Monoclonal antibody-based ELISAs for part-per-billion determination of polycyclic aromatic hydrocarbons: Effects of haptens and formats on sensitivity and specificity

Article abstract of DOI:10.1021/ac980765d

As a first step toward developing sensitive enzyme-linked immunosorbent assays (ELISAs) for multianalyte detection of polycyclic aromatic hydrocarbons (PAHs), haptens with different lengths of carboxylic acid spacers at various positions were derived from naphthalene, fluorene, anthracene, phenanthrene, pyrene, fluoranthene, chrysene, and benzotalpyrene (BaP). These haptens were coupled with bovine serum albumin (BSA) to form competitor conjugates. All of these haptens were recognized to different extents by monoclonal antibodies (MAbs) 4D5 and 10C10 originally derived by Gomes and Santella (Chem. Res. Toxicol. 1990, 3, 307-310). The most sensitive indirect ELISAs were obtained by coating wells with the least competitive conjugates. Direct ELISAs using horseradish peroxidase conjugates of pyrene and BaP were less sensitive. The MAbs bound BaP with spacers at either C1 or C6. The cross-reactivity profiles of the eight PAHs were different with each PAH-BSA conjugate used as coating antigen. The ELISA results for BaP closely correlated with those by gas chromatography (GC), but the detection limit of the ELISA was ～150-fold more sensitive than that of GC, with 2-600 nM spike recoveries of 80-127% from human urine and canal and tap water.

Full text of DOI:10.1021/ac980765d

Anal. Chem. 1999, 71, 302-309
Articles
Mo n o c lo n a l An t ib o d y-Ba s e d ELIS As fo r
P a rt -p e r-Billio n De t e rm in a t io n o f P o lyc yc lic
Aro m a t ic Hyd ro c a rb o n s : Effe c t s o f Ha p t e n s a n d
Fo rm a t s o n S e n s it ivit y a n d S p e c ific it y
†
†
‡
†
‡
§
Ka i Li, Ronglia ng Che n, Bita o Zha o, Me i Liu, Ale xa nde r E. Ka ru, Vic toria A. Robe rts , a nd
Qing X. Li*,
†
Department of Environmental Biochemistry, University of Hawaii at Manoa, Honolulu, Hawaii 96822, College of Natural
Resources Immunochemistry Facility, University of California, Berkeley, California 94720, and Department of Molecular
Biology, The Scripps Research Institute, La Jolla, California 92037
2
As a first step toward developing sensitive enzyme-linked
immunosorbent assays (ELISAs) for multianalyte detec-
tion of polycyclic aromatic hydrocarbons (P AHs), haptens
with different lengths of carboxylic acid spacers at various
positions were derived from naphthalene, fluorene, an-
thracene, phenanthrene, pyrene, fluoranthene, chrysene,
and benzo[a ]pyrene (BaP ). These haptens were coupled
with bovine serum albumin (BSA) to form competitor
conjugates. All of these haptens were recognized to
different extents by monoclonal antibodies (MAbs) 4 D5
and 1 0 C1 0 originally derived by Gomes and Santella
wastes, and groundwater. Benzo[a]pyrene (BaP) is one of the
most potent carcinogens of the PAH family, BaP content correlates
well with total PAH content in contaminated environmental
samples, and BaP adducts of protein and DNA are biomarkers of
exposure to PAHs.3
Determination of PAHs in environmental matrixes typically
involves several steps. These include extraction of the sample by
either liquid-liquid or solid-phase extraction, concentration and
cleanup of the extracts followed by analysis with high-performance
liquid chromatography (HPLC), gas chromatography (GC) or GC/
4
mass spectrometry (GC/ MS). To avoid these complex proce-
(
Ch em . Res. Toxicol. 1 9 9 0 , 3, 3 0 7 -3 1 0 ). The most
dures, immunochemical analyses employing either polyclonal or
monoclonal antibodies have been developed for PAHs.5-9 Com-
sensitive indirect ELISAs were obtained by coating wells
with the least competitive conjugates. Direct ELISAs using
horseradish peroxidase conjugates of pyrene and BaP
were less sensitive. The MAbs bound BaP with spacers
at either C1 or C6. The cross-reactivity profiles of the eight
P AHs were different with each P AH-BSA conjugate used
as coating antigen. The ELISA results for BaP closely
correlated with those by gas chromatography (GC), but
the detection limit of the ELISA was ∼1 5 0 -fold more
sensitive than that of GC, with 2-600 nM spike recoveries
of 8 0 -1 2 7 % from human urine and canal and tap water.
mercially available immunoassay kits for PAHs, total petroleum
hydrocarbons (TPHs), and carcinogenic PAHs in various matrixes
are marketed by Strategic Diagnostics, Inc. (Newark, DE). U.S.
EPA has compiled a series of immunochemical methods for
environmental analysis. Method 4035 detects several PAHs to
different degrees and measures their cumulative responses to
(
1) International Agency for Research on Cancer Monographs on the Evaluation
of Carcinogenic Risk of Chemicals to Humans. Polynuclear Aromatic
Compounds, Part 3; International Agency for Research on Cancer: Lyon,
France; 1984; Vol. 34.
(
(
2) Patnaik, P. In Handbook of Environmental Analysis; Patnaik, P., Ed.; CRC
Press: Boca Raton, FL, 1997; pp 165-169.
3) International Agency for Research on Cancer Monographs on the Evaluation
of Carcinogenic Risk of Chemicals to Humans. Polynuclear Aromatic
Compounds, Part 1; International Agency for Research on Cancer: Lyon,
France; 1983; Vol. 32.
Polycyclic aromatic hydrocarbons (PAHs) are environmental
contaminants produced mainly by incomplete combustion of
organic matter. PAHs in smoke, soot, and dust particles in the
air are carried into water, soil, and crops to contaminate the
environment. Many PAHs are carcinogens in animals and are
suspected carcinogens in humans.1 The U.S. Environmental
Protection Agency (EPA) has listed 16 PAHs as priority pollutants
in wastewater and 24 PAHs in soils, sediments, hazardous solid
(
4) Osborne, M. R.; Crosby, N. T. In Benzopyrenes; Cambridge University
Press: New York, 1987; pp 251-300.
(5) Gomes, M.; Santella, R. M. Chem. Res. Toxicol. 1 9 9 0 , 3, 307-310.
(
(
6) Knopp, D.; Vaananen, V.; Zuhlke, J.; Niessner, R. Immunochemical Technology
for Environmental Applications; ACS Symposium Series 657; American
Chemical Society: Washington, DC, 1997; pp 71-76.
7) McDonald, P. P.; Almond, R. E.; Mapes, J. P.; Friedman, S. JAOAC Int. 1994,
*
Corresponding author: (tel) 808-956-2011; (fax) 808-956-5037; (e-mail)
77, 466-472.
qingl@hawaii.edu.
(8) Roda, A.; Bacigalupo, M. A.; Ius, A.; Minutello, A. Environ. Technol. 1 9 9 1 ,
12, 1027-1035.
(9) Roda, A.; Pistillo, A.; Jus, A.; Armanino, C.; Braldini, M. Anal. Chim. Acta
†
University of Hawaii at Manoa.
University of California.
‡
§
The Scripps Research Institute.
1 9 9 4 , 298, 53-64.
302 Analytical Chemistry, Vol. 71, No. 2, January 15, 1999
10.1021/ac980765d CCC: $18.00 © 1999 American Chemical Society
Published on Web 12/15/1998

Fig u re 1 . Structures of haptens used in the preparation of antigens.
determine the total PAH in the sample.10 These assays provide
an easy and fast alternative for PAH screening.
analysis. The ELISA development approach is generally applicable
to developing immunoassays for other small molecules of envi-
ronmental significance.
Antibodies are the primary agents that determine immunoassay
performance, and properties of the antibody are largely deter-
mined by the immunogen.11 The similar shape, hydrophobicity,
EXPERIMENTAL SECTION
Reagents and Instruments. All PAH standards and deriva-
lack of hydrogen-bonding atoms, and other characteristics of PAHs
make it difficult to derive antibodies that are highly selective for
individual PAHs. Cross-reactivity is expected and will have to be
exploited in PAH immunoanalysis of environmental samples.
Monoclonal antibodies (MAb), 4D5 and 10C10, originally derived
tives were obtained from Aldrich (Milwaukee, WI). PAH stock
solutions were prepared in methanol or DMSO in sealed glass
vials. Immunochemicals, enzyme substrates, and buffer capsules
were obtained from Sigma (St. Louis, MO). The ELISA was carried
out in 96-well polystyrene plates (MaxiSorp F96, Catalog No.
5
and characterized by Gomes and Santella, cross-react with several
4
(
39454, Nalge Nunc International, Denmark). The optical density
OD) was measured using a Vmax microplate reader (Molecular
Devices Corp., Sunnyvale, CA). Hybridoma cell lines 4D5 and
0C10, kindly provided by R. M. Santella (Columbia University,
PAHs of major concern as environmental pollutants. In this paper,
we describe several new PAH haptens we prepared for a
systematic study of cross-reactivity by these MAbs. In addition,
we recently cloned and expressed recombinant Fab antibodies
1
New York), were grown, and IgG was purified as previously
described.14 Precoated TLC plates were silica gel F254 and were
purchased from E. Merck (Darmstadt, Germany). NMR spectra
were recorded with a Nicolet NT 300-MHz instrument. Mass
spectra were obtained with an AEI-MS 30 or VG Trio 2 (VG
BioTech) spectrometer with a direct insertion probe and fast atom
bombardment (FAB) ionization.
Safety Considerations. Syntheses were done in chemical
fume hoods with charcoal filters. PAH-impermeable nitrile gloves
were used. All byproducts and wastes from the syntheses and
primary solutions of PAH analytes and haptens from assays were
disposed of as hazardous materials.
(
rFabs) with similar cross-reactivity from these two MAbs.12
A
model based on the rFab sequences indicates that the PAH
binding sites accommodate a variety of PAHs.13 The experiments
we report here were done to maximize the differences in PAH
cross-reactivity of these two antibodies in the direct and indirect
formats of enzyme-linked immunosorbent assays (ELISAs). The
half-maximal inhibition (I50) of the PAHs and haptens was
measured with MAbs 10C10 and 4D5. Suitable conjugates were
further examined to optimize the assays. In addition, the most
sensitive assay was tested for the analysis of BaP spiked in tap
and canal water and human urine and was compared with GC
Synthesis of Haptens. All haptens except 4-(1-pyrene)butyric
acid (PYR-1a), which was a commercial product from Aldrich
(
(
10) US EPA method 4035. In SW-846 Test Methods for Evaluating Solid Waste
Physical/ Chemical Methods, 3rd ed.; US EPA, Office of Solid Waste:
Washington, DC, 1996; Chapter 4.
11) Tijssen, P. In Laboratory Techniques in Biochemistry and Molecular Biology;
Burdon, R. H., Knippenberg, P. H., Eds.; Elsevier Verlag: Amsterdam; 1985;
Vol. 15, pp 128-135.
12) Li, K.; Bell, C.; Chen, R.; Zhao, B.; Karu A.; Li, Q. X. 215th ACS national
meeting. Dallas, TX, March 29-April 2, 1998; Abstr. AGRO-62.
13) Roberts, V. A.; Pellequer, J.-L.; Bell, C. W.; Karu, A. E.; Li, K.; Li, Q. X. 81st
Canadian Soc. for Chem. Conf., Exhib. Whistler BC, Canada, June 1998.
(
Milwaukee, WI), were synthesized from PAH derivatives and
characterized by NMR and MS (Figure 1, Table 1).
6 -(2 -Naphthoxy)hexanoic Acid (NAP -2 a). NAP-2a was
prepared by reaction of 2-naphthol with ethyl bromohexanoate
(
(
(14) Karu, A. E.; Goodrow, M. H.; Schmidt, D. J.; Hammock, B. D.; Bigelow, M.
W. J. Agric. Food Chem. 1 9 9 4 , 42, 301-309.
Analytical Chemistry, Vol. 71, No. 2, January 15, 1999 303

Ta b le 1 . Ch e m ic a l Ch a ra c t e riza t io n o f Ha p t e n s b y 1 H NMR a n d MS
chemical
NAP-2a
spacer arm
1H NMR (300 MHz)a
MS m/ z (rel int, %)
258 (M , 12), 144 (100), 115 (30)
+
O(CH2)5COOH
4.15 (t, 2H), 2.36 (t, 2H), 1.88 (m, 2H),
1
.74 (m, 2H), 1.55 (m, 2H)
+
b
FLR-3a
CH2NHCH2CH2COOH 3.70 (s, 2H), 2.75 (t, 2H), 2.30 (t, 2H)
268 (MH , 34)
+
PHE-10a CH2NHCH2CH2COOH 3.58 (s, 2H), 2.53 (t, 2H), 2.12 (t, 2H)
279 (M , 38), 218 (8), 206 (32), 191 (100), 178 (12)
279 (M , 66), 218 (10), 206 (18), 191 (100), 178 (35)
317(M , 13), 299 (25), 217 (100), 189 (25)
328 (M , 40), 255 (100), 226 (71), 113 (46)
+
ANT-9a
FLA-3a
CHR-6a
PYR-1a
PYR-1b
BaP-1a
BaP-6b
CH2NHCH2CH2COOH 4.25 (s, 2H), 2.89 (t, 2H), 2.35 (t, 2H)
+
NHCOCH2CH2COOH
COCH2CH2COOH
CH2CH2CH2COOH
CH2NHCH2CH2COOH 3.58 (s, 2H), 2.99 (t, 2H), 2.45 (t, 2H)
COCH2CH2COOH 3.60 (t, 2H), 3.02 (t, 2H)
CH2NHCH2CH2COOH 3.75 (s, 2H), 2.80 (t, 2H), 2.30 (t, 2H)
3.20 (t, 2H), 2.80 (t, 2H)
3.60 (t, 2H), 3.00 (t, 2H)
3.42 (t, 2H), 2.52 (t, 2H), 2.24-2.18 (m, 2H) 288 (M , 36), 239 (100), 215 (55), 135 (65)
+
+
+
303 (M , 37), 242 (10), 230 (21), 215 (100), 202 (25)
+
352(M , 38), 279 (60), 251 (30), 250 (26), 149 (55), 125 (75)
+
353 (M , 20), 292 (8), 280 (12), 265 (100), 252 (18)
a
Only signals from spacer arm are listed. b Obtained using FAB mode.
followed by hydrolysis.15 Briefly, 2-naphthol (1.0 mmol) and ethyl
6 -Benzo[a ]pyreneisocyanate (BaP -6 a) was synthesized as
described by Griffin et al.18 and was generously provided by T.
Vo-Dinh.
6-bromohexanoate (1.0 mmol) were dissolved in acetone (50 mL).
Potassium iodide (40 mg) were added to the acetone solution.
The heterogeneous mixture was refluxed for 24 h. After filtering
off the salt and evaporating acetone, the crude residue was purified
with silica flash chromatography (methylene chloride-ethyl
acetate-hexane 1:1:8). The purified ethyl 6-(2-naphthoxy)hex-
anoate (200 mg) was hydrolyzed in ethanol under the catalysis of
lithium hydroxide (2 mL, 1 N). Upon addition of 4 N HCl, the
product precipitated out as a white solid.
Hapten-P rotein Conjugates. A carboxylic hapten (0.20
mmol), N-hydroxysuccinimide (0.20 mmol), and 1-[3-(dimethy-
lamino)propyl]-3-ethylcarbodiimide (0.22 mmol) were dissolved
in dimethylformamide (1 mL). The mixture was stirred at room
temperature for 3.5 h and centrifuged to remove the precipitate.
The active ester in the supernate (0.25 mL) was added very slowly
to a solution of bovine serum albumin (BSA) (50 mg) in 15 mL of
0.05 M borate buffer, pH 9. After it was stirred gently overnight
at 4 ˚C, the solution was dialyzed against 0.02 M phosphate-
buffered saline, pH 7.5 (PBS), over 24 h, with six changes of 4 L
each. To prepare horseradish peroxidase (HRP) conjugates, 130
µL of active ester solution was added slowly to a HRP solution
N-(3-Fluorenemethyl)-3-alanine (FLR-3a), N-(10-Phenan-
threnemethyl)-3-alanine (PHE-10a), N-(9-Anthracenemethyl)-
3
1
-alanine (ANT-9 a), N-(1 -P yrenemethyl)-3 -alanine (P YR-
b), and N-(6 -Benzo[a ]pyrenemethyl)-3 -alanine (BaP -6 b).
A PAH aldehyde (1.0 mmol) and â-alanine (1.0 mmol) were
dissolved in a mixture of methanol (10 mL) and triethylamine (1
mL). The reaction mixture was refluxed for 2 h. After cooling to
room temperature, the reaction product was reduced with an
excess amount of sodium borohydride (100 mg). The solution
was adjusted to pH 2 using 2 N HCl to precipitate the product,
which was then recrystallized with a mixture of methanol and
water (2:1). BaP-6-carboxaldehyde (BaP-6-CHO) was used as an
intermediate for the synthesis of BaP-6b and was prepared by
reaction of BaP and N-methylformanilide in the presence of
phosphoryl chloride according to the procedures described by
Dewhurst and Kitchen.16
N-(3 -Fluoranthene)succinamic Acid (FLA-3 a). 3-Aminof-
luoranthene (0.5 mmol) and succinic anhydride (0.6 mmol) were
dissolved in pyridine (1 mL), and the mixture was stirred at room
temperature for 24 h. After all starting materials were consumed
as indicated by TLC (ethyl acetate-hexane 1:3), water (20 mL)
was added and the precipitate was collected. The product was
purified on silica gel column using ethyl acetate-hexane (1:5) as
eluant.
(0.67 mg/ mL 0.13 M NaHCO
overnight at 4 °C, and the conjugate was then dialyzed against 4
L of 0.13 M NaHCO at 4 °C for 3 days with buffer changes twice
3
). The solution was stirred gently
3
a day. BaP-6a-BSA was prepared by covalently coupling BaP-6a
with BSA.19
Enzyme-Linked Immunosorbent Assays. Direct Competitive
ELISA. Microplate wells were coated with purified MAb 10C10
(100 µL/ well, 0.2 µg/ mL 0.05 M carbonate-bicarbonate buffer,
pH 9.6). The coating solution was evaporated to dryness at 37 °C
overnight. The plate was washed four times with PBS containing
0.05% Tween 20 (PBST) and blocked with 0.3% BSA in PBST
(PBST-BSA) (200 µL/ well) at 37 °C for 1 h. Alternatively, the
plate was precoated overnight at 37 °C with affinity-purified goat
anti-mouse IgG (Boehringer Mannheim, 1 mg/ mL, 1:2000 in
PBST-BSA) and subsequently incubated with 0.2 µg/ mL purified
MAb 10C10 (100 µL/ well) at 37 °C for 1 h. The coated wells were
washed three times with PBST. Samples containing different
amounts of analyte and 25 ng of PYR-1a-HRP competitor in
PBST-BSA were added to coated microwells (100 µL/ well) and
incubated at 37 °C for 1 h. After another washing, 100 µL of
o-phenylenediamine (OPD) solution (1.0 mg/ mL in 0.05 M
citrate-phosphate with 0.03% sodium perborate, pH 5.0) was
added in each well. After 15-30 min at room temperature, the
(
1 -Benzo[a ]pyrene)succinic acid (BaP -1 a) and (6 -chry-
sene)succinic acid (CHR-6 a) were prepared by reaction of
succinic anhydride with BaP and chrysene, respectively, using
aluminum chloride as catalyst following the procedures of Buu-
Hoi and Lavit.17
(
18) Griffin, G. D.; Ambrose, K. R.; Thomason, R. N.; Murchison, C. N.; McManis,
M.; St. Wecker, P. G. R.; Vo-Dinh, T. Production and characterization of
antibodies to benzo[a]pyrene. Proceedings of 10th International Symposium
on Polycyclic Aromatic Hydrocarbons, October 21-23, 1985.
(
(
(
15) Chiu, Y. W.; Carlson, R. E.; Marcus, K. L.; Karu, A. E. Anal. Chem. 1 9 9 5 ,
7, 3829-3839.
16) Dewhurst, F.; Kitchen, D. A. J. Chem. Soc., Perkin Trans. 1 1 9 7 2 , 710-
12.
17) Buu-Hoi, N. P.; Lavit, D. Tetrahedron 1 9 6 0 , 8, 1-6.
6
7
(19) Vo-Dinh, T.; Tromberg, B. J.; Griffin, K. R.; Ambrose, M. J.; Sepaniak, M.
J.; Gardenhire, E. M. Appl. Spectrosc. 1 9 8 7 , 41, 735-738.
304 Analytical Chemistry, Vol. 71, No. 2, January 15, 1999

reaction was stopped with 2 M sulfuric acid (50 µL/ well). The
OD at 490 nm was then read on a Vmax microplate reader, and
the data were fitted using Softmax software (Molecular Devices,
Sunnyvale, CA).
are carcinogenic PAHs found in urban air.21 These and fluoran-
thene are mutagenic in the Ames test.20 Naphthalene, phenan-
threne, fluorene, and fluoranthene as well as others are found in
cigarette smoke and exhaust gases.21 All are present in different
amounts in oils, fuels, lubricants, cooked foods, and a wide variety
of biological and environmental matrixes.22,23 Metabolites of
chrysene, pyrene, and BaP form protein and DNA adducts that
are biomarkers of PAH exposure.24,25
The FLR-3a, PHE-10a, ANT-9a, PYR-1b, and BaP-6b haptens
were prepared by reacting PAH aldehyde intermediates with
â-alanine. FLA-3a was synthesized by reaction of 3-aminofluoran-
thene with succinic anhydride. Friedel-Crafts acylation of chry-
sene and BaP with succinic anhydride gave CHR-6a and BaP-1a,
respectively. BaP-6a differs in linkage position and spacer from
BaP-1a. BaP-6b differs from BaP-6a only in the spacer. These
carboxylic haptens were conjugated to carrier protein or enzyme
by the active ester method.26 The MAbs bound both the haptens
and PAH-BSA conjugates with very different relative affinities,
and this affected the sensitivity and selectivity of direct and indirect
ELISAs for various PAHs.
Indirect Competitive ELISA. Microplates were coated with PAH
hapten-BSA conjugate (1-5 ng/ 100 µL per well in 0.05 M
carbonate-bicarbonate buffer, pH 9.6) by drying at 37 °C
overnight and blocked with PBST-BSA as described above.
Hybridoma culture supernate (1:500-1:2000 in PBST-BSA) was
mixed with different amounts of analyte, the solution was added
to the plate (100 µL/ well) and incubated at 37 °C for 1 h. After
washing, the plate was incubated for 1 h at 37 °C with HRP-goat
anti-mouse IgG (1:12000 in PBST-BSA, 100 µL/ well), the plate
was washed, and the procedure for the color development was
as described above. Where indicated, the signal was amplified
with biotinylated goat anti-mouse IgG and avidin-HRP conjugate.
After the incubation with the mixture of PAH antibody and analyte,
the plate was washed, incubated with biotinylated goat anti-mouse
IgG (1:12000 in PBST-BSA, 100 µL/ well) at 37 °C for 1 h, followed
by another wash and incubation with avidin-HRP (1:12000 in
PBST-BSA, 100 µL/ well) at 37 °C for 1 h. The rest of the
procedures were the same as described above. This protocol is
referred to as labeled avidin-biotin ELISA (LAB-ELISA).
GC Analysis. Water samples from the Ala Wai canal (Hono-
lulu, HI) were spiked with BaP at concentrations of 20, 40, 100,
Enzymes and Assay Formats. While OPD was used as a
substrate for goat anti-mouse IgG-HRP, p-nitrophenyl phosphate
(
NPP) was used for alkaline phosphatase (AP). Indirect ELISAs
using BaP-6a-BSA or PYR-1a-BSA as coating antigen were used
to compare these enzymes and substrates. There was no signifi-
cant difference in I50 or limit of detection (LOD; I20) for BaP
regardless of whether commercial goat anti-mouse IgG-HRP or
2
00, and 400 nM (i.e., 5-100 ppb). A 400-mL aliquot of the 20
nM spiked sample and 200 mL of the 40-400 nM spiked samples
were extracted with 3 × 100 mL of ethyl acetate, the extracts were
combined and dried over anhydrous sodium sulfate, concentrated
by rotary evaporation, and reconstituted in ethyl acetate to 2.0
-
AP conjugate was used (I50 ) 852 and 1008 nM, respectively
with BaP-6a-BSA coating; both at 128 nM with PYR-1a-BSA
coating). Goat anti-mouse IgG-HRP conjugate and the substrate
OPD were used for the subsequent work because they gave much
faster color development.
(
20 nM spiked sample) or 5.0 mL (40-400 nM spiked sample).
Analysis was done on a Hewlett-Packard (HP) 5890 GC equipped
with a DB-5 capillary column (J & W, 30 m × 0.25 mm, 0.25-µm
film thickness) and flame ionization detector. The injector and
detector temperatures were 275 and 300 °C, respectively. The oven
temperatures were increased from 65 to 140 °C at a rate of 25
In the direct ELISA, the I50 for BaP was 2952 nM when affinity-
purified MAb 10C10 IgG (20 ng/ well) was directly coated on the
plate and the enzyme tracer PYR-1a-HRP was used at 25 ng/
well. When MAb 10C10 was captured from culture supernate on
wells coated with 20 ng/ well of affinity-purified goat anti-mouse
IgG, the I50 for BaP decreased to 2015 nM. Wells could not be
coated directly with unpurified hybridoma culture supernate
because it contains a large excess of serum proteins.
The direct ELISA could be run in the shortest time among
the three ELISA formats, but it was less sensitive than the indirect
ELISA. Conditions optimized for indirect ELISA detection by goat
anti-mouse IgG-HRP (each well coated with 2.7 ng of BaP-6a-
BSA, and a 1:500 dilution of MAb 10C10 culture supernate) gave
an I50 of 980 nM for BaP. The sensitivity was improved 3-4-fold
(I50 ) 330 nM BaP) when conditions were optimized for detection
°C/ min and then to 290 °C at a rate of 10 °C/ min. The temperature
was maintained at 290 °C for 15 min. Peak areas recorded with a
HP 3396A integrator were quantified relative to an external
standard of BaP.
RESULTS AND DISCUSSION
Selection and Synthesis of Haptens. The MAbs 10C10 and
4
D5 used in this study were originally derived by Gomes and
Santella using a 6-amino-BaP-BSA (a BaP 6-isocyanate hapten)
5
as immunogen. These antibodies recognized BaP and cross-
reacted with a number of BaP metabolites as well as several other
5
(21) Pucknat, A. W. Health impacts of polynuclear aromatic hydrocarbons; Noyes
PAHs. As a possible means of expanding the usefulness of these
Data Corp.: Park Ridge, NJ, 1981; p 271.
MAbs, we synthesized and characterized the new PAH haptens
in Figure 1 and examined the performance of the MAbs with them
in direct and indirect ELISAs. These 11 haptens represent the
major two- to five-ring PAHs with aqueous solubilities (0.002-30
(
(
(
22) Neff, J. M. Polycyclic Aromatic Hydrocarbons in the Aquatic Environment;
Applied Science Publishers Ltd.: London, 1979.
23) Lee, M. L.; Novotny, M. V.; Bartle, K. D. Analytical chemistry of polycyclic
aromatic compounds; Academic Press: New York, 1981; p 462.
24) Biomarkers and risk assessment: concepts and principles; World Health
Organization: Geneva, Switzerland, 1993.
-
2
-7
µg/ mL) and vapor pressures (4.92 × 10 -6.9 × 10 Torr at 20
C) that are most amenable to immunoassay.2 BaP and chrysene
0
(25) Schnell, F. C. Protein adduct-forming chemicals and molecular dosimetry:
potential for environmental and occupational biomarkers. In Reviews in
Environmental Toxicology; Hodgson, E. Ed.; Toxicology Communications,
Inc.: Raleigh, NC, 1993; Vol. 5, pp 51-60.
°
(
20) Sims, R. C.; Overcash, M. R. Fate of polynuclear aromatic compounds in
soil-plant systems. In Residue Reviews; Gunther, F. A., Gunther, J. D. Eds.;
Springer-Verlag: New York, 1983; Vol. 88, pp 2-68.
(26) Klibanov, A. L.; Slinkin, M. A.; Torchlin, V. P. Appl. Biochem. Biotechnol.
1 9 8 9 , 22, 45-58.
Analytical Chemistry, Vol. 71, No. 2, January 15, 1999 305

formed by metabolites of pyrene, chrysene, BaP, and other PAHs
for exposure monitoring without hydrolysis of the adducts, which
is often required for instrumental analysis.
Ta b le 2 . S e n s it ivit y o f In d ire c t LAB-ELIS A fo r Ba P a n d
Re la t ive Bin d in g o f MAb 1 0 C1 0 t o P AH-BS A
Co n ju g a t e s
In the second experiment, an indirect competitive LAB-ELISA
for BaP was done on plates coated with a PAH-BSA conjugate.
The I50 values for BaP indicated the relative sensitivity. The results
of both experiments, shown in Table 2, led to several conclusions.
First, all of the listed two-, three-, four-, and five-ring PAH hapten-
BSA conjugates in solution competitively bound MAb 10C10 in
the presence of immobilized BaP-6a-BSA, underscoring the broad
cross-reactivity of this MAb. Second, MAb 10C10, which was
raised against BaP-6a-BSA, bound BaP haptens with the spacer
arm at either C1 or C6 on the ring. Third, differences of only 2-fold
or less in sensitivity for BaP were observed using different linkers
between the hapten and BSA (PYR-1a and -1b, BaP-6a and -6b),
supporting the notion that aromatic ring binding was the pre-
dominant factor in recognition. Fourth, the RI was an approximate
predictor of the I50 for BaP, although only the largest and smallest
RI and BaP I50 values correlated well. When the weakest cross-
reactive conjugate, PYR-1a-BSA, was used as coating antigen, it
gave the most sensitive ELISA for BaP. Conversely, the NAP-
PAH-BSA
recognition index
I50 (ppb) for BaP with
a
indicated coating conjugateb
conjugate
(I50, ppb)
PYR-1a-BSA
PYR-1b-BSA
FLA-3a-BSA
CHR-6a-BSA
ANT-9a-BSA
BaP-6a-BSA
BaP-6b-BSA
BaP-1a-BSA
PHE-10a-BSA
FLR-3a-BSA
NAP-2a-BSA
0.55 (51.6)
0.61 (47.0)
0.70 (40.5)
0.75 (38.0)
0.95 (29.9)
1.0 (28.5)
1.2 (23.7)
1.9 (14.9)
2.1 (13.4)
9.2 (3.1)
11.7 (46.8 nM)
28.7
47.2
17.9
14.3
25.5
35.0
14.5
23.4
24.7
134.0 (536 nM)
95 (0.3)
a
Recognition index, I50 of BaP-6a-BSA/ I50 of a conjugate. I50 values
in parentheses were obtained by indirect LAB-ELISAs of the corre-
sponding PAH-BSA conjugate in the left column using MAb 10C10
(
1:6000) on plates coated with BaP-6a-BSA (2.0 ng/ 100 µL per well).
b
Data were means of four wall replicates. Indirect LAB-ELISAs of BaP
using MAb 10C10 on plates coated with the corresponding hapten-
BSA in the left column. Data were means of four well replicates.
2
a-BSA conjugate, which was most strongly bound, gave an assay
using biotinylated goat anti-mouse IgG and avidin-HRP (LAB-
ELISA: each well coated with 1.0 ng of BaP-6a-BSA, and a 1:2000
dilution of MAb 10C10 culture supernate). The LAB-ELISA
obviously consumed less reagents. We subsequently found that
one incubation step could be eliminated by adding the biotinylated
goat anti-mouse IgG along with the MAb in the incubation with
the analyte. The assay performance was the same, but the time
required for the LAB-ELISA was significantly shortened. There-
fore, we used the indirect LAB-ELISA for all subsequent work.
Selection of Coating Antigens. The sensitivity and specificity
of competitive ELISAs depend on relative affinities of antibody
for the competitor haptens and the analytes. The competitor
conjugates may vary in hapten structure and density. The most
sensitive assays are usually those in which the antibody has a
lower affinity for the competing hapten (i.e., haptenated conjugate)
than it does for the analyte.27 Two indirect competitive LAB-ELISA
experiments were done to identify which haptenated conjugates
improved the sensitivity and gave adequate binding for signal
detection. First, to compare the relative binding of MAb 10C10
to the PAH-BSA conjugates, each hapten-BSA conjugate was
used as the inhibitor in solution on plates coated with BaP-6a-
BSA. The conjugates that were recognized best by MAb 10C10
had the least MAb bound to the coating and thus gave the lowest
for BaP that was roughly 10-fold less sensitive. Presumably,
recognition of the PAH-BSA conjugates was affected primarily
by the MAb’s affinity for the hapten and the hapten density on
the BSA. For example, FLR-3A conjugate had a recognition index
∼
17-fold higher than PYR-1a, but the I50 for BaP was only 2.1-fold
higher when FLR-3A, rather than PYR-1a, was the coating antigen.
Although hapten densities on the conjugates were not determined
in this study, they were presumably similar because of similar
chemical properties of the PAH haptens and parallel synthesis of
these conjugates using the same mole ratio of hapten to BSA
under the same condition. However, hapten densities must be
considered when they vary significantly.
Because PYR-1a-BSA gave the lowest I50 for BaP, it was used
as coating antigen in all subsequent LAB-ELISAs. The assay was
optimized to use 0.5 ng/ 100 µL per well of PYR-1a-BSA and
1
:6000 dilution of MAb 10C10 culture supernate. Standard curves
fitted well with the four-parameter logistic model over the entire
range or with a semilog model over the linear portion (percent
inhibition, % I ) 29.04 + 20.28 log X, r ) 0.999) (Figure 2). The
average I50 for BaP was 10.8 nM (2.7 ppb), with a LOD defined
as the I20, of ∼0.4 nM (0.1 ppb). The linear range of the curve
was about 0.4-1000 nM of BaP. The assay is able to measure
BaP accurately when its concentrations are above 0.4 nM in
samples. The assay sensitivity was improved ∼4-fold relative to
that reported by Gomes and Santella and ∼89-fold compared to
that obtained in our initial assays using BaP-6a coating conjugate
following the procedure of Gomes and Santella.5
I
50 value. The conjugates that were weakly recognized by the MAb
left more MAb bound to the coating and gave a higher I50 value.
These results were expressed as a “recognition index” (RI) and
calculated as ratio of the I50 value of BaP-6a-BSA to that of the
test conjugate, i.e., RI ) I50 of BaP-6a-BSA/ I50 of a conjugate
Cross-Reactivity. Because of the planarity and similar shapes
and chemical properties of PAHs, antibodies such as 4D5 and
(
Table 2). Thus, conjugates to which MAb 10C10 bound more
strongly than to BaP-6a-BSA had a RI greater than 1. Using such
a conjugate as coating antigen may reduce assay sensitivity. In
addition, screening the PAH conjugate competitors in solution
suggested that MAb 10C10 may be able to bind protein adducts
1
0C10 with broad cross-reactivity are likely to occur as a rule
rather than an exception. To test whether assay conditions and
reagents could be manipulated to obtain consistent and interpret-
able reaction patterns with different PAHs, indirect competitive
LAB-ELISAs were performed to obtain dose-response curves for
eight PAHs on plates coated with each of six PAH-BSA conju-
gates that gave classical sigmoid dose responses. The I50 values
(
27) Carlson, R. E. Hapten versus competitor design strategies for immunoassay
development. In Immunoanalysis of Agrochemicals: Emerging Technologies;
Nelson, J. O., Karu, A. E., Wong, R. B. Eds.; ACS Symposium Series 586;
American Chemical Society: Washington, DC, 1995; pp 140-152.
306 Analytical Chemistry, Vol. 71, No. 2, January 15, 1999

Ta b le 3 . Cro s s -Re a c t ivit y o f P a re n t An a lyt e s a n d
Ha p t e n s a
analyte
I50 (nM)
hapten
I50 (nM)
I50 ratiob
naphthalene
anthracene
fluoranthene
pyrene
370
46
336
34
NAP-2a
ANT-9a
FLA-3a
PYR-1a
PYR-1b
CHR-6a
BaP-6b
120
742
36
18
2
58
14
29
99
3.1
0.2
9.3
1.9
17
0.3
0.8
0.4
0.1
chrysene
benzo[a]pyrene
20
11
BaP-6-CHO
BaP-1a
a
Indirect LAB-ELISAs using MAb 10C10 culture supernate diluted
1
:6000 and the coating antigen PYR-1a-BSA at 0.5 ng/ 100 µL per well.
b
Data were means of four well replicates. I50 ratio, (I50 of a PAH/ I50 of
the corresponding hapten).
Fig u re 2 . Standard curve of inhibition by BaP using MAb 10C10 in
the indirect LAB-ELISA. The plate was coated with 0.5 ng/100 µL
per well PYR-1a-BSA, and MAb 10C10 was diluted 1:6000. Error
bars, ( one standard deviation. Inhibition percent, 100(A0 - A)/A0,
where A0 is the absorbance with no BaP present and A is the
absorbance with BaP present. The data were means of four replicates.
assays for the largest number of different PAHs were obtained
with the PYR-1a-BSA conjugate that was most weakly recognized
in the experiment of Table 2. The least sensitive assays were
obtained for the two- and three-ring PAHs using the FLA-3a- and
NAP-2a-BSA conjugates as competitors. However, some of the
largest differences between PAHs were evident from the less
sensitive assays. New antibodies with greater sensitivity for the
two- and three-ring PAHs should provide information that could
be used to identify PAHs in conjunction with the results obtained
using MAb 10C10.
Similar indirect LAB-ELISAs were done to compare recognition
of the haptens and the parent analytes (Table 3). The results show
enhanced recognition by MAb 10C10 of NAP-2a, FLA-3a, PYR-1a
and PYR-1b (I50 ratio >1), suggesting some binding interaction
with the spacers on these haptens. By contrast, haptens ANT-9a,
CHR-6a, BaP-6b, BaP-6-aldehyde, and BaP-1a were not recognized
as well as the corresponding parent compounds (I50 ratio <1),
indicating that the spacers on these haptens reduced the affinity
of MAb 10C10 binding. However, the haptens ANT-9a, PYR-1b,
and BaP-6b had the same â-alanine spacer, which suggests that
the observed affinity differences arise from different fitting of the
PAH moiety in the antibody’s binding pocket.
The cross-reactivity data (Figure 3, Table 3) demonstrate the
well-established principle that changing the competing hapten is
a way to modulate the sensitivity and cross-reactivity pattern of a
particular antibody in an ELISA.2
7-30
Another way to acquire the
Fig u re 3 . Half-maximal inhibition (I50) as a measure of relative
binding of MAb10C10 to eight PAHs in indirect LAB-ELISAs with six
different competitor conjugates (coating antigens). The shortest bars
are the most sensitive responses. The plate was coated with 2.0 ng/
cross-reactivity data is to use antibodies with differing cross-
PAH detection appears to be a good model system
to explore the use of both methods. Selection of an appropriate
reactivities.3
1-35
100 µL per well PAH-BSA except NAP-2a-BSA (2.0 pg/100 µL per
well), and MAb 10C10 was diluted 1:6000. Overall standard deviations
averaged 10.8% of mean values. The numerical data for this figure
are available from the authors upon request. The data were means
of four replicate experiments.
(
28) Van Weemen, B. K.; Schuurs, A. H. W. M. Sensitivity and specificity of hapten
enzyme-immunoassays. In First International Symposium on Immunoenzy-
matic Techniques; Feldmann, M., et al., Eds.; North-Holland Publishing
Co.: Amsterdam, 1976; Vol. 2, pp 125-133.
29) Szurdoki, F.; Bekheit, H. K. M.; Marco, M.-P.; Goodrow, M. H.; Hammock,
B. D. Important Factors in Hapten Design and Enzyme-Linked Immunosor-
bent Assay Development. In New Frontiers in Agrochemical Immunoassay;
Kurtz, D. A., Skerritt, J. H., Stanker, L. Eds.; AOAC International: Arlington,
VA, 1995; pp 39-63.
30) Goodrow, M. H.; Sanborn, J. R.; Stoutamire, D. W.; Gee, S. J.; Hammock,
B. D. Strategies for Immunoassay Hapten Design. In Immunoanalysis of
Agrochemicals: Emerging Technologies; Nelson, J., Karu, A. E., Wong, R.,
Eds.; ACS Symposium Series 586. American Chemical Society: Washington,
DC, 1995; pp 119-139.
(
were determined as a measure of the assay sensitivity. The results,
shown in Figure 3, illustrate the differences in cross-reactivity
pattern for each coating antigen. The shorter the bar, the more
sensitive was the detection. The I50 values varied from 42 nM to
(
(
3
05 µM for different PAHs with two to five fused rings. BaP was
the most sensitively detected PAH regardless of the coating
conjugate, probably because MAb 10C10 was originally evoked
by BaP-6a-BSA. The graph also shows that the most sensitive
31) Muldoon, M. T.; Fries, G. F.; Nelson, J. O. J. Agric. Food Chem. 1 9 9 3 , 41,
322-328.
Analytical Chemistry, Vol. 71, No. 2, January 15, 1999 307

Ta b le 4 . Co m p a ris o n o f P AH Cro s s -Re a c t ivit y o f MAb s
Ta b le 5 . Va ria t io n s a m o n g ELIS A Ru n s fo r t h e An a lys is
o f Ba P a
a
1
0 C1 0 a n d 4 D5
MAb 10C10
MAb 4D5
CV (%)
analyte
I50 (nM)
CR (%)
I50 (nM)
CR (%)
BaP conc (nM)
intraassay
interassay
benzo[a]pyrene
chrysene
pyrene
phenanthrene
anthracene
fluorene
fluoranthene
naphthalene
11 ( 4
20 ( 4
100
55
32
28
24
8
11 ( 5
24 ( 4
100
46
52
17
16
8
0
2.28
9.13
36.5
146
584
6.5
6.5
8.6
6.5
12.5
14.0
6.9
4.7
9.0
12.9
25.9
27.8
34 ( 6
21 ( 5
39 ( 5
46 ( 5
132 ( 17
336 ( 43
370 ( 52
64 ( 8
71 ( 9
144 ( 21
313 ( 48
292 ( 37
3
3
4
4
a
Indirect LAB-ELISAs using MAb 10C10 culture supernate diluted
1
:6000 and the coating antigen PYR-1a-BSA at 0.5 ng/ 100 µL per well.
a
Data were mean values from four plates and eight replicate wells for
each concentration.
Indirect LAB-ELISAs with culture supernate diluted 1:6000 for MAb
1
0C10 and 1:4000 for MAb 4D5. PYR-1a-BSA (0.5 ng/ 100 µL per well)
was used as coating antigen. The I50 values were means ( standard
deviations for 4 plates and eight wells per data point per plate.
these three. Thus, MAbs 10C10 and 4D5 have essentially the same
PAH recognition patterns, despite the amino acid sequence
differences, and only MAb 10C10 was further investigated.
Effect of Organic Solvents. The effects of methanol, acetone,
and DMSO on the indirect competitive LAB-ELISA were exam-
ined. Samples containing up to 2% (v/ v) acetone or DMSO or up
to 5% methanol could be used with no significant change in OD
readings compared with controls in PBS (data not shown). This
solvent tolerance is sufficient for analyses of solvent extracts of
many environmental samples.
Assay Variation. Reproducibility and precision are important
criteria of an immunoassay. Standard curves for the indirect
competitive LAB-ELISA of BaP from four consecutive days were
compared and variations were calculated (Figure 2). Detection
was linear with the log of BaP concentration from 0.4 to 1000 nM.
The variations in replicates from well to well (intraassay) and plate
to plate (interassay) were measured (Table 5).36 Intraassay
variations were generally lower than interassay variations. Varia-
tions in the standards were greatest at the lowest BaP concentra-
combination of antibody-antigen is an essential step to develop
a sensitive ELISA. After an antibody is developed, the assay
sensitivity and specificity are primarily governed by the relative
affinity of antibody-antigen and antibody-analyte. A relatively
low affinity of 10C10 to PYR-1a-BSA (I50 ) 52 ppb) was
demonstrated to afford a sensitive assay for PAHs (Table 2, Figure
3
). A linker substituent of PAH haptens and the linker and BSA
moieties of PAH-BSAs influenced antibody recognition as shown
by I50 changes (Tables 2 and 3). Such changes probably resulted
from factors such as linker type and length of linker, position on
the hapten, and hapten density on the conjugates. Interestingly,
among the PAHs and their corresponding haptens and hapten-
BSA conjugates, little I50 changes of pyrene, PYR-1a, and PYR-
1
a-BSA (34 nM, 18 nM, and 52 ppb, respectively) indicated
negligible contribution of the linker and BAS moieties on 10C10
recognition. I50 values of PAH conjugates are more direct and
reliable criteria than those of free haptens or parent PAHs to guide
the selection of coating antigen for assay development. In addition,
relative inhibitions of analytes and the corresponding haptens and
hapten conjugates should be considered as illustrated in Tables
tions. This nonlinearity of variance with concentration is common
in competitive ELISA.3
7-39
However, the CVs for samples increased
with increasing BaP concentrations (Table 5). Factors that
contribute variations include quality of the hapten, coating, plate
wells and multichannel pipettor, edge effects due to evaporation,
uneven temperature during incubation, and day-to-day variations
in the preparation of reagents. A standard curve should be created
every time an assay is performed to reduce the variations.
Analysis of Spiked Samples. PBS buffer, tap water, Ala Wai
canal water, and human urine from a nonsmoker were spiked with
BaP to 2-600 nM (i.e., 0.5-150 ppb), and BaP concentrations
were determined by the LAB-ELISA using MAb 10C10 culture
supernate. The assay was shortened by adding both the MAb
2
and 3.
Comparison of MAbs 1 0 C1 0 and 4 D5 . In addition to MAb
0C10, a second MAb, 4D5, developed by Gomes and Santella
1
from BaP-6a-BSA immunogen, was examined. MAbs 4D5 and
0C10 differ by 23 amino acids but have similar equilibrium
1
dissociation constants for BaP (C. W. Bell and A. E. Karu, in
preparation). To determine whether the sequence differences
affected selectivity of these two MAbs, indirect competition LAB-
ELISAs were done with PYR-1a-BSA as coating antigen, and the
I
50 values for various PAHs were compared (Table 4). Both
(
1:3000) and biotinylated goat anti-mouse IgG (1:6000) to the
antibodies showed similar patterns of recognition. Analysis of
variance using one-way ANOVA (R ) 0.05) revealed no significant
difference between I50 values obtained with 10C10 and 4D5 except
for pyrene, phenanthrene, and anthracene. The differences in
percent cross-reactivity are too small to be of practical use for
analyte mixtures, combining two incubation steps into one.
Recoveries of the BaP spikes ranged from 81 to 127% in all samples
(
Table 6).
(36) Bookbinder, M. J.; Panosian, K. J. Clin. Chem. 1 9 8 6 , 32, 1734-1737.
(
37) Wellington, D. Statistical Aspects of Enzyme Immunoassay. In Enzyme
Immunoassay; Maggio, E. T., Ed.; CRC Press: Boca Raton, FL, 1980; pp
249-273.
(
(
(
(
32) Cheung, P. Y. K.; Kauvar, L. M.; Engqvist-Goldstein, A. E.; Ambler, S. M.;
Karu, A. E.; Ramos, L. S. Anal. Chim. Acta 1 9 9 3 , 282, 181-192.
33) Jones, G.; Wortberg, M.; Hammock, B. D.; Rocke, D. M. Anal. Chim. Acta
(38) Canellas, P. F.; Karu, A. E. J. Immunol. Methods 1 9 8 1 , 47, 375-385.
(39) Karu, A. E.; Perman, M.; McClatchie, I. R. T.; Speed, T. P.; Richman, S. J.
AutoElisa: A data management system for regulatory and diagnostic
immunoassays, using parallel fitting for data evaluation. In New Frontiers
in Agrochemical Immunoassay; Kurtz, D. A., Skerritt, J. H., Stanker, L., Eds.;
AOAC International: Arlington, VA, 1995; pp 261-282.
1
9 9 6 , 336, 175-183.
34) Wortberg, M.; Kreissig, S. B.; Jones, G.; Rocke, D. M.; Hammock, B. D.
Anal. Chim. Acta 1 9 9 5 , 304, 339-352.
35) Karu, A. E.; Lin, T. H.; Breiman, L.; Muldoon, M.; Hsu, J. Food Agric.
Immunol. 1 9 9 4 , 6, 371-384.
308 Analytical Chemistry, Vol. 71, No. 2, January 15, 1999

Ta b le 6 . Re c o ve rie s o f Ba P in S p ik e d S a m p le s a
CONCLUSION
The new approach of hapten screening used in this study
should be applicable to new immunoassay development and
optimization for other small molecules of environmental signifi-
cance. The LAB-ELISA was optimized by cross-screening PAH-
BSA conjugates as competitors in solution and as coating antigens
immobilized on the ELISA wells. A recognition index was defined
to compare relative affinities of the antibodies to the PAH
conjugates and roughly to predict assay sensitivities with a given
conjugate as coating antigen. PYR-1a-BSA, which was identified
as the least competitive inhibitor among 11 PAH conjugates, gave
the most sensitive assay for BaP and other PAHs when it was
used as coating antigen. The indirect LAB-ELISA format was very
sensitive because the biotinylated second antibody could accom-
modate multiple copies of the avidin-HRP probe. The resulting
amplification allowed us to reduce the concentrations of coating
conjugate and MAb.
recovery determined by LAB-ELISA (% ( SD b)
spiked level
(
nM)
buffer
tap water
canal water
urine
2
4
84 ( 8
96 ( 7
96 ( 12
86 ( 9
97 ( 5
81 ( 9
84 ( 8
85 ( 8
87 ( 9
94 ( 10
96 ( 9
118 ( 17
113 ( 11
85 ( 10
89 ( 10
127 ( 19
127 ( 9
118 ( 13
2
4
2
4
6
0
0
00
00
00
105 ( 7
83 ( 9
86 ( 7
92 ( 5
85 ( 7
106 ( 5
100 ( 7
80 ( 9
98 ( 5
117 ( 15
av
95
88
93
111
a
BaP in spiked samples was directly analyzed by LAB-ELISA without
sample preparation. MAb 10C10 culture supernate was diluted 1:6000
and the coating antigen was PYR-1a-BSA at 0.5 ng/ 100 µL per well.
b
Values were means ( standard deviations for four replicates.
Although absolute immunospecificity for a single analyte is
desirable, it may be impractical to achieve for highly similar
molecules such as the two- to five-ring PAHs. The present work
explored the potential of using antibodies that cross-react and
changing the sensitivity and cross-reactivity by using different
competing haptens. MAbs 4D5 and 10C10 might be used to
complement the reaction patterns of other PAH antibodies as
components of more complex multianalyte PAH assays and
sensors with the new PAH haptens described here. The reactivity
profiles with the new haptens suggest additional combinations of
antibodies and conjugates (for example, antibodies with strong
preference for naphthalene, fluorene, or fluoranthene) that would
be useful together with MAbs 4D5 and 10C10.
Fig u re 4 . Correlation of BaP concentrations measured by GC and
those directly by ELISA using MAb10C10 for fortified canal water
samples (100 nM of BaP ) 25 ppb). The plate was coated with 0.5
ng/100 µL per well PYR-1a-BSA, and MAb 10C10 was diluted
ACKNOWLEDGMENT
1:6000. The data were means of four replicate experiments.
This work was supported by DOE Grant DE-FG07-96ER-62316,
NIEHS Grant ES01896, and ONR Grant N00014-94-1-0312. We
especially thank R. M. Santella of Columbia University, New York,
for providing the 4D5 and 10C10 hybridoma lines, T. Vo-Dinh
Canal water samples fortified with 20-400 nM BaP (i.e., 5-100
ppb) were split for ELISA and GC determination. The portions
for ELISA were analyzed with no concentration or cleanup. The
portions for GC analysis were extracted with ethyl acetate and
concentrated. The concentrations determined by ELISA (26-400
nM) agreed well with those by GC (24-411 nM) (Figure 4). The
least-squares fit had a slope of 0.976 and a coefficient of determi-
(
Oak Ridge National Laboratory) for BaP-6a, M. David in Q.X.L.’s
laboratory for MS analysis, and D. Vogel in A.E.K.’s laboratory
for preparing the hybridoma culture supernates and purified IgG.
2
nation (r ) of 0.998. The ELISA detected BaP as low as ∼10 pg/
Received for review July 13, 1998. Accepted October 29,
1998.
well. The GC method can measure ∼1.5 ng/ injection. Therefore,
the ELISA was ∼150-fold more sensitive than GC for these
analyses.
AC980765D
Analytical Chemistry, Vol. 71, No. 2, January 15, 1999 309