Development of a Panel of Immunoassays for Monitoring DDT, Its Metabolites, and Analogues in Food and Environmental Matrices

Article abstract of DOI:10.1021/jf9802934

A panel of antisera was prepared using analogues and derivatives of metabolites of the organochlorine insecticide, p,p′-DDT as haptens. The assays developed exhibited differing cross-reactions for different DDT analogues and metabolites, and the choice of

Full text of DOI:10.1021/jf9802934

J. Agric. Food Chem. 1998, 46, 3339−3352
3339
Develop m en t of a P a n el of Im m u n oa ssa ys for Mon itor in g DDT, Its
Meta bolites, a n d An a logu es in F ood a n d En vir on m en ta l Ma tr ices
Helen L. Beasley,† Thipsavanh Phongkham,‡ Margaret H. Daunt,‡ Simone L. Guihot,† and
J ohn H. Skerritt*,‡
CSIRO Plant Industry, North Ryde, NSW 2113, Australia, and CSIRO Plant Industry, Canberra,
ACT 2601, Australia
A panel of antisera was prepared using analogues and derivatives of metabolites of the organochlorine
insecticide, p,p′-DDT as haptens. The assays developed exhibited differing cross-reactions for
different DDT analogues and metabolites, and the choice of hapten for the detecting enzyme conjugate
had almost as much effect on assay specificity and sensitivity as the structure of the hapten used
for antibody production. Those assays developed using hapten I, based on esters of bis(p-
chlorophenyl)acetic acid (DDA), typically detected DDA with greater sensitivity than p,p′-DDT or
p,p′-DDE. The most sensitive assay for p,p′-DDT (lower limit of detection of 0.3 µg/L) was obtained
using an immunogen based on bis(p-chlorophenyl)ethanol (hapten IV), although a significant cross-
reaction with dichlorodiphenyltrichloroethane (DDD) and DDE was obtained. The most specific
assay for p,p′-DDT was obtained using an immunogen (hapten VI) that includes all elements of the
DDT structure, except that one of the p-chloro groups was replaced by â-alanine carboxamide for
coupling to carrier proteins. Antibodies based on a similar DDE hapten (V) exhibited specificity
for p,p′-DDE over p,p′-DDT. Greater specificity and sensitivity for dicofol were obtained by using
an immunogen derived from ester hydrolysis of chlorbenzilate (hapten II). The assays provided
methods for detection of p,p′-DDT plus p,p′-DDE either by using the antibody raised to hapten IV
with conjugate based on hapten Ib or by using the assay based on hapten V, with treatment of
samples with warm alcoholic KOH, which converted DDT to DDE. Some of the immunoassays
were applied to the detection of DDT and DDE in water, soil, and selected foods.
Keyw or d s: DDT; DDE; organochlorine; ELISA; immunoassay; water; soil; food
INTRODUCTION
countries restricting, and eventually abandoning, the
use of DDT. Despite not being used by many countries
for some years, the persistence of DDT and its metabo-
lites in soil [including DDE and 1,1-dichloro-2,2-bis(p-
chlorphenyl)ethane (DDD); Guenzi and Beard, 1967;
Agarwal et al., 1994] means that residues remain of
concern when landuse changes for potentially contami-
nated sites are considered. In humans, and other
mammals, the major metabolite of DDT is bis(p-
chorophenylacetic acid) (DDA). DDA itself does not
accumulate but is a marker for exposure to DDT
(Hassall, 1990). Some developing countries such as
India and Mexico have continued to use DDT, especially
for public health applications (WHO, 1969), leading to
the potential for considerable buildup in the environ-
ment and hazards in foodstuffs (Mukherjee et al., 1993;
Boul, 1995; Waliszewski et al., 1996). The use of dicofol
[the 2-hydroxy analogue of DDT, namely, 1,1,1-trichloro-
2-hydroxy-2,2-bis(p-chlorophenyl)ethane], an effective
miticide and acaricide, has not been subject to the same
restrictions, and it is widely used on horticultural crops.
However, a considerable problem existed with many
production batches of dicofol, especially those manufac-
tured before the mid-late 1980s, with several of these
containing a significant percentage of DDT as a con-
taminant (Gillespie et al., 1994). In addition, the
biodegradation patterns of dicofol suggest that it too can
persist as residues in foods and environmental samples
(Liapis et al., 1995; Domaglaski, 1996), and dicofol has
been associated with egg thinning in wild bird species
The insecticidal properties of 1,1,1-trichloro-2,2-bis-
(p-chlorophenyl)ethane (DDT) were first recognized by
Muller of Switzerland in 1939, who was awarded the
Nobel Prize in Medicine and Physiology for his discov-
ery. DDT was patented as the first synthetic multipur-
pose insecticide the same year and applauded for its
lethal effects, persistence, and low nontarget toxicity
(Waterhouse, 1972). DDT proved economical and ver-
satile for use in both agricultural and public health
applications. In the 1960s, sensitive analytical tech-
niques such as gas-liquid chromatography (de Faubert
Maunder et al., 1964; Edwards, 1973) revealed the
bioaccumulation of DDT in the fat of animals higher in
the food chain (Brewerton, 1969). This problem of
biological accumulation is aggravated by the natural
conversion of DDT to the even more stable noninsecti-
cidal product, 1,1-dichloro-2,2-bis(p-chlorophenyl)eth-
ylene (DDE). This compound has recently been asso-
ciated with hormone-related cancers in humans (Kelce
et al., 1995). Detection of DDT and metabolites in milk
and meat for human consumption, together with the
appearance of insect resistance to DDT, led to many
* Address correspondence to this author at CSIRO Plant
Industry, G.P.O. Box 1600, Canberra, ACT 2601, Australia
(telephone 61 2 6246 5350; fax 61 2 6246 5351/5000; e-mail
J .Skerritt@pi.csiro.au).
† CSIRO, North Ryde.
‡ CSIRO, Canberra.
S0021-8561(98)00293-3 CCC: $15.00 © 1998 American Chemical Society
Published on Web 07/28/1998

3340 J. Agric. Food Chem., Vol. 46, No. 8, 1998
Beasley et al.
F igu r e 1. Chemical structures of the four target analytes.
(Schwarzbach et al., 1988; Clark et al., 1995). As a
result, several food-importing countries are concerned
about its continuing use (U.S. EPA, 1985).
Immunological methods for detecting insecticide resi-
dues in food and environmental samples offer several
advantages over gas chromatographic methods. While
maintaining comparable sensitivity, they are simple and
cost-effective and can be adapted for field use (Gee et
al., 1995; Lee et al., 1995). Earlier, several groups have
attempted to develop antibodies for detection of DDT
and metabolites by forming esters or amides of DDA
(Haas and Guardia, 1968; Centero et al., 1970; Furuya
and Urasawa, 1981; Banerjee, 1987); however, typically
these antibodies detected DDT much more poorly than
they did DDA. Burgisser et al. (1990) designed a hapten
based on dicofol, which, unlike the DDA-derived hap-
tens, retained the -CCl3 moiety of DDT, and developed
a competitive radioimmunoassay of low to moderate
sensitivity. Banerjee et al. (1996) utilized diamino
derivatives of DDT, DDE, and DDA as haptens and
developed several competitive ELISAs with IC50 values
in the range of 180-360 µg/L for standards of each of
these pesticides. In the late stages of completion of this
manuscript, a report was published describing the
development of quite sensitive monoclonal antibodies
to DDT and DDE (Abad et al., 1997), although in none
of these reports was application to samples described.
In the present study, we describe the development of a
panel of antibodies to different DDT analogues and
metabolites using a series of haptens, and their incor-
poration into ELISA assays. Some of these assays
exhibited cross-reactions with other stable DDT me-
tabolites, making them useful as a screening tool for
determining total DDT loads, whereas others were more
specific for either DDT, DDE, DDA, or dicofol. We
report initial results demonstrating the application of
assays for residue analysis in selected food and envi-
ronmental matrices.
F igu r e 2. Chemical structures of the haptens prepared
(carboxylic acid forms shown).
ethyl ester] to develop an analogue of dicofol (hapten II); (3)
utilization of dicofol to develop a derivative that retains the
1,1,1-trichloro-2,2-bis(p-chlorophenyl) functionality of DDT
(hapten III); (4) utilization of bis(p-chlorophenyl)ethanol (hap-
ten IV); (5) a DDE derivative (hapten V); and (6) a DDT
derivative through synthesis of an analogue bearing a substi-
tution at one of the p-chloro atoms (hapten VI) (Figure 2).
1. Synthesis of DDA-Derived Haptens (Haptens Ia , Ib, and
Ic). Bis(2-p-chlorophenyl)acetic acid (0.32 mmol, 78 mg) was
reacted with thionyl chloride (2 mL, caution) and refluxed at
90 °C for 1 h. Excess thionyl chloride was then removed by
rotary evaporation, toluene was added to remove the thionyl
chloride residue, and the solvent was removed by a second
evaporation step. The product was dissolved in 2 mL of
benzene and cooled to 0 °C; â-alanine (72 mg) in 2 mL of 1 M
NaOH solution was added, and the mixture was allowed to
react overnight. Unreacted acid chloride was removed by
solvent extraction with benzene and the aqueous layer acidi-
fied, then partitioned into ethyl acetate, and washed with
water and then brine. The product was dried over magnesium
sulfate and filtered under vacuum. It was reconstituted in
acetone and purified by flash chromatography on silica (eluted
in 50% ethyl acetate/50% petroleum ether with 0.01% acetic
acid). The fractions identified as containing the acid by TLC
EXPERIMENTAL METHODS
Ha p ten P r ep a r a tion a n d Con ju ga tion . We aimed to
develop immunoassays with differing specificities for p,p′-DDT,
DDA, p,p′-DDE, and dicofol (Figure 1). Five classes of haptens
were produced based on (1) formation of various amide
derivatives of DDA (haptens Ia , Ib, and Ic); (2) the utilization
of chlorbenzilate [2-hydroxy-2,2-bis(p-chlorophenyl)acetic acid

Immunoassays for DDT, Its Metabolites, and Analogues
J. Agric. Food Chem., Vol. 46, No. 8, 1998 3341
analysis (ethyl acetate/petroleum ether, 1:1) were combined,
and toluene was added to remove the acetic acid residue as
an azeotrope, followed by addition of chloroform to remove
residual toluene, to provide the â-alanine derivative (48%
yield). The product was detected on TLC using 2% CeSO4 in
2 M sulfuric acid. The â-alanine amide (0.13 mmol, 46 mg)
was combined with N-hydroxysuccinimide (NHS; 0.3 mmol,
32 mg) and 5 mL of dry methylene chloride, at 0 °C. Dicyclo-
hexylcarbodiimide (DCC; 0.16 mmol, 40 mg) was added,
followed by (dimethylamino)pyridine (DMAP; 1 mg) and al-
lowed to react overnight. The reaction mixture was filtered
and the white precipitate discarded; the filtrate was shaken
in slightly acidified water, saturated with sodium hydrogen
carbonate, water, and brine, and dried over magnesium
sulfate; the product was filtered off and reconstituted in
chloroform for preparative TLC (in ethyl acetate/petroleum
ether, 1:1) on silica gel (Rf 0.62, 80% yield, mp 95-97 °C)
[hapten Ia , 3-[bis(4-chlorophenyl)acetylamino]propionic acid]:
1H NMR (CD3OD) of acid δ 2.33 (t, 2H, -CH2-COO), 3.17
(m, 2H, -CH2-NHCO), 4.78 (s, H, -CH-), 7.15 (4H, Ar, J )
Ar, J ) 6.45 Hz). The resultant carboxylic acid was converted
to the NHS ester using the method described above to yield
an oil: 1H NMR (CDCl3) δ 1.32 (t, 2H, -CH2-), 3.62 (q, 2H,
-CH2-NH-), 2.85 [t, 4H, -CO(CH2)2-CO-], 5.1 (s, br, 1H,
-NH), 7.33 (d, 4H, -Ar, J ) 9.0 Hz), 7.38 (d, 4H, Ar, J ) 9
Hz).
3. Utilization of Dicofol To Produce Hapten III. This
synthesis was similar to that described by Burgisser et al.
(1990). Dicofol (3.5 g, 9.5 mmol) was combined with 3-bro-
mopropionitrile (11.4 g, 85 mmol) in 2.5 mmol of concentrated
sulfuric acid, and the reagents were stirred at room temper-
ature for 2 days. The mixture was extracted into ethyl acetate
and washed with 1 M sodium carbonate solution and brine,
before drying over magnesium sulfate. The bromo derivative
product was recrystallized from hot ethanol and shown to be
chromatographically pure by TLC in 10% ethyl acetate in
petroleum ether. This bromo compound was converted to the
nitrile by reaction of 0.6 mmol of I with 0.8 mmol of NaCN in
2 mL of dimethyl sulfoxide at 75 °C for 2 h (Friedman and
Shechter, 1960). The mixture was washed with ethyl acetate,
water, 1 M NaHCO3, and then brine. The yellow residue from
filtration and evaporation was taken up in chloroform and
analyzed by TLC in 40% ethyl acetate in petroleum ether. The
presence of the nitrile group in this intermediate was con-
firmed by IR spectroscopy. The nitrile was hydrolyzed to the
carboxylic acid by dissolving in a 15 mol excess of concentrated
HCl and then stirred at room temperature for 5 days. Two
liters of water was then added, and the product was extracted
into ethyl acetate; after washing in water and brine, and
drying over MgSO4, solvent was removed to produce an oil,
hapten III [3-[2,2-bis(4-chlorophenyl)-3,3,3-trichloropropanoy-
lamino]propanoic acid]: 1H NMR (CDCl3) of acid δ 2.51 (t, 2H,
-CH2-CO-N-), 2.97 (t, 2H, -CH2-COO-), 7.28 (d, 4H, Ar,
1
8.21 Hz), 7.24 (4H, Ar, J ) 8.3 Hz); H NMR (CDCl3) of NHS
ester δ 2.55 (t, 2H, -CH2-COO), 2.85 [m, 4H, CO(CH2)2-CO],
3.52 (m, 2H, -CH2-NH), 4.79 (s, 1H, CH), 7.19 (d, 4H, Ar, J
) 9.0 Hz), 7.33 (d, 4H, Ar, J ) 9.0 Hz).
Related haptens were prepared using similar methods, and
a similar molar excess of amino acid used for the spacer, with
the following: (1) A γ-aminobutyric acid spacer [hapten Ib,
4-[bis(4-chlorophenyl)acetylamino]butanoic acid]: 1H NMR
(CDCl3) of acid δ 1.8 (m, 2H, -CH2-), 2.34 (t, 2H, -CH2-
COO), 3.32 (m, 2H, -CH2-NHCO), 4.8 (s, H, -CH-), 5.93 (t,
1H, NH), 7.15 (4H, Ar, J ) 8.34 Hz), 7.29 (4H, Ar, J ) 8.4
Hz); 1H NMR (CDCl3) of NHS ester δ 2.0 (m, 2H, -CH2-),
2.61 (t, 2H, -CH2-COO), 2.85 (m, 4H, CO(CH2)2-CO), 3.4 (m,
2H, -CH2-NH), 4.82 (s, 1H, CH), 7.17 (d, 4H, Ar, J ) 9 Hz),
7.30 (d, 4H, Ar, J ) 9 Hz); NHS mp 110-113 °C. (2)
4-(Aminomethyl)cyclohexylcarboxylic acid [hapten Ic, 4-[[bis-
(4-chlorophenyl)acetylamino]methyl]cyclohexylcarboxylic acid],
using the methods of Hill et al. (1993), instead of â-alanine:
1H NMR (CDCl3) of acid δ 0.95 (m, 2H, CH2-ax), 1.4 (m, 3H,
CH + CH2-ax), 1.75 (m, 2H, CH2-eq), 2.01 (m, 2H, CH2-eq),
2.25 (m, 1H, CH-eq), 3.15 (t, 2H, -CH2-N), 4.72 (s, 1H, -CH-
), 5.60 (t, 1H, -NH-), 7.17 (d, 4H, Ar, J ) 8.4 Hz), 7.30 (d,
4H, Ar, J ) 10.5 Hz); 1H NMR (CDCl3) of NHS ester δ 1.0 (m,
2H, CH2-ax), 1.53 (m, 3H, CH + CH2-ax), 1.76 (m, 2H, CH2-
eq), 2.45 (m, 2H, CH2-eq), 2.55 (m, 1H, CH-eq), 2.82 (s, 4H,
CO-CH2CH2-CO), 3.14 (t, 2H, -CH2-N), 4.81 (s, 1H, -CH-
), 5.78 (t, 1H, -NH-), 7.17 (d, 4H, Ar, J ) 8.46 Hz), 7.31 (d,
4H, Ar, J ) 8.37 Hz); NHS mp 102-103 °C.
2. Hydrolysis of Chlorbenzilate To Produce Hapten II.
Chlorbenzilate (50 mg, 0.154 mmol) was added to 2 mL of 2
M KOH in ethanol and stirred for 2 h at room temperature.
After the reaction appeared to go to completion by TLC in ethyl
acetate/petroleum ether (30:70), a further 10 mL of KOH
solution was added and unreacted chlorbenzilate removed by
extraction with dichloromethane. The aqueous layer was
removed and then acidified to pH 4, and the acid product was
extracted into dichloromethane and dried over magnesium
sulfate. The product (40 mg, 0.135 mmol, yield 88%) in 2 mL
of dichloromethane was stirred at 0 °C with DCC (33.5 mg,
0.16 mmol) and (dimethylamino)pyridine (1 mg, 0.007 mmol)
for 30 min before addition of tert-butyl-â-alanine ethyl ester
(21.5 mg, 0.15 mmol), and the mixture stirred for 4 h. The
dicyclohexylurea was removed by filtration and the dichlo-
romethane solvent removed under reduced pressure. The
product was extracted into ethyl acetate and washed with 1
M sodium hydrogen carbonate, water, and brine before being
dried over MgSO4; the ethyl acetate was removed under
reduced pressure. The product was isolated by radial chro-
matography using 30% ethyl acetate in petroleum ether and
then hydrolyzed to the acid by dissolving in 1 mL of trifluo-
roacetic acid and stirring for 2 h at room temperature, to
produce the acid, hapten II [3-[[bis(4-chlorophenyl)hydroxy-
acetyl]amino]propanoic acid] in 55% yield: 1H NMR (CDCl3)
of acid δ 1.27 (t, 2H, -CH2-), 4.32 (q, 2H, -CH2-NH-), 5.1
(s, br, 1H, -NH), 7.30 (d, 4H, -Ar, J ) 9.0 Hz), 7.35 (d, 4H,
1
J ) 9.0 Hz), 7.49 (d, 4H, Ar, J ) 9.0 Hz); H NMR (CDCl3) of
NHS ester δ 2.69 (t, 2H, -CH2-CO-N-), 2.85 (m, 4H, CO-
CH2CH2-CO), 3.15 (m, 2H, -CH2-CO-O-), 7.29 (d, 4H, Ar,
J ) 7.8 Hz), 7.53 (d, 4H, Ar, J ) 6.66 Hz).
4. Utilization of Bis(p-chlorophenyl)ethanol To Produce
Hapten IV. 2,2′-Bis(4-chlorophenyl)ethanol (100 mg, 0.37
mmol) was reacted with succinic anhydride (375 mg, 3.74
mmol) in 5 mL of dry pyridine with 5 mg of DMAP for 16 h at
room temperature. After addition of 20 mL of water, the
water/pyridine mixture was removed by rotary evaporation
and then the residue rinsed with toluene and solvent removed
by evaporation. The residue was then dissolved in ethyl
acetate and washed with 1 M HCl, water, and brine, before
evaporation and drying over MgSO4 to produce the hemisuc-
cinate in 92% yield. This acid was then activated with NHS,
and the product, hapten IV [mono bis(2,2-(4-chlorophenyl)ethyl
ester of butanedioate] was isolated using radial chromatog-
raphy in ethyl acetate/petroleum ether (40:60) to provide the
NHS ester in 78% yield: 1H NMR (CDCl3) of acid δ 2.58 (dd,
4H, -CO2-CH2-CH2-CO2-), 4.31 (t, 1H, -CH), 4.58 (d, 2H,
-O-CH2-), 7.12 (d, 4H, Ar, J ) 8.46 Hz), 7.27 (d, 4H, Ar, J
) 8.52 Hz); 1H NMR (CDCl3) of NHS ester δ 2.67 (t, 2H,
-CH2-), 2.83 (m, 6H, 3 × -CH2), 4.38 (t, 1H, -CH-), 4.60
(d, 2H, CH2-OCO-), 7.13 (d, 4H, Ar, J ) 8.31 Hz), 7.28 (d,
4H, Ar, J ) 8.52 Hz); NHS mp 98-100 °C.
5. Synthesis of Analogues of DDE (Hapten V) and DDT
(Hapten VI), Coupled through the Ring. Initially, 2-(p-chlo-
rophenyl)-2-(p-tolyl)-1,1,1-trichloroethane was synthesized ac-
cording to the method of Chattaway and Muir (1934).
A
p-tolyltrichloromethylcarbinol (Dinesman, 1905) was first
synthesized by slow addition of 0.1 mol of toluene to a
vigorously shaken solution of chloral hydrate (0.3 mol) in 35
mL of concentrated sulfuric acid. The mixture was shaken
for 2 h, and the emulsion so formed was poured into excess
ice-water. The oil that separated was washed with water and
steam-distilled to yield a colorless oil [bp 155-157 °C, 13.5
mmHg (lit. 155 °C)], which solidified to give a solid with mp
60-61 °C (lit. 63 °C) in 25% yield. The carbinol was dissolved
in a 1.1 molar excess of chlorobenzene and an equal volume
of concentrated sulfuric acid added with vigorous shaking. An
emulsion formed, which was poured into ice-water. An oil
separated and was washed with warm (40 °C) water until the

3342 J. Agric. Food Chem., Vol. 46, No. 8, 1998
Beasley et al.
wash was no longer acidic. The oil was then dissolved in
boiling ethanol and cooled, and after 2-3 days, 2-(p-chlorophe-
nyl)-2-(p-tolyl)-1,1,1-trichloroethane crystallized, mp 79-80 °C
(lit. 81 °C). The tolyl (methyl) group was oxidized to form the
carboxylic acid (Haskelberg and Lavie, 1949) by adding 2.7 g
of the p-tolyl compound to a mixture of 7.7 g of potassium
dichromate and 15 mL of water. Concentrated sulfuric acid
(17 mL) was added with stirring. This caused the mixture to
boil, and after 60 min of sitting, a product precipitated. The
reaction mix was diluted with 100 mL of water, and the
precipitate was washed with 30 mL of 5% sulfuric acid in
water. The precipitate was dissolved in 20 mL of 5% NaOH
in water and then reprecipitated by addition of excess dilute
sulfuric acid before being recrystallized from 50% ethanol. The
product had an mp of 92 °C (lit. 94-96 °C), and its identity
was confirmed by TLC analysis in benzene/dioxane/acetic acid
(200:6:3) (lit. Rf 0.33; Kapoor et al., 1973). For the production
of the DDE analogue [hapten V, 3-[4-[1-(4-chlorophenyl)-2,2-
dichloroethen-1-yl]benzoylamino]propanoic acid], it was then
dehydrohalogenated by treatment for 2 h at 60 °C in 10% KOH
in ethanol. This step was not performed in the synthesis of
the DDT analogue [hapten VI, 3-[4-[1-(4-chlorophenyl)-2,2,2-
trichloroethyl]benzoylamino]propanoic acid].
Subsequently, identical manipulations were performed in
the syntheses of haptens V and VI. Eighty milligrams (0.24
mmol) of the acid was dissolved in 2 mL of thionyl chloride
and refluxed at 90 °C for 1 h. The thionyl chloride was
removed by rotary evaporation, toluene was added to remove
the thionyl chloride residue, and the azeotrope was removed
by evaporation. The yield was 78%, and the product was used
without further purification. All of the acid chloride was
dissolved in 2 mL of benzene, and then 72 mg of â-alanine in
2 mL of 1 M NaOH solution was added and the mixture
allowed to react overnight. The product was obtained by
solvent extraction, benzene was removed, the aqueous layer
was acidified, and ethyl acetate was added followed by washing
with water and then brine; the product was dried using
magnesium sulfate and filtered under vacuum. It was recon-
stituted in acetone and purified by flash chromatography on
silica (eluted in 20% acetone/chloroform/0.01% acetic acid). The
fractions containing the acid by TLC analysis [ethyl acetate/
chloroform (1:1), Rf 0.13 for hapten V synthesis and 0.15 for
hapten VI synthesis] were combined, and toluene was added
to remove the acetic acid residue by evaporation of the
azeotrope, followed by addition of ethanol to remove residual
toluene as an azeotrope. The NHS esters were formed as
follows. The â-alanine amide (0.19 mmol, 75 mg) was com-
bined with 0.4 mmol (49 mg) of NHS and 5 mL of dry
methylene chloride, at 0 °C. DCC (0.23 mmol, 51 mg) was
added, followed by DMAP (1.2 mg, 0.01 mmol) and allowed to
react overnight. The reaction mixture was filtered and the
precipitate discarded; the filtrate was evaporated and recon-
stituted in ethyl acetate. The mixture was shaken in slightly
acidified water, saturated with sodium hydrogen carbonate,
water, and brine, and dried over magnesium sulfate; the
product was filtered off and reconstituted in chloroform for
preparative TLC, performed in ethyl acetate/chloroform (1:1)
(Rf 0.46 for hapten V synthesis and 0.51 for hapten VI
synthesis).
ArH, J ) 8 Hz), 7.51 (d, 2H, ArH, J ) 8 Hz), 7.65 (d, 2H, ArH,
J ) 8 Hz), 7.81 (d, 2H, ArH, J ) 8.1 Hz).
The NHS esters of each hapten were then coupled to
ovalbumin (OA), keyhole limpet hemocyanin (KLH), and
horseradish peroxidase (HRP) in 0.1 M potassium phosphate
buffer, pH 9.1 (McAdam et al., 1992), using 40-, 30-, and 13-
fold molar excess of haptens, respectively.
An tibod y P r od u ction . The OA and KLH conjugates of
each immunogen (1 mg/mL) were emulsified with an equal
volume of Freund’s complete adjuvant (Sigma, St. Louis, MO)
and injected half-subcutaneously, half-intramuscularly into
New Zealand White rabbits. Following two booster injections,
each 0.5 mg/mL, of immunogen in Freund’s incomplete adju-
vant (Sigma) 4 weeks apart, blood was collected from the
marginal ear vein 8-10 days after the last injection and clotted
to form serum. Boosting and bleeding were continued monthly;
IgG fractions from sera were purified by Protein G-Sepharose
(Pharmacia, Uppsala, Sweden) affinity chromatography (Ak-
erstrom et al., 1985)
P esticid e Sta n d a r d s. For the development of enzyme
immunoassays and specificity studies, the following analytical
grade standards were obtained from either Riedel-de-Haen,
Seelze, Germany, or the Australian Government Analytical
Laboratory, Sydney: bis(4-chlorophenyl)acetic acid (DDA),
p,p′-DDD, o,p′-DDT, o,p′-DDE, and 1,1,1-trichloro-2,2-bis(4-
chlorophenyl)ethane (p,p′-DDT). 2,2-Bis(4-chlorophenyl)-1,1-
dichloroethylene (p,p′-DDE) and the analogue 2,2-bis(4-
chlorophenyl)ethanol (DDT-ethanol) were obtained from Aldrich.
The following analytical grade compounds were obtained from
ChemService (West Chester, PA) for the cross-reaction study:
fluometuron, metobromuron, chlorbromuron, linuron, metoxu-
ron, monolinuron, diuron, neburon, bensulfuron-methyl, lin-
dane, heptachlor, aldrin, dieldrin, endrin, methoxychlor, chlor-
benzilate, phenothrin. Cypermethrin stock was prepared from
a 75% pure technical grade product (ICI, Australia). Cross-
reactivity (% ×) is calculated as the concentration of analyte
that causes a reduction of 50% in the assay color relative to a
pesticide-free control (IC50), expressed as a percentage of the
IC50 of the cross-reaction compound. Pesticide stocks were
prepared by dissolving in methanol at 1 mg/mL; assay
standards were prepared by diluting the stock solutions in
disposable borosilicate glass tubes and used within 30 min to
avoid pesticide loss through adhesion to glass surfaces.
ELISA Meth od s. All steps were performed at room tem-
perature (18-23 °C). Microwell plates (Maxisorp, Nunc,
Roskilde, Denmark) were coated for 16 h with antibody (100
µL per well) after dilution to 10 µg/mL in 50 mM sodium
carbonate buffer, pH 9.6. Microwells were washed three times
in 50 mM sodium phosphate/0.9% NaCl, pH 7.2 (PBS)/0.05%
Tween 20 (PBS-T), and then nonspecific antibody binding was
blocked with 150 µL per well of 1% bovine serum albumin
(BSA) in PBS for 1 h. The assay was performed by the
addition of 100 µL/well organochlorine standard in methanol
or methanol sample extract [after 1:10 dilution in 0.1%
Teleostean fish skin gelatin (FG, Sigma) in PBS] and 100 µL/
well enzyme conjugate (diluted in 0.5% FG/PBS, providing a
final oncentration of 0.3% FG) and incubation for 1 h. Other
sample and conjugate diluents were also assessed (see Results
and Discussion). For water analysis, the standards were
diluted in distilled water to below 1% methanol concentration
and the samples added directly to the microwells. The
conjugate concentrations used were the lowest concentrations
that provided an optical density (OD) in the assay of 0.7-1.2
and are listed in Table 1. After washing the microwells five
times with distilled water, color was developed by the addition
of 150 µL/well 3,3′,5′-tetramethylbenzidine (Sigma)/peroxide-
based substrate (Hill et al., 1991) for 30 min. Color develop-
ment was stopped by addition of 50 µL/well 1.25 M sulfuric
acid, and OD values were read at 450 nm using a microplate
reader, interfaced with a personal computer and data fitted
to a four-parameter logistic plot using SOFTmax software
(Molecular Devices, Menlo Park, CA).
The hapten V acid had an mp of 134-136 °C: 1H NMR
(CDCl3) δ 2.72 (t, 2H, CH2), 3.72 (m, 2H, CH2), 6.81 (t, 1H,
NH), 7.20 (d, 2H, ArH, J ) 10.2 Hz), 7.32 (d, 2H, ArH, J ) 8.8
Hz), 7.34 (d, 2H, ArH, J ) 8.4 Hz), 7.74 (d, 2H, ArH, J ) 8.0
Hz). The hapten V NHS ester had an mp of 150-152 °C: 1H
NMR (CDCl3; confirmed the product) δ 2.89 (t, 2H, CH2), 3.90
(m, 2H, CH2), 7.0 (t, H, NH), 7.21 (d, 2H, ArH, J ) 10 Hz),
7.34 (d, 2H, ArH, J ) 10 Hz), 7.36 (d, 2H, ArH, J ) 10 Hz),
7.82 (d, 2H, Ar, J ) 8 Hz).
The hapten VI acid had an mp of 103-106 °C: 1H NMR
(acetone-d6) δ 2.64 (t, 2H, -CH2-COOH), 3.62 (m, 2H, -CH2-
NHCO), 5.47 (s, 1H, -CH-), 7.43 (d, 2H, Ar, J ) 7.8 Hz), 7.57
(d, 2H, Ar, J ) 6.3 Hz), 7.83 (d, 2H, Ar, J ) 6.6 Hz), 7.89 (d,
2H, Ar, J ) 9.0 Hz). The hapten VI NHS ester had an mp of
131-133 °C: 1H NMR (CDCl3) δ 2.90 (m, 2H, CH2), 3.90 (m,
2H, CH2), 5.12 (s, 1H, -CH), 6.98 (t, 1H, -NH), 7.31 (d, 2H,
An a lysis of F ood a n d E n vir on m en t a l Sa m p les. The
extent of matrix interference was initially determined by
extracting a pesticide-free matrix, preparing standards with

Immunoassays for DDT, Its Metabolites, and Analogues
J. Agric. Food Chem., Vol. 46, No. 8, 1998 3343
Ta ble 1. Detection of DDT, DDE, DDA, a n d Dicofol by An tibod y/En zym e-Ha p ten Com bin a tion sa
a
Data shown are concentrations (in µg/L) providing 50% inhibition of antibody binding; conjugate concentrations are shown in italics.
> denotes control OD <0.5; * denotes <50% inhibition at 1000 µg/L; nt, not tested.
known concentrations of pesticide, and comparing the standard
curve to a similar one produced in the particular extractant
only. These experiments were followed by spike and recovery
studies.
River Water. Initial experiments demonstrated that river
water did not interfere with the DDT + DDE assay or the DDE
assay (see Results and Discussion) and could be analyzed
directly without the need for cleanup procedures. Both
distilled and river waters (from Lane Cove River, NSW,
Australia) were spiked to 50, 25, 12.5, 6.25, 3.1, 1.6, and 0.8
µg/L from 1 mg/mL p,p′-DDE and p,p′-DDT stock solutions.
All spikes were sealed and left at room temperature overnight
and analyzed the following day with respect to DDE and DDT
standard curves prepared freshly in distilled water. A set of
spikes was also prepared using both p,p′-DDT and p,p′-DDE
(50:50) to form a combined total of 50, 25, 12.5, 6.25, 3.1, 1.6,
and 0.8 µg/L in the sample. These spikes were analyzed
separately by the DDT + DDE and DDE assays, using the
degree of cross-reaction with the nontarget pesticide (Table
1) to calculate the total recoveries obtained.
Soil, Tomato Puree, and Apple Puree. The soil (gray loam,
Canberra, Australia) was sieved using a 2 mm sieve and
contained 12% moisture. Twenty grams of either soil or
tomato or apple puree (obtained from a local supermarket) or
dried sultana grapes (after dipping in a K2CO3/ethyl oleate
emulsion in water before drying, obtained from CSIRO Hor-

3344 J. Agric. Food Chem., Vol. 46, No. 8, 1998
Beasley et al.
ticulture Research Laboratory, Merbein, Australia) was ex-
tracted by shaking at 100 cycles/min in a reciprocal shaker
for either 1 or 16 h in 100 mL of pure methanol [or methanol/
water (90:10) in the case of soil]. The extracts were allowed
to settle for 15 min and then analyzed in the indicated assays
after a 1:10 dilution of sample extracts in 0.5% FG/PBS.
Milk. Fresh, homogenized, and pasteurized full-cream milk
was obtained for matrix and spike and recovery studies from
a commercial supplier. For overnight incubation of spikes,
amber glass bottles were used for storage and the lids covered
with aluminum foil before sealing. Milk had a significant
effect on the DDE immunoassay (see Results and Discussion)
and the following methods, derived from the instrumental
analysis literature, for potentially removing interference were
examined:
1. Partitioning into N,N′-Dimethylformamide (DMF; de
Faubert Maunder, 1964). Forty milliliters of milk was ex-
tracted with 80 mL of acetone and 80 mL of hexane in a
Waring blender for 2 min. The hexane layer was removed and
dried with 2 g of sodium sulfate. Forty milliliters of DMF was
added and partitioned; this layer was collected for immunoas-
say after 1:10 dilution.
result in significant improvements in sensitivity com-
pared with dilution of standards in purified water from
a concentrated stock in methanol. Diluents containing
1% BSA and/or 0.05% Tween 20 lowered the sensitivi-
ties of several of the assays, although use of BSA
typically increased color development. For example, for
detection of DDE with the assay utilizing antibody to
hapten IV and hapten Ib-HRP (Figure 3A), there was
little change in assay sensitivity (IC50 ) 5 µg/L in all
cases) when 1% BSA was substituted for 0.3% FG, but
the color development was doubled. The color develop-
ment remained elevated when the BSA concentration
was reduced to 0.3% (data not shown). In contrast, for
detection of DDE with the assay utilizing antibody to
hapten V and hapten V-HRP (Figure 3B), the color only
slightly increased with BSA but the assay sensitivity
was poorer, changing from an IC50 ) 8-10 µg/L to IC50
) 40-50 µg/L. The effects of BSA were unlikely to be
due to binding of DDE or DDE-HRP by BSA and not
FG since different phenonema were observed with
different assays for the same analyte. Therefore, 0.5%
FG/PBS was used routinely as the diluent for enzyme
conjugate and for methanol extracts of food or soil
matrices, unless otherwise specified.
Assa y Selection . Each antiserum was screened
against the panel of eight enzyme-labeled haptens to
assess its detection sensitivity for DDT, DDE, DDA, and
dicofol (Table 1). Consistent with previous immunoas-
says for pesticides developed in our laboratories, com-
petitive, solid-phase antibody ELISAs were shown to be
more sensitive than immobilized antigen formats; there-
fore, only results obtained using direct competitive
assays are discussed (Hill et al., 1993). To select the
most suitable combination of antibody and hapten-
labeled enzyme for assaying the target analytes (DDT,
DDE, DDA, and dicofol), each combination was assessed
for its relative sensitivity, cross-reactivity, and perfor-
mance at low enzyme-labeled hapten concentrations.
Although two antisera were prepared to each hapten
(except for haptens Ib and Ic, which were only used to
prepare detection conjugates), only results with the
antiserum that exhibited superior detection sensitivity
for the target compound(s) are shown. Because this was
from the haptens coupled to OA and KLH in an equal
number of cases, neither protein carrier appeared to be
clearly superior.
As expected, antibodies prepared using the hapten
derived from DDA (with a â-alanine spacer arm) usually
exhibited selectivity for DDA, with 50% inhibition
values for DDA as low as 2 µg/L, although it was clear
that the both the structure of the hapten used in the
detecting enzyme conjugate and the particular spacer
(for the DDA haptens) influenced both the sensitivity
and the specificity of the assay. Use of the same
â-alanine spacer for the enzyme conjugate actually
provided a more sensitive assay for DDA than the
assays using either a γ-aminobutyric acid spacer or the
aminocyclohexane carboxylic acid (“bulky”) spacer. The
result with the bulky spacer arm contrasts with results
obtained for the pyrethroid, bioresmethrin (Hill et al.,
1993), and the carbamate insecticide, aldicarb (Brady,
1989), and the common finding that spacer heterology
can increase immunoassay sensitivity (Harrison et al.,
1990). It is possible that the failure of spacer heterology
to increase assay sensitivity arises because the two
chlorinated aromatic rings in the DDA hapten structure
are sufficiently large to comprise most of the epitope
2. Sulfuric Acid Treatment (Waliszewski et al., 1982).
Twenty milliliters of concentrated sulfuric acid was added to
25 mL of milk, mixed gently, and allowed to cool to room
temperature. Forty milliliters of petroleum ether was added
and mixed for a further 3 min before 2 mL of the petroleum
ether layer was removed, evaporated, and resuspended in 625
µL of methanol for analysis.
3. Alcohol and Alkali (Single Drying; Tuinstra et al., 1980).
Ten milliliters of milk was extracted in 10 mL of petroleum
ether and 20 mL of saturated potassium hydroxide in ethanol
and left overnight at room temperature. After 5 mL of distilled
water was added, the petroleum ether layer was removed and
dried with sodium sulfate. A 2 mL extract was evaporated to
dryness and resuspended in 1 mL of methanol for analysis.
4. Alcohol and Alkali (Double Drying). Fifty milliliters of
milk was extracted with an equal volume of petroleum ether
and left overnight. The petroleum ether layer was removed,
evaporated, and resuspended in alcoholic potassium hydroxide.
After heating to 70 °C for 30 min, and cooling to room
temperature, 30 mL of petroleum ether and 20 mL of distilled
water was added. The petroleum ether layer was collected
and dried with 2 g of sodium sulfate and a 2 mL extract
evaporated and then resuspended in 1 mL of methanol.
5A,B. Selective Solvent Extraction (Prapamontol and Steven-
son, 1991). Four milliliters of milk was extracted in 20 mL of
a mixture of ethyl acetate/methanol/acetone (2:4:4), vortex
mixed for 2 min, and then either incubated at room temper-
ature overnight (A, overnight extraction method) or vortex
mixed for 1 min and ultrasonicated for 10 min (B, rapid
extraction method). In both methods the extracts were
centrifuged at 2000 rpm for 15 min, and the upper phase was
collected for pesticide analysis.
6. Solid-Phase Extraction Using Florisil or Silica (Sapp,
1989). Ten milliliters of milk was extracted in 5 mL of hexane
and mixed, and 1 mL of the hexane layer was collected to be
run through Florisil or silica Sep-Pak Vac 6 cc (1 g) (Waters,
Milford, MA) prerinsed with 10 mL of 1% diethyl ether in
hexane. The eluate was collected, and 2 mL was evaporated
and resuspended in 2 mL of methanol for analysis.
7. Addition of Tween 20. Tween 20 (1-10% final concen-
tration) was added to milk and heated to 60 °C for 30 min.
All milk extracts were diluted 1:10 into 0.5% FG/PBS before
analysis by immunoassay. Methods 5A and 5B were of
greatest use (see Results and Discussion).
RESULTS AND DISCUSSION
P r ep a r a t ion of Assa y St a n d a r d s. Although the
solubility of DDT and its related compounds in aqueous
solutions is low (Bowman et al., 1960; Worthing, 1987),
use of solvents such as methanol or dimethyl sulfoxide
at 10-20% in the preparation of DDT standards did not

Immunoassays for DDT, Its Metabolites, and Analogues
J. Agric. Food Chem., Vol. 46, No. 8, 1998 3345
F igu r e 3. Comparison of assay performance for DDE standards diluted in 1% BSA/PBS/10% methanol and 0.3% FG/PBS/10%
methanol, for assays (A) using antibody to hapten V and hapten Ib-HRP and (B) using antibody to hapten V and hapten V-HRP.
The experiment was repeated three times in duplicate with similar results.
recognized by the antibody. However, assay specificity
was affected, with the assay using the bulky spacer in
the conjugate cross-reacting more equally with DDT and
dicofol. Heterology in the hapten itself also greatly
affected the specificity of the assay. Use of hapten II-
HRP (identical to the immunogen hapten Ia except for
an -OH group on carbon 1) gave similar sensitivity for
DDA to the assay using hapten Ia , although cross-
reaction with DDT, DDE, and dicofol was slightly
greater. Even though their structures were different,
when hapten III-HRP or hapten IV-HRP was used,
the assay was virtually specific for DDA with little cross-
reaction with DDT, DDE, or dicofol. Remarkably, with
hapten VI-HRP, which differs from haptens I-IV by
being coupled through one of the aromatic rings, the
antibody to the DDA-â-alanine hapten failed to recog-
nize DDA but recognized DDE with high sensitivity and
DDT with moderate sensitivity.

3346 J. Agric. Food Chem., Vol. 46, No. 8, 1998
Beasley et al.
F igu r e 4. Standard pesticide response curves of the four selected assays for their target analytes: ([) DDA; (9) DDT + DDE
(for DDT); (0) DDT; (2) DDE and (b) dicofol, showing standard deviations of two to four experiments.
The hapten structure for hapten II was derived by
formation of an amide from chlorbenzilate, thus retain-
ing the key functional groups of dicofol. Most of the
assays with the different detection conjugates were
selective for dicofol, with the exception of the assay
using hapten V-HRP, which required very high con-
jugate concentrations. Unlike the DDA assay, detection
of dicofol (and other organochlorines) using the hapten
II-KLH antibody/hapten II-HRP conjugate was of low
sensitivity. The most sensitive and specific assay for
dicofol used the hapten II-KLH antibody and hapten
VI-HRP conjugate, in which dicofol had an IC50 of 0.01
µg/mL, and detection of DDA, DDT, and DDE was
10 000-40 000-fold less sensitive. However, this assay
was not used routinely as it had some disadvantages
in that it required relatively high concentrations of
conjugate (Table 1), and a maximal inhibition of only
65% by free dicofol was obtained, even at concentrations
10 000-fold higher than the IC50. The latter result is
possibly due to two populations of antibodies of differing
affinities in the polyclonal antiserum to this hapten. The
preferred assay for analysis of dicofol used either a
hapten with a hydrogen atom replacing the -OH group
on the central carbon (hapten II) in the immunogen
conjugate or one with a trichloromethane group in the
detection conjugate (hapten III). This assay combinan-
tion was not as specific or sensitive, but it did not have
the same serious limitations. Furthermore, the selected
assay gave a steeper pesticide concentration-response
curve, with a 20% change in inhibition for a 10-fold
change in pesticide concentration (Figure 4). In general,
the antisera from the second rabbit (two rabbits were
used per group) that had been immunized with the same
hapten (but coupled to the alternate protein carrier)
exhibited similar specificity, differing only in sensitivity
for the target compound. However, the antiserum from
the hapten II-OA-immunized rabbit had quite different
specificity. Dicofol could be detected with high sensitiv-
ity in the homogeneous assay (using hapten II-HRP);
in addition, the assay using hapten III, which was
selective for dicofol with the hapten II-KLH antiserum
was selective for DDE with the hapten II-OA antise-
rum.
Hapten III produced a relatively poor antibody re-
sponse, although it provided useful HRP conjugates in
several of the assays. Only two conjugates gave sig-
nificant (OD > 0.2) color development at 2 µg/mL, and
in no case did DDT, DDE, DDA, or dicofol at 1000 µg/
mL produce >10% inhibition. This result was initially
thought to be surprising, given that it was the same
hapten used by Burgisser et al. (1990) to develop a
radioimmunoassay with moderate sensitivity for DDT
and that it retained the -CCl3 group as well as both of
the 4-chloro groups of DDT. However, hapten 2 of Abad
et al. (1997), which had a similar but not identical
structure, also failed to generate polyclonal or mono-
clonal antibodies that were displaceable by DDT.
Hapten IV differed from haptens I-III by having a
carbon-carbon bond between the first two atoms in the
spacer; the carbon atom linked to the central carbon is
thus tetrahedrally substituted rather than being in a
planar amide bond. In this way hapten IV resembles
the DDT structure. This hapten was successfully used
to generate antibodies and develop assays using several
conjugates with high sensitivities for DDT, including
haptens Ib-HRP (most sensitive assay), Ic-HRP, and
VI-HRP. However, in each case the assay also detected
DDE with similar sensitivity. Hapten V retained the
dichloroethylene group of DDE, and one of the 4-chloro
groups on the aromatic rings was substituted by a
spacer. This hapten produced antibodies and assays
with high selectivity for DDE. The homologous assay
was quite sensitive (IC50 for DDE of 9 µg/L); whereas
the assay using the corresponding DDT analogue hapten
(hapten IV) was slightly more sensitive for DDE, it was
not favored since this assay cross-reacted to a much
greater extent with DDT. Clearly, the antisera only
poorly recognized conjugates derived from the haptens
that were substituted through the central carbon atom,
because the only other assay that displayed significant

Immunoassays for DDT, Its Metabolites, and Analogues
J. Agric. Food Chem., Vol. 46, No. 8, 1998 3347
Ta ble 2. Cr oss-Rea ction of Isom er s a n d Meta bolites of DDT a n d Rela ted Com p ou n d s in Selected Assa ysa
assay, antibody/conjugate
DDA assay,
DDT assay,
DDT + DDE assay,
IV/Ib
DDE assay,
dicofol assay,
Ia /Ia
VI/VI
V/V
II/III
compound
IC50
% ×
2
14
IC50
% ×
100
40
3
20
6
IC50
% ×
IC50
130
0.9
9
-
3
-
580
-
% ×
IC50
% ×
p,p′-DDT
p,p′-DDD
p,p′-DDE
o,p′-DDT
o,p′-DDE
DDA
45
7
65
350
70
0.8
0.6
-
13
30
400
70
200
-
2
0.3
3
60
3
40
0.3
-
100
700
70
3
70
5
7
900
100
-
300
-
2
500
30
25
-
-
100
8
0.6
10
12
-
-
3
2
0.3
1.4
100
135
-
-
2,2′-bis(4-chlorophenyl)ethanol
N(1,1′)-bis(4-chlorophenyl)-2,2,2-
(trichloroethyl)acetamide
dicofol
chlorbenzilate
methoxychlor
-
-
-
700
-
40
-
-
-
-
14
1.0
-
7
70
-
200
-
6
-
-
100
-
-
8
6
-
90
-
-
25
30
-
2
-
-
-
-
-
80
-
-
-
3
2
-
-
-
-
100
150
-
-
-
-
thiobencarb
-
-
13
-
0.4
-
-
lindane
HCH
-
-
-
-
-
-
-
-
a
- denotes <20% inhibition at 1000 µg/L: aldrin, endrin, endosulfan, dieldrin, heptachlor, linuron, neburon, chlorbromuron,
metobromuron, diuron, monolinuron, chlortoluron, metoxuron, fluometuron, cypermethrin, phenothrin, 2,4-dichlorophenoxyacetic acid.
DDE displacement utilized hapten Ia -HRP, although
at high concentrations.
high homology to the hapten (which is an ester at carbon
1, rather than a free acid), was detected with slightly
greater sensitivity than DDA itself. Chlorbenzilate
shares the ester group at carbon 1 and was detected
with sensitivity similar to that for DDA. The DDT
assay, using antisera to hapten VI, cross-reacted strongly
with p,p′-DDT and o,p′-DDT (both of which contain the
-CCl3 moiety) and p,p′-DDD (which has a -CHCl2
group). Substitution at carbon 2 with an -OH group
or a double bond between carbons 1 and 2 decreased
recognition by the antibody. The p-chloro group ap-
peared to be very important in hapten recognition in
all of the assays, because methoxychlor, which is identi-
cal to DDT except for two -OCH3 groups, was not
detected by any of the assays. Intriguingly, this assay
recognized thiobencarb, a major rice herbicide, with
sensitivity similar to that for DDT. This herbicide
shares some degree of isosterism to p,p′-DDT, because
it has a p-chlorophenyl group linked to a tetrahedrally
substituted carbon atom. There is not a second aro-
matic ring, but the other half of the molecule is also
electron-rich. Other p-chloro herbicides such as the
phenoxyacetic acids and the substituted ureas were not
detected, presumably because the aromatic group is
linked to an oxygen or a nitrogen atom rather than a
carbon atom.
The DDT + DDE assay detected 2,2′-bis(4-chlorophe-
nyl)ethanol and p,p′-DDD with higher sensitivity than
it did DDT or DDE. In the former case, this is probably
because the hapten used for antibody production (IV)
is based on this compound. DDD is also more similar
to the hapten at carbon 1 in that it is less spatially
hindered. The assay cross-reacted 25-30% with dicofol
and chlorbenzilate (apparently the -OH substitution on
C2 did not markedly hinder antibody binding) and
2-5% with thiobencarb, o,p′-DDT, and DDA. The DDE
assay, using antibodies to immunogen prepared with
hapten V, showed preferential recognition of the DDE
and DDD compounds and was 10-80 times less sensi-
tive for the p,p′-DDT parent compound. Thiobencarb,
a herbicide used in rice cultivation, was detected slightly
by the DDT assay (2.2% cross-reaction) and only at very
high levels (>1000 µg/L) by the DDE, DDA, and dicofol
assays (0.4, <0.1, and <0.3% cross-reaction, respec-
tively). Finally, the dicofol assay, using a hapten
derived from chlorbenzilate, actually detected this
Hapten VI retained the trichloroethane structure of
DDT, but with substitution through one of the aromatic
rings. The only assay that performed at acceptably low
HRP conjugate concentrations utilized hapten VI; it
detected DDT almost 40 times more sensitively than it
did DDE or dicofol and 200 times more sensitively than
it did DDA. Intriguingly, whereas hapten V antisera
could be combined with hapten VI-HRP to form a
sensitive assay, the antisera to hapten VI did not
recognize hapten V-HRP. The reasons for this are
unclear. In common with the antisera to hapten V, the
antiserum very poorly recognized HRP conjugates de-
rived from the haptens that were substituted through
the central carbon atom, including hapten III-HRP,
even though it retained the C-CCl3 structure. Abad
et al. (1997) obtained a similar result; none of the
monoclonal antibodies they generated to 4-{4-[1-(4-
chlorophenyl)-2,2,2-trichloroethyl]phenyl}butanoic acid
(hapten 5 of their study) detected conjugates based on
other haptens that had been derivatized at the central
carbon, although they functioned well in a homogeneous
ELISA. In selecting a format, selectivity for the target
analyte and a low enzyme concentration requirement
were important considerations. On the basis of these
results, five assays were selected (boxed results, Table
1) for further characterization, using the following
combinations: (1) DDA assay, antibody raised to hapten
Ia -OA with hapten Ia -HRP; (2) DDT + DDE assay,
antibody to hapten IV-KLH and hapten Ib-HRP; (3)
DDT assay, antibody to hapten VI-KLH and hapten
VI-HRP; (4) DDE assay, antibody to hapten V-KLH
and hapten V-HRP; and (5) dicofol assay, antibody to
hapten II-OA and hapten III-HRP.
Sp ecificities of Selected Assa ys. The specificities
of the five selected assays were evaluated using a fuller
range of DDT analogues, structurally related com-
pounds, and other agrochemicals (Table 2). The DDA
assay was 50-100 times more sensitive to free DDA
than to p,p′-DDT, p,p′-DDE, and their o,p′ analogues.
Dicofol was detected slightly more sensitively. p,p′-
DDD, which lacks one chlorine substituent on the
carbon atom beta to the aromatic rings (carbon 1) and
is thus less hindered, exhibited greater cross-reaction,
whereas 2,2′-bis(4-chlorophenyl)ethanol, which shares

3348 J. Agric. Food Chem., Vol. 46, No. 8, 1998
Beasley et al.
5B). These high recoveries indicated there was not a
significant loss of pesticide by adhesion to glass. Spikes
in river water were prepared using a 50:50 mixture of
DDT and DDE (at seven levels between 1.6 and 60 µg/L
total DDT + DDE) and the mixtures analyzed using the
DDT + DDE assay. Calculated recoveries, based on this
assay having a 67% cross-reaction with DDE relative
to DDT, showed good correlations between the level
spiked and the recoveries, averaging between 71 and
93% (Figure 6). Similarly, the recovery for these DDT
+ DDE spiked samples analyzed by the DDE assay
(based on this assay having a cross-reaction with p,p′-
DDT of 3%) showed recoveries between 88 and 128%.
In all cases the lowest spiked concentration gave slight
overestimates, with the higher spiked concentrations
being much closer to 100% recovered.
An a lysis of F ood a n d Soil Sa m p les. The assays
were next applied to the analysis of soil and selected
foods (tomato puree, milk custard) in the DDT + DDE
and DDE assays and to dried sultana grapes in the
dicofol assay (because dicofol is registered for use in
viticultural crops in many countries). In these assays,
methanol extracts of the sample matrix were diluted
1:10 before analysis. Using the DDT + DDE assay,
initial examinantion of pesticide standard curves pro-
duced in pesticide-free extracts of each sample matrix
showed that none of the extracts affected the color
developed in the absence of added pesticide standard.
Both of the soil extracts decreased the sensitivity of the
assay, altering the IC50 from 13 µg/L to 15 µg/L (soil)
and 20 µg/L (custard) DDT for the 1 h extracts and to
20 µg/L (soil) and 50 µg/L (custard) DDT for the 16 h
extracts. This suggests that the longer extraction period
either extracted a larger proportion of interfering
substances or allowed binding of the pesticide to com-
ponents in the sample matrix. The recoveries of DDT
spiked at 0.01-10 mg/kg in each of these matrices, using
the DDT + DDE assay and determined with reference
to standards prepared in methanol, are shown in Figure
6A. Slightly higher recoveries were noted for the two
food samples when 16 h of extraction was used, probably
due to more complete extraction of the residue, although
no false positive results were observed. However, in
soil, the apparent recovery increased from 109 to 189%.
DDE spikes into soil, tomato puree, and custard were
also analyzed with respect to DDE standards prepared
in methanol, using both the DDT + DDE assay and the
DDE assay (Figure 6B). Pesticide-free extracts of soil
and tomato did not significantly alter the sensitivity of
the DDE assay, but custard extract decreased the
sensitivity of this assay 3-fold. In contrast to the DDT
+ DDE assay, quantitative recoveries were obtained
using soil, but low (25-65%) recoveries were obtained
for tomato puree and custard. This would suggest that
sample cleanup strategies (or, alternatively, analysis of
samples with respect to standards prepared in residue-
free matrix extracts) may be required if the DDE assay
were to be used for quantitative analysis rather than
screening.
F igu r e 5. (A) Recovery of p,p′-DDE spikes prepared in river
water [[, dashed line, DDE (recovered) ) 0.85 (spiked) + 1.2
µg/L, r ) 0.99] and distilled water [9, solid line, DDE
(recovered) ) 0.99 (spiked) + 0.4 µg/L, r ) 0.99] as measured
by the DDE assay. (B) Recovery of p,p′-DDT spikes prepared
in river water [[, dashed line, DDT (recovered) ) 0.80 + 3.7
µg/L, r ) 0.97] and distilled water [9, solid line, DDT
(recovered) ) 0.96 + 1.8 µg/L, r ) 0.99].
compound slightly better than it did dicofol. The hapten
is an amide, and the assay detected 2,2′-bis(4-chlo-
rophenyl)ethanol rather well (40% cross-reaction),
whereas an approximately 10% cross-reaction was noted
with DDE and DDD. DDT and DDA were detected only
at high concentrations. Methoxychlor is an analogue
of DDT with O-methyl groups in the place of the chlorine
substituents on the aromatic rings. It was not detected
by any of the assays, suggesting that the chlorine
substituents play a key role in influencing antibody
specificity. None of the assays detected (at 1000 µg/L)
lindane or HCH, cyclodiene organochlorines (endrin,
aldrin, endosulfan, dieldrin, heptachlor), or linuron,
monolinuron, diuron, chlortoluron, chlorbromuron, me-
toxurin, fluormeturon, 2,4-D, cypermethrin, and phe-
nothrin.
Recover y of DDE a n d DDT Resid u es fr om River
Wa ter . The performance of the DDE and DDT + DDE
assays was assessed in spike and recovery studies.
Using the DDT + DDE assay, analysis of DDT spikes
in purified and river water showed recoveries of between
98 and 145% and between 80 and 161%, respectively,
for samples spiked with DDT between 3 and 50 µg/L
(Figure 5A). Over the same pesticide concentration
range, the DDE assay also gave good recoveries of DDE
spikes in purified and river water of between 100 and
113% and between 86 and 135%, respectively (Figure
This latter concept was explored using the DDT +
DDE and dicofol assays applied to the analysis of dried
sultana grapes. These comprise a complex matrix
comprising grape and grape skin components, oil, and
traces of alkali. Again the curves for pesticide stan-
dards prepared in methanol and in a methanol extract
of the fruit differed. Using the DDT assay, the absor-
bance value in the absence of pesticide decreased from

Immunoassays for DDT, Its Metabolites, and Analogues
J. Agric. Food Chem., Vol. 46, No. 8, 1998 3349
in the absence of pesticide also decreased [from 0.97
(methanol) to 0.80 (fruit extract)], and the assay IC50
value increased from 17 µg/L in methanol standards to
36 µg/L in fruit extract. Recoveries for a series of DDT
spikes were analyzed using the DDT + DDE assay and
dicofol spikes with respect to the dicofol assay, and
results were obtained by reference to the two different
types of standard curve. Because of the different effects
of the dried grape matrix on the two assays, analysis
with respect to a methanol standard curve (Figure 6C)
provided significant overestimates for DDT and under-
estimates for dicofol, whereas in both cases, recoveries
close to 100% were provided when results were analyzed
with reference to standards prepared in a pesticide-free
extract of the matrix. These results suggest that if
analyzed pesticide-free reference samples are available,
by reference to a standard curve prepared in an extract
of the sample matrix, it is possible to obtained accurate
recovery data without the need for sample cleanup.
Milk had a severe effect on the DDE assay, both
completely eliminating inhibition by DDE color develop-
ment when analyzed without dilution and reducing
sensitivity by 10-fold when diluted to 1% (Figure 7). As
milk is composed of approximately 4% fat and DDE is
known to be highly fat soluble (Worthing, 1987), several
cleanup methods that had been reported for gas chro-
matographic analysis were assessed (see Experimental
Methods), initially by evaluating the effect of a sham
cleanup procedure on pesticide-free matrix. Loss of
inhibition sensitivity by pesticide standards was noted
for many of the treatments; for example, standards
prepared in milk after Florisil treatment provided only
30% inhibition at 0.1 mg/L DDE. One treatment, the
alcohol and alkali (double drying) procedure, gave
superimposable curves for DDE standards prepared in
methanol and pesticide-free milk extract (Figure 7).
Unfortunately, subsequent spiking experiments showed
that pesticide was removed by the cleanup procedure
together with the interfering fat. Milk spiked with 0.01,
0.1, and 1 mg/L DDE and processed through the alcohol
and alkali (double drying) cleanup procedure gave
pesticide recoveries in the range of 6-23%; thus, the
method was of little practical use. The “selective solvent
cleanup” method did not provide a pesticide standard
curve that was superimposable with the methanol curve
(Figure 7) but enabled good pesticide recovery from milk
samples when samples were read with reference to
pesticide standards prepared in pesticide-free milk that
had been processed in the same manner. For p,p′-DDE
spikes of 5, 1, 0.5, 0.2, and 0.1 mg/L, the method gave
mean recoveries (two experiments) of 75, 112, 210, 125,
and 110%, respectively. A rapid extraction format,
which did not use an overnight extraction step, simpli-
fied the procedure and also showed good (but lower)
recoveries at the same spike levels (60, 85, 60, 75, and
80%). The sample required a 1 in 50 dilution during
the cleanup process and was analyzed with respect to a
less sensitive milk extract curve (Figure 7); hence, the
lower limit of detection of this method is 0.1 mg/L in
the sample.
F igu r e 6. (A) Recoveries of DDT spiked at six concentrations
(0.01-10 mg/kg) in soil, tomato puree, and custard, extracted
for either 1 or 16 h, and analyzed using the DDT + DDE assay
with reference to DDT standards prepared in methanol.
(B) Recoveries of DDE spiked at six concentrations (0.01-10
mg/kg) in soil, tomato puree, and custard, extracted for 16 h,
and analyzed using either the DDT + DDE assay or the DDE
assay with reference to DDE standards prepared in methanol.
(C) Recoveries of DDT and dicofol spiked at five concentrations
(0.2-5 mg/kg) in dried sultana grapes, extracted for 16 h, and
analyzed using the DDT + DDE and dicofol assay, respectively,
with reference to standards prepared in either methanol or
methanol extract of dried sultana grapes. Data shown are
means of two to three experiments.
CONCLUSIONS
A number of competitive immunoassays have been
developed for the sensitive analysis of p,p′-DDT, p,p′-
DDE, DDA, and the miticide dicofol using a series of
DDT metabolites and analogues as the basis for hapten
design and synthesis. From a panel of possible antibod-
1.57 (methanol) to 1.27 (fruit extract), although the
assay IC50 values were similar (17 and 16 µg/L, respec-
tively). Using the dicofol assay, the absorbance value

3350 J. Agric. Food Chem., Vol. 46, No. 8, 1998
Beasley et al.
F igu r e 7. Effect of cleanup processes on the determination of DDE residues in milk using the DDE assay showing pesticide
response curves prepared in the following: milk-free methanol ([); milk without cleanup (9); milk after alcohol and double-
drying cleanup (0); and milk after selective solvent cleanup (2) by the overnight (solid line) or rapid (dashed line) method. Standard
deviations from three to four experiments are indicated, with the error bars on the rapid selective solvent curve showing the
range of data from two experiments.
ies and enzyme-labeled hapten combinations, five assays
were selected and evaluated for application to food and
environmental matrices. Other immunoassays have
been developed for DDA (Haas and Guardia, 1968;
Centero et al., 1970; Furuya and Urasawa, 1981),
although much of this work predated the development
of sensitive nonisotopic methods, such as the ELISA.
In the current study, DDA-derived immunogens pro-
duced highly sensitive assays for DDA. The antibodies
were usually most sensitive for DDA when used in
combination with enzyme conjugates that also utilized
DDA-derived haptens. The slightly longer spacer arm
of hapten Ib-HRP gave the lowest limit of detection.
By utilizing chlorbenzilate, an analogue of dicofol, very
specific and sensitive antibodies to dicofol cross-reacting
with chlorobenzilate were raised. The DDT + DDE
assay, based on antibodies raised to a bis(p-chlorophe-
nyl)ethanol derivative, could detect as little as 0.3 µg/L
DDT in water, with significant detection of the DDE
metabolite (67%), allowing for the first time the ability
to directly detect trace levels of these organochlorines
in water without sample concentration, at levels rel-
evant to environmental analytical programs (Pham et
al., 1993). An immunoassay described recently by
Banerjee et al. (1996) had broad specificity, but is
approximately 10 times less sensitive for DDT than the
DDT + DDE assay described in this paper. The DDT
+ DDE assay described herein is also 10 times more
sensitive (and the DDT assay twice as sensitive) as the
competitive solid-phase radioimmunoassay described by
Burgisser et al. (1990) using haptens based on dicofol
derivatives. In the late stages of preparation of this
paper, Abad et al. (1997) described the development of
monoclonal antibodies of similar detection sensitivity
to those described in the present paper. Typically, the
major determinant of assay specificity is the hapten
used for antibody production (Gee et al., 1995). Whereas
we have earlier observed that the choice of hapten for
the enzyme label used in the development of panels of
antibodies for structurally related haptens (Edward et
al., 1993; Lee et al., 1995, 1998) can affect assay
specificity, in the present study the hapten in the
detecting conjugate had remarkable effects on specific-
ity.
Selected assays were applied to different sample
matrices. Organochlorine residues in river water and
extracts of two foods (tomato puree and dried sultana
grapes) could be quantitatively analyzed directly with-
out the need for extract concentration or sample cleanup.
Since the 1996/1997 harvest, assays based on some of
these antisera have been routinely use for residue
screening by the Australian dried fruit industry; full
details of their performance and validation will be
described elsewhere (J . H. Skerritt and T. Phongkham,
unpublished results). With the introduction of a simple
alkali dehydrohalogenation step, which converts DDT
residues in a sample extract to DDE, we have adapted
the DDE assay for the large-scale quantitative analysis
of total DDT plus DDE residues in cotton cropping soils
(Shivaramiah, Kennedy, and Skerritt, 1998). Milk (and
milk-based custard) proved a difficult matrix for analy-
sis of organochlorine residues due to its fat content and
the high fat solubility of these compounds, with proce-
dures that remove the interfering fat often removing the
pesticide strongly associated with it. A simple method
involving treating the milk with a selection of solvents
gave a less sensitive assay than the standard format,
but allowed samples to be analyzed against a standard
curve prepared in pesticide-free milk. The method had
a lower limit of detection of 0.1 mg/mL DDE in milk; a
further modification such as a concentration step would
be required to increase the detection sensitivity.

Immunoassays for DDT, Its Metabolites, and Analogues
J. Agric. Food Chem., Vol. 46, No. 8, 1998 3351
ABBREVIATIONS USED
Clark, D. R.; Flickinger, E. L.; White, D. H.; Hothem, R. L.;
Belisle, A. A. Dicofol and DDT residues in lizard carcasses
and bird eggs from Texas, Florida and California. Bull.
Environ. Contamn. Toxicol. 1995, 54, 817-824.
de Faubert Maunder, M. J .; Egan, H.; Godly, E. W.; Hammond,
E. W.; Roburn, J .; Thomson, J . Cleanup of animal fats and
dairy products for the analysis of chlorinated pesticide
residues. Analyst 1964, 89, 168-174.
Dinesman, A. Condensation du chloral avec les hydrocarbures
aromatiques sous l’influence du chlorure d’aluminium. C.
R. Hebd. Seances 1905, 141, 201-203.
Domagalski, J . Occurrence of dicofol in the San J oaquin river,
California. Bull. Environ. Contam. Toxicol. 1996, 57, 284-
291.
BSA, bovine serum albumin; DCC, dicyclohexylcar-
bodiimide; DDA, bis(p-chlorophenyl)acetic acid; DDD,
dichlorodiphenyltrichloroethane; DDE, dichlorodiphe-
nyldichloroethylene; DDT, dichlorodiphenyltrichloro-
ethane or 1,1,1-trichloro-2,2-bis(p-chlorophenyl)ethane;
DMAP, (dimethylamino)pyridine; FG, Teleostean fish
skin gelatin; KLH, keyhole limpet hemocyanin; NHS,
N-hydroxysuccinimide; OA, ovalbumin. Note: Where
not specified, DDT and DDE refer to the p,p′ isomers.
ACKNOWLEDGMENT
Edward, S. L.; Skerritt, J . H.; Hill, A. S.; McAdam, D. P. An
improved immunoassay for chlorpyrifos-methyl (Reldan(R))
in grain. Food Agric. Immunol. 1993, 5, 129-144.
Edwards, C. A. Persistent pesticides in the Environment; CRC
Press: Cleveland, OH, 1973; 170 pp.
Friedman, L.; Shechter, H. Preparation of nitriles from halides
and sodium cyanide. An advantageous nucleophilic displace-
ment in dimethyl sulfoxide. J . Org. Chem. 1960, 25, 877-
879.
Furuya, K.; Urasawa, S. Gas-liquid chromatography demon-
stration of the specificity of rabbit IgG antibody to the
pesticide DDT and its metabolites. Mol. Immunol. 1981, 18,
95-102.
We are grateful to Mr. Akmal Pasha (CFTRI, Mysore,
India) for assistance with interpretation of NMR spec-
tra, Dr. Rob Allan (Department of Pharmacology, Uni-
versity of Sydney) for assistance with nomenclature,
Drs. Wayne Best and Neil Rothnie (WA Chemistry
Centre, Perth) for advice and assistance with resynthe-
sis of some of the haptens, and Dr. I. Ferris and Ms. B.
Maestroni (International Atomic Energy Agency Labo-
ratories, Seibersdorf, Austria) for advice on target
matrices.
Gee, S. J .; Hammock, B. D.; Skerritt, J . H. Diagnostics for
plant agrochemicalssa meeting of chemistry and immu-
noassay. In New Diagnostics in Crop Sciences; Skerritt, J .
H., Appels, R., Eds.; CAB International: Wallingford, U.K.,
1995; pp 243-276.
Gillespie, M. J .; Lythgo, C. M.; Plumb, A. D.; Wilkins, J . P. G.
A survey comparing the chemical composition of dicofol
formulations sold in the UK before and after the introduc-
tion of the EC “Prohibition Directive 79/117/EEC”. Pestic.
Sci. 1994, 42, 305-314.
Guenzi, W. D.; Beard, W. E. Anaerobic biodegradation of DDT
to DDD in soil. Science 1967, 156, 1116-1117.
Haas, G. J .; Guardia, E. J . Production of antibodies against
insecticide-protein conjugates Proc. Soc. Exp. Biol. Med.
1968, 129, 546-551.
Harrison, R. O.; Goodrow, M. H.; Gee, S. J .; Hammock, B. D.
Hapten synthesis for pesticide immunoassay development.
In Immunoassays for Trace Chemical Analysis; Vanderlaan,
M., Stanker, L. H., Watkins, B. E., Roberts, D. W., Eds.;
ACS Symposium Series 451; American Chemical Society:
Washington, DC, 1990; pp 14-27.
Haskelberg, L.; Lavie, D. Investigations in the DDT Series. J .
Org. Chem. 1949, 14, 498-503.
Hassall, K. A. The Biochemistry and Uses of Pesticides, 2nd
ed.; VCH: Weinheim, Germany, 1990; p 169.
Hill, A. S.; Mei, J . V.; Yin, C.-M.; Ferguson, B. S.; Skerritt, J .
H. Determination of the insect growth regulator, methop-
rene, in wheat grain and milling fractions using enzyme-
immunoassay. J . Agric. Food Chem. 1991, 39, 1882-1886.
Hill, A. S.; McAdam, D. P.; Edward, S. L.; Skerritt, J . H.
Quantitation of bioresmethrin, a synthetic pyrethroid grain
protectant, by enzyme-immunoassay. J . Agric. Food Chem.
1993, 41, 2011-2018.
Kapoor, I. P.; Metcalf, R. L.; Hirwe, A. S.; Coats, J . R.; Khalsa,
M. S. Structure-activity correlations of biodegradability of
DDT analogues. J . Agric. Food Chem. 1973, 21, 310-315.
Kawano, M.; Ramesh, A.; Thao, V. D.; Tatsukawa, R.; Subra-
manian, A. N. Persistent organochlorine insecticide residues
in some paddy, upland and urban soils of India. Int. J .
Environ. Anal. Chem. 1992, 48, 163-174.
LITERATURE CITED
Abad, A.; Manclus, J . J .; Mojarrad, F.; Mercader, J . V.;
Miranda, M. A.; Primo, J .; Guardiola, V.; Montoya, A.
Hapten synthesis and production of monoclonal antibodies
to DDT and related compounds. J . Agric. Food Chem. 1997,
45, 3694-3702.
Agarwal, H. C.; Singh, D. K.; Sharma, V. B. Persistence and
binding of p,p′-DDE in soil. J . Environ. Sci. Health 1994,
29, 87-96.
Akerstrom, B.; Brodin, T.; Reis, K. J .; Lawman, M. J . P.
Protein G: A powerful tool for binding and detecting
monoclonal and polyclonal antibodies. J . Immunol. 1985,
135, 2589-2592.
Banerjee, B. D. Development of an enzyme-linked immun-
osorbent assay for the quantification of DDA (2,2-bis[p-
chlorophenyl]acetic acid) in urine. Bull. Environ. Contamn.
Toxicol. 1987, 38, 798-804.
Banerjee, B. D.; Pasha, S. T.; Koner, B. C. Hapten synthesis
and production of rabbit antibodies with reactivity to DDT
and its metabolites for the development of an immunoassay.
Med. Sci. Res. (India) 1996, 24, 553-555.
Boul, H. L. DDT residues in the environmentsa review with
a New Zealand perspective. N. Z. J . Agric. Res. 1995, 38,
257-277.
Bowman, M. C.; Acree, F., J r.; Corbett, M. K. Solubility of
carbon-14 DDT in water. J . Agric. Food Chem. 1960, 3, 406-
408.
Brady, J . F.; Fleeker, J . R.; Wilson, R. A.; Mumma, R. O.
Enzyme immunoassay for aldicarb. In Biological Monitoring
for Pesticide Exposure; Wang, R. G. M., Franklin, C. A.,
Honeycutt, R. C., Reinert, R. C., Eds.; ACS Symposium
Series 382; American Chemical Society: Washington, DC,
1989; pp 262-284.
Brewerton, H. DDT in fats of Antarctic animals. N. Z. J . Sci.
1969, 12, 194-199.
Burgisser, D.; Frey, S.; Gutte, B.; Klauser, S. Preparation and
characterisation of polyclonal and monoclonal antibodies
against the insecticide, DDT. Biochem. Biophys. Res. Com-
mun. 1990, 166, 1228-1236.
Centero, E. R.; J ohnson, W. J .; Sehon, A. H. Antibodies to two
common pesticides, DDT and malathion. Int. Arch. Allergy
Appl. Immunol. 1970, 37, 1-13.
Kelce, W. R.; Stone, C. R.; Laws, S. C.; Gray, L. E.; Kemp-
painen, J . A.; Wilson, E. M. Persistent DDT metabolite p,p′-
DDE is a potent androgen receptor antagonist. Science 1995,
375, 581-585.
Chattaway, F. D.; Muir, R. J . K. The formation of carbinols in
the condensation of aldehydes with hydrocarbons. J . Chem.
Soc. 1934, 701-703.
Lee, N.; Skerritt, J . H.; McAdam, D. P. Hapten synthesis and
development of ELISAs for detection of endosulfan in water
and soil. J . Agric. Food Chem. 1995, 43, 1730-1739.

3352 J. Agric. Food Chem., Vol. 46, No. 8, 1998
Beasley et al.
Lee, N.; McAdam, D. P.; Skerritt, J . H. Development of
immunoassays for “Type II” synthetic pyrethroids: 1. Hap-
ten design and application to heterologous and homologous
assays. J . Agric. Food Chem. 1998, 46, 520-534.
Liapis, K. S.; Miliadis, G. E.; Aplada-Sarlis, P. Dicofol residues
on field-sprayed apricots and in apricot juice. Bull. Environ.
Contam. Toxicol. 1995, 54, 579-583.
McAdam, D. P.; Hill, A. S.; Beasley, H. L.; Skerritt, J . H. Mono-
and polyclonal antibodies to the organophosphate, feni-
trothion. 2. Antibody specificity and assay performance. J .
Agric. Food Chem. 1992, 40, 1471-1474.
Mukherjee, I.; Gopal, M. Organochlorine pesticide residues in
dairy milk in and around Delhi. J . AOAC Int. 1993, 76, 283-
286.
Pham, T.; Lum, K.; Lemieux, C. The occurrence, distribution
and sources of DDT in the St. Lawrence river, Quebec
(Canada). Chemosphere 1993, 26, 1595-1606.
Prapamontol, T.; Stevenson, D. Rapid method for determina-
tion of organochlorine residues in milk. J . Chromatogr.
1991, 552, 249-257.
Sapp, R. E. Recoveries of chlorinated hydrocarbon pesticides
from fat using Florisil and Silica Sep-Paks. J . Agric. Food
Chem. 1989, 37, 1313-1317.
Schwarzbach, S. E.; Shull, L.; Grau, C. R. Eggshell thinning
in ring doves exposed to p,p′-dicofol. Arch. Environ. Contam.
Toxicol. 1988, 17, 219-227.
Tuinstra, L. G. M. T.; Traag, W. A.; Keukens, H. J . Quantita-
tive determinantion of individual chlorinated biphenyls in
milkfat by splitless glass capillary gas chromatography. J .
Assoc. Off. Anal. Chem. 1980, 63, 952-958.
U.S. EPA. Dicofol Special Review position document 2/ 3; Office
of Pesticide Programs, Office of Pesticides and Toxic Sub-
stances; U.S. EPA: Washington, DC, 1985.
Waliszewski, S. M.; Szymczynnski, G. A. Simple low-cost
method for the determination of selected chlorinated pes-
ticides in fat samples. J . Assoc. Off. Anal. Chem. 1982, 65,
677-679.
Waliszewski, S. M.; Pardio, V. T.; Waliszewski, K. N.; Chantiri,
J . N.; Infanzon, R. M. Levels of organochlorine pesticides
in Mexican butter. J . AOAC Int. 1996, 79, 784-787.
Waterhouse, D. The Use of DDT in Australia; Report 14 of the
Australian Academy of Science; Griffin Press: South Aus-
tralia, 1972.
World Health Organization (WHO). The present place of DDT
in world operations for public health. Symposium for the
Biological Impact of Pesticides in the Environment; WHO:
Geneva, 1969.
Worthing, C. R., Ed. The Pesticide ManualsA World Compen-
dium, 8th ed.; British Crop Protection Council: Thornton
Heath, U.K., 1987.
Shivaramaiah, H. M.; Odeh, I. O.; Kennedy, I. R.; Skerritt, J .
H. Analysis by enzyme-linked immunosorbent assay of the
distribution of DDT residues in soils of the Macintyre and
Gwydir valleys of New South Wales, Australia. In Seeking
agricultural produce free of pesticide residues. Proceedings
of an international workshop held at Yogyakarta, Indonesia,
Feb 17-19, 1998; ACIAR Proceedings 85; Kennedy, I. R.,
Skerritt, J . H., J ohnson, G. I., Highley, E., Eds.; ACIAR:
Canberra, 1998; in press.
Received for review March 20, 1998. Revised manuscript
received May 22, 1998. Accepted May 26, 1998. This work
was funded in part by the Australian Centre for International
Agricultural Research and by the Dried Fruits R&D Council
of Australia.
J F9802934