However, it has been proposed that this mechanism does not apply to carbaryl because its structure differs from that of typical aryl hydrocarbon receptor ligands.. By contrast, carbaryl
Trang 1by carbaryl, a nonconventional ligand
Susanna Boronat1, Susana Casado2, Jose´ M Navas2 and Benjamin Pin˜a1
1 Institut de Biologia Molecular de Barcelona, Consejo Superior de Investigaciones Cientı´ficas, Barcelona, Spain
2 Department of Environment, Instituto Nacional de Investigacio´n y Tecnolog’a Agraria y Alimentaria (INIA), Madrid, Spain
The known or suspected deleterious effects of global
pollution by different chemical species, ranging from
industrial by-products to pesticides, has developed into
a major public concern in recent decades Each year,
thousands of new chemicals are released into the envi-ronment at a pace that makes impossible the precise characterization of their acute and⁄ or chronic impact, both on human health and on exposed ecosystems
Keywords
bioassays; dioxin-like; endocrine disruptors;
recombinant yeast assays; transcriptional
response
Correspondence
B Pin˜a, IBMB-CSIC, Jordi Girona,
18, 08034 Barcelona, Spain
Fax: +34 93 204 59 04
Tel: +34 93 400 61 57
E-mail: bpcbmc@cid.csic.es
(Received 2 March 2007, revised 2 May
2007, accepted 3 May 2007)
doi:10.1111/j.1742-4658.2007.05867.x
Carbaryl (1-naphthyl-N-methylcarbamate), a widely used carbamate insecti-cide, induces cytochrome P450 1A gene expression in mammalian cells This activity is usually mediated by the interaction of the compound with the aryl hydrocarbon receptor However, it has been proposed that this mechanism does not apply to carbaryl because its structure differs from that of typical aryl hydrocarbon receptor ligands We show here that carb-aryl promotes activation of target genes in a yeast-based bioassay expres-sing both aryl hydrocarbon receptor and aryl hydrocarbon receptor nuclear translocator By contrast, carbaryl acted as a competitive inhibitor, rather than as an agonist, in a simplified yeast system, in which aryl hydrocarbon receptor nuclear translocator function is bypassed by fusing aryl hydrocar-bon receptor to a heterologous DNA binding domain This dual action of carbaryl, agonist and partial antagonist, was also observed by comparing carbaryl response in two vertebrate cell lines A yeast two-hybrid assay showed that the mammalian coactivator cAMP response element-binding protein readily interacts with aryl hydrocarbon receptor bound to its canonical ligand b-naphthoflavone, but not with the carbaryl–aryl hydro-carbon receptor complex We propose that carbaryl interacts with aryl hydrocarbon receptor, but that its peculiar structure imposes a substandard configuration on the aryl hydrocarbon receptor ligand-binding domain that prevents interaction with key coactivators and activates transcription with-out the need for aryl hydrocarbon receptor nuclear translocator This effect may be relevant in explaining its physiological effects in exposed animals, and may help to predict its effects, and that of similar compounds, in humans Our data also identify the aryl hydrocarbon receptor⁄ cAMP response element-binding protein interaction as a molecular target for the identification and development of new aryl hydrocarbon receptor antagonists
Abbreviations
AhR, aryl hydrocarbon receptor; ARNT, AhR nuclear translocator; BNF, b-naphthoflavone; CBP, cAMP response element-binding protein; DBD, DNA binding domain; GUS, b-glucuronidase; HAT, histone acetyltransferase; LBD, ligand-binding domain; LOEC, lowest observed effect concentration; RTL, rainbow trout liver; RYA, recombinant yeast assay; TCDD, 2,3,7,8-tetrachlorodibenzo(p)dioxin; XRE, xenobiotic responsive element.
Trang 2Among the different types of pollutants, those
interact-ing with cell receptors constitute an even greater risk
as they become toxic at very low concentrations,
some-times at, or under, the limits of detection by
conven-tional analytical procedures [1]
The aryl hydrocarbon receptor (AhR) belongs to
the basic helix-loop-helix-PAS family of transcription
regulators [2] This family roots itself on the
prokary-otic kingdom; however, the capacity to bind specific
ligands and to modulate the transcriptional activity
according to this binding has apparently only evolved
in chordates [3] The physiological role of AhR in
vertebrates has not yet been completely elucidated,
but it is known to regulate specific phase I and II
metabolic enzymes, among others [4,5] Ectopic
acti-vation of AhR constitutes an initial step leading to
toxic effects of a variety of harmful pollutants, such
as 2,3,7,8-tetrachlorodibenzo(p)dioxin (TCDD) and
benzo[a]pyrene [6,7], which include immune
dysfunc-tion, endocrine disruption, reproductive toxicity,
developmental defects, and cancer in vertebrates
[8–12]
The use of yeast systems to monitor the interaction
of different chemicals with vertebrate receptors has
become a common tool to detect the presence of
recep-tor-binding activity in the environment [13–16] These
yeast-based bioassays, known as recombinant yeast
assays (RYAs), have been used to correlate the
pres-ence of suspected or bona-fide endocrine disruptors
and estrogenic activity in environmental samples [17–
19], and to establish relationships between chemical
structures and affinity for vertebrate hormone
recep-tors [20–24]
It is generally accepted that ligand-free AhR
mole-cules are mainly cytoplasmic, and that binding to the
ligand triggers the translocation of the receptor–ligand
complex to the cell nucleus During this process, the
receptor–ligand complex binds to an auxiliary cofactor,
the AhR nuclear translocator (ARNT), to form a
tern-ary complex, which is capable of recognizing specific
DNA sequences (xenobiotic responsive elements;
XRE) in the promoter of target genes, increasing their
transcription rates [25] Both AhR and ARNT by
themselves are capable of triggering transcription when
tethered to upstream regions of reporter genes by
heterologous DNA binding domains (DBD) [26,27]
More specifically, the ligand-binding domain (LBD) of
AhR, when fused to an heterologous DBD, produces
a ligand-dependent, ARNT-independent activator
maintaining most pharmacological features of the
AhR⁄ ARNT ⁄ XRE system [27] These chimeric systems
have been used mostly in yeast, but they also work in
mammalian cell lines [28]
The mechanisms by which transcriptional activation occurs upon binding of the ligand⁄ AhR ⁄ ARNT com-plex to XRE are still unclear, but they probably include recruiting of different coactivators and general tran-scription factors, which ultimately promote trans-cription initiation by interacting with the RNA polymerase II [29] A key component of this mechanism
is the cAMP response element-binding protein (CBP)⁄ p300 complex, which is assumed to have a major role on the transcriptional activation by AhR⁄ ARNT
in mammals by interacting with histone acetyltrans-ferases (HATs) [29,30] In yeast, a key coactivator for ligand-dependent transcriptional activation by AhR is the Spt-Ada-Gcn5-acetyltransferase (SAGA) complex [31], a HAT complex required for function in yeast of many, but not all, transcriptional activators, including Gcn4p and VP16 [32,33] It is also required for ligand-dependent activation mediated in yeast by several ver-tebrate receptors, including the glucocorticoid, estrogen and retinoic acid receptors [34–36]
Carbaryl (1-naphthyl-N-methylcarbamate) is a wide-spectrum carbamate insecticide that has been applied for approximately 40 years as a contact and ingestion insecticide on a wide variety of crops, as well as on poultry, livestock and pets It is also used as acaricide and as molluscicide in aquaculture facilities It has been reported that this compound is an inducer of cyto-chrome P450 1A gene expression [37,38], a biomarker
of ectopic activation of the AhR receptor by exogenous ligands present in the environment [11,39] However, carbaryl differs structurally from typical AhR ligands, which are aromatic compounds with two or more rings in the same plane that can be accommodated within a rectangular binding site of approximately
14 A˚· 12 A˚ · 5 A˚ [40] Carbaryl does not fit easily into these structural constraints (Fig 1) Nevertheless, both activation of the AhR system by carbaryl in cul-tured mammalian cells and specific binding in vitro of carbaryl to AhR has been demonstrated [41] The present study intended to further characterize the inter-action of carbaryl with AhR by using a combination of
Fig 1 Chemical structures of b-naphthoflavone (left) and carbaryl (right).
Trang 3mammalian cell culture and yeast-based systems The
data obtained suggest that the peculiar structure of
carbaryl imposes a nonstandard structure of the
AhR-LBD, which in turns modulates the capacity of the
complex to interact with CBP⁄ p300 or SAGA This
property of carbaryl may be relevant in the explanation
of its physiological effects, and provide an explanation
for the largely contradictory current data on the effects
of carbaryl in different cell lines and tissues
Results
Differential response of pLMAX and YCM
systems to carbaryl
Addition of increasing concentrations of carbaryl to
YCM cells resulted in a bell-shaped activation⁄ toxicity
curve, as described for many receptor agonists that
become toxic or inhibitory at high concentrations [22]
(Fig 2) The calculated EC50 value for carbaryl in
YCM cells was 124.3 ± 9.6 lm (Table 1), which
cor-responds to a weak agonist At higher concentrations,
carbaryl becomes an inhibitor, with an apparent IC50
value of 578.0 ± 36.2 lm (Table 1) By contrast,
carb-aryl was unable to activate the LMAX-RYA system at
any concentration (not shown) In this system, carbaryl
acted as an antagonist because simultaneous addition
of 1 lm of a typical AhR ligand, b-naphthoflavone
(BNF), and increasing concentrations of carbaryl
resul-ted in a typical inhibition curve, with an apparent IC50
of 256.3 ± 38.2 lm (Fig 2 and Table 1; see BNF and
carbaryl structures in Fig 1) Therefore, the response
to carbaryl depended on the RYA system used and, presumably, on the transcriptional activation mechan-ism predominant in each of them
Carbaryl as a competitive inhibitor of AhR
in yeast
To elucidate the mechanisms causing inhibition of carbaryl in LMAX-RYA, a number of dose–response assays with increasing concentrations of BNF were per-formed in the presence of different concentrations of carbaryl As shown in Fig 3A,B, the presence of carba-ryl affected both the maximal activation at saturating concentrations of BNF and the position of the sigmoi-dal curve Whereas the latter is consistent with a com-petition for binding to AhR-LBD by carbaryl and BNF, the decrease of the maximal activation value is more consistent with a noncompetitive inhibition, either reversible or irreversible IC50values were obtained for both effects separately (Table 1) from the analysis of the experimental data shown in Fig 3A,B Figure 4A shows the adjustment of the apparent EC50 values
at different carbaryl concentrations to a competitive inhibition model [Eqn (2)] An IC50 value of 77.3 ± 15.3 lm can be calculated from the slope of the regression line (Table 1) Similarly, the decrease of maximal activation at increasing concentrations of carbaryl can be adjusted to a noncompetitive binding model [Eqn (3); Fig 4B] In this case, the correspond-ing IC50 value obtained from the slope of the regres-sion line was significantly higher, 461.5 ± 102.3 lm (Table 1), which is compatible to the toxic effect observed in YCM-RYA Therefore, we conclude that the behaviour of carbaryl in both RYA systems is sim-ilar, with a binding constant of approximately 100 lm and a toxic effect at concentrations higher than 400 lm
Analysis of carbaryl toxicity in yeast Irreversible inhibition can also be explained as a conse-quence of cell inactivation by the inhibitory ligand In this case, the phenomenon would not be related to the
Fig 2 Dose–response curve for carbaryl in YCM-RYA (s) and in
LMAX-RYA with simultaneous addition of 1 l M BNF (d) Data are
the average of four independent determinations; bars represent
standard errors.
Table 1 Adjustments to the different activation ⁄ inhibition models.
a 95% confidence margins.
Competitive inhibition LMAX-RYA Eqn (2) 77.3 ± 15.3 Irreversible inhibition LMAX-RYA Eqn (3) 461.5 ± 102.3
Trang 4characteristics of the receptor, but to the sensitivity of
the particular cell strain used in the assay This effect
can be monitored by measuring the effect of the
com-pound to the activation of galactose-responsive genes,
an endogenous yeast activation mechanism completely
unrelated to AhR [42] Figure 5 shows the decrease of
cell response to galactose in the presence of increasing concentrations of carbaryl The decrease follows a sig-moidal curve with an IC50 value of 459 ± 74 lm for carbaryl, similar to the Ki value obtained for the
A
B
Fig 3 Dose–response of BNF in the presence of increasing
con-centrations of carbaryl in pLMAX-RYA (A) Showing data relative to
the maximal activity of BNF in the absence of carbaryl Data were
adjusted to the maximal concentration in each series (relative
expression) (B) Showing unadjusted data (GUS arbitrary units).
Dots represent replicas for each series; curves are calculated from
the observed EC 50 for each carbaryl concentration (all replicas
com-bined).
Fig 4 Data adjustment for the competitive-reversible (A) and irre-versible (B) models for the experiments in Fig 2; IC50values were calculated from the slopes of the regression lines (- - -).
Fig 5 Inhibition of galactose response by carbaryl in GAL-GUS sys-tem The discontinuous curve represents a nonlinear fitting to a logistic function.
Trang 5negative effect at high concentrations of carbaryl in
both RYA systems (irreversible model; Table 1)
Therefore, we consider that the noncompetitive⁄
irreversible component of carbaryl inhibition in
LMAX-RYA was likely due to cytotoxicity rather than to a
putative second site for carbaryl binding in the AhR
Carbaryl as a competitive inhibitor of AhR
in vertebrate cell cultures
Two vertebrate cell lines were tested for their
sensitiv-ity to the presence of carbaryl in dose–response assays
using BNF as agonist The CALUX cell line,
com-monly used for testing AhR agonists, showed
essen-tially identical dose–response curves in the presence
and absence of 200 lm carbaryl, with EC50 values of
8.40 ± 1.32 lm and 9.04 ± 1.34 lm for BNF,
respect-ively (Fig 6) By contrast, the rainbow trout liver
(RTL) cell line showed an EC50 for BNF of
0.75 ± 0.28 lm, approximately one tenth of the
corresponding value for CALUX These values
increased to 4.25 ± 0.87 lm when the dose–response
curve was performed in the presence of 200 lm
carba-ryl, indicating an antagonistic effect in these cells
sim-ilar to the one observed for the yeast YCM-RYA
system (Fig 6; compare with Fig 3)
Modulation of the interaction of AhR-LBD with
CBP by AhR ligands
AhR activation is at least partially mediated by the
recruitment of CBP to the target promoters [29] To
assess the influence of different ligands in this
interac-tion, we performed two-hybrid assays in yeast, using
the pLMAX as a DNA binding domain and
CBP-Gal4AD as activation domain Addition of different
concentrations of BNF to the triple-transformant
resulted in a significant, 50% increase of maximal
tran-scription level relative to an isogenic strain lacking the
CBP-Gal4AD plasmid (Fig 7A) This effect was
min-imal, if any, when carbaryl was added to the same
strain, indicating that the interaction between
carbaryl-loaded AhR-LBD and CBP did not occur (Fig 7B)
The effect of the presence of CBP-Gal4AD in the
two-hybrid system was obscured by the strong
activa-tion signal of pLMAX-RYA in the presence of BNF
This activity can be strongly reduced by the disruption
of the endogenous gene ADA2 in yeast [32] (Fig 7C,
compare fluorescence units with Fig 7A) Dada2 strains
expressing pLMAX showed a limited response to the
presence of BNF, and no response whatsoever to
carba-ryl (Fig 7C,D) In this specific genetic background,
the presence of the CBP-Gal4AD construct increased
transcriptional response to BNF by four to five-fold, whereas no significant response was observed when carbaryl was added (Fig 7C,D) As the activation potential of Gal4p activation domain present in the CBP-Gal4AD chimera is completely unrelated to the presence of AhR ligands, we conclude that the lack of response to carbaryl in the two-hybrid system was due
to the inability of the carbaryl–AhR complex to interact with CBP At this point, it should be remembered that deletion of ADA2 affects very little the transcriptional activation by Gal4AD [43] Co-expression of CBP-Gal4AD In Dada2 strains did not increase expression
of LexA-AhR-LBD The amount of LexA-AhR-LBD mRNA in Dada2 cells was calculated at 1.9· 108± 8.5· 107copies per cell (an average of 12 independent
Fig 6 Dose–response curves for BNF in the presence (s) and absence (r) of 200 l M carbaryl in (A) RTL cells and (B) DR-CA-LUX system Values are average of three independent determina-tions; bars represent standard deviations.
Trang 6measurements) The corresponding figure for Dada2
co-transformed with pLMAX and pGADT7-mCBP was
1.3· 108± 7.1· 107mRNA copies per cell (12
deter-minations) These two values were not statistically
different (P > 0.05, Student’s t-test); therefore, we
attributed the increased response to BNF in
CBP-Gal4AD expressing Dada2 strains to a higher efficiency
to promote transcription of the LexA-AhR-LBD⁄
CBP-Gal4AD complex relative to the LexA-AhR-LBD
alone, rather than to a differential expression of the
LexA-AhR-LBD fusion protein From these data, we
conclude that there is ligand-dependent interaction
between AhR and CBP upon addition of BNF, and
that this interaction did not occur when carbaryl,
instead of BNF, was added to the medium
Discussion
The molecular mechanisms underlying the activation of
genes under the control of XREs by carbaryl have
been object of controversy, especially because of
contradictory reports on its ability to bind AhR
[37,44–46] However, some recent determinations using
computer modeling, together with experimental data
from cell culture assays with DR-CALUX (dioxin
responsive-chemically activated luciferase) cells and from an immunoassay detecting activated AhR com-plexes, demonstrated that carbaryl can interact with the AhR and trigger transcriptional activation [41] The data presented here intend to further elucidate this mechanism by using a combination of mammalian cell culture and yeast-based systems, allowing the dissection
of transcriptional activation pathways by genetic tools Carbaryl appears to be a better activator in YCM-RYA [lowest observed effect concentration (LOEC), approximately 20 lm) than in the vertebrate DR-CA-LUX system, with LOEC values of 100 lm [41] By contrast, carbaryl acted as a competitive antagonist, instead of as an agonist, in LMAX-RYA A similar antagonistic effect of carbaryl was observed in the mammalian RTL cell line, but not in the DR-CA-LUX system, in which it is know to act as an agonist [41] We propose the peculiar structure of carbaryl as the main reason of this dual role as agonist and antag-onist in yeast and cell culture mammalian systems
In silico studies showed that carbaryl adopts preferen-tially nonplanar conformations, which, in principle, are less likely to interact with AhR, whereas even the most stable planar conformations are energetically slightly less favourable (less than 7 kJÆmol)1) than
Fig 7 Dose–response curves for (A,C) BNF and (B,D) carbaryl in yeast strains transformed with pLMAX (—) or with pLMAX and pGADT7-mCBP plasmids (- - - ), using pRB1155 plasmid as a reporter (A,B) Wild-type yeast strains; (C,D), Dada2 strains.
Trang 7noncoplanar ones [41] Assuming that interaction of
carbaryl (as other AhR ligands) with AhR-LBD
should occur through planar and close-to planarity
conformers [41,47], the structural constraints of the
resulting complexes could modify key surfaces of
inter-action with coactivators and preclude transcriptional
activation This specific configuration of the
AhR-LBD would allow the translocation of the
receptor-ligand complex to the nucleus and its interaction with
ARNT, but not the interaction of AhR-LBD with
transcriptional coactivators, including CBP, which are
required for transcriptional activation In YCM-RYA,
ARNT would provide for the missing interactions and
therefore the system behaves as an agonist; in
LMAX-RYA these additional interactions would be missing
and the resulting effect is competitive inhibition A
simplified scheme of this model is depicted in Fig 8
There are several reports in the literature of
antago-nists of AhR, including flavonoids [16,48,49] and
sev-eral phenolic compounds, like resveratrol [50], either
by inhibiting translocation of AhR to the nucleus and
to stabilize the inactive AhR⁄ hsp90 complex [48,49]
or by inducing a inactive configuration to the
ligand⁄ AhR ⁄ ARNT ⁄ XRE complex [50] This latter
mechanism may partially apply to carbaryl, with the
difference that its binding to AhR may result in
agon-istic or antagonagon-istic effects depending on the cofactors
prevalent in each cell type, as illustrated by the
differ-ent effects on RTL and DR-CALUX cell lines The
model proposed here is similar to the one proposed for
some partial agonists of the estrogen receptor, such as
tamoxifen [51] but, to our knowledge, ours is the first
report indicating that it may also apply to AhR
antag-onists It is also the first one on proposing a specific
AhR⁄ coactivator interaction (CBP) as a target for
AhR inactivation by a ligand
The results presented here, together with other avail-able data concerning gene activation by carbaryl in cell lines and in test animals, are relevant in predicting the effects of carbaryl when it is released into the environ-ment Carbaryl exposure will likely result in the ectopic activation of the P450 system in vertebrates, although with less potency than other known pollutants How-ever, this effect may vary in different tissues, and per-haps in different organisms, as the activation potential
of activators may depend on the relative importance of key coactivators in different cell systems As the ecto-pic activation of P450 systems is considered to be det-rimental in many biological systems [11], this argues for a stringent control of the release of carbaryl into the environment
Experimental procedures
Chemicals
Carbaryl (Riedel-de Hae¨n, Seelze, Germany) was obtained
at a purity of 99.7% and BNF (used as positive control and considered to be a model ligand compound of the AhR) was obtained from Sigma (St Louis, MO, USA) at a minimum purity of 95% Stock solutions of both com-pounds were prepared by dissolving them in dimethyl sulf-oxide (Sigma)
Plasmids PLMAX
Plasmid pLMAX contains a fusion construct between the LexA protein DNA binding domain (amino acids 1–202) and the 1914 bp EcoRI-XhoI fragment of the mouse AhR (amino acids 167–805) in the expression plasmid pLexA202 from Clontech (BD Biosciences, Palo Alto, CA, USA)
TCDD/ βNF
Carbaryl
AhR AhR
ARNT ARNT
LexA-AhR LBD
LexA-AhR LBD
Fig 8 Model of transcriptional activation for
TCDD ⁄ BNF (upper) and carbaryl (lower) in
YCM-RYA (left) and LMAX-RYA (right) Note
the difference on DNA binding domains and
DNA sequences between both systems as
well as the absence of ARNT in LMAX-RYA.
The proposed differential conformation of
TCDD ⁄ BNF and carbaryl complexes with
the AhR-LBD is also shown.
Trang 8Plasmid pGADT7-mCBP
Plasmid pGADT7-mCBP was kindly provided by H Jiang
[52] It contains the N-terminus of mouse CBP (amino acids
1–464) fused at the C-terminus of the GAL4 protein
activa-tion domain in the yeast expression vector pGADT7 from
Clontech
Plasmid pRB1155
Plasmid pRB1155 is a high copy number yeast reporter
plasmid encompassing lexA-binding sites driving the
expres-sion of the lacZ reporter gene [53]
Yeast strains and RYA systems
AhR⁄ ARNT system (YCM-RYA)
Strain YCM4 was a generous gift from C A Miller (Tulane
University, New Orleans, LA, USA) [54] This strain is a
derivative of W303a (MATa, ade2-1, can1-100, his3-11, 15,
leu2-3, 112, trp1-1, ura3-1), which harbours two foreign
gen-etic elements: one of them is chromosomally integrated and
coexpresses human aryl hydrocarbon receptor and ARNT
genes under the GAL1-10 promoter The second construct is
the pDRE23-Z reporter, encompassing three XRE5 sequence
and the CYC1-lacZ fusion (more information in the original
paper [54]) To perform the RYA assay, YCM4 and YCM4
derived cells were grown in galactose overnight to express
both AhR and ARNT
LMAX-system (pLMAX-RYA)
YSB7 (MATa, leu2, his3, met 15, URA3::lexA-GUS) is a
Germany) and contains the 2l plasmid pLMAX and
eight copies of the LexA DNA recognition sequence in
front of the b-glucuronidase (GUS) reporter gene
integra-ted into the genome Strains YSB37 and YSB39 were
obtained from Euroscarf, with plasmids pLMAX and
pRB1155 Yeast strains YSB52 and YSB53 were obtained
by transformation of YSB37 and YSB39 with plasmid
pGADT7
GAL-GUS system
A yeast reporter strain was constructed, in which GUS
transcription was controlled by the GAL1-10 promoter [42]
Briefly, yeast strain BY474 was transformed by one-step
double homologous recombination using two overlapping
PCR fragments that allowed both GUS integration at the
GAL1,10 site and nourseothricin selection Details about
the strategy and the characterization of the strain are
provi-ded elsewhere [55]
Recombinant yeast assay
Yeast strains were grown overnight in minimal medium
ammonium sulfate; DIFCO, Basel, Switzerland)
with either glucose or galactose as a carbon source When cells were at the appropriate attenuance (0.1–0.2) they were mixed with carbaryl or with BNF dissolved in dimethyl sulfoxide Some 50–100 lL of this mix were added in tripli-cates in a 96-well siliconized polypropylene microtiter plate
same plate in wells containing cell culture without the
shaking Permeabilization of yeast cells and fluorogenic quantitation of either lacZ or GUS activity was performed
as described [16] EC50values were calculated by fitting the data to a noncooperative version of the Hill equation using SPSS for Windows package (version 11.01, SPSS Inc Chi-cago, IL, USA), as described in [16] For general toxicity testing, the GAL-GUS system strain was grown overnight
proto-trophic markers as required and with raffinose as a carbon source When cells were at the appropriate attenuance (0.1–0.2), 2% galactose was added and they were mixed with carbaryl or BNF and treated as described above
RNA extraction and real time RT-PCR
(Madison, WI, USA) and used according to the manufac-turer’s instructions cDNAs were prepared with
USA) using oligo-dT primers and 2 lg of total RNA as template 1 lL of each cDNA and further 1 : 10 dilutions were used for real time RT-PCR using SYBRGREEN PCR Master Mix (Applied Biosystems, Warrington, UK) with 300 nm of each primer in a final volume of 20 lL
Detection System (Applied Biosystems), using the following primers: 5¢-AGTTTTCCGGCTTCTTGCAA-3¢ (forward) and 5¢-TTGGACTGGACCCACCTCC-3¢ (reverse), from Roche (Basel, Switzerland) LexA-mAhR-LBD mRNA copy numbers were calculated by interpolation in a stand-ard curve using plasmid pLMAX as standstand-ard
Vertebrate cell lines culture and enzymatic measurements
The rainbow trout liver cell line, RTL-W1, was grown as outlined in the original description of this cell line [56] Cells were grown in Leibovitz’s L-15 cell culture medium (Cambrex, North Brunswick, NJ, USA) supplemented with
Trang 95% fetal bovine serum (Cambrex) and
detached from confluent flasks using trypsin (Sigma), and
then, seeded in 96 well Falcon plates (Becton Dickinson,
Oxnard, CA, USA) at a density of 20 000 cells in 200 lL
of culture medium per well and allowed to grow to
conflu-ency for 1 day Subsequently, medium was substituted and
new medium with the corresponding concentrations of
BNF (0.2–100 lm) and carbaryl (100 lm) was added The
maximal concentration of dimethyl sulfoxide in the culture
medium was 0.2% Control cells received only solvent
After 48 h of treatment, medium was removed, cells washed
with phosphate buffered saline (pH 7.5) and the plates
until analysis of ethoxyresorufin-O-deethylase activity and
protein following the methodology previously described
[57,58]
The DR-CALUX bioassay performed in this study is
based on the use of a rat hepatoma (H4IIE) cell line stably
transfected with a construct containing the luciferase
repor-ter gene under direct control of DRE (Dioxin Responsive
Element) (BioDetection Systems, Amsterdam, the
Nether-lands)
Cells were maintained in aMEM (Cambrex) with
phe-nol red and supplemented with 10% fetal bovine serum
(Cambrex), 1% 2 mm l-glutamine (Cambrex) and
humidified incubator For the assay, cells grown in
bottles were trypsinized and plated in 96 well plates at a
density of 2.5· 104
cells per well After 24 h, the cells were cotreated with different concentrations of BNF
(200 lm)
Carbaryl or BNF stock solutions were diluted in culture
medium at a maximal solvent concentration of 0.2%
Control cells received the maximal dimethyl sulfoxide
con-centration used in the treated cells Cells were exposed to
the xenobiotics for 48 h Subsequently, culture plates were
washed with phosphate buffer and the luminescence
emit-ted by the cells was quantified by means of the
Steady-Glo Luciferase assay System from Promega (Madison,
WI, USA) following the manufacturer’s instructions in a
Tecan Genios (Maennedorf, Switzerland) luminescence
detector
Mathematical modelling
The equations and definitions used in this work are
derived from standard ligand-receptor mathematical
mod-els, as previously described [59] A more detailed
descrip-tion of the models can be found in the Supplementary
material
Interaction of a receptor with a single ligand
assumes an equilibrium between hormone-free and hor-mone-loaded hormone receptor molecules in solution:
Rþ h1 Kd
Rh1 where R represents the concentration of hormone-free receptor molecules, h is the hormone concentration, Rh is the concentration of the hormone-loaded receptor molecule,
single agonist molecule binding to each receptor molecule, and an hormone concentration much larger than the recep-tor concentration The fraction of receprecep-tor bound to the
Ur¼½Rh
Ro
1þKd
½h
½1
lig-and concentration at which 50% of receptor molecules
hor-mone concentration at which the physiological effect (i.e the reporter activity in our case) reaches 50% of its max-imal value at saturating hormone concentration When applied to inhibitory effects, such as a decrease on tran-scription rates upon addition of a compound, the
the measured physiological activity is reduced to 50%
Interaction of a receptor with two ligands
Below, we considered three mechanisms of mutually
BNF) and an inhibitor (h2, in our case carbaryl)
Reversible binding, competitive inhibition
This model proposes an equilibrium between free receptor, R, and two ligands that bind alternatively to a single site of the receptor molecule, with dissociation constants Kd1and Kd2:
Rþ h1
Kd 1
Rh1;Rþ h2
Kd 2
Rh2
would show a fraction of its maximal activation at satur-ating concentration of h1A⁄ Amaxthat could be expressed as:
A
Amax
¼½Rh1
Ro
1þKd1þ
Kd1½h2
Kd2
½h 1
from h2, whereas the apparent EC50for h1(EC50app) equals
Trang 10to Kd1only when h2¼ 0 Kd2(identical to IC50) can be
cal-culated by measuring EC50appat different concentrations of
h2following the equation:
EC50ðh 2 Þ
EC50ðh 2 ¼0Þ
¼ 1 þ½h2
Kd2
½2
in which EC50 2 Þand EC50 2¼0Þ correspond to the EC50for
h1in the presence and in the absence of a given
concentra-tion of h2
Noncompetitive, reversible inhibition
This model postulates the binding of h1and h2to two
inde-pendent binding sites in the receptor, and that binding of
h2allows binding of h1but precludes transcriptional
activa-tion The model predicts three ligand-receptor complexes:
The fraction of the active complex relative to the total
amount of receptor molecules, Rocan be calculated as:
½Rh1
1þ½h2
Kd 2þKd1ð1þ
½h2 Þ
½h 1
equa-tion:
Amax;h2¼0
Amax;h 2
¼ 1 þ½h2
Kd2
½3
In which Amax, h 2 ¼0 corresponds to the maximal activation
in the absence of h2and Kd2 equals to IC50
Noncompetitive, irreversible inhibition
This model proposes that binding to h2 irreversibly
inacti-vates the receptor, reducing the amount of available
the proportion of receptor molecules becoming inactivated
follows a typical logistic function with an inhibitory
con-stant Ki:
Rh2
Ro
1þ K i
½h 2
; R¼ Ro Rh2
In this model, the maximal activity at saturating concentra-tions of h1depends on the concentration of h2, as follows:
h1! 1;Amax; h2 ¼0
Amax; h2
¼Ro
R ¼ 1 þ
½h2
Ki
Kiis therefore equivalent to IC50.This equation is identical
to Eqn (3), and therefore Kican be calculated as Kd2in the previous model
Acknowledgements
This work has been supported by the Spanish Ministry for Science and Technology (BIO2005-00840) and INIA (RTA2006-00022-00-00) The contribution of the Centre de Refere`ncia en Biotecnologia de la Generali-tat de Catalunya is also acknowledged
References
1 Colborn T (1995) Environmental estrogens: health implications for humans and wildlife Environ Health Perspect 103, 135–136
2 Hahn ME, Karchner SI, Shapiro MA & Perera SA (1997) Molecular evolution of two vertebrate aryl hydrocarbon (dioxin) receptors (AHR1 and AHR2) and the PAS family Proc Natl Acad Sci USA 94, 13743– 13748
3 Hahn ME (2002) Aryl hydrocarbon receptors: diversity and evolution Chem Biol Interact 141, 131–160
4 Nebert DW, Puga A & Vasiliou V (1993) Role of the Ah receptor and the dioxin-inducible toxicity, cancer, and signal transduction Ann NY Acad Sci 685, 624–640
5 Sutter TR, Tang YM, Hayes CL, Wo YY, Jabs EW,
Li X, Yin H, Cody CW & Greenlee WF (1994) Com-plete cDNA sequence of a human dioxin-inducible mRNA identifies a new gene subfamily of cytochrome P450 that maps to chromosome 2 J Biol Chem 269, 13092–13099
6 Gonzalez FJ & Fernandez-Salguero P (1998) The aryl hydrocarbon receptor: studies using the AHR-null mice Drug Metab Dispos 26, 1194–1198
7 Shimizu Y, Nakatsuru Y, Ichinose M, Takahashi Y, Kume H, Mimura J, Fujii-Kuriyama Y & Ishikawa T (2000) Benzo[a]pyrene carcinogenicity is lost in mice lacking the aryl hydrocarbon receptor Proc Natl Acad Sci USA 97, 779–782
8 Abbott BD, Perdew GH & Birnbaum LS (1994) Ah receptor in embryonic mouse palate and expression Toxicol Appl Pharmacol 126, 16–25
9 Poland A & Knutson JC (1982) 2,3,7,8-tetrachloro-dibenzo-p-dioxin and related hydrocarbons: examination