Performance assessment of the TEMPO® system :
an automated method for Escherichia coli enumeration in foods
Danièle SOHIER1*, Maryse RANNOU1, Florence GORSE2
1
ADRIA Développement, Quimper, France
2
bioMérieux, Marcy l’Etoile, FRANCE
* Corresponding author: [email protected]
INTRODUCTION
Culture methods used to determine Escherichia coli levels in food products are time consuming and
labor intensive. The TEMPO system and reagents have been developed to replace plate and tube
counting methods. In this study the TEMPO system has been used to compare the performance of
the TEMPO EC test for the enumeration of E. coli in food products to the ISO 16649-2 plate count
method (TBX) according to ISO 16140 technical rules for performance assessment, as part of the
AFNOR validation.
The difference between TBX and PCA shows in all cases (except for E. coli 1) a negative
value due to an underestimation of E. coli pure cultures in TBX.
The enumeration obtained with TEMPO EC is more in agreement with that obtained on non
selective plates (PCA) than with selective medium TBX, which signifies that TEMPO EC
ensures a better recovery of E. coli cells than the ISO 16649-2 method.
Table 2 : Enumeration of non E. coli pure cultures performed by ISO 16649-2 and TEMPO EC
Non E. coli strains
TEMPO EC PRINCIPLE
The TEMPO EC test consists of a vial of culture medium and an innovative card (Figure 1), which
are specific to this test. The culture medium is inoculated with the sample to be tested by the
TEMPO Filler (Figure 2) into the card containing 48 wells of three different volumes. The TEMPO
Filler seals the card to avoid any contamination during subsequent handling.
During incubation, the E. coli strains present in the card reduce the medium substrate,
MU-β-glucuronide, and cause a fluorescent signal to appear, detected by the TEMPO Reader
(Figure 3). Depending on the number and size of the positive wells, the TEMPO system deduces
the number of Escherichia coli present in the original sample according to a calculation based on
the MPN method.
Card reading, interpretation and reporting are managed by the system after a 24h incubation at
37°C. From a 1:10 diluted sample, the TEMPO EC method allows E. coli enumeration from 10 to
49,000 CFU/g of food.
Bacillus subtilis 630
Escherichia vulneris 132
Citrobacter diversus 140
Hafnia alvei 111
Citrobacter freundii 23
Klebsiella oxytoca CIP 7932
Enterobacter aerogenes CIP6086
Klebsiella pneumoniae 47
Enterobacter cloacae 10
Kluyvera ascorbata CIP 8295T
Enterobacter sakazakii D7
Proteus vulgaris 56
Enterococcus faecalis 25
Providencia stuartii 46
Erwinia carotovora CIP 103762
Rhanella aquatilis 69
Escherichia hermanii 395
Serratia liquefaciens 26
Escherichia vulneris 127
Shigella flexneri CIP 8248
None of the 20 non E. coli strains belonging to 19 species were detected by either method.
MATERIAL AND METHODS
Two methods, TEMPO EC and ISO 16649-2 on TBX
medium, used for enumeration of β-glucuronidase positive
E. coli were compared with pure cultures, and on artificially
and naturally contaminated food products.
This comparison was performed according to the ISO 16140
rules and to robust statistics, the choice of the regression
depending on the SDwy/SDwx ratio : GMFR (orthogonal
regression) and OLS (linear regression).
TEMPO EC is sensitive and specific.
RELATIVE LINEARITY
Figure 4
Figure 5
Figure 6
Figure 1 : TEMPO reagents
SENSITIVITY AND SPECIFICITY
The sensitivity and specificity of TEMPO EC were assessed
using pure cultures of 30 E. coli strains and 20 non-E. coli
strains encountered in food products (Tables 1 and 2).
R = 0.99
N = 10
RELATIVE LINEARITY
The relative linearity was investigated by mixing pure culture
serial dilutions with 5 food categories : dairy , meat,
vegetable, seafood and pet food (Figures 4 to 8).
RELATIVE EXACTITUDE
50 artificially and naturally contaminated food products were
tested in parallel with the ISO 16649-2 method and TEMPO
EC (Figure 9) including:
-
Figure 7
R = 0.96
N = 10
Figure 8
All
linear
correlation
coefficients are contained
between 0.96 to 1.00
(Figures 4 to 8).
Figure 2 : Preparation
station with TEMPO Filler
R = 0.96
N = 10
Meat and poultry: raw and processed, fresh or frozen.
Dairy products: frozen, dry, fermented.
Fish and seafood: smoked, raw, fresh and frozen.
Fruits and vegetables: fresh, dehydrated or frozen
Pet food: raw, processed or dehydrated
R = 0.97
N = 10
R = 1.00
N = 10
Figure 9
Figure 3 : Reading station with
TEMPO® Reader
RELATIVE EXACTITUDE
Figure 9 : Relative exactitude
according the analysis
of 50 food products
RESULTS
SENSITIVITY AND SPECIFICITY
Table 1 : Enumeration of E. coli pure cultures performed by ISO 16649-2 and TEMPO EC
Control
PCA
log CFU/ml
ISO 16649-2
TBX
log CFU/ml
E.coli 1
E.coli 17
E.coli 19
E.coli 101
8.87
9.28
8.76
9.15
8.89
8.53
8.34
7.81
E.coli 108
E.coli 2B
E.coli 118
E.coli 12
9.32
8.81
8.89
9.08
Strains
TEMPO
log CFU/ml
log difference
TEMPO-PCA
log CFU/ml
log difference
TBX-PCA
log CFU/ml
log difference
TEMPO-TBX
log CFU/ml
8.99
9.40
8.97
8.93
0.12
0.12
0.21
-0.22
0.02
-0.75
-0.42
-1.34
0.10
0.87
0.63
1.12
9.00
8.96
9.00
<6
-0.32
0.15
0.11
NA
-0.66
-0.05
-0.18
NA
0.34
0.20
0.29
NA
8.75
9.32
9.21
9.08
8.83
-0.17
0.02
0.06
-0.20
-0.37
-0.21
-0.62
-0.45
-0.58
-1.24
0.04
0.64
0.505
0.38
0.87
8.96
-0.19
-0.39
0.20
8.70
<6
-0.26
NA
-0.35
NA
0.09
NA
9.10
9.11
6.42
9.21
8.93
9.25
9.14
9.11
9.11
8.93
9.06
8.96
9.14
9.36
-0.19
-0.12
-2.94
-0.03
0.16
0.21
-0.06
-0.09
-0.07
-0.19
-0.85
-0.20
-0.15
0.10
-0.64
-0.55
-2.71
-0.47
-0.47
-0.40
-0.57
-0.44
-0.47
-0.85
-0.95
-0.97
-0.72
-0.54
0.46
0.43
-0.23
0.45
0.63
0.61
0.51
0.35
0.40
0.67
0.10
0.78
0.58
0.64
Light blue colonies
8.66
8.76
8.71
8.88
Non characteristic
white colonies
E.coli 121
E.coli 123
E.coli 13
E.coli 14
E.coli 143
8.92
9.30
9.15
9.28
9.20
8.71
8.68
8.70
8.70
7.96
white colonies
with a blue outline
E.coli 144
9.15
E.coli 21
E.coli ATCC 43888
8.96
8.93
E.coli 70
E.coli 9
E.coli 91
E.coli 94
E.coli 96
E.coli 97
E.coli Ad 217
E.coli Ad 222
E.coli Ad 223
E.coli Ad 228
E.coli 18
E.coli CIP 54127
E.coli CIP 54117
E.coli CIP 7624
9.28
9.23
9.36
9.23
8.77
9.04
9.20
9.20
9.18
9.11
9.91
9.15
9.28
9.26
8.76
with a blue outline
8.61
8.04
R = 0.90
N = 50
All TEMPO EC results match with those obtained with the reference method. No false
positive or false negative results are observed, confirming the specificity and sensitivity of
TEMPO EC. According to the regression equation of the relative exactitude, TEMPO EC
favours the recovery of E. coli cells compared to ISO 16649-2 method.
TEMPO PRACTICABILITY
The practicability of the two methods was compared according to AFNOR validation rules. TEMPO
system offers important economic savings by standardizing the analysis and by minimizing the
training time, the volume of waste and the number of operations.
Only 4 minutes are required for the analysis of 1 sample, of a large series, by the TEMPO system,
instead of 11 minutes by the plate count method (Table 3).
Table 3 : Timing required for samples analysis
Steps
Sampling
Weighting
Petri dishes preparation (*)
Serial dilutions and Inoculation
Reading
Total time
Time required for analysis
of 1 sample
NF ISO 16649-2
method
20 samples
45
15
13
64
90
227
11.4
1 sample
4
2
0,3
4
2.5
12.8
12.8
TEMPO EC
1 sample
4
2
/
2
0.45
8.45
8.45
20 samples
45
15
/
20
2
82
4.1
Non characteristic
white colonies
8.64
8.68
6.65
8.76
8.30
8.64
8.63
8.76
8.71
8.26
8.96
8.18
8.56
8.72
All E. coli strains, characteristic on the reference method, have been detected on TEMPO EC.
(*) The timing indicated does not include media preparation and this should be added for the
reference method. TEMPO EC is a ready to use medium.
TEMPO significantly improves the productivity by at least a three fold reduction in the time to
perform an analysis.
Moreover, the unique bar-code on each TEMPO card ensures complete traceability and fully
automated reading, thereby limiting the risk of errors.
CONCLUSION
TEMPO EC :
- is sensitive and specific
- gives comparable results to the ISO 16649-2 method
- shows very high practicability
TEMPO EC represents a valuable and innovative alternative
automated method for E. coli enumeration in 24 hours in food
products.