figure 1. Samples C3, C5, B1, B2,
B4, B5, G2, G4 and G5 showed amplification using primers specific for E. coli.
Approximately 150 bp fragment was amplified by the primers. Amplification was
observed in two samples of cow milk, four samples of buffalo milk and three
samples of goat milk. Pasteurized milk samples did not show any amplification.
This could be due to the pasteurization time and temperature employed at which
majority of the heat sensitive microorganisms get destroyed.
Results and Discussion
183
Figure 1. Samples from different milch animals showing amplification by E. coli
specific primers
M1- 100 bp marker; M2- 10 bp marker; C- control;
Results and Discussion
184
Pathogenic bacteria in milk have been a major public health concern since the
early days of the dairy industry. Many diseases are transmissible via milk
products. Traditionally raw or unpasteurised milk has been a major vehicle for
transmission of pathogens (Vasavada, 1988). The health of dairy herd, milking
conditions etc. are basic determinants of milk quality. Another source of
contamination by microorganisms is unclean teats. The use of unclean milking
and transport equipment also contribute to the poor hygienic quality (Bonfoh et
al., 2003). In the present study, the samples of raw milk examined contained
coliform group of microorganisms. This indicates that the analyzed milk samples
can prove to be a potential risk for public health when consumed or when used in
the production of dairy products such as cheese, butter, cream and ice cream
without being pasteurized or when subjected to insufficient heat processing.
From the results of the present study, it was found that majority of the samples
were contaminated by coliform bacteria. Coliform bacteria were found in milk
samples of different origin of milk. In operational conditions, mainly a failure to
observe hygienic rules of milking process contributes to the impairment of
microbial quality of milk (Jayarao et al., 2004). Tando et al. (2000) investigated
more than reported that 35.2 % of food handlers were asymptomatic carriers of
staphylococcus aureus, and that 90.4 % of raw milk samples among more than
3200 investigated dairy products by them.
Oksuz et al. (2004) reported E. coli 0157:H7 at the rate of 1% in 100 samples of
raw milk. Soomro et al. (2002) isolated E. coli in 57% of the 100 raw milk
samples. Coliforms and S. aureus are good indicators of the standard of hygiene
and handling. According to Harrigan and McCance (1976), coliform bacterial
count should be less than 200 cfu/g in food. The existence of the Coliforms has
been considered as leading to the fact that the product was subject to process
under inefficient hygienic conditions (Harrigan and McCance, 1976; Altug and
Bayrak, 2003).
Results and Discussion
185
A high level of coliform of the fresh farm milk can indicate the evidence of
unhygienic conditions of the product (Altug and Bayrak, 2003). Collins et al.
(1995) reported that E. coli and coliform bacilli which belong to the family of
Enterobacteriaceae may indicate evidence of contamination or pollution
especially of fecal nature. Enterobacteriaceae include other organisms, like
important pathogens such as salmonella and various non-lactose fermentors that
may be present in human and animal faeces. The bacterial count of milk is used
to measure its sanitary quality and most grading of milk is on the basis of some
method for estimating numbers (Collins et al., 1995).
Post pasteurization contamination has received most of the attention and is
considered to be the factor, which limits shelf life in the majority of cases (Waes,
1982). Waes (1982) found out that Pasteurized milk, which was collected from
the local shops, showed different values for standard bacterial counts. The higher
count of coliform bacteria in the milk might be due to improper handling, poor
cleaning and storage of equipments etc. as stated by Hayes et al. (2001).
The total viable count of fresh milk samples in the present study showed a mean
value of 4.572 log cfu/ml. Milk can be contaminated with different kinds of
microorganisms due to direct or indirect contact with any source of external
contamination during the steps of milking, collection, packing and transport.
Direct physical contact of milk with unclean surfaces such as those of milking
utensils, udders and teats, and the hands of milkers besides environmental
factors such as the design and cleanliness of buildings and installations, the
adequacy of the water supply, the manner in which the manure and other wastes
are disposed of, and the amount of dust in the immediate surroundings are
important in so far as they may contribute to the microbial contamination of
surfaces with which milk comes into contact (Hayes et al., 2001).
During milking operation, however, milk may be exposed to contamination from
the animal, especially the exterior of the udder and adjacent areas. Bacteria
found in manure, soil, and water may enter from this source. Such contamination
can be reduced by clipping the cow, and washing the udder with water or a
germicidal solution before milking. Contamination of cow with manure, soil, and
Results and Discussion
186
water may also be reduced by paving and draining barnyards, keeping cows from
stagnant pools, and cleaning manure from the barns or milking parlors.
Pasteurization kills pathogens that may enter the milk and improves the keeping
quality of milk (Hayes et al., 2001).
PCR methods are mostly used for the detection of microorganisms in different
types of food materials. These methods often allow better specificity compared to
traditional biochemical identification methods. In the present study, colonies
growing on the Nutrient agar plate were given pre-enrichment. This pre-
enrichment step was performed to achieve appropriate sensitivity. Another critical
component of the PCR assay is the inclusion of an internal positive control that
indicates PCR failures, e.g., through carry over of PCR inhibitors. Differentiation
of bacterial foodborne pathogens beyond the species level also provides exciting
opportunities to better understand the biology of bacterial strains and subtypes,
including differences in their ability to cause human foodborne disease (Maurer,
2006).
Samples analyzed in the present study can contribute a potential risk for public
health in cases when it is consumed or used in the production of dairy products
such as cheese, butter, cream and ice cream without being pasteurized or being
subjected to insufficient heat processing. Moreover, PCR is less labor intensive
and more rapid than bacterial culturing followed by conventional methods of
bacterial identification (Maurer, 2006).
The obtained results indicate that strict hygienic measures should be applied
during production, processing and distribution of milk and it’s products to avoid
contamination. Periodical inspection must be done by specialists on the dairy
farms to minimize milk contamination with different types of microorganisms.
Efficient cleaning and sanitization of farm dairy utensils must be done to improve
the quality of raw milk and consequently the related dairy products. Milk and milk
products should be kept under refrigeration at all times and the practice of
storage at room temperature should be discouraged.
Results and Discussion
187
3. Microbiological and molecular detection of E. coli, S. typhi
and S. aureus from milk samples
This part of study was carried out to detect the presence of three organisms
namely E.Coli, S. aureus and S. typhi in different milk samples available in the
local market.
10 different raw milk samples were collected from the local areas of V.V. Nagar
under aseptic conditions, samples were serially diluted and after pre trials, the
aliquots from 10-4 dilution was plated on nutrient agar, VRBA and EMB, for total
plate count, coliform count and E. coli count, respectively. From nutrient agar
loopful of colony was preserved in either glycerol or sterile D/W. Prior to analysis,
the culture was transferred to enrichment broth (Luria broth) for 18 hours for
activation.
For PCR analysis, preserved colonies were transferred to sterile distilled water,
vortexed for 5 to 10 sec and boiled at 95°C for 10 min and immediately
transferred to ice for 5 min, centrifuged and supernatant was directly taken for
PCR reaction.
Bacterial counts and PCR results are discussed in this section.
Microbiological analysis of milk samples
Collected milk samples were analyzed for total plate count, EMB count, VRBA
and yeast and mold count.
Total plate count
The total plate count results of milk samples are presented in Table No. 1. Total
plate count in all samples ranged from 2.31 × 104 to 5.12 × 106 cfu/ml. Sample
No. 5 showed highest total plate count while sample No. 1 showed lowest plate
count.
EMB count
The EMB (Eosin Methylene Blue) plate count results of milk samples are
presented in Table No. 2. EMB count in all samples ranged from 2.13 × 101 to
Results and Discussion
188
3.14 × 102 cfu/ml. Sample No. 1 showed highest plate count while sample No. 7
showed lowest plate count on EMB plates.
VRBA count results
The VRBA (Violet Red Bile Agar) plate count results of milk samples are
presented in Table No. 3. VRBA count in all samples ranged from 1.61 × 101 to
7.24 × 101 cfu/ml. Sample No. 2 showed highest plate count while sample No. 10
showed lowest plate count results in VRBA plates.
Yeast and Mold count results
The yeast and mold count results of milk samples are presented in Table No. 4.
Yeast and mold count in all samples ranged from 2.21 × 101 to 1.29 × 102 cfu/ml.
Sample No. 4 showed highest count while sample No. 8 showed lowest count on
Potato Dextrose Agar plates.
Results and Discussion
189
Table 1. Total plate count of milk samples
Sample No. CFU/ml log CFU/ml
1 2.31 × 104 4.23
2 7.92 × 104 4.79
3 1.21 × 106 6.121
4 4.15 × 104 4.415
5 5.12 × 106 6.512
6 8.27 × 105 5.827
7 5.83 × 105 5.583
8 1.79 × 106 6.179
9 8.14 × 105 5.814
10 3.86 × 105 5.386
Results and Discussion
190
Table 2. EMB plate count of milk samples
Sample No.
CFU/ml
EMB plate count
(log CFU/ml)
1 3.14 × 102 2.314
2 1.12 × 102 2.112
3 7.61 × 101 1.761
4 8.27 × 101 1.827
5 8.55 × 101 1.855
6 1.83 × 102 2.183
7 2.13 × 101 1.213
8 3.37 × 101 1.337
9 3.20 × 101 1.320
10 2.72 × 101 1.272
Results and Discussion
191
Table 3. VRBA plate count of milk samples
Sample No.
CFU/ml
VRBA plate
count (log
CFU/ml)
1 5.21 × 101 1.521
2 7.24 × 101 1.724
3 2.85 × 101 1.285
4 3.32 × 101 1.332
5 2.91 × 101 1.291
6 1.97 × 101 1.197
7 1.77 × 101 1.177
8 2.28 × 101 1.228
9 2.39 × 101 1.239
10 1.61 × 101 1.161
Results and Discussion
192
Table 4. Yeast and mold count of milk samples
Sample No. Yeast and mold
count (CFU/ml)
Yeast and
mold count
(log
CFU/ml)
1 8.24 × 101 1.824
2 6.12 × 101 1.612
3 7.31 × 101 1.731
4 1.29 × 102 2.129
5 4.96 × 101 1.496
6 5.18 × 101 1.518
7 7.91 × 101 1.791
8 2.21 × 101 1.221
9 3.83 × 101 1.383
10 6.38 × 101 1.638
Results and Discussion
193
PCR results of milk samples
Figures 1, 2 and 3 show the PCR results of milk samples. Figure 1 shows PCR
results for E. coli. It can be seen from the figure that sample numbers 2,4,5,6, 8,
9 and 10 were found to be positive for the presence of E. coli. Figure 2 shows the
PCR results for S. aureus. Sample number 2,4,5,6 and 10 were found positive
for the presence of S. aureus. Figure 3 shows the PCR results for S. typhi. It can
be seen from the figure that none of the samples were found to be positive for
salmonella.
Results and Discussion
194
Figure 1. Detection of E. coli from milk samples by PCR
C: Control; M: Marker, 1-10: milk samples
Results and Discussion
195
Figure 2. Detection of S. aureus from milk samples by PCR
C: Control; M: Marker, 1-10: milk samples
Results and Discussion
196
Figure 3. Detection of S. typhi from milk samples by PCR
C: Control; M: Marker, 1-10: milk samples
Results and Discussion
197
The quality of milk is determined by aspects of composition and hygiene. Due to
it’s complex biochemical composition and high water activity, milk serves as an
excellent culture medium for the growth and multiplication of many kinds of
microorganisms. Therefore in the processing of milk, some of the
microorganisms may produce undesirable effects and some micro-organisms
produce food infections which can increase the likelihood of infection of the
consumer‘s food. The contamination of milk and milk products is largely due to
the human factor and unhygienic conditions. Usually milk is contaminated with
different kinds of microorganisms at milk collecting places. Milk is a major part of
human food and plays a prominent role in the Indian diet. Approximately 50
percent of the milk produced is consumed as fresh or boiled, one sixth as yoghurt
or curd and the remaining is utilized for manufacturing of indigenous varieties of
milk products such as Ice cream, Butter, Khoa, Paneer, Rabri, Kheer, Burfi and
Gulabjaman (Anjum et al., 1989). The manufacture of these products is based on
traditional methods without any regard to the quality of raw material used and/ or
the hygienic quality of the products. Under such conditions many microorganisms
can find access to the milk products. Among all micro-organisms, Escherichia
coli is the frequently contaminating organism and is a reliable indicator of fecal
pollution generally in insanitary conditions of water, food, milk and other dairy
products (Diliello,1982). Martin et al., (1986) reported two cases of hemolytic
uraemic syndrome which provide evidence that raw milk may be a vehicle of
transmission of E.coli O157: H7, both affected persons consumed raw milk.
Recovery of E. coli from food is an indicative of possible presence of
enteropathogenic and/or toxigenic micro-organism which could constitute a
public health hazard. Enteropathogenic E. coli (EEC) can cause severe diarrhoea
and vomiting in infants and young children (Anon, 1975). In 1971 USA faced an
outbreak of food poisoning in which 387 persons were suffered with
Enteropathogenic E. coli due to the consumption of imported French cheese
(Marrier, 1973).
Results and Discussion
198
In the present study, the microbiological and molecular analysis of the milk
samples revealed that the highest number of samples was contaminated with E.
coli as compared to S. aureus. However, S. typhi was found to be absent in all
the milk samples.
According to Mosupye and Van Holy (1999), the method of production, handling,
transportation and marketing of milk is entirely dependent upon the traditional
system. Such a system could pose a favourable environment for bacterial
contamination. The unclean hands of workers, poor quality of milk, unhygienic
conditions of manufacturing unit, inferior quality of material used and water
supplied for washing the utensils could be sources of accelerating the bacterial
contamination of milk products and the post manufacturing contamination
(Grewal and Tiwari, 1990; Kulshrestha, 1990).
Although E. coli is a frequently occurring organism in milk and its products, the
incidence of the species of E. coli itself in milk and milk products as a possible
cause of food borne disease is insignificant because E. coli normally is a
ubiquitous organism (Hahn, 1996). However, the occurrence of pathogenic
strains of E. coli in milk products can be unhygienic, which could be hazardous
for consumers.
Results and Discussion
199
4. A comparison of methods for the detection of Escherichia
coli O157:H7 from artificially-contaminated dairy products
using PCR
The present study was planned to evaluate various DNA extraction methods for
detection of E. coli O157:H7 from artificially contaminated dairy products (liquid
skim milk, skim milk powder, cheese) with the help of PCR. The methods
evaluated were (i) Solvent method and (ii) Concentration method
Comparison of PCR detection limits for the solvent method and the
bacterial concentration method
Solvent method and concentration method were evaluated for their efficacy of E.
coli DNA extraction from liquid skim milk, skimmed milk powder and cheese. The
extracted DNA was amplified by PCR using E. coli specific primers. The
amplified PCR products were run on agarose gel electrophoresis to evaluate the
efficacy of the methods used for DNA extraction from food materials.
Results are expressed in Table 1. When both the methods were applied to liquid
skim milk, the solvent-based method provided higher PCR detection limits
compared to the bacterial concentration technique. In this case, the final
detection limits were 103 cfu/ ml using the solvent technique and 105 cfu /ml for
the concentration method (Table 1 and Figure 1). It can be seen that PCR
amplicon was not obtained on the inoculum level of 101 and 102 cfu/ml in the
solvent method, whereas in bacterial concentration method PCR amplicons were
not obtained in the inoculum level upto 105 cfu/ml. This indicates the better
detection limits of the solvent method compared to the concentration method.
Results and Discussion
200
Table 1. Detection of PCR products in artificially contaminated dairy
products with two DNA extraction methods
Food
Material
Method Inoculum level of E. coli O157:H7
101 102 103 104 105 106 107
Liquid
Skim milk
Solvent - - + + + + +
concentration - - - - - + +
Skim milk
powder
Solvent - - - - - + +
Concentration - - - - + + +
cheese solvent - - - + + + +
concentration - - - - + + +
+ : presence of amplicon; - : absence of amplicon
Results and Discussion
201
Fig. 1. PCR amplification of Escherichia coli O157:H7 DNA isolated from
liquid skim milk using the (a) solvent-based method (b) the bacterial
concentration method
(a)
(b)
The initial inoculum levels are given above each gel lane. They are in 10x mode (where x = 1,2,3,4,5,6,7).
The PCR product in this and all other foods is a 1.5 kb segment of slt-II from E. coli O157:H7.
- and + indicate negative (no template) and positive controls, respectively.
Results and Discussion
202
Fig. 2. PCR amplification of Escherichia coli O157:H7 DNA isolated from
powder skim milk using the (a) solvent-based method (b) the bacterial
concentration method
(a)
(b)
The initial inoculum levels are given above each gel lane. They are in 10x mode (where x = 1,2,3,4,5,6,7).
The PCR product in this and all other foods is a 1.5 kb segment of slt-II from E. coli O157:H7.
- and + indicate negative (no template) and positive controls, respectively.
Results and Discussion
203
Fig. 3. PCR amplification of Escherichia coli O157:H7 DNA isolated from
cheese using the (a) solvent-based method (b) the bacterial concentration
method
(a)
(b)
The initial inoculum levels are given above each gel lane. They are in 10x mode (where x = 1,2,3,4,5,6,7).
The PCR product in this and all other foods is a 1.5 kb segment of slt-II from E. coli O157:H7.
- and + indicate negative (no template) and positive controls, respectively.
Results and Discussion
204
Similarly, when applied to cheese, the solvent-based extraction method allowed
a higher improvement in PCR detection limits with visible amplicons obtained at
contamination levels of 104 cfu/ ml while the bacterial concentration method
could detect only upto 105 contamination levels (Table 1). Both sample
preparation methods performed equally well for PCR amplification of DNA
extracted from cheese sample. Visible amplification bands were obtained from
samples with initial E. coli O157:H7 levels of 104 cfu/ ml or higher (Table 1 and
Figure 3).
Likewise, for skim milk powder, PCR detection limits were ≥ 106 cfu /ml for both
the solvent and concentration methods. Surprisingly, attempts at using the
solvent method on skim milk powder resulted in no detectable PCR amplicon,
even when skim milk powder was seeded with 105 cfu/ ml. Whereas, inoculum
level of 105 cfu/ ml was detectable by bacterial concentration method.
Using the concentration technique for skim milk powder, PCR amplicons from E.
coli O157:H7 could be visually detected at initial contamination levels of 105 cfu
ml-1 and above (Table 1 and Figure 2).
Overall, it can be concluded that the Solvent method showed better results than
the bacterial concentration method for detection of PCR amplicons from
artificially contaminated dairy products.
The detection limits of PCR-based pathogen screening methods for foods are
directly dependent on the efficiency of the nucleic acid extraction method
employed. Direct DNA extraction from a variety of foods has been applied
recently, with varying degrees of detection sensitivity (Dickinson et al. 1995;
Drake et al. 1996 and Abolmaaty et al. 1998). For example, the direct DNA
extraction and PCR detection method described by Dickinson et al. (1995)
reported detection limits between 103 and 104 cfu/ml. L. monocytogenes or
Aerococcus viridans in Camembert cheese. Lantz et al. (1994) achieved a
Results and Discussion
205
detection sensitivity of 104 cfu/ml when detecting L. monocytogenes from soft
cheese using an aqueous two-phase system.
A number of recent studies for detecting pathogens in food have made use of
bacterial concentration strategies, such as immunomagnetic separation (IMS) to
sequester cells for subsequent detection by PCR (Okrend et al., 1991; Jinneman
et al., 1995; Gooding and Choudary 1997; Tomoyasu 1998; Onoue et al., 1999).
Lucore et al. (2000) described a novel method for concentrating Salmonella
enteritidis and L. monocytogenes from food samples using absorption by metal
hydroxides prior to RNA extraction and subsequent detection by RT-PCR. Cell
recovery efficiencies of 65-96 % were obtained from non-fat dry milk
artificially contaminated with these two pathogens. RT-PCR detection limits were
in the order of 101-102 cfu /25 ml of non-fat dry milk. As the target template used
in the Lucore et al. study was 16S rRNA, detection limits were high, partly
because of the high initial copy number of this macromolecule in viable cells.
The present findings indicate that in almost all cases, the solvent method
performed equally well or better than the bacterial concentration method. For the
milk and cheese sample used in this study, the quantity of DNA obtained using
the solvent extraction method was considerably higher than that obtained with
the concentration method, perhaps because bacterial concentration prior to
nucleic acid extraction may limit the amount of food-related DNA that is co-
extracted. The results for skim milk powder were not in accordance with these
findings. It was hypothesized by McKillip et al., (2000) that the lactose
component of skim milk powder either decreased the efficiency of adsorption to
metal hydroxides and/or the extraction of DNA.
Other investigators (Rossen et al., 1992) have attempted to improve PCR
detection limits by procedural changes and PCR additives and were largely
unsuccessful. Many of these strategies are commonly employed to reduce the
effect of carry-over inhibitors on the PCR reaction, and/or function to maximize
primer-template annealing, template availability and/or Taq polymerase activity.
Results and Discussion
206
Of these approaches, few had any positive impact on DNA yield, and none
increased PCR sensitivity. Gel-based detection of PCR products is undoubtedly
less sensitive than real-time fluorescent-based detection systems (McKillip and
Drake 2000).
The advantages and disadvantages of either the solvent extraction or the
concentration method vary depending on time constraints, PCR detection limits
desired, level of contamination and specific project objectives. The bacterial
concentration method involves fewer steps than the solvent method and does not
require the extensive use of organic solvents. It has the added advantage of
allowing an assessment of bacterial recovery/viability by plate counts
immediately following metal hydroxide adsorption. The solvent extraction method
offers no such opportunity for bacterial recovery prior to cell lysis. In contrast, the
solvent method takes less time to complete than concentration, and provides
better end-point detection limits for some foods. While some minor procedural
modifications were needed when adapting both methods to different dairy
commodities, it is expected that either method could readily be modified for a
variety of food products. It is crucial; however, that such procedural modifications
evaluate the efficiency of DNA extractions, as this appears to be an extremely
important and frequently overlooked variable impacting the overall detection
limits of PCR-based detection strategies (Mckillip et al., 2000).
The study indicates that the Solvent method showed better results than the
bacterial concentration method for detection of PCR amplicons from artificially
contaminated dairy products. The detection limits of PCR-based pathogen
screening methods for foods are directly dependent on the efficiency of the
nucleic acid extraction method employed.
Results and Discussion
207
5. Assessment of viability of probiotic bacteria and competitive
growth of Lactobacillus acidophilus in yoghurt during
refrigerated storage
In the present study, survival of lactic acid bacteria was evaluated using
molecular methods like PCR during refrigerated storage of three experimental
yoghurt samples. Competitive growth of Lactobacillus acidophilus was also
monitored in the presence of other lactic acid bacteria during refrigerated storage
of yoghurt for a period of 30 days.
Microbial Analysis of Yoghurt samples
Microbial analysis of yoghurt samples was carried out by analyzing for total
Lactobacillus count, L. acidophilus count, Yeast and Mold count and coliform
count. Monitoring the viability of 7 probiotic strains in yoghurt over 30 days has
indicated trends that are related to the different species of organisms tested. In
control yoghurt, count of total lactobacilli increased during first 15 days of storage
and then it decreased with increased storage time (during 15 to 30 days) at 4˚C
(Table 1 and Figure 1), although the increase and decrease was nonsignificant
(p>0.05). Similar trend was also observed in experimental yoghurt B and yoghurt
C. In experimental yoghurt A, total lactic acid bacteria decreased significantly
(p
Results and Discussion
208
during 30 days of storage at 4˚C. In this sample, L. acidophilus was inoculated
with all other lactic acid bacteria. In Exp. B and Exp. C yoghurt samples, L.
acidophilus count was also found to be decreasing during last 15 days of 30 days
of storage but it was nonsignificant (p>0.05).
Yeast and Mold are one of the most common groups of microbes responsible for
spoilage of fermented dairy products (Pitt & Hocking, 1997). These microbes
have the ability to reduce the shelf life of dairy products even after refrigerated
storage. There was no significant (p>0.05) growth of yeast and mould during the
storage of yoghurt for 30 days at 4˚C (Figure 3). Coliform count increased during
15 to 30 days, however it was also nonsignificant (p>0.05) (Figure 3).
Results and Discussion
209
Table 1. Total lactic acid bacteria count in yoghurt samples during storage
period of 30 days at 4˚C
Yoghurt Samples
Day 0
(CFU/ml)
Day 15
(CFU/ml)
Day 30
(CFU/ml)
Control 6.72 x 106 7.41 x 106 6.93 x 106
Experimental A 4.13 x 107 3.72 x 106 1.25 x 106
Experimental B 5.02 x 108 6.29 x 108 2.51 x 108
Experimental C 7.38 x 108 8.42 x 108 3.17 x 108
Results are mean of three observations
Results and Discussion
210
1
2
3
4
5
6
7
8
9
Control Experimental A Experimental B Experimental C
Yoghurt Samples
log
10
(CF
U/m
l)
0 day
15 day
30 day
Fig. 1. Total lactic acid bacteria count in yoghurt samples during storage period of
30 days at 4˚C
Results and Discussion
211
Table 2. Lactobacillus acidophilus count in yoghurt samples during
storage period of 30 days at 4˚C
Yoghurt Samples 0 Day (CFU/ml) 15 Day (CFU/ml) 30 Day (CFU/ml)
Experimental I 4.17 x 103 3.72 x 102 8.24 x 101
Experimental II 5.7 x 103 6.28 x 103 2.17 x 103
Experimental III 7.4 x 103 8.49 x 103 3.15 x 103
Results are mean of three observations
Results and Discussion
212
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0 day 15 day 30 day
Days
log
10
(CF
U/m
l)
Experimental A
Experimental B
Experimental C
Fig. 2. Lactobacillus acidophilus count in yoghurt samples during storage period of
30 days at 4˚C
Results and Discussion
213
0.8
0.9
1
1.1
1.2
1.3
1.4
Day 0 day 30
Days
log
10(C
FU
/ml)
Yeast andMold
Coliform
Fig. 3. Yeast and mold and coliform count of yoghurt samples during storage period
of 30 days at 4˚C
Results and Discussion
214
PCR Analysis
Enzymatic nucleic acid amplification methods such as the polymerase chain
reaction (PCR) offer several advantages for the rapid and reliable detection of
microorganisms. DNA extraction and PCR analysis was performed from the
yoghurt samples to confirm the survival of each individual bacteria during the
storage period of 30 days at 4˚C. The results are presented in Table 3.
All the bacterial strains were found to be present i.e. amplification was
observed in all the strains by their respective primers in all the yoghurt samples
at the initial level i.e. on 0 day and at final level i.e. on day 30. In spite of the low
count of L. acidophilus in yoghurt samples (Exp. A, Exp. B and Exp. C), there
was sufficient DNA which was amplified by PCR cycles. It indicates that L.
acidophilus survives in the presence of other lactic acid bacteria, but it’s viability
gets reduced.
Studies indicate that, in the development of rapid detection methods,
fermented dairy products have been particularly challenging because they are
compositionally complex and contain food-associated components and high
level of background microflora that often interfere with detection assay, resulting
in less than optimal detection limits. In general, different authors have found
differences of 10 fold or more in detection limits when results from seeded dairy
matrices are compared to results from pure cultures, even after the incorporation
of procedural modifications such as increased Mg++ concentration to improve
amplification efficiency and prior DNA purification (Wernars, Heuvelman,
Chakraborty, & Notermans, 1991; Fluit, Torensma, Visser, Aarsman, Keller, &
Verhoef, 1993 and Wilson, 1997).
Results and Discussion
215
Table 3. Polymerase chain reaction amplification results of different bacterial
species in yoghurt samples during storage period of 30 days at 4˚C
Bacterial
species
Yoghurt samples
Control Experimental A Experimental B Experimental C
0
day
30
day
0 day 30 day 0 day 30 day 0 day 30 day
Lactobacillus
delbrueckii
subsp.
bulgaricus
+ + + + + + + +
Lactobacillus
plantarum
-- -- + + + + + +
Lactobacillus
acidophilus
-- -- + + + + + +
Lactobacillus
casei
-- -- + + + + + +
Streptococcus
thermophilus
+ + + + + + + +
Lactobacillus
fermentum
-- -- + + + + + +
Lactobacillus
paracasei
-- -- + + + + + +
+: Presence of amplification; --: not applicable
Results and Discussion
216
Proteolytic activity of Yoghurt samples during storage period
All the three experimental yoghurt samples showed lower proteolytic activity than
the control yoghurt. No significant change (p>0.05) in the proteolytic activity
(zone of clearance on the milk agar plate) of the yoghurt samples was observed
during 30 days of storage (Figure 4). Among all yoghurt samples, control sample
showed highest proteolytic activity, which was on day 30. Among the
experimental yoghurt samples, experimental sample C showed highest
proteolytic activity, which was on Day 0. L. acidophilus along with the other
thermophilus rods is adjudged to be more proteolytic and releases more amounts
of free amino acids (Alm, 1982). However, Amar & Lammerding (1980) on testing
the L. acidophilus strain found it to be relatively less proteolytic compared with
other lactic acid bacteria.
The proteolytic activity of yoghurt is mainly due to the action of Lactobacillus
bulgaricus (Tamine & Robinson, 1999 and Thomas & Mills, 1981). Lactobacillus
bulgaricus, a lactic acid bacterium with complex growth requirements, is
extensively used in the manufacture of cheese and yogurt (Law & Kolstad, 1983).
The pool of free amino acids and peptides present in milk is not enough to
ensure optimal bacterial growth (Mills & Thomas, 1981). The main source of
nitrogen for this species in milk is provided by the hydrolysis of caseins by the
action of L. bulgaricus proteases (Tamine & Robinson, 1999; Thomas & Mills,
1981). Expression of proteolytic activity is important in relation to symbiotic
growth with Streptococcus thermophilus during the production of yoghurt (Rasic
& Kurman, 1981). Results concerning the proteolytic activity of L. bulgaricus have
been obtained in rich media such as MRS broth (Argyle, Mathison, & Chandan,
1976), in which bacteria utilized free amino acids present in the broth. In contrast,
the proteolytic activity in milk has not been extensively studied. A recent study by
Laloi, Atlan, Blanc, Gilbert, & Portalier, 1991 regarding proteases of cells grown
in milk and in MRS broth showed identical patterns of hydrolytic products of α-
and β-caseins.
Results and Discussion
217
10.6
10.8
11
11.2
11.4
11.6
11.8
12
12.2
12.4
Control Experimental A Experimental B Experimental C
Yoghurt samples
Zo
ne o
f cle
ara
nce i
n m
m
Day 0
Day 30
Fig. 4. Proteolytic activity (Zone of clearance on milk agar plate) of yoghurt samples
during storage period of 30 days at 4˚C
Results and Discussion
218
Sensory Evaluation
Sensory evaluation was carried out from all yoghurt samples and it is presented
in Table 4. The results indicated that overall acceptability of experimental yoghurt
A obtained a score of 7.97 on a scale of 10 points which indicates very good
acceptability. Control yoghurt showed an overall acceptability score of 7.78. The
average flavour score was higher for experimental yoghurt B, although the
differences were non-significant (p>0.05). Overall acceptability was higher in the
experimental yoghurt A. Gupta, Mital & Garg (1997) reported no significant
differences in the textural characteristics of both the control and acidophilus
yoghurt. In the present study, organoleptic evaluation revealed that all the three
yoghurt samples were almost identical with respect to colour, flavour, texture and
overall acceptability with a score ranging from 7.72 to 7.97 on a 10 point scale.
According to Gilliland & Speck, 1975; the acidic nature of L. acidophilus
enables it to withstand storage in an acidic environment for a reasonable time
without loss in viability. The situation is different when L. acidophilus is mixed
with a medium, such as yoghurt, containing the metabolic products of other
microorganisms. In spite of reports that indicate that L. acidophilus can be added
to yoghurt successfully, no supporting data have been found for these
presumptions. By the use of a medium which can enumerate differentially L.
acidophilus in a mixture with yoghurt culture, we have shown that L. acidophilus
indeed is damaged markedly with respect to viability during storage with the
products contained in yoghurt. The microorganism in yoghurt responsible for the
antagonism for L. acidophilus was L. bulgaricus. The microorganism produces
some substance(s), other than acid during it’s growth which is the antagonistic
agent(s) (Servin & Coconnier, 2003). Gilliland & Speck, 1975 reported that
hydrogen peroxide seems to be the main agent responsible for the loss in
viability of L. acidophilus when mixed in yoghurt. Studies have shown that L.
bulgaricus produces hydrogen peroxide in milk at 5˚C (Gilliland & Speck, 1975;
Premi & Bottazii, 1972 and Ito, Sato, Kudo, Sato, Nakajima, & Toba, 2003). They
also reported that sufficient peroxide was produced to inhibit the growth of
psychrotrophic bacteria.
Results and Discussion
219
Table 4. Sensory evaluation of yoghurt samples on 0 day
Sensory Score ( out of 10 )
Product Colour Flavour Texture Overall
acceptability
Control
7.97 ±
1.02
7.17 ±
1.36
8.22 ±
1.26
7.78 ± 0.96
Experimental
Experimental
Aa
8.25 ±
0.94
7.33 ±
0.91
8.33 ±
1.28
7.97 ± 0.91
Experimental
Bb
7.80 ±
1.13
7.36 ±
1.03
8.03 ±
1.31
7.73 ± 0.99
Experimental
Cc
7.87 ±
1.13
7.22 ±
1.03
8.08 ±
1.31
7.72 ± 0.82
Mean of 10 judges ± S.D.
a: Addition of L. acidophilus along with other bacterial cultures
b: addition of L. acidophilus 2 hours after the inoculation of other bacterial cultures
c: Addition of L. acidophilus 2 hours before the inoculation of other bacterial cultures
Results and Discussion
220
PCR amplification was observed in the 0 day samples and 30 day samples.
Hence, all lactic acid bacteria were viable at the end of the storage study period
indicating survival of lactic acid bacteria during synbiotic growth in yoghurt. All
the prepared yoghurt samples had good overall acceptability. No significant
growth was observed in yeast and coliform count during storage. No significant
change in the proteolytic activity of the yoghurt samples was observed upon
storage. The study of the proteolytic activity of L. bulgaricus in milk can enhance
the knowledge base required for selection of starter cultures. The present study
indicates that yoghurt product is a suitable career for a variety of probiotic
bacteria but survival of L. acidophilus will need to be improved to provide
optimum health benefits to consumers. Further research can be carried out to
determine genetic relationship of antagonistic effect of microorganisms during
synbiotic growth in a model system.
Results and Discussion
221
6. Effect of nisin on growth and survival of selected food
pathogens
The present study was planned to evaluate the effect of the bacteriocin nisin on
selected food pathogens namely S. aureus, L. monocytogenes and S. typhi
during yoghurt fermentation and storage. Effect of nisin was also evaluated on
the growth of yoghurt starter cultures namely L. bulgaricus and S. thermophilus.
Parameters studied were:
1. Effect of Nisin on yoghurt fermentation
2. Effect of Nisin on S. aureus, S. typhi and L. monocytogenes
3. Effect of Nisin on S. aureus, S. typhi and L. monocytogenes in Yogurt
1. Effect of Nisin on yoghurt fermentation
As yogurt starter cultures i.e. L. bulgaricus and S. thermophilus is reported to be
sensitive to nisin (Kumar and Prasad, 1994; Kuma and Prasad, 1994b;
Vandenbergh, 1993), this preliminary experiment was carried out to determine
the concentration of nisin which can be used without affecting significantly normal
yogurt processing and acid production.
Nisin was added to the cups containing yoghurt to a final concentration of 10, 50
and 100 IU/ml. A sample not containing nisin was used as control. The
containers were incubated at 43˚C for 24 hours and then transferred to the
refrigerator (4 -7°C).
The pH and acidity were measured every 4 hours upto 24 hours. The change in
pH and acidity of control sample is shown in Table 1. Table 2 shows change in
pH and acidity of yoghurt sample having nisin concentration 10 IU/ml. Table 3
shows change in pH and acidity of yoghurt sample having nisin concentration 50
IU/ml, whereas Table 4 shows change in pH and acidity of yoghurt sample
having nisin concentration 100 IU/ml.
Results and Discussion
222
Table 1. pH of the prepared yoghurt during fermentation
Fermentation
Hours
pH of yoghurt samples
C Sample A Sample B Sample C
0 6.5 6.6 6.5 6.5
4 5.6 5.4 5.9 5.6
8 4.7 4.9 4.9 5.1
12 4.2 4.5 4.5 4.9
16 3.7 3.9 4.3 4.7
20 3.5 3.6 4.2 4.3
24 3.2 3.4 4.2 4.2
Sample A: Nisin Concentration. 10 IU/ml
Sample B: Nisin Concentration 50 IU/ml
Sample C: Nisin Concentration 100 IU/ml
C: Control
Results and Discussion
223
Table 2. Acidity of the prepared yoghurt during fermentation
Fermentation
Hours
Acidity of yoghurt samples (gm/Lit)
C Sample A Sample B Sample C
0 2.2 2.2 2.1 2.1
4 2.2 3.9 3.4 3.7
8 6.6 7.1 6.3 5.8
12 7.0 7.2 7.0 6.0
16 7.6 7.6 7.6 6.0
20 7.9 7.8 7.7 6.1
24 8.1 8.0 7.9 6.1
Sample A: Nisin Concentration. 10 IU/ml
Sample B: Nisin Concentration 50 IU/ml
Sample C: Nisin Concentration 100 IU/ml
C: Control
Results and Discussion
224
Table 1 and Table 2 show the decrease in pH and the increase in acidity during
fermentation of yogurt containing different concentrations of nisin. In the control
yoghurt sample, pH dropped and acidity increased as the time of fermentation
increased.
In the samples having 10 IU/ml and 50 IU/ml, the pH dropped and the acidity
increased in the same way as in the control. Milk coagulation in all samples was
normal; it occurred between 6 to 7 hours and the curd was firm and without
syneresis. It can be seen from Table 1 and Table 2 that a nisin concentration of
50 IU/ml or less had no noticeable effect on yoghurt fermentation.
However, in the samples containing 100 IU/ml of nisin, fermentation was greatly
retarded and the curd had an abnormal, viscous body. Yoghurt sample having
nisin concentration of 100 IU/ml showed 4.2 and 6.1 gm/Lit pH and titrable acidity
respectively at the end of 24 hours fermentation.
pH and titrable acidity in control sample were 3.2 and 8.1 gm/lit respectively, at
the end of 24 hours fermentation. This indicates that nisin has an inhibitory effect
on L. bulgaricus and S. thermophilus (yoghurt starter cultures). This result was
also observed by several other scientists (Benkerroum et al., 2003, Kumar and
Prasad, 1994).
Benkerroum et al. (2003) and Kumar and Prasad (1994) tested several strains of
lactobacilli for their sensitivity to nisin and found that the MIC ranged between 35
and 100 IU/ml for Lb. delbueckii subsp. bulgaricus strains at optimal growth
conditions and when an inoculum of 10 ml/Lit was used. These values doubled
for an inoculum of 20 ml/l (Benkerroum et al., 2003; Kumar and Prasad, 1994).
As matter of fact, a moderate delay in yogurt acidification may be suitable in
yogurt technology since in the conventional process the product usually develops
too much acidity towards the end of the storage and wheys-off. According to
Bayoumi (1991), nisin addition to the level of 50 IU/ml prevents such a defect
Results and Discussion
225
resulting in a 7 day increase of the shelf-life without affecting sensory
characteristics.
2. Effect of Nisin on S. aureus, S. typhi and L. monocytogenes
(A) Varying pH with constant Nisin concentration
Table 3 (Figure 1, Figure 2, Figure 3, Figure 4) shows the effect of nisin on S.
aureus growth at different pH values. It may be seen that numbers of S. aureus
increased at pH 6.8 in the controls (without nisin), while they decreased steadily
in test samples containing 50 IU/ml of nisin. Similar results were obtained at pH
5.5. At pH 4.5, a decrease in S. aureus counts was observed in both test and
control samples; however, the pathogen was eliminated from the test samples
within 48 hours while, in the control, few cells were still viable at 48 hours. An
almost similar trend was observed at pH 5.0.
Table 4 (Figure 5, Figure 6, Figure 7, Figure 8) shows the effect of nisin in
yoghurt on S. typhi at different pH values. At pH 6.8, S. typhi count in test
samples decreased steadily at the end of 48 hours, while it increased in control
samples. Similar trend was observed in pH 5.5 and 5.0. At pH 4.5, S. typhi was
eliminated from the test sample after 24 hours, while it took 48 hours to eliminate
S. typhi from control samples
Growth of L. monocytogenes in listeria selective agar plates at different pH
values in the absence or presence of 50 IU/ml of nisin is shown in Table 5
(Figure 9, Figure 10, Figure 11 and Figure 12). It may be seen that numbers of L.
monocytogenes increased at pH 6.8 in the control (without nisin), while they
decreased steadily in test samples containing 50 IU/ml of nisin. Similar results
were obtained at pH 5.5. At pH 4.5, a decrease in Listeria counts was observed
in both test and control samples; however, the pathogen was eliminated from the
test samples within 24 hours while, in the control, few cells were still viable even
at 48 hours. The same behavior was observed at pH 5.0 (Table 1). Similar
results were previously reported in TSB but with a higher nisin concentration
(Benkerroum et al., 2003). These data suggest that 50 IU/ml of Nisin is effective
in the control of L. monocytogenes in milk and dairy products. A concentration of
Results and Discussion
226
10 to 500 IU/g has been recommended in food preservation in general (Eapen et
al., 1983).
The use of nisin, the bacteriocin produced by Lactococcus lactis subsp. lactis, is
successfully used nowadays as an antibacterial agent in various food products.
Nisin affects several Gram-positive bacteria such as Listeria spp.,
Staphylococcus spp. but does not inhibit the majority of Gram-negative bacteria
(Abee et al., 1994; Martinis et al., 1997). Nisin has shown to be efficient in
inactivating Gram-negative bacteria when used together with chelating agents
(EDTA), causing an aberration in cell membrane lipopolisaccharide component
(Stevens et al., 1992).
EDTA was added in the tubes having S. aureus and S. typhi to enhance the
inhibitory activity of nisin. Nisin when used in combination with the chelating
agent EDTA; inhibits a wide variety of Salmonella and Staphylococcal species
(Grisi and Gorlach-Lira, 2005). Stevens et al. (1991), reported that inhibition of
Salmonella species by a combination of nisin and EDTA, is a time-dependent
phenomenon and the method of application (simultaneous versus sequential) is
critical to achieving the desired effect. Increasing the nisin concentration above
50 ug/ml will, most likely, increase the magnitude-of inactivation. Furthermore,
the observed inactivation by nisin can be extended to other gram-negative
bacteria. The observed population reductions by nisin are facilitated by the
chelation of magnesium ions, present in the outer membrane, by EDTA. The
removal of magnesium ions from the lipopolysaccharide layer of the outer
membrane results in the loss of lipopolysaccharide and an increase in cell
permeability (Nikaido et al, 1987).
This increase in outer membrane permeability to nisin is proposed to facilitate
inactivation of the cell via bactericidal action at the cytoplasmic membrane.
Applications involving simultaneous treatment with nisin and an outer membrane
modifying-chelating agent such as EDTA may be of value in controlling food-
borne Salmonella species as well as other gram-negative pathogens in foods.
Leonides et al (2008) reported that nisin is an efficient alternative to antibiotics for
the treatment of staphylococcal mastitis.
Results and Discussion
227
The use of the Generally Recognized As Safe (GRAS) lactic acid bacteria (LAB),
or the antimicrobial compounds they produce (bacteriocins), is a promising
ongoing development in food preservation (smid and Gorris, 2007). Bacteriocins
are antimicrobial peptides with activity mainly against Gram-positive bacteria has
gained great attention in recent years. Although the efficacy of bacteriocins can
be limited in food systems if applied alone, several bacteriocins have shown
additive or synergistic effects when used in combination with other antimicrobial
agents or processes such as chelating agents, heat, modified atmosphere
packaging and high hydrostatic pressure (HHP).
Results and Discussion
228
Table 3. Effect of different pH and Nisin (50 IU/ml) on the growth of S. aureus (log
CFU/ml) at 37˚C
pH
6.8 5.5 5.0 4.5
Hours + nisin - Nisin + Nisin - Nisin + Nisin - Nisin + Nisin - Nisin
0 6.00 6.00 6.00 6.00 6.00 6.00 6.00 6.00
4 6.18 6.07 6.23 6.09 5.89 6.26 5.24 5.44
8 5.89 7.81 6.93 7.20 5.72 5.84 4.47 4.81
24 5.02 8.64 4.62 7.61 4.29 5.36 2.70 3.42
48 3.28 9.23 3.38 8.27 2.28 3.80 0.00 1.33
Nisin was used in combination with 20 mM EDTA
Media used: Baird-Parker Agar media
Results and Discussion
229
Table 4. Effect of different pH and Nisin (50 IU/ml) on growth of S. typhi (log
CFU/ml) at 37˚C
pH
6.8 5.5 5.0 4.5
Hours + nisin - Nisin + Nisin - Nisin + Nisin - Nisin + Nisin - Nisin
0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0
4 6.32 6.74 6.26 6.38 5.78 6.26 5.37 5.72
8 6.18 7.61 6.79 7.22 5.20 5.49 4.12 4.96
24 5.29 8.47 4.92 8.04 4.06 4.20 0 2.69
48 3.77 9.58 3.95 9.09 2.22 3.48 0 0
Nisin was used in combination with 20 mM EDTA
Media used: Salmonella differential agar plates
Results and Discussion
230
Table 5. Effect of different pH and Nisin (50 IU/ml) on growth of L. monocytogenes
(log CFU/ml) at 37˚C
pH
6.8 5.5 5.0 4.5
Hours + nisin - Nisin + Nisin - Nisin + Nisin - Nisin + Nisin - Nisin
0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0
4 6.64 7.15 6.33 6.24 5.7 6.3 4.6 5.3
8 7.63 8.86 7.22 7.61 5.8 5.92 4.4 4.8
24 4.21 9.76 3.31 7.38 3.2 5.80 0 2.92
48 1.72 9.71 1.4 8.52 0 2.3 0 0.65
Media used: Listeria selective agar plates
Results and Discussion
231
0
1
2
3
4
5
6
7
8
9
10
0 4 8 24 48
Time (in hours)
Lo
g C
FU
/ml
Nisin
Control
Figure 1. Growth of S. aureus in Nisin added (50 IU/ml) at pH 6.8
Results and Discussion
232
0
1
2
3
4
5
6
7
8
9
0 4 8 24 48
Time (in hours)
Lo
g C
FU
/ml
Nisin
Control
Figure 2. Growth of S. aureus in Nisin added (50 IU/ml) at pH 5.5
Results and Discussion
233
0
1
2
3
4
5
6
7
0 4 8 24 48
Time (in hours)
Lo
g C
FU
/ml
Nisin
Control
Figure 3. Growth of S. aureus in Nisin added (50 IU/ml) at pH 5
Results and Discussion
234
0
1
2
3
4
5
6
7
0 4 8 24 48
time (in hours)
Lo
g C
FU
/ml
Nisin
Control
Figure 4. Growth of S. aureus in Nisin added (50 IU/ml) at pH 4.5
Results and Discussion
235
0
2
4
6
8
10
12
0 4 8 24 48
Time (in hours)
Lo
g C
FU
/ml
Nisin
Control
Figure 5. Growth of S. typhi in Nisin added (50 IU/ml) at pH 6.8
Results and Discussion
236
0
1
2
3
4
5
6
7
8
9
10
0 4 8 24 48
Time (in hours)
Lo
g C
FU
/ml
Nisin
Control
Figure 6. Growth of S. typhi in Nisin added (50 IU/ml) at pH 5.5
Results and Discussion
237
0
1
2
3
4
5
6
7
0 4 8 24 48
Time (in hours)
Lo
g C
FU
/ml
Nisin
Control
Figure 7. Growth of S. typhi in Nisin added (50 IU/ml) at pH 5
Results and Discussion
238
0
1
2
3
4
5
6
7
0 4 8 24 48
Time (in hours)
Lo
g C
FU
/ml
Nisin
Control
Figure 8. Growth of S. typhi in Nisin added (50 IU/ml) at pH 4.5
Results and Discussion
239
0
2
4
6
8
10
12
0 4 8 24 48
Time (in Hours)
Lo
g c
fu/m
l
Nisin
Control
Figure 9. Growth of L. monocytogenes in nisin added (50 IU/ml) at pH 6.8
Results and Discussion
240
0
1
2
3
4
5
6
7
8
9
0 4 8 24 48
Time (in Hours)
Lo
g c
fu/m
l
Nisin
Control
Figure 10. Growth of L. monocytogenes in Nisin added (50 IU/ml) at pH 5.5
Results and Discussion
241
0
1
2
3
4
5
6
7
0 4 8 24 48
Time (in hours)
Lo
g c
fu/m
l
Nisin
Control
Figure 11. Growth of L. monocytogenes in Nisin added (50 IU/ml) at pH 5
Results and Discussion
242
0
1
2
3
4
5
6
7
0 4 8 24 48
Time (in hours)
Lo
g c
fu/m
l
Nisin
Control
Figure 12. Growth of L. monocytogenes in Nisin added (50 IU/ml) at pH 4.5
Results and Discussion
243
B. Effect of varying pH and varying Nisin concentration
Effect of varying pH and varying Nisin concentration are presented in Table 6,
Table 7 and Table 8 for S. typhi, S. aureus and L. monocytogenes, respectively.
Considering the fact that in practice there is a tendency to minimize the amount
of an additive in food preservation, this experiment was carried out in TSB using
different combinations of pH and nisin concentrations. At pH 6.8, no
concentration was inhibitory to L. monocytogenes; no significant difference (P
>0.05) between O.D. reached after 48 hours in samples containing up to 200
IU/ml and the control, was observed (Table 5). At pH 6.0, 5.5 and 5.0; however,
50, 100 and 200 IU/ml were all effective. A concentration of 10 IU/ml had no
significant (P> 0.05) effect on L. monocytogenes at pH 6.8 and 5.5 but was
significantly (p
Results and Discussion
244
concentration or to the polymerization of the molecule (Kumar and Prasad,
1994). In fact, nisin activity was completely and irreversibly lost after 4 days of
storage at pH 7.0 and ambient temperature. Results show that nisin action at a
given pH depends on the concentration and the medium used. In effect, 50 IU/ml
resulted in more than 4 log units reduction in Listeria counts in listeria selective
agar plates at pH 6.8 after 48 hours.
Results and Discussion
245
Table 6. Effect of varying pH and varying Nisin concentration on
growth of S. typhi in luria broth and incubated at 37˚C
pH Nisin (IU/ml)
O.D. at different time intervals
0 hr 2 hr 4 hr 8 hr 24 hr 48 hr
6.8 0 0 0.132 0.184 0.238 0.653 0.935
10 0 0.141 0.195 0.319 0.77 0.872
50 0 0.223 0.238 0.398 0.822 0.848
100 0 0.217 0.235 0.317 0.790 0.824
200 0 0.196 0.240 0.298 0.793 0.811
6 0 0 0.115 0.232 0.241 0.911 0.932
10 0 0.122 0.127 0.272 0.902 0.917
50 0 0.228 0.214 0.298 0.353 0.427
100 0 0.218 0.218 0.311 0.341 0.382
200 0 0.168 0.179 0.305 0.320 0.344
5.5 0 0 0.089 0.177 0.328 0.732 0.777
10 0 0.169 0.183 0.281 0.584 0.642
50 0 0.211 0.231 0.251 0.279 0.297
100 0 0.219 0.239 0.263 0.288 0.305
200 0 0.230 0.235 0.242 0.261 0.289
5 0 0 0.113 0.175 0.202 0.306 0.378
10 0 0.26 0.182 0.212 0.276 0.308
50 0 0.209 0.263 0.281 0.288 0.302
100 0 0.069 0.122 0.202 0.233 0.242
200 0 0.077 0.106 0.169 0.183 0.192
Results and Discussion
246
Table 7. Effect of varying pH and varying Nisin concentration on
growth of S. aureus grown in luria broth and incubated at 37˚C
pH Nisin (IU/ml)
O.D. at different time intervals
0 hr 2 hr 4 hr 8 hr 24 hr 48 hr
6.8 0 0 0.139 0.181 0.248 0.642 0.922
10 0 0.149 0.178 0.332 0.717 0.861
50 0 0.238 0.242 0.376 0.822 0.859
100 0 0.229 0.246 0.321 0.790 0.838
200 0 0.185 0.233 0.283 0.782 0.837
6 0 0 0.123 0.224 0.241 0.911 0.932
10 0 0.134 0.122 0.262 0.902 0.917
50 0 0.227 0.227 0.246 0.353 0.427
100 0 0.229 0.189 0.324 0.341 0.382
200 0 0.179 0.150 0.341 0.322 0.351
5.5 0 0 0.098 0.177 0.328 0.732 0.807
10 0 0.144 0.201 0.226 0.521 0.589
50 0 0.213 0.219 0.242 0.257 0.270
100 0 0.231 0.217 0.244 0.263 0.331
200 0 0.234 0.239 0.242 0.261 0.289
5 0 0 0.128 0.178 0.232 0.328 0.407
10 0 0.211 0.172 0.236 0.263 0.337
50 0 0.228 0.253 0.274 0.262 0.323
100 0 0.066 0.128 0.216 0.253 0.232
200 0 0.065 0.118 0.152 0.171 0.198
Results and Discussion
247
Table 8. Effect of varying pH and varying Nisin concentration on
growth of L. monocytogenes in TSB incubated at 37˚C
pH Nisin (IU/ml)
O.D. at different time intervals
0 hr 2 hr 4 hr 8 hr 24 hr 48 hr
6.8 0 0 0.72 0.197 0.316 0.766 0.917
10 0 0.214 0.239 0.382 0.753 0.878
50 0 0.168 0.228 0.351 0.717 0.778
100 0 0.188 0.205 0.279 0.639 0.741
200 0 0.62 0.188 0.251 0.651 0.722
6 0 0 0.131 0.166 0.27 0.783 0.933
10 0 0.43 0.169 0.203 0.742 0.849
50 0 0.188 0.212 0.417 0.481 0.509
100 0 0.174 0.188 0.356 0.408 0.431
200 0 0.161 0.84 0.321 0.361 0.393
5.5 0 0 0.057 0.17 0.203 0.763 0.781
10 0 0.049 0.101 0.213 0.641 0.664
50 0 0.136 0.212 0.267 0.328 0.34
100 0 0.118 0.233 0.241 0.281 0.315
200 0 0.114 0.227 0.281 0.319 0.342
5 0 0 0.042 0.063 0.127 0.228 0.276
10 0 0.055 0.058 0.089 0.131 0.165
50 0 0.117 0.157 0.171 0.149 0.117
100 0 0.125 0.141 0.147 0.158 0.128
200 0 0.061 0.112 0.123 0.118 0.083
Results and Discussion
248
Nisin, due to its legal status (European parlament of council, 1995) and
commercial availability is the bacteriocin that has been more widely studied,
although other bacteriocins such as sakacin K (produced by Lactobacillus sakei
CTC494) and enterocins A and B (produced by Enterococ-cus faecium CTC492)
have also been shown to reduce the counts of artificially inoculated L.
monocytogenes, Salmonella or S. aureus in meat products (Garriga et al., 2002;
Aymerich et al., 2005; Chung et al., 2005; Yuste et al., 1998; Ananou et al.,
2005) and to provide extra protection when breaks in the cold chain occurred
(Marcos et al., 2008).
3. Effect of Nisin on S. aureus, S. typhi and L. monocytogenes in Yogurt
Table 9 shows the growth of L. monocytogenes, S. aureus and S. typhi in yogurt
in the presence and absence of nisin during storage at 7˚C. Although a
significant decrease in Listeria counts was observed in yogurt without nisin, the
pathogen survived manufacture and 10 days of storage at 7˚C. However, in
yogurt containing 10 IU/ml of nisin, no Listeria was found after 24 hours. Table 5
shows the behaviour of S. aureus and S. typhi in yoghurt in the presence and
absence of nisin during storage at 7˚C. S. aureus and S. typhi were not able to
survive for 24 hours in the presence of Nisin in yoghurt. However in control
samples, they were able to survive until 10 days.
Results and Discussion
249
Table 9. Effect of nisin (10 IU/ml) on the growth of Listeria
monocytogenes, S. aureus and S. typhi in yogurt
L. monocytogenes S. aureus S. typhi
Days Log
CFU/ml
(sample)
Log
CFU/ml
(Control)
Log
CFU/ml
(sample)
Log
CFU/ml
(Control)
Log
CFU/ml
(sample)
Log
CFU/ml
(Control)
0 6.0 6.0 6.0 6.0 6.0 6.0
1 0 4.2 0 4.7 0 5.3
2 0 3.3 0 3.2 0 4.4
3 0 3.1 0 2.6 0 3.5
10 0 2.4 0 1.3 0 2.1
15 0 0 0 0 0 0
Nisin (10 IU/ml) was used in combination with 20 mM EDTA in samples inoculated with
S. aureus and S. typhi
Results and Discussion
250
Survival of pathogens in fermented dairy products, in spite of the antagonistic
effect of lactic acid bacteria used as starter cultures is well documented
(Benkerroum et al., 2000; De Buyser et al., 2001; Donnely, 1990). In yogurt and
in other dairy products fermented with the same starter [e.g. Lb. bulgaricus and
Str. thermophilus (ST:LB::1:1)] such as some cultured milks and Feta cheese, L.
monocytogenes survives manufacture and storage (Ribeiro and Carminati, 1996;
Rocourt, 1996). Papageorgiou and Marth (1989) showed that L. monocytogenes
could survive the manufacture and more than 90 days of storage at 4°C in Feta
cheese. This bacterium survived in cultured milk fermented with STLB and in
yoghurt for 1 to 12 weeks and 1 to 12 days, respectively (Piccinin and Shelef,
1995).
The same authors showed that survival of L. monocytogenes in yogurt depended
on the size of Listeria and starter culture inocula, the final pH reached the
temperature and duration of the fermentation, and Listeria strain. Shaack and
Marth (1988) showed that L. monocytogenes survives only between 9 to 15
hours during the actual fermentation process of yogurt. However, in a typical
yogurt fermentation of 4 to 6 hours, Listeria was able to grow during fermentation
and then survive during storage at 4 °C. Lammerding and Doyle (1989) also
could recover L. monocytogenes from yoghurt after 7 days of storage at 4˚C,
although the initial inoculum was relatively low (about 32 x 102
CFU/ml).
Results and Discussion
251
7. Bacteriocin production by free and immobilized Lactococcus
lactis
In the present study, bacteriocin production was carried out using Lactococcus
lactis. Free bacterial cells and immobilized bacterial cells were evaluated for
bacteriocin production. Bacteriocin production was carried out from free bacterial
cells using two different media: Media 1 (MRS broth) and Media 2 (MRS broth +
sucrose). The supernatants of the two different media and the immobilized
organism were assessed for bacteriocin activity.
Bacteriocin activity of all the three culture filtrates was compared using agar
diffusion assay. Firstly standard nisin concentrations over a range (10-50 IU/ml)
were used for assessing activity against each sensitive bacterial strain like M.
luteus, S. aureus and B. subtilis. Diameter of inhibition zones were measure
against each organism and a standard curve were generated.
Micrococcus luteus which is used as the indicator strain did show increase in
zone of inhibition with increase in the number of transfers. Maximum zone of
inhibition were obtained by supernatant of MRS + sucrose while only MRS did
show lower levels of inhibition. The results are presented in Figure 1.
Results and Discussion
252
0
0.5
1
1.5
2
2.5
3
3.5
1 2 3 4 5
No. of Transfers
Zo
ne o
f in
hib
itio
n (
cm
)
MRS
MRS + Sucrose
Immobilized
Fig. 1. Zone of inhibition of three different media at different transfers against
M. Luteus
Results and Discussion
253
In the case of the immobilized supernatant, good inhibition was achieved upto
two transfers but it decreased gradually in later transfers. S. aurues which is
used for assessing antimicrobial activity showed a good zone of inhibition using
different supernatants. In the case of MRS supernatant, increased inhibition with
an increase in the no. of transfers was observed. In the case of MRS + sucrose
supernatant, maximum zone of inhibition compared to other samples was
observed but it decreased in the 5th transfer. Immobilized culture filtrate showed
good inhibition zones in the 1st two transfers then remained stationary from the
third transfer. Results are presented in Figure 2. B. subtilis which was used for
assessing antimicrobial activity did not show good zone of inhibition compared to
S. aurues & M. luteus. B. subtilis further did not show any inhibition by
supernatant of MRS and MRS + sucrose supernatant in the 1st transfer.
In the 2nd transfer, both MRS and MRS + sucrose supernatants showed a slight
zone of inhibition. The immobilized filtrate showed a very large zone of
Recommended