ROSEPROMILK QLK1 CT 2001 - 01617
ROBUST CHEMICAL SENSORS AND BIOSENSORS FOR RAPID ON-LINE
IDENTIFICATION OF FRESHLY COLLECTED MILK
FIRST YEAR INDIVIDUAL ANNUAL REPORT
Reporting period: 1 December 2001 – 30 November 2002
Coordinator: University of Rome “Tor Vergata”
Department of Science and Chemical Technology
1
1. Objectives and Expected Achievements
Mycotoxin contamination
Milk is susceptible to contamination from external sources; one key analyte is aflatoxin
M1 (CRM). Aflatoxin M1 is a hepatocarcinogenic mycotoxin, that could occur in milk
of cows fed with aflatoxin B1- contaminated feedstuffs. It is proposed to develop a
sensor for aflatoxin M1. The procedure we have to develop is totally innovative
consisting in the development of an electrochemical procedure, which allows a sensitive
detection of Aflatoxin M1 directly on a probe surface using commercially available
antibodies to aflatoxin M1.
This procedure requires simple steps, can be carried out directly in the milking parlour
with the use of inexpensive instrumentation, also the probes are strips which will be
disposable to avoid problems with probe regeneration and calibration.
Expected achievements
1. Immunosensor for Aflatoxin M1.
2. Laboratory validation for faecal contamination and Aflatoxin M1 detection.
3. Integration of (i) and (ii) into robotic milking machine.
4. Validation of (iv) in operational milking parlour.
2
ROSEPROMILK
FIRST ANNUAL REPORT
ROLE OF PARTICIPANTS:
Participant n. 2.
University of Roma “Tor Vergata”. Dipartimento di Scienze e Tecnologie Chimiche
Team Leader:
Prof. Giseppe Palleschi
Scientist involved in the project:
Prof. Danila Moscone
Dr Laura Micheli
Objectives
Immunodetection for aflatoxin M1
Work plan
Task 2.1 Development of an immunosensor for Aflatoxin
Subtask 2.1.1.
(Months 0-6)
The first step will be the conjugation of commercially available antibodies anti-aflatoxin
with appropriate enzymes, as alkaline phosphate or peroxidase or other label enzyme
compatible with the immunoanalysis with milk matrix. Also aflatoxin itself will be
conjugated with enzyme.
Subtask 2.1.2.
(Months 6-12)
Direst ELISA competition tests or ELISA sequential analysis will be carried out using
spectrophotometric detection. Buffer, temperature, pH will be selected to obtain the best
signal/noise ratio.
Subtask 2.1.3 (Months 12-18)
Study of selectivity, interference, cross-reaction and stability of the conjugate will be
carried out and the lower and upper detection limits defined. Aflatoxin determination
using standard solution and ELISA with spectrophotometric detection will be carried
out.
Task 2.2 Screen printed electrode evaluation
Subtask 2.2.2
(Months 18-21)
After the selection of the best electrochemical transducer, conjugated enzyme-toxins or
enzyme antibody will be fixed on the electrode strip using immobilisation procedures in
which almost all participants are expert.
Subtask 2.2.3
(Months 21-24)
Then ELISA procedures (direct or sequential) will be performed according with well
know protocols. Also the optimisation of the analytical parameter will be carried out
and the results obtained with the electrochemical procedure in standard solutions of
aflatoxin will be compared with those obtained spectrophotometrically.
3
Deliverables
!!!!!!!!!!!!!!!!!!!!!!!!!
Research activities during the first reporting period
TASK 2.1 Development of an immunosensor for aflatoxin
Immunosensors are analytical devices, which selectively detect analytes and
provide a concentration - dependent signal. Electrochemical immunosensors employed
either antibodies or their complementary binding partners (antigens) as biorecognition
elements in combination with electrochemical transducers.
Immunoassay techniques are based on the ability of the antibodies to form
complexes with corresponding antigens. The property of highly specific molecular
recognition of antigens by antibodies leads to highly selective assays based on immune
principles. The extreme affinity of antigen-antibody interactions results also in a great
sensitivity of immunoassay methods.
The most common type of immunoassay is known as enzyme linked
immunosorbent assay or ELISA. This method is a conventional solid phase
immunoassay technique, where the antigen-antibody interaction occurs at a solid phase
surface.
The first task of this work concerns the study of an electrochemical immunosensor for
the determination of AFM in milk using “screen printed electrodes”.
Subtask 2.1.1 months: 0-6
Person months: !!!!!!!!!!!!!!!
The first step of this work has been the evaluation of commercial AFM1 labelled
with Peroxidase (AFM1-HRP). This product was purchased by R-Biopharm
(Ridascreen kit - Germany) and was used for testing the affinity and the quality of the
disposable antibody against AFM1.
With this conjugate, a direct ELISA competition test with spectrophotometric
and electrochemical detections has been developed, then results obtained with the two
procedures have been compared.
Subtask 2.1.2: months 6-12
Person months: !!!!!!!!!!!!!
4
The development of spectrophotometric ELISA for AFM1 before the
electrochemical study allowed the definition of the working ranges, limit of detection
and the characteristic of the immunological reagents (antigen and antibody) prior the
additional testing with the electrode. The test was performed in a 96-well microplate
according with a direct ELISA format. This kind of test was based on a competition
between antigens labelled (added) with enzyme and unlabelled (sample or standard) for
binding sites of antibodies, immobilised on the preactivated support (scheme A). The
preactivation consists of a preimmobilisation of antibodies anti-IgG (mouse) for
improvement the amount and the orientation of the antibodies specific for the analyte to
be analysed. The amount of labels associated with the solid phase is inversely related to
the concentration of antigen.
Ab
IMMOBILISATION
PRECOATING
PRECOATING
E
AFM1 and AFM1-enzyme
ADDITION
DETECTION
E
Anti-IgG (mouse)
Solid Support
Monoclonal
Primary Ab
Enzyme-antigen
conjugate
S
P
Antigen
Assay optimisation
For improvement the amount and the orientation of the primary antibodies (antiAFM1), the study of a preimmobilisation of antibody against anti-IgG has been carried
out. Fig. 1 shows the precoating results and it is possible to observe that the better
working conditions were obtained when a precoating step with 10 µg/mL of anti-IgG in
100 mM carbonate buffer pH 9.6 was carried out.
5
PRECOATING
350
Fig. 1. Precoating study
ABS @ (655 nm)
300
250
 coating, blocking, AFM1-HRP
NO
PRECOATING
200
 precoating, coating, blocking and AFM1-HRP
150
100
 black: no coating, only blocking and AFM1-HRP
50
0
0
1
2
3
4
5
The study of the buffer (carbonate and phosphate saline buffers), the time
(overnight, 2 h), and the concentration (0.3 – 40 µg/mL) for the coating step had been
carried out. Results showed that a good immobilisation was achieved when the coating
step was carried out for 2 h at 37°C, with a solution of 5 µg/mL of primary monoclonal
antibody (anti-AFM1) in PBS (fig. 2, 3, 4). We have chosen 2 h for the coating
incubation time because the results were comparable with those obtained with overnight
incubation time, but shorter.
PBS
350
ABS@ (655nm)
300
250
Fig. 2. Coating Buffer Study
CB
200
CB = 100 mM carbonate buffer pH 9.6
PBS = phosphate saline buffer pH 7.4
150
100
50
0
160
ABS@ (655nm)
2h
OVERNIGHT
140
2
h
120
100
Fig. 3. Coating Time Study
 overnight black: no coating, only blocking and AFM1-HRP
 2 h black: no coating, only blocking and AFM1-HRP
80
60
40
20
0
0
1
2
3
4
5
6
Coating Curve
ABS @ 655 nm
0,25
0,20
Fig. 4. Coating Curve
0,15
Black: no coating
0,10
Blank
0,05
1
10
[MAb] µg/mL
The binding study was carried out adding several dilutions of AFM1–HRP
conjugate into the microtiter plate. The 80% of the curve answer supplies the labelled
antigen to use in the competition study to have the maximal sensitivity and the 1:30 v/v
dilution of this was chosen (fig. 5).
The competition study was carried out in two ways: in one where free and
labelled AFM1 were mixed together directly on the support where the specific antibody
was immobilised, while, in the second case, different dilutions of free toxin were added
on the support before the addition of the labelled antigen. In this second case we tried to
promote the binding of the AFM1 with its antibody to increase the sensitivity of the
analysis. Different intervals (10, 30, 60 min) of the incubation time of the free antigen
were studied and better results were archived when the antigen was added 10 min before
the labelled AFM1 (data not showed). In both cases the total time of the competition
step was 2 h.
Standard curves were obtained using AFM1 standard solutions prepared in PBS (0 –
480 pg/mL) for spectrophotometric ELISA.
Calibration curves (absorbance at 655 nm vs. competitor concentration) were fitted
using “non-linear 4 parameter logistic calibration plots” [i].
The four parameter logistic function is:
f(x) = [a – d]/ [1 + (x/c)b] + d.
The parameter a and d are the asymptotic maximum and minimum values,
respectively; c is the value at the inflection point (IC50) and b is the slope.
7
To allow a direct comparison of some calibration curves, absorbance values were
converted into their corresponding test inhibition values (% A/A0) as follows:
% A/A0 = 100 x (A - Asat)/(A0 – Asat)
where A is the absorbance value of competitors, Asat and A0 are the absorbance values
corresponding to the saturating analyte and to the non-competition concentration
respectively (as evaluated by the four-parameter logistic function).
The detection limit was defined as the concentration of AFM1 equivalent to the three
times the value of the standard deviations (s) measured in the absence of AFM1 (mean
value – 3 s). The midpoint value (IC50) was evaluated as the concentration of AFM1 at
50% A/A0. The working range was evaluated as the toxin concentration that gives test
inhibition values of 90 % and 10% of A/A0.
The data obtained for each curve were plotted and fitted using a SigmaPlot software
(SPSS). Regression analysis on the linear portion of the sigmoidal curves was also
performed.
Results showed different working ranges:
-
in the first case, the working range was 0.2 – 7.4 ng/mL (ppb);
-
in the second case, with a free AFM1 incubation time, was 30 -160 pg/mL (ppt).
Only in the second case the maximal level (50 ppt) of AFM1 was detectable.
Then we report the optimised final protocol of spectrophotometric competitive ELISA:
the wells were coated with 200 µL of 10 µg/mL of Anti-IgG (mouse) solution in
100mM carbonate buffer pH 9.6 overnight at 4°C. After rising with 250 µL of PBS-T
(PBS + 0.01% Tween 20), the wells were incubated with 100 µL of primary antibody (5
µg/mL of monoclonal anti-AFM1 in PBS). After washing, the competition step was
carried out as follow: several AFM1 standard solution were let to react into the wells for
10 min and then mixed with the AFM1-HRP conjugate at constant concentration (1:15
v/v). The competition reaction was allowed to proceed for 2 h at room temperature at
dark. The microtiter plate was, then, rinsed with PBS-T and, finally the
chromogen/substrate solution was added to the wells and the enzymatic reaction was
stopped after 30 min at room temperature. Absorbance was read at 655 nm.
8
120
100
% A/A0
% A/A
0
Current
(µA)
80
Current (µA)
80
100
60
40
20
60
0
100
Fig. 6. Competition curve for directly
mixing of free and labelled AFM1
101
[AFM1] ng/mL
40
20
0
10-1
100
101
102
[AFM1] ng/mL
120
120
100
100
ABS 655nm
80
60
40
100 x A/A0
20
100 x A/A0
100
80
80
60
40
20
0
0
60
-20
Fig. 7. Competition curve with adding of free toxin
1h before the AFM1-HRP conjugate.
100
-20
[AFM1] pg/mL
100
[AFM1] pg/ml
40
20
0
101
102
103
[AFM1] pg/ml
Sample Analysis
The determination of AFM1 in milk has been carried out with spiking samples
(before or after extraction), in order to determine the matrix effect and the efficiency of
the extraction.
The preparation of the sample consisted in a centrifugation for defatting for 15 min at
6000 rpm. After centrifugation the two phases were completely separated into layers of
fat cream and skimmed milk from top to bottom, respectively. The defatted supernatant
was recovered and tested directly.
In a preliminary study, the matrix effect of blank samples on both ELISA formats
was tested. Centrifuged blank milks were fortified with AFM1 standard solutions (8 –
960 pg/mL) and used directly in plate (fig. 8).
Results showed a good working range when the AFM1 in milk was added 60 min
before the conjugated toxin, comparable with that obtained in buffer (addition of AFM1
9
10 min before the labelled toxin), also the detection limit was lower the maximum level
allowed, which is 50 pg/mL (ppt) (fig. 9).
competition in milk
blank in milk
competition in buffer
blank in buffer
100
% A/A0
80
Fig. 9. Matrix Effect: comparison between
competition curve in buffer and competition curve
in milk without dilution.
60
40
20
0
101
102
103
[AFM1] ppt
120
100 x A/A0
120
100
80
60
40
80
% A/A0
100
20
0
60
102
Fig. 9. Competition curve in milk
[AFM1] pg/mL
40
20
0
101
102
103
[AFM1] pg/mL
In order to evaluate the extraction efficiency, calibration curves in milk, prepared
spiking a blank tissue with AFM1 standard solutions before extraction, were performed.
Results showed a good working range. The extraction was carried out as reported in the
sample extraction procedure. Each experiment was performed in triplicate and the mean
of each value was used for curve fitting. Results showed a working range of 24 – 218
pg/mL, comparable with the results obtained in buffer (fig. 9) and a high extraction
efficiency (tab. 1).
10
Tab. 1
Added AFM1
(pg/ml)
Found AFM1
(pg/ml)
30
60
240
720
27
55
230
720
RDS % % Recovery
(n=6)
4
6
6
8
90
92
97
100
The optimised direct ELISA and the commercial available Ridascreen kit, were
compared. The latter is based on a competitive immunoassay for quantitative analysis of
AFM1 in milk. The microtitre plate with 48 wells is coated with antibodies directed
against AFM1. The mean lower detection limit of the kit was approximately 10 pg/mL
(ppt), comparable with our results. Compared with the kit method our assay gave
similar results but with the advantages of simpler (no washing step after the addition of
the sample).
Task 2.2 Screen printed electrode evaluation
Subtask 2.2.2 months 18-21
Subtask 2.2.3 months 21-24
Electrochemical ELISA
An amperometric immunosensor for the determination of AFM1 in milk has been
constructed and based on a screen-printed electrode (SPE), coated with a monoclonal
antibody, using the same competitive assay developed for the spectrophotometric
system. In our study, we used the SPE, produced by thick film technology, for the massproduction of disposable electrochemical sensors having low cost, small size, and good
reproducibility. The system comprises graphite working electrode, a graphite and a
silver electrode as auxiliary and reference respectively. Reagents were added in small
volumes as a drop of sample onto the electrode surface.
11
The electrochemical technique, chosen for the detection of the enzymatic product, is
the chronoamperometry, performed at an applied potential of –100 mV, where TMB
product undergoes reduction.
The operative conditions, such as the amount of antibodies and labelled antigen,
buffer and pH, length of time and temperature of each steps were evaluated and
optimised experimentally.
Assay optimisation
Milk study
To establish the way, in which milk affects the behavior of the immunosensor, its
response was studied by cyclic voltammetry from 600 mV to – 400 mV. In fig. 10 the
cyclic voltammograms of milk, centrifuged milk (see “Analysis of sample” in
spectrophotometric part) and buffer were compared. An electrochemical answer from
non-identified compound of milk around -150 mV was observed, while it was absence
when we worked with centrifuged milk. Also we observed that, in this second case, the
electrochemical behavior of the milk was comparable with that of buffer and chosen this
treatment of the milk sample for our analysis.
Also the study for the determination of AFM1 in milk has been based on the
characterization of the enzymatic product in presence of the milk in order to evaluated
of the matrix effect. In fig. 11 you can see the results obtained with the TMB oxidase
product before and after 30 min of incubation of the working electrode in milk. There
was not a big interference by the electrochemical compound present in milk, but we can
observe a reduction in the TMB signal. However the reduction peaks, useful for the
electrochemical reduction, are still present.
Fig. 10. Milk study: Cyclic voltammograms
of milk on SPE for the behaviour
study
--- Milk
--- Centrifuged milk
--- Buffer
12
0.100x10 -5
i/A
Fig. 11. Cyclic voltammograms of TMB-ox
before (pale blue) and after (red)
30 min incubation in milk
I/A
0.050x10 -5
0
-0.050x10 -5
-0.100x10 -5
-0.750
-0.500
-0.250
0
0.250
0.500
0.750
E/ V
Assay optimisation
The first step of the optimisation, as the spectrophotometric assay, was the study
of the concentration for the precoating step. Different concentrations (5, 10, 20, 40,
80g/mL) of the precoating (anti-IgG mouse) were tested in order to find the better
working condition. The experiment was carried out with the electrode surface covered
with a fixed concentration of primary antibody (against AFM1) and the AFMA1-HRP
conjugate. Results showed that a good orientation of the primary antibody and better
reproducibility was archived when a precoating step carried out overnight at 4°C with a
solution 10 g/mL of anti-IgG in carbonate buffer pH 9.6 (fig. 12).
0,9
No precoating
0,8
Precoating
Current (A)
0,7
Fig. 12. Precoating study
0,6
0,5
RSDprec = 4%, RSDnoprec = 16%
0,4
0,3
0,2
0,1
0,0
0
1
2
3
4
5
13
In order to decrease the non specific binding of the reagent added after the
coating step PVA 0.5, 1 and 2% and BSA 1, 2% were tested as blocking buffer (data not
showed). The lower non-specific binding was showed when the electrode surface was
covered with a solution of PVA 1% in PBS for 30 min at room temperature (data not
showed).
Next parameter to be optimised was the concentration of the primary antibody
(coating step). Several concentration of this were tested (50, 20, 10, 5 and 1 g/mL of
monoclonal anti-AFM1) and 20 g/mL was chosen for a good analysis sensitivity (fig.
13). The experiment was carried out with the electrode surface covered with a fixed
concentration of AFM1-HRP conjugate.
0
,
7
0
0
,
6
5
Current (µA)
0
,
6
0
0
,
5
5
Fig. 13. Coating curve
0
,
5
0
0
,
4
5
0
,
4
0
1
1
0
1
0
0
[
M
A
b
]
µ
g
/
m
L
The binding study had carried out adding several dilution of AFM1–HRP
conjugate into the working electrode and 1:20 v/v dilution of this was chosen, because
offered a good sensitivity for the competition analysis (fig. 14).
1
,
5
Fig. 14. Binding curve
Curent(µA)
1
,
0
0
,
5
0
,
0
0
,
0
1
0
,
1
[
A
F
M
1
H
R
P
]
v
/
v
1
14
The combination of this results (precoating, coating and binding concentration)
gave an adequate signal, with a signal/blank ratio, showing that the signal was due to
complementary binding between the anti-IgG-anti-AFM1 and AFM1-HRP complex.
The last parameter investigated was the condition to use in competition step for having
a good sensitivity.
The competition study had carried out in two way (fig. 15, 16), as
spectrophometric assay (see before). In the case in which the free antigen was added
before, different incubation time (5, 10, 15 min) were studied and the better results was
carried out when the antigen was added 5 min before the labelled AFM1. In both cases
total time of the competition step was 50 min.
Standard curves were obtained using AFM1 standard solutions prepared in PBS for
electrochemical ELISA
Calibration curves (current vs. competitor concentration) were fitted using “nonlinear 4 parameter logistic calibration plots” [1].
Results showed different working ranges:
-
in the first case, the working range was 1 - 10 ng/mL (ppb)
-
in the second case, with a free AFM1 incubation time, was 40 - 250 pg/mL
(ppt).
Then we report the preliminary protocol of electrochemical competitive ELISA:
the working electrodes of SPEs were coated with 7 µL of 10 µg/mL of Anti-IgG
(mouse) solution in 50 mM carbonate buffer pH 9.6 overnight at 4°C. After rising with
100 µL of PBS-T (PBS + 0.01% Tween 20), the SPEs were incubated with 10 µL of 1%
PVA in PBS, as blocking solution, for 30 min at room temperature primary antibody (5
µg/mL of monoclonal anti-AFM1 in PBS). After washing, the competition step was
carried out as follow: several AFM1 standard solution were let to react into the wells for
10 min and then mixed with the AFM1-HRP conjugate at constant concentration (1:15
v/v). The competition reaction was allowed to proceed for 2 h at room temperature at
dark. The microtitre plate was, then, rinsed with PBS-T and, finally the
chromogen/substrate solution was added to the wells and the enzymatic reaction was
stopped after 30 min at room temperature.
15
120
120
Current (µA)
110
Current (µA)
100
100
90
80
70
80
60
1
10
[AFM1] ng/mL
60
40
Fig. 15. Competition curve for directly
mixing of free and labelled AFM1
20
0
0,1
1
10
[AFM1] ng/mL
120
100
80
100 x I/I0
100
100 x I/I0
80
60
40
20
60
0
102
40
[AFM1] pg/mL
Fig. 16. Competition curve with adding of
free toxin 1h before the AFM1HRP conjugate.
20
0
101
102
103
[AFM1] pg/mL
The first year of activity has been quite successful. We reached all the objectives
proposed and we are now pursuing a major milestone. He enhance the sensitivity of our
system and to make our system working in the parlour in collaboration with DeLaval
and SILSOE research.
1. Brian Law (1996) (ed) Taylor &Francis Ltd (UK) 160
16