Biacore biosensor immunoassay for 4-nonylphenols: Assay optimization and applicability for shellfish analysis

Article abstract of DOI:10.1016/j.chemosphere.2004.07.028

A rapid Biacore biosensor immunoassay of 4-nonylphenols was developed. Two types of antibodies were used in the study: polyclonal antibodies with high cross-reactivity towards technical 4-nonylphenol and a monoclonal antibody very specific to 4-n-nonylphenol. 9-(p-Hydroxyphenyl)nonanoic acid was immobilized onto surface of a sensor chip. The best assay sensitivity was achieved using a flow rate of 50 μl min^-1 and injection time of 2 min. For the assay incorporating monoclonal antibodies a limit of detection 2 ng ml ^-1 for 4-n-nonylphenol was achieved. With polyclonal antibodies one order lower sensitivity was observed for 4-nonylphenols. High background level of calibration curve for technical 4-nonylphenol was decreased by using IgG fraction of polyclonal antibodies in combination with lower amount of immobilised 9-(p-hydroxyphenyl)nonanoic acid. Sensitivity of the assay was improved by using a chip with a new derivative on a surface-N-aminobutyl [2-(4-hydroxyphenyl)ethylamine] (limit of detection - 5 ng ml^-1). Applicability of the developed assays to ecological monitoring was checked in experiments using shellfish samples. 4-n-Nonylphenol from spiked samples was extracted into hexane followed by clean-up on NH2 SPE columns. Calibration curves generated for cockles, mussels and oyster samples were identical (limit of detection about 10 ng g^-1) whereas for scallop samples a slight decrease (about 5-10%) of absolute response was observed. In the assay using the monoclonal antibody specific to 4-n-nonylphenol 31 shellfish samples were found to be negative. Results obtained with polyclonal antibodies indicated that two scallop samples contained a quantity of 4-nonylphenols. The developed biosensor assay could be applied for shellfish analysis as a preliminary screening method.

Full text of DOI:10.1016/j.chemosphere.2004.07.028

Chemosphere 57 (2004) 975–985
www.elsevier.com/locate/chemosphere
Biacore biosensor immunoassay for 4-nonylphenols:
assay optimization and applicability for shellfish analysis
a,
a
b
Jeanne V. Samsonova *, Natalya A. Uskova , Alexey N. Andresyuk ,
c
Milan Franek , Christopher T. Elliott
d
a
Department of Chemical Enzymology, Chemistry Faculty, M.V. Lomonosov Moscow State University, 119992 Moscow, Russia
b
‘‘Chimmed’’, Kashirskoe shosse 9, Bld. 3, 115230 Moscow, Russia
Department of Biotechnology, Veterinary Research Institute, Hudcova 70, 621 32 Brno, Czech Republic
c
d
Department of Veterinary Science, QueenÕs University of Belfast, Stoney Road, Stormont, BT4 3SD Belfast, UK
Received 23 September 2003; received in revised form 4 June 2004; accepted 13 July 2004
Abstract
A rapid Biacore biosensor immunoassay of 4-nonylphenols was developed. Two types of antibodies were used in the
study: polyclonal antibodies with high cross-reactivity towards technical 4-nonylphenol and a monoclonal antibody
very specific to 4-n-nonylphenol. 9-(p-Hydroxyphenyl)nonanoic acid was immobilized onto surface of a sensor chip.
The best assay sensitivity was achieved using a flow rate of 50 ll minÀ1 and injection time of 2 min. For the assay incor-
porating monoclonal antibodies a limit of detection 2 ng mlÀ1 for 4-n-nonylphenol was achieved. With polyclonal anti-
bodies one order lower sensitivity was observed for 4-nonylphenols. High background level of calibration curve for
technical 4-nonylphenol was decreased by using IgG fraction of polyclonal antibodies in combination with lower
amount of immobilised 9-(p-hydroxyphenyl)nonanoic acid. Sensitivity of the assay was improved by using a chip with
a new derivative on a surface-N-aminobutyl [2-(4-hydroxyphenyl)ethylamine] (limit of detection––5 ng mlÀ1). Applica-
bility of the developed assays to ecological monitoring was checked in experiments using shellfish samples. 4-n-Nonyl-
phenol from spiked samples was extracted into hexane followed by clean-up on NH2 SPE columns. Calibration curves
generated for cockles, mussels and oyster samples were identical (limit of detection about 10 ng gÀ1) whereas for scallop
samples a slight decrease (about 5–10%) of absolute response was observed. In the assay using the monoclonal antibody
specific to 4-n-nonylphenol 31 shellfish samples were found to be negative. Results obtained with polyclonal antibodies
indicated that two scallop samples contained a quantity of 4-nonylphenols. The developed biosensor assay could be
applied for shellfish analysis as a preliminary screening method.
ꢀ 2004 Elsevier Ltd. All rights reserved.
Keywords: 4-nonylphenols; Endocrine disruptors; Biacore biosensor; Shellfish
1. Introduction
Non-ionic synthetic detergents––polyethoxylated
alkylphenols––are widely used in different domestic
Corresponding author. Tel.: +7 095 9393407; fax: +7 095
*
and industrial products. Huge amounts of these
detergents are directly released into the environment
9392742.
E-mail address: jvs@enz.chem.msu.ru (J.V. Samsonova).
0045-6535/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.chemosphere.2004.07.028

976
J.V. Samsonova et al. / Chemosphere 57 (2004) 975–985
where they degrade to more toxic and persistent metab-
olites than the parent compounds (Ying et al., 2002).
Alkylphenols such as nonylphenol (NP) and octylphenol
(OP) are the most dangerous metabolites because of
theirs enhanced resistance towards biodegradation, toxi-
city, strong estrogenic effects (Fig. 1). They can mimic
natural hormones and interact with the estrogen recep-
tor causing different adverse effects such as feminisation
in fish (Schwaiger et al., 2002). There are a large number
of investigations reported with regard to the toxic and
estrogenic properties of 4-NP. Alkylphenols occurrence
in the environment is quite common, e.g., 4-NP residues
were discovered in water sources, sediments and aquatic
organisms even far away from a cost (Tsuda et al.,
2000b; Bester et al., 2001; Ferrara et al., 2001; Guenther
et al., 2001; Petrovich et al., 2001). Recent studies
showed that 4-NP residues are ubiquitous in food
(Guenter et al., 2002). It was also reported about migra-
tion of 4-NP from polyvinyl chloride food packaging
films into food (Inoue et al., 2001). These findings have
raised increasing concern of 4-NP impact on the envi-
ronment and human health.
properties of biological recognising molecules such as
antibodies, receptors, etc., with sensitive signal transfor-
mation allowed to develop faster and simpler methods
which found application in different fields (Nakamura
and Karube, 2003).
Recently polyclonal and monoclonal antibodies with
a range of different specificities to alkylphenols were pro-
duced (Franek et al., 2001; Zeravik et al., 2004). The
polyclonal antibodies were used for establishment of
an indirect ELISA, polarisation fluoroimmunoassay
and flow injection immunoassays for 4-NPs operating
in the mg lÀ1 range (Franek et al., 2001). In the recently
published study a more sensitive direct ELISA was
developed (Zeravik et al., 2004). The polyclonal anti-
bodies and new monoclonal antibody showed IC50 val-
ues around 40 lg lÀ1 for technical 4-NP and 11.5
lg lÀ1 for the linear isomer 4-n-NP. Cross-reactivity pat-
tern of the antibodies was thoroughly investigated. Two
polyclonal antibodies showed a strong tendency to rec-
ognise a total amount of 4-alkylphenols whereas other
two polyclonal antibodies and monoclonal antibody
allowed sensitive detection of linear forms of 4-alkyl-
phenols (Zeravik et al., 2004). Very sensitive chemilumi-
nescent ELISA and simple dipstick test for 4-n-NP were
also recently developed (Samsonova et al., 2003).
The purpose of the present study was to develop a
rapid and sensitive biosensor method for 4-NPs determi-
nation. For the assay development two 4-NP derivatives
for immobilization on a surface of a sensor chip and two
types of antibodies with different cross-reactivity pattern
were used. Applicability of the developed assays for
shellfish analysis was investigated.
For detection of nonylphenol ethoxylates and 4-NP
in the environmental and food samples, the chromato-
graphic methods such as gas, liquid or high pressure liq-
uid chromatography are used (Ferrara et al., 2001;
Guenther et al., 2001; Corsi and Focardi, 2002). How-
ever, simple and quick analytical methods are required
to detect 4-NP presence in the environment. Chromato-
graphic methods are usually time-consuming and de-
mand sophisticated equipment. Immunoassays based
on very specific antibody-antigen interaction are usually
cheaper and more sensitive then chromatographic
methods and are very suitable for use in routine moni-
toring programmes. Biosensor techniques combined
2. Materials and methods
2.1. Materials
HO(
O)n
C9H19
4-n-Nonylphenol (99.7% purity) was purchased from
Supelco (Bellefonte, PA, USA); technical 4-nonylphenol,
4-(tert-octyl)phenol (97%), glycidyl 4-nonylphenol ether
were supplied by Sigma–Aldrich (Milwaukee, USA), 3-
(p-hydroxyphenyl) propionic acid and 2-(4-hydroxyphe-
nyl)ethylammonium chloride were from Merck
(Hohenbrunn, Germany). Ready-to-use HBS-EP buffer
(0.01 M HEPES, 0.15 M NaCl, 3 mM EDTA, 0.005%
polysorbate 20 (v/v), pH 7.4), CM5 sensor chips (re-
search grade), Amine Coupling Kit were obtained from
Biacore AB (Uppsala, Sweden). Bond Elute NH2 SPE
columns (500 mg/3 ml) were supplied by Jones Chroma-
tography (Hengoed, UK). All other chemicals and rea-
gents were purchased from Sigma (Poole, Dorset, UK).
Stock solutions of alkylphenols (1 mg mlÀ1) were
prepared in methanol and stored at À20 ꢁC. Standard
solutions in concentration range 0–10 lg mlÀ1 were
prepared in HBS-EP buffer and stored at 4 ꢁC.
n = 1-22
nonylphenol ethoxylates
HO
4-n-nonylphenol
HO
4-tert-octylphenol
HO
COOH
9-(p-hydroxyphenyl)nonanoic acid
Fig. 1. Structure of nonylphenols ethoxylates, alkylphenols and
9-(p-hydroxyphenyl)nonanoic acid.

J.V. Samsonova et al. / Chemosphere 57 (2004) 975–985
977
Production of monoclonal and polyclonal antibodies
was described earlier (Franek et al., 2001; Zeravik et al.,
2004).
was heated with occasional shaking for 12 h. The mate-
rial which crystallised from supernatant on standing
was washed with ethanol and dried under vacuum
(m.p. = 80 ꢁC).
Synthesis of 9-(p-hydroxyphenyl)nonanoic acid was
described by Mikhura et al. (2000).
An optical biosensor BIACORE Q was obtained
from Biacore AB (Uppsala, Sweden) and operated by
BIACORE 3.0 control software.
2.2.2. Synthesis of N-phthalimidobutyl-2-(4-hydroxyphe-
nyl) ethylamine
2-(4-hydroxyphenyl)ethylammonium chloride (1.0 g,
5.8 mmol) was neutralised in solution 0.61 g (5.8 mmol)
NaHCO3 in 10 ml water for 30 min under stirring. Neu-
tralised amine precipitated and water was evaporated
under vacuum. A mixture of the neutralised amine and
1.95 g (6.9 mmol, 20% access) N-(4-bromobutyl)-pht-
halimide in 10 ml dry acetonitrile was refluxed for 5 h.
Thin layer chromatography (TLC) in the solvent system
petroleum ether/methanol/ethyl acetate (30:20:50, v/v)
2.2. Synthesis of N-aminobutyl [2-(4-hydroxyphenyl)eth-
ylamine] (Fig. 2)
2.2.1. Synthesis of N-(4-bromobutyl)-phthalimide was
performed as described by Mizzoni et al. (1954)
Briefly, a mixture of 65 g (0.35 mol) potassium pht-
halimide and 200 g (0.93 mol) tetramethylene dibromide
O
Br
NK
Br
+
O
I
O
N
O
Br
KBr
+
O
Br
HO
N
+
NH2 HCl
O
II
O
N
HO
N
H
O
III
NH2
HO
N
H
N-aminobutyl [2-(4-hydroxyphenyl)ethylamine]
Fig. 2. Synthesis of N-aminobutyl [2-(4-hydroxyphenyl)ethylamine].

978
J.V. Samsonova et al. / Chemosphere 57 (2004) 975–985
indicated the appearance of the products with Rf = 0.3
which slightly reacted with ninhydrin. The reaction mix-
ture was fractioned by column chromatography. A glass
column (450 · 35 mm) was filled with silica gel using the
same solvent system as a mobile phase. Elution was
accomplished by gravity. After solvent evaporation
0.61 g of cream crystalline substance (N-phthalimidobu-
thyl-2-(4-hydroxyphenyl) ethylamine) was obtained
(m.p. = 140 ꢁC).
onto chip surface and incubated for 2 h. After incuba-
tion the solution was removed, the chip was washed with
HBS-EP buffer and dried under a stream of nitrogen.
When not in use the chips were stored at 4 ꢁC in plastic
tubes containing silica gel desiccant.
2.4. Immobilisation of N-aminobutyl [2-(4-hydroxyphe-
nyl)ethylamine] onto a sensor chip surface
A sensor chip was equilibrated to room temperature
and the chip surface was activated as described in the pre-
vious section. After incubation the chip was washed with
HBS-EP buffer and dried under a stream of nitrogen. The
solution of 2 mg N-aminobutyl [2-(4-hydroxyphenyl)eth-
ylamine] in 1 ml DMF was prepared and 50 ll was ap-
plied onto chip surface and incubated for 2 h. After
incubation the solution was removed, the chip was
washed with HBS-EP buffer and dried under a stream
of nitrogen. When not in use the chips were stored at 4
ꢁC in plastic tubes containing silica gel desiccant.
2.2.3. Synthesis of N-aminobutyl [2-(4-hydroxy-
phenyl)ethylamine]
N-phthalimidobuthyl-2-(4-hydroxyphenyl) ethylam-
ine was dissolved in 15 ml absolute ethanol then 160 ll
hydrazine hydrate was added and the reaction mixture
was refluxed for 2 h. Then 8 ml 1 N hydrochloric acid
was added and the reaction mixture was refluxed for an-
other hour. The mixture was cooled for 1 h at À20 ꢁC and
precipitate was discarded. Supernatant was adjusted to
pH 9 with 2 N sodium hydroxide and solvent was evap-
orated under vacuum. TLC in the solvent system petro-
leum ether/methanol/ethyl acetate (20:50:30, v/v)
indicated the appearance of the spot corresponding to
the product (amine with long spacer) with Rf = 0.1. After
preparative TLC 0.16 g pale cream powder was obtained.
N-aminobutyl [2-(4-hydroxyphenyl)ethylamine] identity
was confirmed by NMR. 1H NMR (DMSO-d6): d
1.60–1.70 (m, 4H, 2 · CH2), 2.75–3.00 (m, 8H, 4 ·
CH2), 6.71 (d, 2H, aromatic), 7.05 (d, 2H, aromatic),
9.40 (s, 1H, OH). 13C NMR (DMSO-d6): d 22.42
(CH2), 24.04 (CH2), 30.62 (CH2), 38.68 (CH2), 45.77
(CH2), 48.03 (CH2), 115.40 (2 · CH, aromatic), 127.22
(C, aromatic), 129.44 (2 · CH, aromatic), 156.21
(Cquat-OH, aromatic).
2.5. Extraction of 4-NPs from shellfish
Shellfish samples were homogenised (whole body)
and weighed (1.0 0.1) g into glass bottles. 3 ml of hex-
ane was added to each sample and bottles were mixed on
a roller mixer for 30 min followed by centrifugation at
3000g for 10 min at 10 ꢁC. The solutions were applied
onto NH2 SPE columns previously preconditioned by
3 ml of hexane. Columns were washed with 3 ml of hex-
ane and 4-NPs were eluted using 3 ml of diethyl ether.
The eluates were evaporated to dryness under a stream
of nitrogen at 40 ꢁC. After addition of 200 ll HBS-EP
buffer and vortex mixing for 30 s solutions were trans-
ferred to glass tubes.
MS m/z (%): 209(100), 192(26), 121(3), 107(12). Melt-
ing point: >300 ꢁC.
2.6. Biosensor procedure
2.3. Immobilisation of 9-(p-hydroxyphenyl)nonanoic acid
onto a sensor chip surface
Mixing of all the reagents, injection and washing
were performed automatically by instrument in accord-
ance with the chosen programme. An aliquot of anti-
bodies solution in HBS-EP buffer was mixed with
standard solution (or extract diluted 1:2 with HBS-EP
buffer) in the ration 1:1 (v/v) (or 4:1 (v/v) for extracts).
This mixture was injected onto the chip surface at a flow
rate of 50 ll minÀ1 for 2 min. Report points were taken
on each sensogram produced before and after sample
injection. Chip surface was regenerated by 30 s pulse
of regeneration solution (100 mM sodium hydroxide
for monoclonal antibodies and 100 mM sodium hydrox-
ide with 10% acetonitrile for polyclonal antibodies). A
typical analytical cycle for each sample took about 10
min to complete. Each calibration point or sample was
analysed in duplicate.
A sensor chip was equilibrated to room temperature.
The chip surface was activated by 30 min incubation
with 50 ll of a 1:1 (v/v) mixture of 0.05M N-hydroxysuc-
cinimide (NHS) and 0.2 M 1-ethyl-3-(3-dimethylamino-
propyl) carbodiimide hydrochloride (EDC). The
solution was removed then 50 ll of 1 M ethylenediamine
(pH 8.5) added and incubated for 1 h. The chip was
washed with HBS-EP buffer (3 · 50 ll). Free carboxyl
groups were blocked with 50 ll 1 M ethanolamine
hydrochloride (pH 8.5) for 20 min. The chip was washed
with 50 ll of HBS-EP buffer. A solution of 5 mg EDC
and 2 mg of NHS in 400 ll 10 mM sodium acetate
(pH 4.5) was added dropwise to the solution of 2 mg
9-(p-hydroxyphenyl)nonanoic acid in 600 ll dimethyl-
formamide (DMF). The mixture (50 ll) was applied

J.V. Samsonova et al. / Chemosphere 57 (2004) 975–985
979
Table 1
Binding of excess of antibodies with different flow cells of 9-
HPNA chips
3. Results and discussion
3.1. Antibodies
Chip (%)
Binding with
monAb (RU)
Binding with
polAb (RU)
In the present study two types of antibodies were used
(Zeravik et al., 2004). In direct ELISA the monoclonal
antibody was found to be very sensitive to linear forms
of 4-alkylphenols and 4-n-NP in particular with negligi-
ble cross-reactivity to technical 4-NP (0.9%). Characteri-
sation of the polyclonal antibodies produced (Franek
et al., 2001) showed a strong tendency to recognise the
total amount of 4-alkylphenols (cross-reactivity: 4-n-
NP, 100%; technical 4-NP, 65.7%; 4-tert-OP, 38.2%).
Technical NP is comprised of a mixture of ring and
branch isomers with 4-NP isomers as a major component
(Wheeler et al., 1997). This makes it difficult to identify
and determine individual compounds in a mixture. In
chromatographic assays all resolved peaks corresponding
to different isomers are usually summed without identifi-
cation to give a total 4-NP concentration (Guenther et al.,
2001). Variability of NP mixture also makes it difficult to
produce either specific or generic antibodies.
9-HPNA 100
9-HPNA 50
9-HPNA 20
9-HPNA 10
3000–4000
1000–1500
500–600
5500–6000
2500–3300
1000–1200
700–760
150–250
Conditions: Monoclonal antibody concentration, 20 lg mlÀ1
polyclonal antibodies dilution, 1/10; flow rate, 5 ll/min; injec-
tion time, 5 min; RU, relative units.
;
surface could have had a significant impact on the assay
performance. Calibration curves with similar sensitivÀit1y
were obtained for flow rates from 25 to 100 ll minÀ1
In contrast, the lower flow rate selected (10 ll min
.
)
produced the lower sensitivity and the highest back-
ground level (Fig. 3). Finally, the following assay condi-
tions were selected for further study: injection time 2
min, flow rate 50 ll minÀ1 and regeneration for 30 s
using 100 mM sodium hydroxide. Technical 4-NP and
4-tert-OP showed no inhibition up to 1000 ng mlÀ1 con-
centrations. For the assay a limit of detection of 2
ng mlÀ1 and IC50 of 15 ng mlÀ1 for 4-n-NP were
achieved.
All antibodies used in the present study were ob-
tained against 9-(p-hydroxyphenyl)nonanoic acid-pro-
tein conjugate (Franek et al., 2001). Using derivatives
with shorter alkyl chains as immunogens did not result
in production of antibodies with generic specificity to
4-NPs. It was also shown that only homologous system
provided the most sensitive assays in indirect ELISA
format and the importance of long lateral chain was pro-
ven. Similar findings were obtained using a direct
ELISA (Zeravik et al., 2004).
4-n-NP and the mixture of branch isomers (technical
4-NP) as target substances were used for assay evalua-
tion with polyclonal antibodies. Similar results (when
compared to monoclonal data) were obtained (data
not shown). The same assay conditions provided a sen-
sitive assay however a stronger regeneration solution
was used to accomplish full chip surface regeneration
(100 mM sodium hydroxide with 10% acetonitrile).
For both target substances similar cross-reactivity was
obtained (Fig. 4). However high background level was
observed for technical 4-NP calibration curve. The
same effect was noted for 4-tert-OP (Fig. 4). The sensi-
tivity of the biosensor assay for both substances (limit
of detection 30 ng mlÀ1) was better then in indirect
ELISA (Franek et al., 2001) but was one order higher
compared to monoclonal antibody biosensor assay
and direct ELISA results (Samsonova et al., 2003; Zera-
vik et al., 2004).
3.2. Biosensor assay in buffer
3.2.1. The chip production and generation of calibration
curves
9-(p-hydroxyphenyl)nonanoic acid (9-HPNA, Fig. 1)
was used for immobilisation onto surface of a sensor
chip. Both monoclonal and polyclonal antibodies
showed good binding with the 9-HPNA chip (Table 1).
For both types of antibodies calibration curves in buffer
were established. For monoclonal antibody and 4-n-
nonylphenol as a target substance the influence of flow
rate and injection time on assay sensitivity was investi-
gated in detail. Each minute of antibody/analyte mixture
injection gave in the region of 250 units of maximum rel-
ative response. This enabled a reduction in antibody
concentration but did not improve assay sensitivity.
An injection time of 2 min was chosen for further devel-
opmental work. The flow rates selected were varied from
10 to 100 ll minÀ1. The higher the flow rate was used the
higher the maximum relative response was obtained
(Fig. 3). Surprisingly a quite high flow rate of 100
ll minÀ1 still provided sensitive assays despite the fact
that at such flow rates dissociation of antibody from a
3.2.2. Assay optimisation: chip surface and purity of
binding protein
From a very practical point of view an assay capable
of determining the total amount of 4-NPs is required.
Only the polyclonal antibodies tested showed the re-
quired generic specificity to the 4-NP group of com-
pounds. The calibration curve obtained during the
preliminary assay development work was insufficiently
sensitive and further investigations were undertaken to
improve the assay characteristics.

980
J.V. Samsonova et al. / Chemosphere 57 (2004) 975–985
220
200
180
160
140
120
100
80
-1
-1
-1
10 µl min
25 µl min
50 µl min
-1
100 µl min
60
40
20
0
1
10
100
1000
[4-n-nonylphenol], ng ml-1
Fig. 3. Influence of flow rate on 4-n-nonylphenol biosensor assay with monoclonal antibody. Monoclonal antibody concentration 0.75
lg mlÀ1, chip 9-HPNA 100%.
220
200
180
160
140
120
4-n-nonylphenol
technical 4-nonylphenol
4-tert-octylphenol
100
80
60
40
10
100
1000
10000
[alkylphenol], ng ml-1
Fig. 4. Calibration curve for different alkylphenols in biosensor assay with polyclonal antibodies. Polyclonal antibodies dilution
1/2000, chip 9-HPNA 100%.
During Biacore biosensor assay development it is
common practise to immobilise the maximum possible
amount of derivative onto a chip surface. Such surfaces
provide the highest binding characteristics with antibod-
ies and increase the chip surface robustness (stability). It
was postulated that analogously to indirect ELISA for
haptens the amount of immobilised derivative could
have an influence on assay performance. To study this
possibility a range of sensor chips with differing amounts
of immobilized 9-HPNA were prepared (50%, 20% and
10%). These surfaces were compared to the initial chip
(100% immobilisation) with regards binding of mono-
clonal and polyclonal antibodies (Table 1). All chip sur-
faces were prepared using the same immobilization
conditions. Solutions for immobilization were obtained
by serial dilutions of the same 9-HPNA stock solution.

J.V. Samsonova et al. / Chemosphere 57 (2004) 975–985
981
100
It could be seen that the antibody binding pattern cor-
mAb - 4-n-NP
responded to an amount of the derivative used for immo-
bilization (Table 1). For all chips calibration curves were
obtained and compared. Calibration curves for 4-n-NP
with monoclonal antibody and chips 9-HPNA (100%)
and 9-HPNA (50%) were almost identical (Fig. 5(a)).
Whereas for polyclonal antibodies the reduced amount
of immobilised 9-HPNA helped to decrease the back-
ground level observed in the calibration curve for 4-n-
NP and technical 4-NP and thus improve sensitivity
for 4-NP (Fig. 5(b) and (c)). For chip 9-HPNA (20%) as-
say sensitivity was low in both cases resulting in higher
background level for technical 4-NP. Reduction in the
amount of the derivative on a chip surface led to the
requirement for high working concentrations of antibod-
ies (dilution 1/100) to get a sufficient analytical signal.
Therefore for this chip surface non-specific binding had
a detrimental impact on assay performance.
(a)
(b)
(c)
80
60
40
20
0
chip 9-HPNA 100%
chip 9-HPNA 50%
chip 9-HPNA 20%
11
0
100
[4-n-nonylphenol], ng ml-1
100
80
60
40
20
0
pAb - 4-n-NP
The problem of high background levels of calibra-
tion curves for technical 4-NP was overcome by using
IgG fraction of polyclonal antibodies (Fig. 5(c)). This
finding clearly indicated a high level of non-specific
binding of serum components was occurring on the
chip surface.
chip 9-HPNA 100%
chip 9-HPNA 50%
chip 9-HPNA 20%
In general chips with immobilised 9-HPNA were
highly stable and the greater the amount of 9-HPNA
which was presented on a surface the longer the time
the chip could be used. For example, 500 injections re-
sulted in loss of about 20% surface activity of 9-HPNA
chip. Therefore a single chip (4 lanes per chip) had the
capability to be used for approximately 2000 injections.
In conclusion, among all produced surfaces the chip
9-HPNA (50%) showed the best assay sensitivity for
all combinations of antibodies and target analytes
(Fig. 5). This result shows that particular attention
should be given to the quantity of analyte immobilised
on a surface to gain a fully optimised assay system. High
background level of calibration curve for technical mix-
ture of 4-NP can be decreased by using purified poly-
clonal antibodies in combination with lower amount of
immobilised 9-HPNA.
10
100
1000
10000
[4-n-nonylphenol], ng ml-1
100
pAb - 4NP
80
60
40
20
0
chip 9-HPNA 100%
chip 9-HPNA 50%
chip 9-HPNA 20%
chip 9-HPNA 50% - IgG
10
100
1000
10000
[4-nonylphenol], ng ml-1
3.2.3. Assay optimisation: standard solutions
Fig. 5. Calibration curve for different chips and target analytes
in biosensor assay. (a) Monoclonal antibody, 4-n-NP, (b)
polyclonal antibodies, 4-n-NP, (c) polyclonal antibodies, tech-
nical 4-NP. B/Bo, %, the ratio of the response at the measured
concentration of alkylphenol (B) to signal obtained at the
absence of the analyte (Bo).
Nonylphenols are lypophilic compounds with low
solubility in water (about 5 mg lÀ1) (Ahel and Giger,
1993; Brix et al., 2001). It was postulated that the sensi-
tivity of the assay for technical 4-NP could be improved
by the introduction of a small quantity of organic sol-
vent into the reaction mixture. Assays on the Biacore
system were found to be very tolerable to the presence
of methanol and ethanol (Fig. 6). At the ratio anti-
body:standard solution 1:1 (v/v) up to 20% methanol
or up to 5–10% ethanol could be introduced into a
standard solution without a significant decrease of the
maximum response. A range of calibration curves with
4-NP standard solutions contained different amount of
methanol were obtained and compared (Fig. 7). It was
found that the higher amount of methanol was intro-
duced the lower assay sensitivity was observed. A simi-
lar, unusual effect of decreasing ELISA sensitivity for
4-n-NP in the presence of methanol and dimethylsulf-
oxide has been reported earlier (Zeravik et al., 2004).

982
J.V. Samsonova et al. / Chemosphere 57 (2004) 975–985
350
300
250
200
150
100
50
MeOH
EtOH
0
0
20
40
60
80
100
[% of organic solvent]
Fig. 6. Influence of organic solvents on a signal at zero concentration of technical 4-NP. Dilution of IgG fraction of polyclonal
antibodies 1/500, chip 9-HPNA 50%.
350
300
250
200
150
buffer
1% MeOH
5% MeOH
100
10% MeOH
20% MeOH
50
0
10
100
1000
[4-nonylphenol], ng ml-1
Fig. 7. Influence of methanol content on calibration curves of technical 4-NP. Dilution of IgG fraction of polyclonal antibodies 1/500,
chip 9-HPNA 50%.
3.2.4. Design of a chip surface and improvement of the
assay sensitivity
(Zeravik et al., 2004) and it was assumed that introduc-
tion of this substance with shorter linear chain onto a
chip surface could improve assay sensitivity. However,
immobilisation of the compound did not give good re-
sults and low binding with polyclonal antibodies was
observed (about 250 RU).
One of the means by which biosensor assay sensitiv-
ity can be improved is to change the chemistry on the
chip surface. It is well known that heterologous system
usually provides better sensitivity compared to homolo-
gous formats. Polyclonal antibodies showed 11.1% of
cross-reactivity to 3-(p-hydroxyphenyl) propionic acid
An alternative means of generating a heterologous
system is to change the conjugation site. Commercially

J.V. Samsonova et al. / Chemosphere 57 (2004) 975–985
983
available glycidyl 4-nonylphenol ether containing a reac-
tive epoxy group on the hydroxyl group side was immo-
bilised onto a chip surface. Again however, low binding
with antibodies was observed.
sodium hydroxide) but full optimisation was not
achieved.
3.3. Analysis of shellfish samples
Synthesis of 4-NP derivatives with a reactive terminal
group suitable for coupling is a difficult task (Mikhura
et al., 2000). A new derivative N-aminobutyl [2-(4-
hydroxyphenyl)ethylamine] (AHPE) was synthesised
with using N-(4-bromobutyl)-phthalimide as an auxil-
iary substance (Fig. 2). The substance was used for the
introduction of alkylamine fragments into polyamines
structure (Mizzoni et al., 1954). AHPE has a different
nature of spacer compared to lipophylic 4-n-NP because
of an electrophylic amine moiety in the middle of a lin-
ear chain (Fig. 2). AHPE was immobilised onto a chip
surface and some binding (590–850 RU) was observed
with polyclonal antibodies but not the monoclonal anti-
body which showed strong tendency to recognise 4-phe-
nol derivatives with long lypophilic alkyl chains.
Because of its low solubility 4-NPs can easily accu-
mulate in aquatic organisms (Ahel et al., 1993). These
organisms are a good target for environmental monitor-
ing and assessment of water quality (pollution level).
Numerous publications showed that 4-NPs are often
found in fish and shellfish (Tsuda et al., 2000b; Ferrara
et al., 2001; Guenther et al., 2001; Corsi and Focardi,
2002). This makes it difficult to get a true negative sam-
ple for immunoassay development. NP free organisms
could be cultured at strictly controlled farms but this
can make their production very expensive.
Applicability of an assay developed for the analysis
of shellfish for 4-NPs residues was examined. The mono-
clonal antibody assay was used as a model for an
extraction procedure development and matrix effect
assessment. Moreover, the monoclonal antibody was
very specific for 4-n-NP which represents only a minor
fraction of 4-NPs mixture. In such a system the vast
majority of samples can be assumed as being negative.
The same assay conditions as for the buffer assay were
used, with only the ratio of antibody standard solution
(extract) being changed to 4:1 to reduce matrix
influence.
The calibration curve of technical 4-NP for poly-
clonal antibodies and AHPE chip was found to be more
sensitive (limit of detection––5 ng mlÀ1) compared to
homologous surface (Fig. 8). A working dilution of
polyclonal antibodies of just 1/100 could be used but
there was not a high background level associated with
the calibration curve. The calibration curves pattern
similar to 9-HPNA chips was obtained for this surface
with different flow rates using the same assay conditions.
However in this instance chip surface instability was ob-
served. The chip quickly lost 50% of its activity over 200
cycles whilst being regenerated by 100 mM sodium
hydroxide. It was found that the chip was substantially
more stable using weak regeneration solution (10 mM
Nonylphenol and nonylphenol ethoxilates are usually
extracted from fish or shellfish by steam distillation/sol-
vent extraction (Guenther et al., 2001), Soxhlet extrac-
tion (Corsi and Focardi, 2002) or solvent extraction
followed by SPE clean-up (Tsuda et al., 2000a). A simple
chip 9-HPNA 50%, pAb - 4-n-NP
chip 9-HPNA 50%, pAb - 4-NP
chip 9-HPNA 50%, IgG pAb - 4-NP
chip 9-HPNA 50%, mAb - 4-n-NP
chip AHPE, pAb - 4-NP
100
80
60
40
20
0
1
10
100
1000
10000
[analyte], ng ml-1
Fig. 8. Calibration curve for different antibody-target substance combinations in biosensor assay.

984
J.V. Samsonova et al. / Chemosphere 57 (2004) 975–985
250
200
150
100
50
scallop
oyster
mussel
cockle
0
10
100
1000
[4-n-nonylphenol], ng ml-1
Fig. 9. Calibration curve for 4-n-NP in biosensor assay for different shellfish samples. Monoclonal antibodies 1 lg mlÀ1, chip 9-HPNA
50%.
and rapid extraction procedure was developed. To gen-
erate a calibration curves samples of different species
(cockles, mussels, oyster, scallop) were homogenized
(whole body) and spiked with 4-n-NP. The substance
was extracted into hexane followed by clean-up on
NH2 SPE columns. Several solvents of different polarity
showed good washing potential and diethyl ether was
chosen as the most suitable in terms of elution speed
and evaporation time. The extracts were diluted two
times with buffer and directly analyzed in biosensor. Cal-
ibration curves generated for cockles, mussels and oyster
samples were identical (limit of detection at about 10
ng gÀ1) whereas for scallop samples a slight decrease
(about 5–10%) of absolute response was observed (Fig.
9). The method showed slightly lower sensitivity com-
pared to published chromatographic methods (Tsuda
et al., 2000a). The method was not evaluated because
of absence of a negative sample. The aim of this part of
the study was to establish a suitable extraction method
and assess the potential influence of shellfish as a matrix
during biosensor based analysis. These results were used
as a base for further investigations.
Results obtained in the polyclonal antibodies based
assay were markedly different. Two different chip sur-
faces were used (9-HPNA 50% and AHPE). Of the 31
shellfish samples tested 29 showed very similar results,
i.e., response units close to each other with CVs between
9 and 12%. The remaining two samples showed a signif-
icant reduction in recorded signal. On the basis of these
data it was concluded that these two shellfish samples
contained a quantity of at least one of 4-nonylphenols.
The 9-HPNA (50%) chip for two scallop samples signal
at about 70% of mean of all samples was observed.
Whereas for the more sensitive assay with AHPE chip
the response for these samples was at about 50% below
mean. The described study showed that developed
method could be applied for shellfish analysis as a pre-
liminary screening method, however verification by
chromatographic methods will be required.
4. Conclusion
Careful choice of a derivative for immobilization,
amount of immobilized derivative and purity of binding
proteins helped to improve assay performance and al-
lowed the development of a range of sensitive biosensor
assays for 4-nonylphenols of different specificity. To the
authors knowledge this is the first publication which has
shown the importance of optimising the amount of
immobilized hapten for Biacore biosensor assay devel-
opment. The assays developed were slightly less sensitive
when compared to conventional ELISA but had unques-
tionable advantages as a fast screening method. Biosen-
Applicability of the assay developed for 4-NP
screening was determined in the following experiment.
Thirty one shellfish samples collected from different
locations along the Irish coast were analyzed by biosen-
sor using monoclonal and polyclonal antibodies. In the
monoclonal antibodies system sensitive to 4-n-NP all
samples were deemed as negative (i.e. concentrations
found were lower than the established limit of detection
for the assay). CV for responses of all samples was
4.0%.

J.V. Samsonova et al. / Chemosphere 57 (2004) 975–985
985
ubiquitous in food. Environ. Sci. Technol. 36 (8), 1676–
1680.
sor assay developed showed good potential for shellfish
analysis for environmental contaminant monitoring.
Inoue, K., Kondo, S., Yoshie, Y., Kato, K., Yoshimura, Y.,
Horie, M., Nakazawa, H., 2001. Migration of 4-nonylphe-
nol from polyvinyl chloride food packaging film into food
simulates and foods. Food Addit. Contam. 18 (2), 157–164.
Mikhura, I.V., Formanovsky, A.A., Nikitin, A.O., Yakovleva,
J.N., Eremin, S.A., 2000. Synthesis of x-(4-hydroxyphe-
nyl)alkanecarboxylic acids. Mendeleev Commun. 5, 193–
194.
Acknowledgments
The authors thank Imelda Traynor (Veterinary Sci-
ences Department, Belfast) for the preparation the first
chip. Dr. J.V. Samsonova gratefully acknowledges IN-
TAS (fellowship Grant YSF 2001/2-0052) and The Well-
come Trust (Travel Award TG02/MEP/LEC/LR) for
providing financial support. Financial support from
the European Commission (EC Contract IC15CT98-
0138) is gratefully acknowledged.
Mizzoni, R.H., Hennessey, M.A., Scholz, C.R., 1954. Polyam-
ine salts with autonomic blocking properties. J. Am. Chem.
Soc. 76 (9), 2414–2417.
Nakamura, H., Karube, I., 2003. Current research activity in
biosensors. Anal. Bioanal. Chem. 377 (3), 446–468.
Petrovich, M., Eljarrat, E., Lopez de Alda, M., Barcelo, D.,
2001. Analysis and environmental levels of endocrine-
disrupting compounds in freshwater sediments. Trends
Anal. Chem. 20, 637–648.
References
Samsonova, J.V., Rubtsova, M.Y., Franek, M., 2003. Deter-
mination of 4-n-nonylphenol in water by enzyme immuno-
assay. Anal. Bioanal. Chem. 375 (8), 1017–1019.
Schwaiger, J., Mallow, U., Ferling, H., Knoerr, S., Braunbeck,
Th., Kalbfus, W., Negele, R.D., 2002. How estrogenic is
nonylphenol? A transgenerational study using rainbow
trout (Oncorhynchus mykiss) as a test organism. Aquat.
Toxicol. 59, 177–189.
Ahel, M., Giger, W., 1993. Aqueous solubility of alkylphenols
and alkylphenol ethoxylates. Chemosphere 26, 1461–1470.
Ahel, M., McEvoy, J., Giger, W., 1993. Bioaccumulation of the
lipophilic methabolites of non-ionic surfactants in freshwa-
ter organisms. Environ. Pollut. 79, 243–248.
Bester, K., Theobald, N., Schroder, H.Fr., 2001. Nonylphenols,
nonylphenol-ethoxylates, linear alkylbenzene sulfonates
(LAS) and bis(4-chlorophenyl)-sulfone in the German Bight
of the North Sea. Chemosphere 45, 817–826.
Tsuda, T., Suga, K., Kaneda, E., Ohsuga, M., 2000a. Deter-
mination of 4-nonylphenol, nonylphenol monoethoxylate,
nonylphenol diethoxylate and other alkylphenols in fish
and shellfish by high-performance liquid chromatography
with fluorescence detection. J. Chromatogr. B. 746, 305–
309.
Brix, R., Hvidt, S., Carlsen, L., 2001. Solubility of nonylphenol
and nonylphenol ethoxylates. On the possible role of
micelles. Chemosphere 44, 759–763.
Corsi, I., Focardi, S., 2002. Nonylphenols in
a lagoon
environment: p-nonylphenol and nonylphenol ethoxylates
in fish tissue. Bull. Environ. Contam. Toxicol. 68, 908–914.
Ferrara, F., Fabietti, F., Delise, M., Piccioli Bocca, A., Funari,
E., 2001. Alkylphenolic compounds in edible molluscs of the
Adriatic sea (Italy). Environ. Sci. Technol. 35, 3109–3112.
Franek, M., Zeravik, J., Eremin, S.A., Yakovleva, J., Badea,
M., Danet, A., Nistor, C., Emneus, J., 2001. Antibody
based methods for surfactant screening. Fresen. J. Anal.
Chem. 371, 456–466.
Tsuda, T., Takino, A., Kojima, M., Harada, H., Muraki, K.,
Tsuji, M., 2000b. 4-Nonylphenols and 4-tert-octylphenol in
water and fish from rivers flowing into Lake Biwa.
Chemosphere 41, 757–762.
Wheeler, T., Heim, J.R., LaTorre, M.R., Janes, A.B., 1997.
Mass spectral characterization of p-nonylphenol isomers
using high-resolution capillary GC–MS. J. Chromatogr. Sci.
35, 19–30.
Ying, G-G., Williams, B., Kookana, R., 2002. Environmental
fate of alkylphenols and alkylphenol ethoxylates––a review.
Environ. Int. 28, 215–226.
Guenther, K., Durbeck, H.W., Kleist, E., Thiele, B., Prast, H.,
Schwuger, M., 2001. Endocrine disrupting nonylphenols––
ultra-trace analysis and time-dependant trend in mussels
from the German bight. Fresen. J Anal. Chem. 371 (6), 782–
786.
Zeravik, J., Skryjova, K., Nevorankova, Z., Franek, M., 2004.
Development of direct ELISA for the determination of 4-
nonylphenol and octylphenol. Anal. Chem. 76, 1021–
1027.
Guenter, K., Heinke, V., Thiele, B., Kleist, E., Prast, H.,
Raecker, T., 2002. Endocrine disrupting nonylphenols are