The components of the AhR‐molecular chaperone complex differ depending on whether the ligands are toxic or non‐toxic

The aryl hydrocarbon receptor (AhR) forms a complex with the HSP90‐XAP2‐p23 molecular chaperone when the cells are exposed to toxic compounds. Recently, 1,4‐dihydroxy‐2‐naphthoic acid (DHNA) was reported to be an AhR ligand. Here, we investigated the components of the molecular chaperone complex when DHNA binds to AhR. Proteins eluted from the 3‐Methylcolanthrene‐affinity column were AhR‐HSP90‐XAP2‐p23 complex. The AhR‐molecular chaperone complex did not contain p23 in the eluents from the DHNA‐affinity column. In 3‐MC‐treated cells, AhR formed a complex with HSP90‐XAP2‐p23 and nuclear translocation occurred within 30 min, while in DHNA‐treated cells, AhR formed a complex with AhR‐HSP90‐XAP2, and translocation was slow from 60 min. Thus, the AhR activation mechanism may differ when DHNA is the ligand compared to toxic ligands.

function regulation, and physiological function activation of various mature proteins in cells in addition to folding, and the client protein of HSP90 of eukaryotic organisms exceeds 300 kinds [4][5][6][7].It is an important protein in vital function maintenance such as signal transduction, and transcription factors such as protein kinase and steroid hormone receptors form the largest interaction group.HSP90 regulates the physiological functions of many client proteins by changing the composition of co-chaperones such as p23, HOP, and CDC-37.
We have studied the structure and physiological function of molecular chaperones mainly HSP60, HSP70, and HSP90 [8][9][10].The aryl hydrocarbon receptor (AhR), one of the client proteins of HSP90, is activated by binding with ligands in the cytoplasm, translocating the nucleus, and functions as a transcription factor [11].In the case of dioxin ligands such as 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), AhR is known to form AhR-HSP90-p23-XAP2 (hepatitis B Virus X-associated protein 2, also known as aryl hydrocarbon receptor interacting protein, AIP) complexes in the cytoplasm and to induce the xenobiotic-metabolizing enzyme as a member of the cytochrome P450 family of drug-metabolizing enzymes, CYP1A1, CYP1A2, and CYP1B1 after nuclear transfer [12][13][14][15][16]. Thus, the mechanism of activation of AhR by molecular chaperone complexes in the case of toxic ligands has been well studied and we have also reported on the AhR-molecular chaperone complex [17][18][19][20].However, the activation mechanism of AhR by molecular chaperone complexes in the case of non-toxic ligands is not fully understood.
Recently, DHNA, 1,4-dihydroxy-2-naphthoic acid (DHNA), major molecules of whey fermented by propionibacteria, has been reported to be a ligand for AhR [21,22].The activation mechanism of AhR, when non-toxic DHNA is liganded, was studied from the viewpoint of a molecular chaperone complex.In this study, we investigated the components of the molecular chaperone complex when the ligand DHNA binds to AhR and the physiological functions of AhR in vivo.

Animal experiments
Six-week-old male BALB/c mouse (CLEA Japan, Tokyo, Japan) was sacrificed and brains were immediately removed, frozen in liquid nitrogen, and used as samples for DHNA-affinity column chromatography and immunoblotting.All mice were housed at a controlled temperature of 25 °C and a daily 12 : 12-h light/dark cycle with 60% relative humidity.The protocols for animal experimentation described in this article were previously approved by the Animal Research Committee, Akita University (Approval number: a-1-2956); the 'Guidelines for Animal Experimentation' of the University were completely adhered to in all subsequent animal experiments.We followed the essential 10 of the ARRIVE 2.0 guidelines for animal experiments.

Molecular modeling
Molecular modeling was performed using an MF MYPRESTO v2.1 (FiatLux, Tokyo, Japan).The structural data of HIF-2a was obtained from the Protein Data Bank (PDB ID: 3H82) [23].3H82 shows a structure of the heterodimer of HIF-2a and ARNT C-terminal PAS domains.Docking simulation with ligand was performed using the structure of PAS-B of HIF-2a extracted from 3H82.Topology data and grid potential were set up using default parameters.The calculation method in the Sievgene docking simulation was performed by Precise.Grid potentials were generated under the conditions.) and incubated with gentle rotation using a rotator (RM-2M, TOHO, Tokyo, Japan) for 1 h at 4 °C.After washing with the same buffer three times, the bound proteins were eluted from the column by adding 5 mM DHNA, 5 mM b-NF, or 5 mM 3-MC.The proteins were analyzed by SDS/PAGE and stained with Coomassie Brilliant Blue R250 or proteins were separated by SDS/PAGE followed by immunoblotting using an anti-AhR-, anti-HSP90-, anti-XAP2-, and anti-p23 antibody.

Immunoblotting
Samples were analyzed by SDS/PAGE, followed by immunoblotting.The PVDF membrane reacted with an anti-rabbit AP IgG (BioRad, Berkeley, CA, U.S.A.) antibody or an anti-mouse AP IgG (BioRad).Samples were treated with BCIP-NBT solution (Nacalai Tesque).The PVDF membrane reacted with an anti-rabbit HRP IgG (BioRad) antibody or an anti-mouse HRP IgG (BioRad).Samples were treated with an ECL Plus Western Blotting Detection System (Cytiva).Image analysis was performed using CHEMIDOCK XRS PLUS (BioRad).

Circular dichroism
Circular Dichroism was performed using a Jasco J-720 spectropolarimeter (Jasco, Tokyo, Japan).A 0.1 (v/v) % concentration of DMSO, 5 mM DHNA in DMSO, or 5 mM 3-MC in DMSO was added to 1.5 lM full-length AhR in 10 mM Tris-HCl buffer (pH 7.4).CD spectra were collected using instrumental parameters: 190 < k < 250 nm, 0.2 nm step size, 100 nmÁmin À1 scan rate, 1.0 s integration time, 2.0 nm bandwidth, in a 0.1 cm cell, with six total scans averaged.Secondary structure was predicted using J-720 for Windows protein secondary structure analysis.
In situ proximity ligation assay (in situ PLA) HeLa cells (2.0 9 10 5 cells per well) were grown on coverslips in 35 mm tissue culture plates for 24 h and stimulated by 3 lM DHNA or 3 lM 3-MC for 30 min.DMSO was used as a control.After incubation, the cells were washed twice with 1 9 PBS, then fixed with 4% paraformaldehyde for 15 min at 37 °C and permeabilized with 0.2% Triton X-100 for 10 min at room temperature.After washing, the cells were incubated with 3% BSA in 1 9 PBS for 1 h at room temperature and incubated with primary antibodies specific for AhR (anti-rabbit AhR) or XAP2 (anti-rabbit XAP2) or p23 (anti-rabbit p23) and HSP90 (anti-mouse HSP90) overnight at 4 °C.PLA Probe anti-rabbit PLUS and anti-mouse MINUS were applied, then Duolink Proximity Ligation Assays (PLA) were performed by using a detection kit (Sigma-Aldrich) according to the manufacturer's instructions.The cells were mounted onto glass slides by ProLong Gold (Invitrogen, Waltham, MA, USA) and visualized by BZ-X700 All-in-one Fluorescence Microscope (Keyence, Tokyo, Japan).The number of AhR-molecular chaperone signals in the nucleus was measured by BZ-X700 All-in-one Fluorescence Microscope and the mean and AESD were calculated.

Immunofluorescence
HeLa cells grown on a cover slip were treated with the vehicle (DMSO), 3 lM 3-MC, or 3 lM DHNA for 2 h, fixed in ice-cold methanol at 4 °C for 15 min, washed three times with PBS, and incubated with 1% BSA in PBS at room temperature for 1 h.After washing with PBS, the primary antibody against HSP90b, AhR, p23, or XAP2 (diluted in 1% BSA/PBS) was mounted on a coverslip at 4 °C for 18 h.The cells were washed three times with PBS and incubated with an Alexa 488-or Alexa 546-conjugated secondary antibody for 3 h at room temperature.Finally, the cells were washed three times with PBS, incubated with DAPI (40, 6-diamidino-2-phenylindole) for 30 min, and mounted onto a slide glass with ProLong Gold antifade reagent (Invitrogen).Immunofluorescence images were obtained by confocal laser microscopy (LSM780, Zeiss, Jena, Freistaat Thuringen, Germany).

Reverse transcription-polymerase chain reaction (RT-PCR)
Total RNA was isolated from cells cultured in 60 mm tissue culture plates using RNeasy Mini Kit (Qiagen, Hilden, Germany).According to the manufacturer's instructions, RNA (1 lg) was used for synthesizing the cDNA with a Super Script III kit (Invitrogen).RT-PCR was performed with Pri-meSTAR GXL Polymerase (Takara-Bio, Shiga, Japan) using the following conditions: initial denaturation at 94 °C for 2 min, followed by 40 amplification cycles (94 °C for 15 s, 55 °C for 30 s, 68 °C for 30 s).The cDNAs were amplified by SimpliAmp (ThermoFisher Scientific, Waltham, MA, USA) with the primers (see PCR Primers).PCR products were separated in 1% agarose gel at 100 V and stained with ethidium bromide.The band ratios were quantified using IMAGE J software and normalized by b-actin.

Molecular modeling
We first performed in silico docking simulations to determine whether DHNA binds to the ligand-binding pocket (PAS-B domain) of the AhR.Topology data and a grid potential were set up using default parameters described previously [20].We first generated grid potentials (front-, rear-, top-, bottom-, side 1-, and side 2 view) under the conditions described in the Materials and methods section (Fig. 1A).We used ligands (DHNA, TCDD, 3-MC, 3-MI, and b-NF), and methotrexate and reserpine were used as negative controls.We then attempted to analyze the fit of these compounds to the model from various angles (Fig. 1B-G).We have expanded and compared the binding of DHNA, TCDD, 3-MC, 3-MI, and b-NF ligands to the ligand pocket of the PAS-B domain and found that DHNA, TCDD, 3-MC, 3-MI, and b-NF mainly bind to the b-sheet of PAS-B, with a helix and loop supporting ligand binding.DHNA, TCDD, 3-MC, 3-MI, and b-NF appear to bind well to the ligand pocket (Fig. 1B-G).On the other hand, methotrexate and reserpine do not fit well in situations where they protrude from the ligand pocket.If the ligand binds well to the receptor, the docking score will be shown as a minus number with higher minus numbers meaning better docking.The docking scores are shown in Table 1.The order of fitting ligands was DHNA, b-NF, TCDD, 3-MI, and 3-MC.Methotrexate and Reserpine had positive scores in that order, indicating that these two compounds did not fit the molecular modeling.An in silico docking simulation indicated that non-toxic DHNA is the ligand for the AhR.
Changes in the secondary structure of AhR upon binding of toxic and non-toxic ligands We investigated the influence of ligands on the secondary structure of full-length AhR on the CD spectrum.
The secondary conformational change of AhR upon binding of the toxic ligand 3-MC or the non-toxic ligand DHNA was shown in Fig. 1H.In the secondary structure of AhR, compared to ligand-free DMSO, DHNA showed positive maxima at 191-193 nm and negative maxima at 195-200 nm and 207 nm, suggesting an increase in a-helix and a slight increase in random structure.On the other hand, in the case of 3-MC, a decrease in the molar ellipticity (h) values at 216-218 nm and a negative maximum at 195 nm were observed, suggesting a decrease in the b-structure [24].Thus, the 3-MC and DHNA ligands are suggested to cause mutually distinct secondary conformational changes in the AhR.Based on the results of ligand-dependent differences in the secondary structure of AhR, we hypothesized that there might be differences in the components of the AhR-molecular chaperone complex depending on the ligand.Most AhR ligands are selective AhR modulators (SAhRMs) in which their binding differences are reflected in their differentially induced responses [23,25,26].The results of this study support the concept of SAhRMs.

DHNA binds to the AhR-HSP90-XAP2 complex
We confirmed the AhR-molecular chaperone complex using the DHNA-, 3-MC, and b-NF affinity columns.Since the non-toxic ligand, DHNA has an OH group, Epoxy-activated Sepharose 6B was selected as the coupling gel.On the other hand, the toxic ligand 3-MC has only a CH 3 group and C=C groups.In the selection of coupling gels for the preparation of affinity columns, no coupling gel with a -CH 3 group was found.The synthetic flavonoid that has been used as an agonist of the AhR b-NF has a C=O group.On the other hand, the C=O or C=C group can be coupled by using Thiopropyl-Sepharose 6B.Therefore, an affinity column using 3-MC as the toxic ligand or AhR agonist b-NF was prepared using the Thiopropyl-Sepharose 6B and investigated.As shown in Fig. 1I, AhR, HSP90, and XAP2 were detected in the eluents from the DHNA-affinity column, but not p23.On the other hand, when a 3-MC affinity column was used, AhR, HSP90, XAP2, and p23 were detected in the eluent from the column (Fig. 1J).The same data were obtained from the b-NF affinity column (Fig. 1K).
No eluting proteins were detected on each DHNA-, 3MC-, or b-NF Mock column (Figs S1-S3).The results showed that the components of the AhR-molecular chaperone complex in the case of toxic ligands, 3-MC, were HSP90, XAP2, and p23 [11].On the other hand, in the case of the non-toxic ligand DHNA, the components of the AhR-molecular chaperone complex were HSP90 and XAP2, and p23 was not included in the complex.The AhR agonist b-NF was also similar to toxic ligand 3-MC, and the molecular chaperone complex of AhR was HSP90, XAP2, and p23.
AhR was translocated to the nucleus with the HSP90-XAP2 complex when DHNA is the ligand Differences in the composition of the AhR-molecular chaperone complexes bound by toxic and non-toxic ligands were also suggested to affect the activation of the AhR.We performed cellular experiments to determine whether DHNA activates the AhR.We added 3 lM DHNA or 3 lM 3-MC to cells and monitored the subcellular localization of AhR using PLA (In situ proximity ligation assay) according to the 'Materials and methods' procedure.In this experiment, we used DMSO as a control.As shown in Fig. 2A, AhR was detected in the cytoplasm complexed with HSP90 in the absence of ligands.When in the presence of 3-MC or DHNA, the AhR-HSP90 complex was detected both in the nucleus and cytoplasm on PLA.The same data were obtained from the immunofluorescence (Fig. S4).Regardless of the type of ligand, both HSP90 and XAP2 localized to the nucleus in the presence of the ligand (Fig. 2B and Fig. S5).On the other hand, the AhR-p23 complex was detected in the cytoplasm and nuclear translocation of the complex could not be confirmed when DHNA was the ligand (Fig. 2C and Fig. S6).However, the AhR-molecular chaperone complex components were suggested to be different from each other.In other words, when 3-MC is present as a ligand, p23 associates with the AhR-HSP90-XAP2 complex, whereas in the presence of DHNA, this association is absent.Although both the toxic ligand 3-MC and the non-toxic ligand DHNA activated the AhR and nuclear translocation was detected, the components of the molecular chaperone complex were different in the case of the toxic and non-toxic ligands.The number of AhR-molecular chaperone signals in the nucleus (Fig. 2A-C) was summarized in Table 2.

3-MC promotes rapid AhR nuclear translocation, whereas DHNA slowly promotes it
We studied whether differences in the ligands affect the time of nuclear translocation of the AhR.In the case of the 3-MC ligand, the nuclear localization was Docking simulation with or without ligand was performed using the structure of PAS-B of HIF-2a extracted from Protein Data Bank (ID: 3H82).(A) Grid potentials (white squires) were generated as described in 'Materials and methods'.The molecular models show the front-, rear-, top-, bottom-, side 1-, and side 2 views without the ligand.(B-G) Molecular model of AhR-PAS-B domain with the ligand.The prediction was performed with 272-342 amino acids of AhR using the Predict Protein.The molecular models in the absence (no ligand) or presence of the ligands (TCDD, DHNA, 3-MC, 3-MI, b-NF, Methotrexate, and Reserpine) show the front view (B), rear view (C), top view (D), bottom view (E), side 1 view (F), and side 2 view (G).(H) Circular dichroism spectrometry of full-length AhR.The circular dichroism spectrum of AhR in the presence or absence of DHNA or 3-MC was measured.The green solid line indicates the absence of a ligand (DMSO), and the blue and red solid lines indicate the presence of DHNA and 3-MC, respectively.The graph shows the molar ellipticity (h 9 10 À6 ) at wavelengths from 190 to 250 nm.Mouse brain cytosol was applied to the DHNA-affinity column (I), 3-MC affinity column (J), or b-NF-affinity column (K).The eluted proteins were analyzed by SDS/PAGE followed by immunoblotting with an antibody against AhR, HSP90b, XAP2, and p23 (I-K).confirmed at 30-60 min and not at 120-180 min (Fig. 3A-D).In the case of toxic ligands, AhR was translocated to the nucleus in a short time.On the contrary, in the case of the DHNA ligand, nuclear translocation was detected at 30 min, was seen to increase through 60-120 min, and subsequently decreased at 180 min.(Fig. 3A-D).These results  2. suggested that in the case of the toxic ligand 3-MC, the AhR is immediately translocated into the nucleus, whereas, in the case of the non-toxic ligand DHNA, the nuclear translocation of the AhR is relatively slow (Fig. 3A-D).The number of AhR-molecular chaperone signals in the nucleus (Fig. 3A-D) was summarized in Table 3.

3-MC strongly induces CYP1A1, while DHNA induces weekly
We next analyzed the CYP1A1 levels after AhR nuclear translocation.In the case of 3-MC, the CYP1A1 mRNA levels increased over time, and 3.7-fold more CYP1A1 mRNA was produced compared to the controls even  3.
8 h after the 3-MC administration (Fig. 4A).We have reported that CYP1A1 mRNA expression was induced approximately 3.2-fold at 8 h after administration of b-NF [17].On the contrary, in the case of DHNA, the CYP1A1 mRNA levels hardly increased, and even 2-6 h after the DHNA addition, the CYP1A1 mRNA levels were 1.4-fold higher than those in the controls (Fig. 4B).We also analyzed the CYP1A1 protein levels.
When DHNA was the ligand, the protein was barely detectable at 48 h (Fig. 4C).

Discussion
Proteostasis is maintained by a network of molecular chaperones, of which HSP90 is a major member.HSP90 regulates the function of more than 300 client proteins by altering co-chaperones such as p23, CDC-37, and HOP.The ligand-free form of AhR is usually found in cytosols composed of the HSP90, p23, and XAP2 [11,27].The AhR-HSP90-XAP2-p23 complex is probably the most stable.As for the toxic ligand of AhR, such as 3-methylcholanthrene (3-MC), benzo (a) pyrene, anthracene, polychlorinated dibenzo-p-dioxins, biphenyls, dibenzofurans have been reported [11,28].DHNA, a key substance in the whey fermentation of propionibacteria, has been reported to be a probiotic AhR ligand [21,22].Secondary structure analysis using CD spectra showed that the structures of the AhR when 3-MC and DHNA were the ligands differed from each other.We therefore focused on the components of the molecular chaperone complex of AhR when 3-MC and DHNA are the ligands.An analysis of binding proteins in mouse brains using the DHNA-affinity column detected AhR-HSP90-XAP2, but not p23.Similar to the affinity column experiments, the cell experiments confirmed that the AhR complex is AhR-HSP90-XAP2 when DHNA is the ligand.The structure of the AhR-molecular chaperone complex has recently been reported from two groups using cryo-electron microscopy [29,30].In the Cryo-EM structure of the intact HSP90-AhR-p23-XAP2 complex, AhR is shaped like a ball with threads hanging down from the lumen of the HSP90 dimer and is considered a mature, stable, ligand-unbound structure [26].Compared to before ligand binding, the structure of AhR upon ligand binding is altered and the NLS becomes more exposed.Upon activation by ligand binding, the cytoplasmic AhR complex translocates to the nucleus.After nuclear translocation, p23 is separated, and the dissociated AhR from the HSP90-AhR-XAP2 complex forms an active heterodimer with ARNT and functions as a transcription factor [29].
On the other hand, the structure of the indirubin-bound HSP90-XAP2-AhR was elucidated by Cryo-EM [30].Indirubin is one of the previously reported endogenous activators of AhR.Gruszczyk et al. prepared the HSP90-AhR-XAP2 complex in the presence of indirubin.They found that the PAS-A domain and the N-terminal portion of the AHR do not fit within the electron density and that this region has reported being highly flexible [30].This portion of the AhR complex was exposed to solvent and was free to interact with nuclear transport mechanisms, including the predicted importin complex; the NLS sequence of the AHR was reported to be within the first 60 amino acid residues of the N-terminal bHLH [11,17,20].
Thus, two AhR complexes, HSP90-AhR-XAP2-p23 and HSP90-AhR-XAP2, are thought to exist in the AhR complex.In our present study, The toxic 3-MC ligand is the former AhR-molecular chaperone complex, while the non-toxic DHNA ligand forms the latter AhR-molecular chaperone complex.AhR agonist b-NF ligand is the former AhR-molecular chaperone complex.Given the high flexibility of the N-terminus, including the NLS of AhR, as reported by Cryo-EM, it is suggested that the 3-MC-bound HSP90-AhR-XAP2-p23 complex is rapidly translocated to the nucleus and functions as a transcription factor for toxic metabolism and biological defense.On the other hand, our present results when DHNA is the ligand are consistent with those of HSP90-AhR-XAP2 and also showed that p23 is not required for ligand binding and target gene expression [31].The relatively slower nuclear migration when DHNA is the ligand may depend on the structural freedom of the AhR N-terminus.That is, the NLS of AhR in the absence of p23 binding is less exposed than when p23 is bound, suggesting that the NLS of AhR in the absence of p23 binding is less exposed than when p23 is bound.
The N-terminus of HSP90 is the ATP hydrolysis domain, and the middle domain is required to activate the ATP hydrolysis of Hsp90-N [32].The co-chaperone p23 binds to the hinge region of the N and middle domains of HSP90 and inhibits the ATP-hydrolyzing function of Hsp90.Wen et al. reported that the structure of the ligand-free AhR, HSP90 2 -AhR-XAP2-p23 2 , is  stable when analyzed by Cryo-EM [30].Based on numerous reports, the HSP90-AhR-AhR-XAP2-p23 structure of AhR does not change when the toxic TCDD or 3-MC ligands are bound, suggesting that the highly exposed NLS of the AhR is instantly nuclear transferred and induces a toxic metabolism.We analyzed the time required for the nuclear translocation of the AhR when 3-MC or DHNA is the ligand.In the case of the toxic ligand, nuclear translocation of the AhR takes place in a short time, whereas in the case of the non-toxic ligand, the AhR translocates relatively slowly.An increase in the CYP1A1 expression was also observed at the mRNA and protein levels when 3-MC was the AhR ligand.On the other hand, when DHNA was the AhR ligand, the expression level of CYP1A1 was increased by about 1.2-fold compared to the control at the mRNA levels and decreased to 1/5 of the control at the protein level after 48 h.DHNA activates AhR but does not perform physiological functions of xenobiotic metabolism such as CYP1A1.DHNA activates AhR but provides physiological functions other than xenobiotic metabolism.Thus, in the case of toxic ligands, AhR forms a complex with HSP90-XAP2-p23 and undergoes instantaneous nuclear translocation and xenobiotic metabolism, whereas, in the case of non-toxic ligands, AhR binds to HSP90-XAP2 and undergoes nuclear translocation relatively slowly, suggesting that it performs physiological functions other than the xenobiotic metabolism.
In this study, we demonstrated that DHNA from propionate-fermented whey is a non-toxic ligand for the AhR.The toxic and non-toxic ligand selectivity models for AhR in HeLa cells are shown in Fig. 5.The role of p23 in the AhR-molecular chaperone complex is supported by Cryo-EM structure studies [29,30].
We found that DHNA-bound AhR has a novel physiological function that is distinct from its drug metabolism function.We will publish somewhere in the near future regarding the new physiological functions of DHNA-bound AhRs.Our study will contribute to the research on non-toxic ligands of AhR.

Fig. 1 .
Fig. 1.Interaction between DHNA and AhR.(A-G) Molecular modeling.The secondary structure prediction of the AhR-PAS-B domain.Docking simulation with or without ligand was performed using the structure of PAS-B of HIF-2a extracted from Protein Data Bank (ID: 3H82).(A) Grid potentials (white squires) were generated as described in 'Materials and methods'.The molecular models show the front-, rear-, top-, bottom-, side 1-, and side 2 views without the ligand.(B-G) Molecular model of AhR-PAS-B domain with the ligand.The prediction was performed with 272-342 amino acids of AhR using the Predict Protein.The molecular models in the absence (no ligand) or presence of the ligands (TCDD, DHNA, 3-MC, 3-MI, b-NF, Methotrexate, and Reserpine) show the front view (B), rear view (C), top view (D), bottom view (E), side 1 view (F), and side 2 view (G).(H) Circular dichroism spectrometry of full-length AhR.The circular dichroism spectrum of AhR in the presence or absence of DHNA or 3-MC was measured.The green solid line indicates the absence of a ligand (DMSO), and the blue and red solid lines indicate the presence of DHNA and 3-MC, respectively.The graph shows the molar ellipticity (h 9 10 À6 ) at wavelengths from 190 to 250 nm.Mouse brain cytosol was applied to the DHNA-affinity column (I), 3-MC affinity column (J), or b-NF-affinity column (K).The eluted proteins were analyzed by SDS/PAGE followed by immunoblotting with an antibody against AhR, HSP90b, XAP2, and p23 (I-K).

Fig. 2 .
Fig. 2. Activation of AhR by DHNA.(A-C) HeLa cells were treated with DMSO (control), 3 lM DHNA, or 3 lM 3-MC for 30 min.Cells were incubated with anti-HSP90b and anti-AhR antibodies (A), anti-HSP90b and anti-XAP2 antibodies (B), and anti-HSP90 b and anti-p23 antibodies (C).Duolink Proximity Ligation Assays (PLA) were performed using a detection kit.Blue staining indicates DAPI staining of the cell nucleus(A-C).BZ-X700 All-in-one Fluorescence Microscope (Keyence Tokyo, Japan) obtained immunofluorescence images.The bar in all panels is 10 lm.The number of AhR-molecular chaperone signals in the nucleus (A-C) was summarized in Table2.

Fig. 3 .
Fig. 3. Comparison of nuclear translocation times of AhR by DHNA or 3-MC.HeLa cells were treated with the DMSO (control), 3 lM DHNA, or 3 lM 3-MC for 30 (A), 60 (B), 120 (C), and 180 min (D).Cells were incubated with anti-HSP90b and anti-AhR antibodies.Fluorescence signals were detected by the PLA method.The bar in all panels is 10 lm.The number of AhR-molecular chaperone signals in the nucleus (A-D) was summarized in Table3.

Fig. 4 .
Fig. 4. CYP1A1 mRNA and protein expression.HeLa cells were cultured in the presence of 3 lM 3-MC (A) or 3 lM DHNA (B) for 1, 2, 4, 6, and 8 h.The CYP1A1 was investigated by RT-PCR or immunoblotting with antibodies against CYP1A1 and b-Actin.The HeLa cells were cultured in the presence of 3 lM 3-MC or 3 lM DHNA for 24 and 48 h.The CYP1A1 was analyzed by immunoblotting with an anti-CYP1A1 antibody and an antibody against b-Actin.Samples were treated with an ECL Plus Western Blotting Detection System (C).The CYP1A1/ b-Actin ratio was quantified using IMAGE J software.The data are expressed as mean AE SD (n = 3) (A-C).DMSO was used as a control (A-C).

Fig. 5 .
Fig.5.Models for selecting toxic and non-toxic ligands for AhR by regulating molecular chaperone complexes in HeLa cells.Upon intracellular exposure to the toxic ligands of dioxins, AhR forms the HSP90, XAP2, and p23 complexes, which immediately translocate to the nucleus due to the high exposure of NLS, dissociate from the chaperone complexes in the nucleus, bind the nuclear AhR-binding protein, Arnt, bind the xenobiotic response element XRE and promote transcription of drug metabolism genes.When intracellularly exposed to non-toxic ligands, such as DHNA, the conformational change of AhR is accompanied by a decrease in NLS exposure.As a result, the structure of HSP90 also changes, and p23 bound to HSP90 dissociates.The AhR forms HSP90, and XAP2 complexes migrate relatively slowly to the nucleus due to the reduced exposure of the NLS in AhR.

Table 2 .
Number of nuclear AhR-molecular chaperone signals.

Table 3 .
Number of nuclear AhR-molecular chaperone signals at different times.