Adjuvant and immuno-suppressive effect of six monophthalates in a subcutaneous injection model with BALB/c mice

Article abstract of DOI:10.1016/S0300-483X(01)00484-X

The prevalence of allergic airway diseases is rapidly increasing in Western Europe and North America. This increase in disease prevalence may be associated with environmental pollutants. The present study investigated the adjuvant and immuno-suppressive effect of a series of monophthalates which are considered to be important metabolites of commonly used phthalate plasticizers. The effects were studied in a screening model. Ovalbumin (OA), used as the model antigen, was injected subcutaneously in the neck region of BALB/cJ mice with or without one of the test substances, mono-n-butyl phthalate (MnBP), monobenzyl phthalate (MBnP), mono-n-octyl phthalate (MnOP), mono-2-ethylhexyl phthalate (MEHP), mono-iso-nonyl phthalate (MiNP) or mono-iso-decyl phthalate (MiDP). The levels of OA-specific IgE, IgG1 and IgG2a in sera were measured by ELISA. Immuno-suppressive effect, defined as a statistically significant reduction in IgE or IgG1 antibody production, was observed with MEHP (1000 μg/ml, IgE and IgG1), MnOP (1000 μg/ml, IgE and IgG1), MiNP (1000 μg/ml, IgE and 10 μg/ml, IgG1) and MiDP (100 μg/ml, IgE and IgG1). Adjuvant effect, defined as a statistically significant increase in IgE or IgG1 antibody level, occurred with MEHP (10 μg/ml, IgE), MnOP (100 μg/ml, and 10 μg/ml, IgG1) and MiNP (100 μg/ml, IgE). No statistically significant immune modulating effect was seen with MBnP and MnBP.

Full text of DOI:10.1016/S0300-483X(01)00484-X

Toxicology 169 (2001) 37–51
www.elsevier.com/locate/toxicol
Adjuvant and immuno-suppressive effect of six
monophthalates in a subcutaneous injection model with
BALB/c mice
Søren Thor Larsen a,b,*, Jitka Stilund Hansen a, Peter Thygesen b,
Mikael Begtrup c, Otto Melchior Poulsen a, Gunnar Damga˚rd Nielsen a
a National Institute of Occupational Health, Lersø Parkalle´ 105, 2100, Copenhagen, Denmark
b Department of Pharmacology, Royal Danish School of Pharmacy, Uni6ersitetsparken 2, 2100, Copenhagen, Denmark
c Department of Medicinal Chemistry, Royal Danish School of Pharmacy, Uni6ersitetsparken 2, 2100, Copenhagen, Denmark
Received 20 December 2000; accepted 16 July 2001
Abstract
The prevalence of allergic airway diseases is rapidly increasing in Western Europe and North America. This
increase in disease prevalence may be associated with environmental pollutants. The present study investigated the
adjuvant and immuno-suppressive effect of a series of monophthalates which are considered to be important
metabolites of commonly used phthalate plasticizers. The effects were studied in a screening model. Ovalbumin (OA),
used as the model antigen, was injected subcutaneously in the neck region of BALB/cJ mice with or without one of
the test substances, mono-n-butyl phthalate (MnBP), monobenzyl phthalate (MBnP), mono-n-octyl phthalate
(MnOP), mono-2-ethylhexyl phthalate (MEHP), mono-iso-nonyl phthalate (MiNP) or mono-iso-decyl phthalate
(MiDP). The levels of OA-specific IgE, IgG1 and IgG2a in sera were measured by ELISA. Immuno-suppressive effect,
defined as a statistically significant reduction in IgE or IgG1 antibody production, was observed with MEHP (1000
mg/ml, IgE and IgG1), MnOP (1000 mg/ml, IgE and IgG1), MiNP (1000 mg/ml, IgE and 10 mg/ml, IgG1) and MiDP
(100 mg/ml, IgE and IgG1). Adjuvant effect, defined as a statistically significant increase in IgE or IgG1 antibody
level, occurred with MEHP (10 mg/ml, IgE), MnOP (100 mg/ml, and 10 mg/ml, IgG1) and MiNP (100 mg/ml, IgE). No
statistically significant immune modulating effect was seen with MBnP and MnBP. © 2001 Elsevier Science Ireland
Ltd. All rights reserved.
Keywords: Monophthalate; Phthalate; Plasticizer; Adjuvant effect; Immuno-suppression; Allergy
1. Introduction
The prevalence of allergic airway diseases has
increased over the last decades. The increase in
allergic diseases is most likely caused by changes
in the environment (Peat and Li, 1999). Diesel
* Corresponding author. Tel.: +45-39165248; fax: +45-
39165201.
E-mail address: stl@ami.dk (S.T. Larsen).
0300-483X/01/$ - see front matter © 2001 Elsevier Science Ireland Ltd. All rights reserved.
PII: S0300-483X(01)00484-X

38
S.T. Larsen et al. / Toxicology 169 (2001) 37–51
exhaust particles (DEP) with polycyclic aromatic
hydrocarbons (PAH) adsorbed to their surface is
an agent of environmental exposure that has at-
tracted much attention. Several studies have
shown that both DEP and PAH are able to
increase the potency of immunogens, i.e. they
posses adjuvent activity (Peterson and Saxon,
1996; Løvik et al., 1997; Ormstad et al., 1998).
Another environmental exposure, phthalate di-
ester liberated from polyvinyl chloride (PVC) ma-
terials, has recently been associated with the
development of asthma in small children (Øie et
al., 1997; Jaakkola et al., 1999). The phthalates
are diesters of phthalic acid and they may be
degraded to monoesters by hydrolysis.
The monoesters share physico–chemical prop-
erties with the anionic surfactants. Recently, our
group has shown that the anionic surfactant
sodium dodecyl sulphate possesses an adjuvant
effect (Clausen et al., 2000). It has been reported
in the literature that the monoester may be the
biologically active metabolite of the phthalates.
Thus, monophthalates are important in relation
to the development of toxic effects on the testes
(Foster et al., 1981; Albro et al., 1989; Heindel
and Powell, 1992; Li et al. 2000), and in the
development of liver cancer (Klaunig et al., 1988).
We, therefore, studied whether the monophtha-
lates possess immune modulating properties using
a screening model with subcutaneous administra-
tion of a model protein, ovalbumin (OA). The
model attempts hazard identification, i.e. to deter-
mine whether or not a substance possesses a given
biological property. If so, it must also be studied
under realistic exposure conditions before it can
be determined whether it possesses a risk to hu-
mans under normal conditions.
mann, 1996). Secondly, the structural diversity of
the monophthalate should be large enough to
allow to search for Structure Activity Relation-
ship (SAR) of the immune modulating properties.
2.2. Structure and nomenclature
The nomenclature of phthalates and monoph-
thalates may give rise to confusion. The name
monooctyl phthalate is sometimes synonymous
with mono-2-ethylhexyl phthalate and sometimes
synonymous with the unbranched mono-n-octyl
phthalate. However, both substances are included
in the present study. In this report, the abbrevia-
tion MEHP is used for mono-2-ethylhexyl phtha-
late and the abbreviation MnOP for the
mono-n-octyl phthalate. Mono-iso-nonyl phtha-
late (MiNP) is not an unambiguous name either,
as it relates to two different compounds, namely
the mono-3,5,5-trimethylhexyl phthalate, which is
used in this study, and the isomer mono-7-methy-
loctyl phthalate. Monobutyl phthalate may refer
to both of the two branched isomers, mono-iso-
butyl phthalate and mono-tert-butyl phthalate as
well as to the unbranched mono-n-butyl phtha-
late. The latter compound is used in the present
study, and is abbreviated MnBP. Finally, mono-
iso-decyl phthalate is a mixture of isomers and
not a single, well-defined substance.
2.3. Chemicals
Mono-2-ethylhexyl phthalat (MEHP) and
monobenzyl phthalate (MBnP) were obtained
from Tokyo Kasei Organic Chemicals (TCI),
Tokyo, Japan. The purities were \90 and \
98%, respectively. Both chemicals were used with-
out further purification. As mono-n-butyl
phthalate (MnBP), mono-n-octyl phthalate
(MnOP), mono-iso-nonyl phthalat (MiNP) and
mono-iso-decyl phthalate (MiDP) were not com-
mercially available, they were synthesised in our
laboratory.
2. Materials and methods
2.1. Selection of monophthalates
The monophthalates used in this study were
selected according to two criteria. First, they
should include degradation products from com-
monly used phthalates. Thus, they were selected
among the most commonly used phthalates (Hoff-
Chemicals used for synthesis of phthalic acid
monoesters were phthalic acid anhydride (:
98%), 1-butanol (:99.5%) and 1-octanol (:
99%) all from Merck, Germany. Isononyl alcohol
(3,5,5-trimethyl-1-hexanol) was from TCI. The

S.T. Larsen et al. / Toxicology 169 (2001) 37–51
39
purity of isononyl alcohol is not given by TCI,
but Proton Nuclear Magnetic Resonance (1H-
NMR) spectroscopic analysis showed a purity
better than 95%. Isodecyl alcohol (:99%), a
mixture of isomers with a boiling point (b.p.) of
215–225 °C, was obtained from Riedel-de-Hae¨n,
Germany.
ties of MnBP, MnOP, MiNP and MiDP were
further confirmed and the purities determined by
1H-NMR spectroscopy. The purities of MnBP,
MnOP and MiNP were all more than 95%. Exten-
sive carbon-NMR analysis showed that the MiDP
was a mixture of at least six isomers. The alcohol
used for the synthesis of MiDP was a mixture of
several alcohols, comprised of two main compo-
nents, two components in lower concentrations
and even lower concentrations of at least two
other alcohols. All alcohols were present in the
MiDP synthesised. Additionally, the MiDP con-
tained low concentrations of impurities, which
could include phthalic acid and phthalic acid
diesters.
2.4. Synthesis of monophthalates
MnBP was synthesised according to Veibel and
Pedersen (1955) dissolving the crude, crystalline
MnBP in 120 ml of diethyl ether. Purification was
performed by extraction with 2×100 ml of 0.2 M
potassium carbonate, addition of 4 M HCl to the
combined aqueous solution until pH 2 and extrac-
tion with 2×125 ml diethyl ether. The purifica-
tion procedure was repeated twice. Finally, the
collected ether phases were dried with 10 g anhy-
drous sodium sulphate and rotary evaporated un-
der vacuum.
The principle method described by Goggans
and Copenhaver (1939) was found to be advanta-
geous for the synthesis of MnOP, MiNP and
MiDP. About 0.03 mole of the alcohol was heated
with 0.06 mole of phthalic acid anhydride at
105–110 °C for 120 min. The reaction mass was
treated with 70 ml of diethyl ether and filtered to
remove excess anhydride. After rotary evapora-
tion, 90 ml of water was added, warming to 65 °C
for 50 min to hydrolyse any residue of phthalic
acid anhydride. The dried residue was dissolved in
45 ml of chloroform and filtered to remove any
phthalic acid. The chloroform was removed under
vacuum. The esters were crystallised by use of
dry-ice, and were recrystallized three times from
petroleum ether (b.p. 60–80 °C). The MnOP was
kept cold during the purification process as it has
a melting point of 21.5–22.5 °C.
2.5. Animals
Inbred BALB/cJ female mice, 6–7 weeks old,
were from Bomholtga˚rd Breeding and Research
Centre, Denmark. The mice were randomly di-
vided into groups of 10–12, and were housed in
polypropylene cages (425×266×150 mm) with
pine tree sawdust bedding (Lignocel S8, Brogaar-
den, Denmark). The cages were sanitised twice
weekly and the mice were allowed to acclimatise 2
weeks before immunisation. At the time of immu-
nisation, the weight of the animals was 19 g with
a S.D. of 1.6 g. The photoperiod was from 07:00
to 19:00 h, and the temperature and relative hu-
midity in the animal room were 2192 °C and
5095%, respectively. Ovalbumin-free food (Al-
tromin no. 1324, Christian Petersen, Denmark)
and tap water were available ad libitum. Treat-
ment of the animals followed procedures ap-
proved by The Animal Experiment Inspectorate,
Denmark.
2.6. Sensitisation procedure
The MiDP formed an oil, which could not be
crystallised.
The animals were sensitised subcutaneously
(s.c.) in the neck region by a standard low-dose
regime with the model allergen ovalbumin (OA,
grade V from Sigma, Germany). The OA control
animals were given 1 mg OA in 100 ml solvent. The
composition of the solvent (vehicle) was sterile
0.9% saline, sterile polyethylenglycol (PEG) 400
(Merck-Schuchardt, Germany) and 99.9% ethanol
(Merck, Germany) in the ratio 89:10:1.
The melting points of the dried MnBP, MnOP
and MiNP were found to be 73.0–73.5, 22.5–23.3
and 50.0–51.0 °C, respectively. In the literature,
the melting points for MnBP and MnOP are
reported to be 73.1–73.5 and 21.5–22.5 °C, re-
spectively (Goggans and Copenhaver, 1939). No
melting point was available for MiNP. The identi-

40
S.T. Larsen et al. / Toxicology 169 (2001) 37–51
Mice in the positive control groups were in-
viewed by Spiegel and Noseworthy (1963) Fur-
thermore, it was found not to possess adjuvant
effect when injected in mice together with OA
(Larsen et al., 2001). Before use, the PEG 400
was heat sterilised at 140 °C for three h.
jected with 1 mg OA together with the known
adjuvant aluminium hydroxide (Al(OH)3), Alhy-
drogel® 2%, which was a kind gift from Super-
fos
Biosector,
Denmark.
The
Al(OH)3
concentrations were 2.7 and 0.27 mg/ml, respec-
tively. The Alhydrogel® was diluted with sterile
water. Mice in the test groups were given 1 mg
OA together with the monophthalates, using the
same solvent as for the OA control group. The
concentrations of monophthalate were 1000,
100, 10 and 1 mg/ml. Each dose consisted of 100
ml solution. After the primary immunisation, the
animals were given one or two booster injec-
tions s.c. in the neck region with 0.1 mg OA in
100 ml 0.9% saline without adjuvant or monoph-
thalate. The first booster injection was given 10
days after primary immunisation, the next one 5
days later.
Blood was collected 4 days after the respective
booster injections, i.e. day 14 and 19. Before
collection of blood by heart puncture, the mice
were anaesthetised with Hypnorm® (fentanyl cit-
rate 0.315 mg/ml and fluanisone 10 mg/ml from
Janssen Pharma, Belgium) and Dormicum® (mi-
dazolam 5 mg/ml from Roche, Schweiz) each
diluted 1:1 in water and then mixed. A volume
of 125 ml per mouse was injected s.c. The blood
samples were centrifuged at 2500×g and the
supernatant (serum) was stored at −80 °C until
analysed.
2.8. Determination of OA-specific antibodies in
serum
The serum was analysed by Enzyme-Linked
Immunosorbent Assay (ELISA) in order to de-
termine the levels of OA-specific IgE, IgG1 and
IgG2a antibodies. IgE antibodies were deter-
mined as described by Ormstad et al. (1998). A
similar procedure was used for the determina-
tion of IgG2a antibodies, using monoclonal rat
anti-mouse IgG2a antibodies (University of Lou-
vain, Belgium).
For the determination of IgG1 antibodies, mi-
crotiter plates (Immulon 129B, Dynatech Labo-
ratories, USA) were coated with 100 ml per well
OA (2 mg/ml) in a Tris–HCl buffer, pH 7.4.
The plates were gently shaken for 1 h at room
temperature and thereafter placed in the refrig-
erator (4 °C) over-night. After washing the
plates three times with washing buffer, consist-
ing of Tris–HCl buffer (pH 7.4) with 0.05%
Tween 20, the plates were blocked with washing
buffer containing 5% skimmed cow milk pow-
der. Blocking was performed by adding 250 ml
blocking buffer with skimmed cow milk powder
to each well and shaking the plate for 1 h at
room temperature, followed by washing three
times. Serum samples were diluted 1:5 and
added to the wells, which were shaken gently
for 5 min and then incubated for 1 h at 37 °C.
After washing another three times, alkaline
phosphatase conjugated goat anti-mouse IgG1
(Southern Biotechnology Associates Inc.) diluted
1:2000 in Tris–HCl pH 7.4 was added, the
plates were gently shaken for 5 min at room
temperature and then incubated 1 h at 37 °C.
After washing another three times, 100 ml of 0.5
mg/ml p-nitrophenyl-phosphate (Sigma-Aldrich,
Germany) was added to each well. The plate
was incubated 30 min in the dark at room tem-
perature, before the reaction between enzyme
and substrate was stopped by adding 50 ml 2 M
2.7. Choice of 6ehicle
It was not possible to dissolve the monoph-
thalates in water. Therefore, a mixture of cosol-
vents was used. The monophthalate was
dissolved in absolute ethanol (99.9%) immedi-
ately before use, to avoid re-esterification with
ethanol. PEG 400 was added to the solution
before 0.9% saline to avoid phase separation.
PEG 400 is a widely used cosolvent and it has
been shown to induces little or no irritation
when applied on the abdominal skin of albino
rabbits (Smyth et al., 1945). It disappears
rapidly from the site of injection (Carpenter and
Shaffer, 1952) i.e. it does not possess a depot
effect. Also, PEG 400 has a low toxicity as re-

S.T. Larsen et al. / Toxicology 169 (2001) 37–51
41
NaOH. After 5 min, the absorbance was read at
405 nm in a ELISA reader (model ELX 808,
Bio Tek Instruments, USA).
For the determination of OA-specific IgE,
IgG1 and IgG2a antibodies, sera were serially
diluted by a factor of two. Standard curves were
constructed based on the absorbance value of
these samples. An absorbance of 1.000 was
defined as 100 arbitrary units for the determina-
tion of IgE and IgG2a antibodies.
by pair-wise comparison by means of the
Mann–Whitney’s U-test. Comparisons were be-
tween the corresponding OA control groups and
the four exposure groups for IgE and the IgG1
levels after one and two booster injections. Fur-
thermore, test groups were compared with the
responder rate in the control groups by means
of Fisher’s Exact test. Where statistically appro-
priate, similar evaluations were performed by
means of the cumulated control groups. Minitab
Statistical Software, Release 13.1 Xtra (Mini-
Inc.), was used for the Kruskal–Wallis and the
Mann–Whitney’s U-tests. P-values less than
0.05 were considered statistically significant.
2.9. No-obser6ed-effect le6el
To
establish
no-observed-effect
levels
(NOELs) for the immuno-suppressive effect of
the tested monophthalates, the levels of IgE and
IgG1 antibodies were used as the endpoint. The
NOEL was accepted as the highest concentra-
tion that did not induce a statistically significant
reduction in antibody production. When differ-
ent suppressive effects were seen after one and
two boosters, the lower of the suppressive con-
centrations was taken as the NOEL.
3. Results
3.1. Systemic effect
A decrease in body weight may be used as an
indicator of general toxicity (Timbrell, 2000).
Therefore, the change in weight of each animal
in the test groups and the Al(OH)3 groups was
followed from the time of primary immunisation
until 24 h later. No statistically significant de-
crease in body weight was observed in any of
the test groups exposed to the highest dose of
monophthalate when these were compared with
the OA control groups. This suggests that no
general toxic effects occurred after a single s.c.
2.10. Statistical analysis
If the corresponding OA control groups from
the different test series were not statistically dif-
ferent (P]0.05, Kruskal–Wallis test), these
were cumulated and used for the statistical eval-
uation. To elucidate effects of
a
certain
monophthalate, the four exposure concentrations
of the monophthalate and the cumulated OA
control group that had received the same num-
ber of boosters were compared by means of the
Kruskal–Wallis test. As the cumulated control
groups have a higher number of animals than
the corresponding OA control groups, this test
procedure is expected to give rise to a high
statistical power of the test and thus minimize
errors of type II. If the corresponding OA con-
trol groups were statistically different (PB0.05,
Kruskal–Wallis test), the corresponding OA
control group was used in testing. This initial
evaluation intend to identify which substances
should be evaluated for exposure-effects. If sig-
nificant effects were detected in the Kruskal–
Wallis test, the exposure-effects were elucidated
injection with up to
monophthalates.
4
mg/kg of the
3.2. Serological response after primary
sensitisation
For three of the six tested monophthalates,
five additional groups were included. These in-
cluded an OA control and the four test concen-
trations. The animals were given a primary
immunisation but not boosted. Blood was col-
lected on day 14. None of the three measured
isotypes of antibodies were formed in detectable
concentrations in any of the groups. In order to
reduce the number of animals, the groups that
received no boosters were excluded when the re-
maining three monophthalates were tested.

42
S.T. Larsen et al. / Toxicology 169 (2001) 37–51
3.3. OA-specific IgG2a antibody production
ried out. The 100 mg/ml concentration induced a
significantly lower IgE level. As an immuno-sup-
pressive effect on the IgE response was absent in
the animals given a 10-fold higher concentration
and no decrease was seen in the IgG1 production
at 1000 mg/ml, the significance is considered spuri-
ous. Furthermore, pair-wise comparison between
IgG1 responses of the 100 mg/ml group and the
corresponding OA control group did not give rise
to significant difference either (P=0.62, Mann–
Whitney’s U-test). Thus, neither an adjuvant ef-
fect nor a consistent concentration-dependent
immuno-suppressive effect was produced by
MnBP. The NOEL for an immuno-suppressive
effect both on IgE and IgG1 production is, there-
fore, considered to be above 1000 mg/ml.
No detectable IgG2a antibody level was seen in
any of the control groups or monophthalate
groups, suggesting the absence of a T helper cell
type 1 (Th1) response.
3.4. Control groups
For monophthalates, which were tested at dif-
ferent times, new OA and Al(OH)3 control groups
were included. In one experiment, the 0.27 mg/ml
Al(OH)3 group was lost due to technical error, so
the cumulated number of animals was 40 in this
group (cf. Table 1). If the equally treated control
groups were significantly different, only the corre-
sponding control groups were used in the statisti-
cal tests.
3.6. Effects of MBnP
3.5. Effects of MnBP
No significant difference in antibody produc-
tion on either the IgE or the IgG1 level occurred
with MBnP after one booster (Tables 2 and 3).
When boosted twice, no statistically significant
effect was seen on IgE production. The MBnP
showed a significant exposure-effect relationship
on the IgG1 response (P=0.034, Kruskal–Wallis
test, cf. Table 3). Further study of the effect of
MBnP on the IgG1 response was carried out by
The Kruskal–Wallis test showed that there was
no significant difference in the antibody produc-
tion either on the IgE or on the IgG1 levels after
one and two boosters (Tables 2 and 3). However,
the IgE response after two boosters was border-
line to significant (P=0.06, cf. Table 2), and,
therefore, further pair-wise comparisons were car-
Table 1
Antibody production in the OA and OA+Al(OH)3 control groups
OA-specific antibody
Number of boosters
OA control group
0.27 mg/ml Al(OH)3
2.7 mg/ml Al(OH)3
IgE
1
2
1
2
104 (0–223)
32/52
553 (254–1745)
50/50
237** (95–460)
36/40**
182* (95–347)
42/51*
331* (211–453)
47/49
397(248–767)
40/40
IgG1
0 (0–19)
23/52
11–51a
39 (28–72)***
47/51***
32/40***
269 (58–338)
40/40
110-308a
51/51
138–579a
49/49
The values are median with the 25 and 75 percentiles given in brackets (arbitrary units). If the corresponding control groups from
each of the tested monophthalates showed no statistically significant difference (P]0.05) in the Kruskal–Wallis test, the groups were
cumulated and the median value of the cumulated group is given together with the 25 and 75 percentiles. Also, the fraction of
animals that respond to the treatment with a detectable level of antibody is given. The Mann–Whitney’s U-test was used for
comparison of the cumulated OA control group and the OA+Al(OH)3 control groups. Comparison of the corresponding responder
rates were performed by Fischer’s exact test. P-values less than 0.05, 0.01 and 0.001 are indicated by *, ** and ***, respectively.
a The range of the mean values of the groups is given, as the groups were significantly different (PB0.05) and thus could not be
combined to a cumulated group.

S.T. Larsen et al. / Toxicology 169 (2001) 37–51
43
Table 2
Effect of monophthalates on IgE production after one booster and two boosters (1B and 2B)
Monophthalate Exposure concentration (mg/ml)
Kruskal–Wallis
test (P-value)
1000
100
10
1
0
1B
MnBP
MBnP
MEHP
MnOP
MiNP
MiDP
122 (11–135)
0 (0–108)
0 (0–0)a
107 (0–203)
74 (0–218)
54 (0–167)
93 (76–175)
0 (0–137)
0 (0–0)a
121 (0–164)
108 (0–222)
0 (0–169)
95 (62–279)
83 (0–160)
89 (0–200)
99 (0–170)
47 (0–106)
0 (0–121)
90 (0–207)
0 (0–168)
77 (0–143)
104 (0–205)
212 (0–229)
87 (0–229)
132 (0–239)
132 (0–239)
0 (0–168)
0.961
0.292
0.01
0.007
0.006
0 (0–0)a
0 (0–0)a
0 (0–0)a
B0.001
2B
MnBP
MBnP
MEHP
MnOP
MiNP
MiDP
485 (220–1423)
833 (159–1772)
88 (0–369)***b
290 (170–596)
252 (84–339)**b
223 (127–640)*b
418 (231–1216)
508 (325–987)
603 (285–1912)
1330 (219–2477)
1302 (465–5185)
1329 (556–3999) 651 (137–1639)
348 (208–3355) 1821 (627–2525)
405 (270–751)
451 (296–1260) 631 (354–2279)
778 (466–1961) 631 (354–2279)
0.060
0.388
0.001
0.185
0.002
1358 (571–4555)*c 595 (171–761)
581 (385–1482)
1553 (681–4940)*d 1382 (701–3606)
211 (101–258)***b 145 (56–242)***b 438 (255–760)
426 (215–2389) 431 (254–1765) B0.001
The values are median with the 25 and 75 percentiles given in brackets (arbitrary units). The 0 mg/ml groups are the corresponding
OA control groups. MnOP and MiNP share corresponding OA control group. The difference between IgE production in the four
exposure groups and the cumulated OA control group (cf. Table 1) was evaluated by use of Kruskal–Wallis test. Further comparison
between cumulated OA control group and monophthalate groups was done by use of Fischer’s exact test or Mann–Whitney’s U-test.
A significantly different IgE antibody response compared with either the corresponding or the cumulated OA control group when
tested by Mann–Whitney’s U-test is indicated by *(PB0.05), **(P50.01) or ***(P50.001).
a Significantly lower number of responders compared with cumulated OA control group (Fischer’s exact test, P=0.0003).
b Significantly decreased compared with the cumulated OA control group (Mann–Whitney’s U-test).
c Significantly increased compared with the corresponding OA control group (Mann–Whitney’s U-test).
d Significantly increased compared with the cumulated OA control group (Mann–Whitney’s U-test).
seen in both IgE and IgG1 levels for the one-
booster group (Tables 2 and 3). None of the 10
mice in the 1000 mg/ml group had formed a
detectable level of IgE after one booster. This was
a significantly lower responder rate (P=0.0003,
Fischer’s exact test) than that of the cumulated
OA control group (32 in 52 mice, cf. Table 1),
indicating an immuno-suppressive effect. To study
the exposure-effect relationship further, the 1000
mg/ml test group was excluded and the Kruskal–
Wallis test was repeated on the remaining three
test groups together with the cumulated OA con-
trol group. In this case, no significant effect was
apparent (P=0.85), indicating that the statisti-
cally significant effect of MEHP on IgE produc-
tion was due to the immuno-suppression in the
1000 mg/ml group. Similarly, the IgG1 antibody
level after one booster showed a significantly re-
pair-wise comparison between the corresponding
OA control group and the test groups. None of
the test groups were significantly different from
the corresponding OA control group, although 10
mg/ml induced a borderline significant adjuvant
effect (P=0.064, Mann–Whitney’s U-test). The
1000 mg/ml group had a lower antibody produc-
tion than the cumulated OA control group. This
was, however, not significant (P=0.19, Mann–
Whitney’s U-test). All in all, neither statistically
significant adjuvant effect nor an immuno-sup-
pressive effect was seen with MBnP (cf. Table 4).
This indicates that the NOEL for an immuno-
suppressive effect is above 1000 mg/ml.
3.7. Effects of MEHP
An exposure-dependent effect of MEHP was

44
S.T. Larsen et al. / Toxicology 169 (2001) 37–51
duced responder rate in the animals given 1000
mg/ml MEHP (P=0.007, Fischer’s exact test).
When the 1000 mg/ml group was excluded from the
Kruskal–Wallis test, no exposure-dependent effect
was seen on IgG1 levels.
Also, after two boosters, an exposure-dependent
effect was apparent for both the IgE and the IgG1
levels. The immuno-suppressive effect of MEHP on
IgE and IgG1 formation was pronounced at the
1000 mg/ml level. When the 1000 mg/ml test group
was excluded from the IgE groups, the Kruskal–
Wallis test was no longer statistically significant
(P=0.15). Nevertheless, 10 mg/ml induced a re-
markably high IgE formation (cf. Table 2). There-
fore, pair-wise comparison was performed.
Mann–Whitney’s U-test showed a significantly
increased IgE level in the 10 mg/ml group compared
with the corresponding OA control group (P=
0.022), suggesting an adjuvant effect. When the 10
mg/ml group was compared with the cumulated OA
control group, a borderline to significant adjuvant
effect was seen (P=0.084, Mann–Whitney’s U-
test). The IgG1 production showed no statistically
significant adjuvant effect.
Altogether, an adjuvant effect due to the 10
mg/ml concentration was suggested from the IgE
response after two boosters, when compared with
the corresponding OA control group. The 100
mg/ml concentration did not induce any statistically
significant immuno-suppressive effect in IgE
Table 3
Effect of monophthalates on IgG1 production after one and two boosters (1B and 2B)
Monophthalate Exposure concentration (mg/ml)
Kruskal–Wallis test
(P-value)
1000
100
10
1
0
1B
MnBP
MBnP
MEHP
MnOP
MiNP
MiDP
0 (0–43)
0 (0–16)
0 (0–0)a
0 (0–0)a
0 (0–0)a
0 (0–0)
22 (0–41)
0 (0–14)
0 (0–5)
0 (0–22)
7 (0–17)
0 (0–19)
0 (0–14)
0 (0–0)
16 (14–24)
0 (0–17)
7 (0–32)
0 (0–21)
0 (0–19)
0 (0–15)
0 (0–19)
0 (0–19)
15 (0–23)
0.474
0.662
0.015
B0.001
0.108
0.187
0 (0–0)
25 (0–54)**b
0 (0–17)
0 (0–0)
21 (10–54)**b*c
0 (0–23)
0 (0–25)
2B
MnBP
MBnP
MEHP
MnOP
MiNP
MiDP
143 (78–178)
36 (15–79)
121 (98–170)
96 (61–255)
98 (69–121)
98 (69–271)
179 (98–266)
90 (44–157)
57 (11–86)
360 (177–665)
148 (81–288)
250 (195–393)
137 (88–261)
61 (36–128)
77 (38–149)
190 (143–318)
190 (143–318)
304 (146–424)
0.889
0.034
B0.001
B0.001
B0.001
0.004
134 (89–212)
141 (96–219)
150 (132–623)
87 (68–135)**d
415 (138–531)
20 (9–30)**d
32 (20–57)***d 177 (126–687)
20 (17–59)***d 256 (120–321)
15 (10–37)**d
22 (14–310)*d
The values are median with the 25 and 75 percentiles given in brackets (arbitrary units). The 0 mg/ml groups are the corresponding
OA control groups. MnOP and MiNP share corresponding OA control group. The difference between IgG1 production in four
exposure groups and the OA control group was evaluated by use of Kruskal–Wallis test. After one booster, comparison was between
test groups and the cumulated OA control group (cf. Table 1), whereas after two boosters, the comparison was between the test
group and its corresponding OA control group. Further comparison between the OA control group and monophthalate groups was
done by use of Fischer’s exact test or Mann–Whitney’s U-test. A significantly different IgG1 antibody response compared with OA
control group when tested by Mann–Whitney’s U-test is indicated by *(PB0.05), **(P50.01) or ***(P50.001).
a Significantly lower number of responders compared with the cumulated OA control group (Fischer’s exact test, P=0.007).
b Significantly increased compared with the cumulated OA control group (Mann–Whitney’s U-test).
c Significantly increased compared with the corresponding OA control group (Mann–Whitney’s U-test).
d Significantly decreased compared with the corresponding OA control group (Mann–Whitney’s U-test).

Table 4
Adjuvant and immuno-suppressive effects of monophthalates
Compound
Number of carbon
atoms in side chain
NOEL for
Concentration(s) (mg/ml) with adjuvant
effect. Also indicated are the antibody
isotype and the number of boosters
Ratios between the antibody production in
the test group and the corresponding OA
control groupb
immuno-suppression
(mg/ml)a
Mono-n-butyl
phthalate
Monobenzyl
phthalate
Mono-2-ethylhexy
l phthalate
Mono-n-octyl
phthalate
Mono-iso-nonyl
phthalate
Mono-iso-decyl
phthalate
4
7
\1000
\1000
100
/
8
10c (IgE, two boosters)
3.4
\1e
2.5
8
100
10c,d and 100d (IgG1, one booster)
100d (IgE, two boosters)
169
(
9
100 (IgE) and 1
(IgG1)
10
10
3
5
a If no antibody isotype is mentioned, the NOEL applies to both IgE and IgG1 antibodies.
b Median values (cf. Tables 2 and 3) are used for calculation of the ratios.
c When compared with corresponding OA control group.
d When compared with cumulated OA control group.
e Not defined, as the OA control group has the median value 0 (cf. Table 3).

46
S.T. Larsen et al. / Toxicology 169 (2001) 37–51
or IgG1 production after one or two boosters and
thus, is accepted as the NOEL.
After two boosters, an exposure-dependent ef-
fect was observed in the IgE level, giving rise to
an immuno-suppressive effect at 1000 mg/ml (P=
0.010, Mann–Whitney’s U-test) and a significant
adjuvant effect at 100 mg/ml (P=0.021, Mann–
Whitney’s U-test). For IgG1, the observations are
more complex, as both 1000 and 10 mg/ml show a
significantly suppressed antibody response (Table
3). Accepting that the immuno-suppressive effect
of MiNP is concentration-dependent, this effect is
expected also to be present at 100 mg/ml. Never-
theless, the IgG1 level at 100 mg/ml is significantly
higher than the IgG1 level at 1000 mg/ml (P=
0.0007, Mann–Whitney’s U-test) and borderline
to significant elevated compared with 10 mg/ml
(P=0.05, Mann–Whitney’s U-test). Therefore,
the IgG1 response at 100 mg/ml may be the result
of the coexistence of an immuno-suppressive and
an adjuvant effect mutually inhibiting each other,
as an adjuvant effect may result from a balance
between toxicity and adjuvanticity (Gupta et al.,
1993). For IgE, no immuno-suppressive effect was
seen at 100 mg/ml after one booster, suggesting
that this is the NOEL. The NOEL for IgG1 is 1
mg/ml.
3.8. Effects of MnOP
Kruskal–Wallis analysis on IgE data showed a
significant exposure-dependent effect after one
booster (Table 2). By pair-wise analysis, this sig-
nificance was found to be due to immuno-sup-
pression at the 1000 mg/ml level (P=0.0003,
Fischer’s exact test, cf. Table 2) as no exposure-
dependent effect was seen in the Kruskal-Wallis
test when the 1000 mg/ml group was excluded
(P=0.92). The highly significant concentration-
dependent effect for the IgG1 antibody level after
one booster was partly due to an immuno-sup-
pressive effect at 1000 mg/ml (P=0.007, Fischer’s
exact test, cf. Table 3) and partly due to an
adjuvant effect. For the 10 mg/ml level, the adju-
vant effect was apparent by comparison to the
corresponding OA control group (P=0.046,
Mann–Whitney’s U-test). Both the 100 mg/ml
level (P=0.01, Mann–Whitney’s U-test) and the
10 mg/ml level (P=0.0073, Mann–Whitney’s U-
test) showed an adjuvant effect when compared
with the cumulated OA control group.
When boosted twice, no statistically significant
concentration-dependent effect was seen for IgE
(Table 2), whereas the immuno-suppressive effect
at 1000 mg/ml remains for IgG1.
All in all, adjuvant effect was highly statistically
significant as seen by the IgG1 antibody produc-
tion after one booster at both the 100 and 10
mg/ml levels. The 100 mg/ml concentration was the
highest concentration that did not induce signifi-
cant suppression of the IgE or IgG1 level, indicat-
ing that this concentration was the NOEL.
On the whole, an adjuvant effect was seen on
IgE at the 100 mg/ml concentration after two
boosters. In addition, a hidden adjuvant effect
might be present for IgG1 at the same concentra-
tion. NOELs were estimated to be 100 and 1
mg/ml for IgE and IgG1, respectively.
3.10. Effects of MiDP
A highly significant, concentration-dependent
effect was seen in IgE levels after one booster
(Table 2). None of the ten animals in the 1000
mg/ml and the 100 mg/ml groups had formed any
detectable level of IgE. This is a significantly
lower responder rate (P=0.0003, Fischer’s exact
test) than in the cumulated OA control group (32
in 52 mice, cf. Table 1). When the 1000 mg/ml and
the 100 mg/ml levels were excluded from the
Kruskal–Wallis test, no statistically significant ex-
posure-dependent effect was apparent (P=0.81).
This indicates that no adjuvant effect is apparent
after one booster regarding the IgE antibody pro-
duction. The IgG1 antibody production is less
3.9. Effects of MiNP
After one booster, a concentration-dependent
suppression occurred in both IgE and IgG1 pro-
duction (Tables 2 and 3), due to the lack of
production of antibodies in the 1000 mg/ml group.
Excluding the highest concentration from the
Kruskal–Wallis tests, the statistically significant
exposure-dependent effect on the IgE level was no
longer apparent (P=0.46).

S.T. Larsen et al. / Toxicology 169 (2001) 37–51
47
4. Discussion
To our knowledge, no studies of the adjuvant
effect of monophthalates has previously been
published. Nor did we succeed in finding studies
dealing with the immuno-suppressive effect of
the monophthalates.
The immuno-modulating effects of six
monophthalates were studied in an animal
model using s.c. injection in mice, following a
previously established protocol (Clausen et al.,
2000). The concentrations were intended to in-
clude both a high, immuno-suppressive concen-
tration and the NOEL with regard to
immuno-suppressive and adjuvant effects. Study-
ing such a broad range of concentrations may
be important for the detection of the effects tak-
ing into account the fact that an adjuvant effect
is expected to occur as a balance between toxic-
ity and adjuvanticity (Gupta et al., 1993).
Fig. 1. Relationship between the length of the side chain of the
monophthalates and their immuno-suppressive effect based on
the IgE and IgG1 productions. The immuno-suppressive effect
is given as the no-observed-effect level (NOEL) defined as the
highest concentration of the monophthalate that did not in-
duce a statistically significant reduction in antibody produc-
tion. An arrow ( ) indicates that the NOEL exceeds the
highest test concentration.
The mouse strain BALB/c was chosen, as sev-
eral types of substances have been shown to
possess an adjuvant effect in this strain with OA
as the model allergen. This applies e.g. to fly
ash (Fujimaki et al., 1986), diesel exhaust parti-
cles (Løvik et al., 1997), detoxified Escherichia
coli enterotoxin (Di Tommaso et al., 1996) and
amorphous silica (Mancino and Ovary, 1980).
Unfortunately, most of the published studies do
not specify the mouse substrain. In a recent
study, the substrain BALB/cA was used
(Clausen et al., 2000). This substrain has a
lower responder rate as regards IgE antibody
formation compared with the BALB/cJ sub-
strain, which is used in the present study. Thus,
BALB/cA mice given 1 mg OA s.c. and boosted
10 days later with 0.1 mg OA s.c. showed a
responder rate of two in 38 animals which is a
significantly lower responder rate compared with
32 in 52 BALB/cJ mice (PB0.001, Fischer’s ex-
act test). Also, after a second booster with 0.1
mg OA s.c., the responder rate among the
BALB/cJ mice was higher, 50 in 50 animals,
compared with 30 in 38 BALB/cA mice (PB
0.01, Fischer’s exact test).
affected by MiDP, as no exposure-dependent ef-
fect is apparent after one booster.
After two boosters, both 1000 and 100 mg/ml
concentrations gave rise to a significant im-
muno-suppression of both IgE and IgG1 levels
(Tables 2 and 3) when studied by the Mann–
Whitney’s U-test. No adjuvant effect was seen
from MiDP either after one or two boosters. All
in all, the NOEL for immuno-suppressive effect
both from the IgE and the IgG1 levels was 10
mg/ml.
3.11. Structure acti6ity relationship
The logarithm of the NOELs was plotted as a
function of the number of carbon atoms in the
monophthalate alkyl side chains (Fig. 1). Chain
lengths of up to seven carbon atoms did not
have any immuno-suppressive effect within the
tested concentration interval, whereas immuno-
suppressive effect increased with increasing chain
length from 7 to 10 carbon atoms.
A high responder rate is an advantage for the
detection of immuno-suppressive effects but a
disadvantage if used for the detection of an ad-

48
S.T. Larsen et al. / Toxicology 169 (2001) 37–51
juvant effect. Thus, in the present study, all the
OA control animals responded after two boosters.
It is, of course, not possible to increase this
responder rate, and, therefore, in this case it was
not possible to show any adjuvant effect using the
responder rate. As a consequence, the adjuvant
effect was studied exclusively from the antibody
levels in BALB/cJ mice, whereas, the suppressive
effect was studied in terms of both the responder
rate and the antibody level.
IV allergies. It is generally accepted that the Th1
and Th2 responses reciprocally inhibit each oth-
er’s development (Mosmann and Moore, 1991;
Holt, 1996; Constant and Bottomly, 1997) even
though this is not always necessarily true (Mu-
raille and Leo, 1998). In the present study, only
IgE and IgG1 antibodies are formed when apply-
ing the adjuvant-active monophthalates, suggest-
ing that they may promote the development of
Type I allergy. The formation of IgG1 antibodies
may also indicate the presence of a Th2 response.
The absence of OA-specific IgG2a antibodies in
all the groups indicates that none of the monoph-
thalates stimulated the development of a Th1
response under the conditions of the test. How-
ever, the use of OA as the model antigen and the
low dose applied both may favor the development
of a Th2 response (Gue´ry et al., 1996; Wang et
al., 1996). Furthermore, the BALB/c mouse may
be more likely to develop a Th2 response than
other mouse strains (Gue´ry et al., 1996). Thus, the
animal model may be sub-optimal for the detec-
tion of a Th1 response. Nevertheless, BALB/cJ
mice produced high levels of IgG2a antibodies
when treated with OA and dimethyldioctadecyl
ammonium bromide or Freunds Complete Adju-
vant, which were used to prepare the IgG2a stan-
dard sera in this study. This shows that BALB/cJ
mice are able to produce Th1 antibodies in the
presence of a Th1 adjuvant.
As IgE and IgG1 antibodies are both related to
the Th2 response, the levels of these may show a
similar trend over time (e.g. Kawabata et al.,
1996). However, this is not always the case, as the
relationship between IgE and IgG1 may be influ-
enced by the dose of administered antigen (Su-
dowe et al., 2000), and the time from
immunisation until blood sampling (Renz et al.,
1992). In the present study, adjuvant effect was
seen on the IgE level for 10 mg/ml MEHP and 100
mg/ml MiNP after two boosters, for example.
These concentrations also gave rise to a non-sig-
nificant higher IgG1 production, i.e. some degree
of co-regulation between the IgE and IgG1 re-
sponses existed.
Exposure-effects were evaluated by comparison
with the corresponding control groups. Also, the
identically treated corresponding control groups
from the different experiments were cumulated, if
not statistically different, and used for evaluation
of the test groups. In this manner, the cumulated
control group resembles historical control data,
which has been found useful, for example, for
evaluation of carcinogenic effects (e.g. Haseman
et al., 1998), mutagenic effects (e.g. Adler et al.,
1998) as well as for comparing different medical
therapies (Holmgren, 1999). The use of historical
data for the evaluation of exposure-effects may
involve several pitfalls (Roe, 1994; Adler et al.,
1998; Haseman et al., 1998). Nevertheless, histori-
cal data may be valuable if the data are from the
same laboratory, the animals are of the same
strain and origin, and the studies are of the same
duration and involve animals of the same age at
the time of testing. Furthermore, the animals
should be fed the same way using the same diet,
and the same diagnostic criteria should be used
for the evaluations (Roe, 1994). All these criteria
were fulfilled in this study and, in addition, the
experiments were carried out within a period of
no more than 5 months. The comparisons using
the cumulated control groups allow evaluations
based on a long span of experiences. Further-
more, the evaluations based on the corresponding
control groups and the cumulated control groups
were in overall accordance.
An adjuvant may stimulate either a Th1 or a
Th2 response, depending on the mechanism by
which it interacts with the immune system (Cox
and Coulter, 1997). IgE antibodies are related to
the Th2 response, which is predominant in Type I
allergies, whereas IgG2a antibodies are related to
the Th1 response, which is predominant in Type
Appearance of an adjuvant effect on the IgE
and the IgG1 levels may occur after a different
number of boosters (Purkerson and Isakson,
1992; Robinson et al., 1996), which is in agree-

S.T. Larsen et al. / Toxicology 169 (2001) 37–51
49
ment with the present study. The adjuvant effect
of aluminium hydroxide was apparent on IgE and
IgG1 production after one booster as was the
adjuvant effect of MnOP on IgG1 production.
MEHP and MiNP enhanced IgE production after
two boosters.
Recently published studies suggested that PVC
with DEHP content was associated with the in-
creased rate in asthma among small children
(Jaakkola et al. 1999; Øie et al., 1997). It was
hypothesised that DEHP was degraded to MEHP
which should possess prostaglandin-like proper-
ties, increasing the risk of inducing inflammation
in the airways and increasing the risk of develop-
ing asthma (Øie et al., 1997).
Monophthalates have a lipophilic part and a
hydrophobic carboxylate group, which is ionised
at physiological pH. This is a general structure of
many surfactants, which are major components in
detergents, for example. As monophthalates and
surfactants both have surface active properties,
this may justify some comparison of their biologi-
cal effects. Surfactants interact with bilayer mem-
branes and their proteins, which may give rise to
toxic effects on cells. This effect depends on the
length of the lipophilic alkyl side chain (WHO,
1996). This may explain the low degree of im-
muno-suppression seen in this study with MnBP
and MBnP, which have the shortest side chains.
The effect of the side chain length on the im-
muno-suppressive effect of the monophthalates is
illustrated in Fig. 1.
The chain length of the monophthalate seems
to affect the adjuvant effect. Neither MnBP nor
MBnP or MiDP show any adjuvant effect, and it
is tempting to suggest that a critical chain length
of about 8 or 9 carbon atoms is required for
monophthalates to have an adjuvant effect. Gall
(1967) observed that the adjuvant property of
quaternary ammonium compounds, a group be-
longing to the cationic surfactants, was strongly
dependent of the alkyl chain length. Those quater-
nary ammonium compounds with short alkyl
chain lengths were weak adjuvants, while adju-
vanticity increased with increasing alkyl chain
length reaching an optimum at about C16–C18.
Gall (1967) also tested guanidines, another group
of strongly charged cationic alkyl compounds.
For these compounds, there was a marked transi-
tion from lack of adjuvant activity to adjuvant
activity at about C10. Furthermore Teerlink et al.
(1987) found that members of the zwitterionic
surfactant N-alkyl-N,N-dimethylammonio-3-pro-
pansulphonate with an alkyl chain length of 10 or
more carbon atoms showed adjuvant activity,
while those with a chain length of EIGHT or less
were totally inactive as regards adjuvant activity.
A certain degree of cytotoxicity may be re-
quired for surfactants to achieve an adjuvant ef-
fect, as suggested for adjuvants used in vaccines
(Gupta et al., 1993). In agreement with this,
monophthalates without immuno-suppressive ef-
fects showed no adjuvant effect in the present
study. One explanation might be that irritation
and toxic effects on the tissue around the injection
site can lead to the liberation of inflammatory
cytokines. However, it is also apparent that there
is no direct relationship between adjuvant and
toxic effects as monoisodecylphthalate showed an
immuno-suppressive effect but lacked adjuvant
properties.
In conclusion, this study has shown that some
monophthalates possess adjuvant activity while
others do not. Of special interest is that MEHP
possessed adjuvant activity which had previously
been proposed (Øie et al., 1997). Also, a relatively
specific structure was required to confer adjuvant
activity as an alkyl side chain with eight or nine
carbon atoms showed optimal adjuvant activity.
However, a final conclusion about the structure-
activity relationship must await further studies
due to the limited number of substances tested.
This study attempts to gain insight into the effects
of monophthalates, and it is not possible to ex-
trapolate directly from the effects of the monoph-
thalates to the effects of the corresponding
diesters. Thus, it can be speculated that differ-
ences in rates of hydrolysis may be important in
relation to exposure-effects. However, if the adju-
vant effects of the phthalate plasticizers are
caused, at least partly, by the effect of their
monophthalates, it follows that the diphthalates
with a side chain length of eight to nine carbon
atoms may possess an adjuvant effect. The
present results should, therefore, be compared

50
S.T. Larsen et al. / Toxicology 169 (2001) 37–51
Fujimaki, H., Hirano, S., Takenaka, S., Murakami, M.,
Watanabe, N., 1986. Enhanced IgE antibody production in
mice injected with fly ash. Int. Arch. Allergy Appl. Im-
munol. 80, 127–131.
Gall, D., 1967. Observations on the properties of adjuvants.
Symp. Series Immunobiol. Standards 6, 39–48.
Goggans, J.F., Copenhaver, E., 1939. Monoalkyl phthalates
from normal aliphatic alcohols. J. Am. Chem. Soc. 61,
2909–2910.
Gue´ry, J.-C., Galbiati, F., Smiroldo, S., Adorini, L., 1996.
Selective development of T helper (Th)2 cells induced by
continuous administration of low dose soluble proteins to
normal and b2-microglobulin-deficient BALB/c mice. J.
Exp. Med. 183, 485–497.
Gupta, R.K., Relyveld, E.H., Lindblad, E.B., Bizzini, B.,
Ben-Efraim, S., Gupta, C.K., 1993. Adjuvants-a balance
between toxicity and adjuvanticity. Vaccine 11, 293–306.
Haseman, J.K., Hailey, J.R., Morris, R.W., 1998. Sponta-
neous neoplasm incidences in Fischer 344 rats and B6C3F1
mice in two-year carcinogenicity studies: a national toxicol-
ogy program update. Toxicol. Pathol. 26, 428–441.
Heindel, J.J., Powell, C.J., 1992. Phthalate ester effects on rat
Sertoli cell function in vitro: effects of phthalate side chain
and age of animal. Toxicol. Appl. Pharmacol. 115, 116–
123.
with effects of the corresponding diphthalates,
which we are currently studying in the same ani-
mal model. This should allow an overall evalua-
tion of the possible adjuvant effects of these
substances.
Acknowledgements
The authors thank laboratory technicians Ulla
Tegner, Maria Hammer and Jette Larsen and
animal technicians Gitte B. Jepsen and Lotte S.
Drifte for excellent technical support. This study
has been produced as part of the research activi-
ties in the Centre for the Environment and Res-
piratory System, which is supported by the
Danish Environmental Research Program.
References
Hoffmann, L., 1996. Mass balance for phthalates, Environ-
mental project nr. 320., Danish Environmental Protection
Agency, in Danish.
Holmgren, E.B., 1999. Establishing equivalence by showing
that a specified percentage of the effect of the active
control over placebo is maintained. J. Biopharm. Stat. 9,
651–659.
Adler, I.-D., Bootman, J., Favor, J., Hook, G., Schriever-
Schwemmer, G., Welzl, G., Whorton, E., Yoshimura, I.,
Hayashi, M., 1998. Recommendations for statistical de-
signs of in vivo mutagenicity tests with regard to subse-
quent statistical analysis. Mutat. Res. 417, 19–30.
Albro, P.W., Chapin, R.E., Corbett, J.T., Schroeder, J.,
Phelps, J.L., 1989. Mono-2-ethylhexyl phthalate, a metabo-
lite of di-(2-ethylhexyl) phthalate, causally linked to testic-
ular atrophy in rats. Toxicol. Appl. Pharmacol. 100,
193–200.
Carpenter, C.P., Shaffer, B., 1952. A study of the polyethylene
glycols as vehicle for intramuscular and subcutaneous in-
jection. J. Am. Pharm. Assoc. 41, 27–29.
Clausen, S.K., Sobhani, S., Poulsen, O.M., Nielsen, G.D.,
2000. Study of adjuvant effect of model surfactants from
the group of alkyl sulfates, alkylbenzene sulfonates, alco-
hol ethoxylates and soaps. Food Chem. Toxicol. 38, 1065–
1074.
Holt, P.G., 1996. Infections and the development of allergy.
Toxicol. Lett. 86, 205–210.
Jaakkola, J.J., Øie, L., Nafstad, P., Botten, G., Samuelsen,
S.O., Magnus, P., 1999. Interior surface materials in the
home and the development of bronchial obstruction in
young children in Oslo, Norway. Am. J. Public Health 89,
188–192.
Kawabata, T.T., Babcock, L.S., Horn, P.A., 1996. Specific IgE
and IgG1 responses to subtilisin Carlsberg (Alcalase) in
mice: development of an intratracheal exposure model.
Fundam. Appl. Toxicol. 29, 238–243.
Klaunig, J.E., Ruch, R.J., DeAngelo, A.B., Kaylor, W.H.,
1988. Inhibition of mouse hepatocyte intercellular commu-
nication by phthalate monoesters. Cancer Lett. 43, 65–71.
Larsen, S.T., Nielsen, G.D., Thygesen, P., 2001. Investigation
of the adjuvant effect of polyethylene glycol (PEG) 400 in
BALB/c mice. Int. J. Pharm., in press.
Li, L.H., Jester, W.F.J., Laslett, A.L., Orth, J.M., 2000. A
single dose of Di-(2-ethylhexyl) phthalate in neonatal rats
alters gonocytes, reduces sertoli cell proliferation, and de-
creases cyclin D2 expression. Toxicol. Appl. Pharmacol.
166, 222–229.
Constant, S.L., Bottomly, K., 1997. Induction of Th1 and Th2
CD4+ T cell responses: the alternative approaches. Annu.
Rev. Immunol. 15, 297–322.
Cox, J.C., Coulter, A.R., 1997. Adjuvants-a classification and
review of their modes of action. Vaccine 15, 248–256.
Di Tommaso, A., Saletti, G., Pizza, M., Rappuoli, R., Dou-
gan, G., Abrignani, S., Douce, G., De Magistris, M.T.,
1996. Induction of antigen-specific antibodies in vaginal
secretions by using a nontoxic mutant of heat-labile entero-
toxin as a mucosal adjuvant. Infect. Immun. 64, 974–979.
Foster, P.M., Lake, B.G., Thomas, L.V., Cook, M.W., Gan-
golli, S.D., 1981. Studies on the testicular effects and zinc
excretion produced by various isomers of monobutyl-o-ph-
thalate in the rat. Chem. Biol. Interact. 34, 233–238.
Løvik, M., Hogseth, A.K., Gaarder, P.I., Hagemann, R., Eide,
I., 1997. Diesel exhaust particles and carbon black have
adjuvant activity on the local lymph node response and

S.T. Larsen et al. / Toxicology 169 (2001) 37–51
systemic IgE production to ovalbumin. Toxicology 121, model. Fundam. Appl. Toxicol. 34, 15–24.
51
165–178.
Roe, F.J.C., 1994. Historical histopathological control data
for laboratory rodents: valuable treasure or worthless
trash? Lab. Anim. 28, 148–154.
Smyth, H.F., Carpenter, C.P., Shaffer, C.B., 1945. The sub-
acute toxicity and irritation of polyethylene glycols of
approximate molecular weights of 200, 300 and 400. J.
Am. Pharm. Assoc. 34, 172–174.
Spiegel, A.J., Noseworthy, M.M., 1963. Use of nonaqueous
solvents in parenteral products. J. Pharm. Sci. 52, 917–
927.
Mancino, D., Ovary, Z., 1980. Adjuvant effects of amorphous
silica and of aluminium hydroxide on IgE and IgG1 anti-
body production in different inbred mouse strains. Int.
Arch. Allergy Appl. Immunol. 61, 253–258.
Mosmann, T.R., Moore, K.W., 1991. The role of IL-10 in
crossregulation of TH1 and TH2 responses. Immunol.
Today 12, A49–A53.
Muraille, E., Leo, O., 1998. Revisiting the Th1/Th2 paradigm.
Scand. J. Immunol. 47, 1–9.
Øie, L., Hersoug, L.G., Madsen, J.O., 1997. Residential expo-
sure to plasticizers and its possible role in the pathogenesis
of asthma. Environ. Health Perspect. 105, 972–978.
Ormstad, H., Gaarder, P.I., Johansen, B.V., Løvik, M., 1998.
Airborne house dust elicits a local lymph node reaction
and has an adjuvant effect on specific IgE production in
the mouse. Toxicology 129, 227–236.
Peat, J.K., Li, J., 1999. Reversing the trend: reducing the
prevalence of asthma. J. Allergy Clin. Immunol. 103, 1–10.
Peterson, B., Saxon, A., 1996. Global increases in allergic
respiratory disease: the possible role of diesel exhaust
particles. Ann. Allergy Asthma Immunol. 77, 263–268.
Purkerson, J.M., Isakson, P.C., 1992. Interleukin 5 (IL-5)
provides a signal that is required in addition to IL-4 for
isotype switching to immunoglobulin (Ig) G1 and IgE. J.
Exp. Med. 175, 973–982.
Sudowe, S., Arps, V., Vogel, T., Ko¨lsch, E., 2000. The role of
interleukin-4 in the regulation of sequential isotype switch
from immunoglobulin G1 to immunoglobulin E antibody
production. Scand. J. Immunol. 51, 461–471.
Teerlink, T., Beuvery, E.C., Evenberg, D., van Wezel, T.L.,
1987. Synergistic effect of detergents and aluminium phos-
phate on the humoral immune response to bacterial and
viral membrane proteins. Vaccine 5, 307–314.
Timbrell, J., 2000. Principles of Biochemical Toxicology, third
ed. Taylor & Francis, London.
Veibel, S., Pedersen, C., 1955. Synthe`se des esters acides
d’acides dicarboxyliques par action de l’alcoolate de
sodium ou de potassium sur les anhydrides d’acide corre-
spondants. Acta Chem. Scand. 9, 1674–1684.
Wang, L.-F., Lin, J.-Y., Hsieh, K.-H., Lin, R.-H., 1996.
Epicutaneous exposure to protein antigen induces a pre-
dominant Th2-like response with high IgE production in
mice. J. Immunol. 156, 4079–4082.
Renz, H., Smith, H.R, Henson, J.E., Ray, B.S., Irvin, C.G.,
Gelfand, E.W., 1992. J. Allergy Clin. Immunol. 89, 1127–
1138.
WHO, 1996. Linear alkylbenzene sulfonates and related com-
pounds. Environmental Health Criteria 169,World Health
Organisation, Geneva.
Robinson, M.K., Babcock, L.S., Horn, P.A., Kawabata, T.T.,
1996. Specific antibody responses to subtilisin Carlsberg
(Alcalase) in mice: development of an intranasal exposure