(Received for publication, November 18, 1996, and in revised form, December 20, 1996)
From the Department of Molecular Pharmacology, Stanford University School of Medicine, Stanford, California 94305-5332
To identify new proteins involved in dioxin-dependent signal transduction and transcriptional regulation, we used a yeast two-hybrid system to identify proteins that interact with the Ah receptor (AhR). We cloned a mouse cDNA, which encodes a novel ~37-kDa protein that binds to AhR; we have designated the protein as Ah receptor-interacting protein (AIP). The amino acid sequence of mouse AIP exhibits homology with members of the FK506-binding protein family. AIP also contains three tetratricopeptide repeat (TPR) motifs; the TPR sequence is present in proteins required for cell cycle control and RNA synthesis and in steroid receptor-binding immunophilins. Coimmunoprecipitation experiments in mouse hepatoma cells reveal that AIP is cytoplasmic and associates with unliganded Ah receptor and with hsp90; 2,3,7,8-tetrachlorodibenzo-p-dioxin treatment disrupts the AhR-AIP-hsp90 interaction. Overexpression of AIP augments the response of the CYP1A1 gene to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Our data suggest that AIP influences ligand receptivity and/or nuclear targeting of AhR.
The aromatic hydrocarbon receptor (AhR)1 is a cytoplasmic, ligand-activated, basic helix-loop-helix (bHLH) transcription factor, which functions together with a second bHLH protein, the Ah receptor nuclear translocator (Arnt), in mediating the induction of CYP1A1 gene expression in response to the environmental contaminant 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) (1-3). CYP1A1 encodes the microsomal cytochrome P450 1A1 enzyme, which oxygenates polycyclic aromatic hydrocarbons, such as the carcinogen, benzo(a)pyrene (4). AhR and Arnt have similar modular structures. The bHLH domains are located toward the amino termini of the proteins; the HLH regions mediate dimerization between AhR and Arnt, whereas the basic regions are involved in DNA recognition by the AhR·Arnt heterodimer (5-8). Both AhR and Arnt contain two inverted repeats (designated as PAS regions), which share homology with Per and Sim, two Drosophila proteins involved in circadian rhythm and central nervous system development, respectively (9, 10). The PAS regions may contribute to AhR·Arnt dimerization, ligand/hsp90 binding, and suppression of AhR's transactivation activity (7, 11, 12). The carboxyl regions of AhR and Arnt contain transactivation domains that contribute to transcriptional control by the AhR·Arnt heterodimer (12-15).
Induction of CYP1A1 gene transcription by TCDD is an interesting response for analyzing AhR function. Biochemical and genetic analyses of induction have revealed a multi-step pathway of AhR action (3, 16). In uninduced cells, the unliganded AhR is localized in the cytoplasm, complexed with hsp90. Upon ligand binding, AhR dissociates from hsp90 and enters the nucleus, where it interacts with Arnt. The AhR·Arnt heterodimer binds to a specific nucleotide sequence within an enhancer upstream of the CYP1A1 gene; this process is accompanied by alterations in chromatin structure, binding of general transcription factors to the promoter, and induction of transcription (16, and references therein).
Genetic and biochemical studies implicate AhR in numerous biological responses to TCDD, in addition to CYP1A1 induction (17-19). In experimental animals, TCDD induces a broad range of adaptive and toxic responses, including induction of other xenobiotic-metabolizing enzymes, tumor promotion, thymic involution, skin disorders, and alterations in endocrine homeostasis (17, 18). Recent studies imply that AhR is also required for normal liver development and immune function in mice (20, 21) and is involved in cell cycle control and differentiation in mouse and rat hepatoma cells (22, 23). Thus, AhR contributes to many complex cellular functions; therefore, we envisioned that AhR-mediated gene regulation may involve additional components that remain to be identified.
To test this hypothesis, we have used a yeast two-hybrid screening procedure to find new proteins that interact with AhR, and we have cloned a protein that we designate as AhR-interacting protein (AIP). Our findings reveal that AIP is a novel cytoplasmic protein, which binds to the unliganded AhR. Furthermore, overexpression of AIP in mouse hepatoma cells increases the response of the CYP1A1 gene to TCDD. Our results imply that AIP plays a positive role in AhR-mediated signaling, possibly by influencing AhR's receptivity for ligand and/or AhR's nuclear targeting.
Vent polymerase was from New England BioLabs (Beverly, MA). Restriction endonucleases and other DNA-modifying enzymes were from Life Technologies, Inc. and Promega (Madison, WI). Radioactive compounds were purchased from Amersham Corp. Cell culture materials were from Life Technologies, Inc. Media for yeast culture were from Bio 101, Inc. (Vista, CA) and Difco. Reduced glutathione and glutathione-agarose were from Sigma. Expression plasmids pGEX-4T3 and pET28a were from Pharmacia Biotech Inc. and Novagen (Madison, WI), respectively. TCDD was from the National Cancer Institute Chemical Carcinogen Reference Standard Repository.
Cell CultureWild-type hepa1c1c7, Ah receptor-defective
(AhR-D) cells, and Arnt-defective (Arnt-D) cells were grown as
monolayers in -minimal essential medium containing 10% fetal bovine
serum, in a 5% CO2 atmosphere, at 37 °C, as described
previously (24).
The full-length coding region of
mouse AhR cDNA was generated by PCR (12) and was fused to the
sequence encoding the Gal4 DNA-binding domain (DBD) in plasmid pGBT9
(Clontech) at the SmaI site. The resulting plasmid expresses
Gal4 DBD-AhR fusion protein under the control of the yeast
PADH1 promoter and was designated as pGBT-AhR. The plasmid
was cotransformed into yeast strain HF7C together with a library in
which cDNAs from HeLa cells were fused to the Gal4 transactivation
domain (AD) in pGADGH (Clontech), using protocols suggested by
Clontech. The interaction of the two hybrid proteins reconstitutes a
functional Gal4 protein that activates transcription of two reporter
genes, lacZ and HIS3, which were integrated into
the genome of the yeast host and respond to the Gal4 transcription
activator. Plasmids were isolated from His- and
-galactosidase-positive clones and were transformed into HB101, an
Escherichia coli strain that is auxotrophic for leucine.
Colonies containing the putative positive cDNAs were selected on
minimal medium lacking leucine. To test the specificity of the
interaction between AhR and the positive cDNA clones, plasmids were
cotransformed into yeast strain SFY526 that carries a Gal4-responsive lacZ reporter gene (Clontech), together with the empty
plasmid pGBT9, pGBT-ArntNT which encodes a fusion protein of Gal4 DBD, and the amino acids 1-470 of Arnt (25) or pGBT-AhR. Positive clones
were identified by their
-galactosidase activity in filter assays.
Using this approach, we isolated a HeLa cDNA fragment (~1 kb),
which encodes a peptide that interacts with AhR.
The nucleotide sequence of the HeLa 1-kb clone was determined by the dideoxynucleotide chain termination method. To obtain a full-length mouse cDNA, the 1-kb fragment was used to screen a mouse hepatoma (hepa1c1c7) cDNA library. A cDNA fragment (~1.35 kb) in the pBK/CMV vector (Stratagene) was isolated, as described previously (25), and was designated as pAIP/BKCMV. Sequence analyses were performed with GCG programs (Genetic Computer Group Sequence Analysis Software Package).
Protein Expression in E. coli and Preparation of Polyclonal AntibodiesThe EcoRI-XhoI fragment of the 1-kb cDNA was excised from the two-hybrid positive clone and was inserted into a bacterial expression vector pGEX-4T3. Expression and purification of the GST fusion protein was carried out as suggested by Pharmacia. The purified GST fusion protein was injected into rabbits to generate polyclonal antibodies, as described previously (22). Antibodies specific for the antigen were purified by using an affinity column that was prepared by coupling purified mouse AIP (AIP-T7, see below) to Affi-Gel (Bio-Rad, Ref. 22). The resulting antibodies recognize a single band of ~37 kDa on immunoblot analysis of whole cell lysate from hepa1c1c7 cells.
To express mouse AIP in E. coli, the coding region of AIP
cDNA was amplified by PCR with pAIP/BKCMV as a template. The
forward primer, 5 TCGATGATCA
3
, includes a BclI restriction site (in bold) and 19 nucleotides of the AIP 5
-end coding sequence (underlined). The T7
sequencing primer, 5
GTAATACGACTCACTATAGGGC 3
, which anneals to the
T7 promoter sequence downstream of the multiple cloning site of the
pBK/CMV vector, was used as the reverse primer. The PCR product was
digested with BclI and XhoI and inserted into two
E. coli expression vectors at BamHI and
XhoI sites. Plasmid pAIP/GEX4T3 expresses a GST-AIP fusion
protein, and plasmid pAIP/ET28a expresses an AIP-T7 fusion protein that
has a His.tag (6 histidine residues), which permits purification of the
fusion protein by using His.tag affinity chromatography, and a T7.tag
epitope (i.e. the initial 11 amino acid residues of the T7
gene 10 protein), which permits immunodetection of the fusion protein
by using commercially available monoclonal anti-T7.tag antibody
(Novagen), at its amino terminus. The plasmids were transformed into
E. coli strain BL21DE3 (Novagen). Both the GST-AIP and the AIP-T7 fusion proteins were expressed at high levels in E. coli. GST-AIP was purified using a glutathione-agarose affinity
column (Sigma), and AIP-T7 was purified with a pET His.Tag system
(Novagen).
AhR and Arnt proteins were expressed using the TNT reticulocyte lysate system (Promega) in the presence of [35S]methionine using pAhR/CMV and pArnt/BK as templates, as described previously (12). Five µl of 35S-labeled AhR or Arnt were mixed with either AIP-T7 bound to anti-T7.tag antibody-agarose beads (Novagen) or GST-AIP bound to glutathione-agarose beads (Sigma) in 1 ml of buffer A (25 mM HEPES, pH 7.5, 1.5 mM EDTA, 1 mM dithiothreitol, 10% glycerol, 135 mM NaCl, 1 mM phenylmethyl sulfonyl fluoride, 1 mM leupeptin, and 0.1% Triton X-100) for 4 h at 4 °C. The beads were washed five times in the same buffer and were boiled for 5 min in SDS sample buffer. The eluted proteins were separated in SDS-polyacrylamide gels and were detected by fluorography.
Overexpression of AIP-T7 in Hepatoma Cells and Interaction of AIP with AhR in SituTo express AIP in mouse hepatoma cells, the
full-length coding region of AIP cDNA with a T7.tag sequence at its
5-end was amplified by PCR with pAIP/ET28a as a template. The forward
primer, 5
GTCGAAGCTTCCAACATGTCG 3
,
contains an AflIII restriction site (in bold) and 20 nucleotides of the T7.tag sequence at its 3
-end (underlined). The
reverse primer, 5
TGCCTCTAGATGA 3
,
includes a BclI restriction site (in bold) and 21 nucleotides of the AIP coding sequence at its 3
-end (underlined). The
resulting PCR product was digested with AflIII and
BclI and inserted into pMFG, a retroviral expression vector,
at NcoI and BamHI sites, to generate pAIP-T7/MFG.
Five µg of pAIP-T7/MFG were transfected into Bosc23, an ecotropic
retrovirus-packaging cell line, using a modified calcium precipitation
method (26). Two days after transfection, 1 ml of the virus-containing
medium was removed and mixed with 2 ml of
-minimum essential medium
containing 10% fetal bovine serum plus Polybrene (5 µg/ml). The
virus-containing mixture was transferred to 60-mm diameter plates
containing wild-type (1 × 105), AhR-D (2 × 105), or Arnt-D (2 × 105) cells. The
plates were centrifuged at 1,000 × g for 90 min at 32 °C and were incubated at 37 °C, 5% CO2,
thereafter. Infection efficiency was greater than 90%, as estimated
from the percentage of cells infected with a retroviral vector that
expresses
-galactosidase. Expression of AIP-T7 fusion protein was
monitored by immunoblotting of the cell lysates with anti-T7 antibodies
as described below.
For coprecipitation experiments, wild-type, AhR-D, and Arnt-D cells were grown in 60-mm diameter plates and were treated with TCDD (1 nM). Dimethyl sulfoxide was the vehicle control. Sixteen hours later, the medium was removed and the plates were washed twice with phosphate-buffered saline. Cells were scraped into 400 µl of buffer A (see above) and were sonicated for 5 s. Aliquots of the whole cell lysate were mixed with 600 µl of buffer A and were incubated with anti-T7.tag antibody-agarose (Novagen) for 1 h at 4 °C with shaking. In other experiments, aliquots of the whole cell lysate were incubated with anti-AIP serum at 4 °C for 1 h, followed by incubation with protein A-agarose beads (Bio-Rad) for an additional 1 h with shaking. The beads were pelleted by brief centrifugation and were washed five times with buffer B (25 mM HEPES, pH 7.5, 1.5 mM EDTA, 1 mM dithiothreitol, 10% glycerol, 400 mM NaCl, and 0.5% Triton X-100). The beads were boiled for 5 min in 40 µl of SDS sample buffer. Equal aliquots were fractionated on three separate SDS-polyacrylamide gels, transferred to nitrocellulose membranes, and probed with different antibodies, as indicated under "Results."
Immunoblot, Northern, and Slot Blot AnalysesFor immunoblotting, total cell lysates, cytosolic and nuclear fractions, or coprecipitated pellets were prepared, fractionated on SDS-polyacrylamide gels, and transferred to nitrocellulose membranes according to established procedures (27). The blots were blocked with 10% dry milk, 0.1% Tween 20 in phosphate-buffered saline for 1 h with shaking. Blots were then incubated with antibodies against T7.tag (Novagen), AIP, AhR (22), or hsp90 (StressGen) for 1 h, followed by incubation with alkaline phosphatase-conjugated second antibodies for an additional 1 h. Nitro blue tetrozolium and 5-bromo-4-chloro-3-indolyl phosphate were used for color visualization (Promega).
For RNA analysis, total RNA was isolated from hepatoma cells using the
Qiagen total RNA purification system. Ten µg of RNA were fractionated
on a 1% agarose gel, transferred to a Nitran membrane, and hybridized
with a radiolabeled 700-base pair probe specific for CYP1A1
mRNA using the Rapid-hyb buffer (Amersham Life Sciences, Inc.).
After autoradiography, the blots were stripped and were reprobed with a
radiolabeled probe specific for mouse -actin, as a control. For slot
blots, 5 µg of total RNA were blotted to a Nitran membrane using a
Minifold II slot-blotter (Schleicher & Schuell). The blots were probed
with the CYP1A1 and the actin probes as described above.
To find new proteins that interact with AhR, we used a yeast two-hybrid system, which identifies proteins that interact with a particular bait protein (28). For bait, full-length AhR cDNA was fused to cDNA encoding the Gal4 DNA-binding domain. The Gal4-AhR fusion protein is incapable of transcriptional activation by itself, possibly due to the existence of an inhibitory domain(s) in AhR (12).
Using Gal4-AhR as bait, we screened a yeast two-hybrid library derived
from HeLa cell cDNAs. After screening 6 million colonies, we
obtained several positive clones that grew on selective media lacking
Leu, Trp, and His and were positive for -galactosidase. Four of the
clones interacted with AhR following transformation into yeast strains
containing various bait constructs (data not shown). Sequence analyses
of the clones implied that they were probably derived from the same
mRNA species. Northern blot analysis of total RNA from HeLa cells,
using the largest cDNA fragment (~1 kb) as a probe, revealed a
mRNA band of ~1.35 kb (data not shown).
To obtain full-length cDNA, we used the HeLa 1-kb fragment as a
hybridization probe to screen a mouse hepatoma (hepa1c1c7) cDNA
library. Ten overlapping clones were isolated and analyzed. The largest
clone contains a 1367-base pair insert (Fig.
1A). Two lines of evidence support the
conclusion that this 1.37-kb cDNA clone represents the full-length
mouse homologue of the HeLa 1-kb clone. (i) The 1.37-kb fragment
exhibits high homology with the HeLa cDNA sequence (~87%
identity). (ii) Northern blot analysis of mouse hepatoma mRNA using
the 1.37-kb fragment as a probe identified a single band of ~1.37 kb
(Fig. 2A), which agrees with the size of the
mRNA transcript recognized by the HeLa 1-kb fragment. We designated
the mouse 1.37-kb clone as "AhR-interacting protein" (AIP).
Primary structure of mouse AIP. A,
the nucleotide and predicted amino acid sequences of mouse AIP. A
region with sequence similarity to FKBP12 is shown in
italics. Three tetratricopeptide repeat (TPR)
regions are underlined. B, alignment of TPR
sequences. The amino acid sequences of AIP's TPR motifs were aligned with the TPR
motifs found in rabbit FKBP59 (44, 46) and a TPR consensus sequence
generated from CDC23, CDC16, NUC2,
SSN6, and SKI3 (41) using the Pileup and Pretty
programs of the GCG Wisconsin Sequence Analysis Package. The consensus
residues are present at or above a frequency of 0.50 in the TPRs shown.
Residues identical with or similar to the consensus are
boxed or circled, respectively. , hydrophobic
residues.
Analyses of the nucleotide sequence of mouse AIP cDNA indicate that
it contains an open reading frame with a putative Kozak sequence (29)
for translation initiation upstream of the first ATG codon of the
reading frame. A 22-nucleotide poly(A) tail is located at the end of
the 3-untranslated region of the cDNA. The predicted amino acid
sequence of mouse AIP (Fig. 1A) reveals 330 amino acid
residues, a molecular mass of 37,602 daltons, and a pI of 5.96. Immunoblot analysis of wild-type, AhR-defective, and Arnt-defective
cells using polyclonal antibodies against AIP revealed a single band of
~37 kDa in wild-type, Arnt-defective, and AhR-defective hepatoma
cells (Fig. 2B), in good agreement with the protein's
predicted molecular weight.
Inspection of AIP's amino acid sequence revealed a region (between residues 12 and 160) that exhibits sequence homology with human FKBP12 (30), with 52% similarity and 32% identity (Fig. 1A). AIP also exhibits homology with human FKBP59 (also known as p59 and hsp56, Ref. 31) (50% similarity and 27% identity) and mouse p51 (32) (50% similarity and 28% identity). These findings indicate that AIP is structurally related to the FKBP family of proteins, which function as molecular chaperones in steroid receptor signaling, heat shock responses, and drug-induced immunosuppression (33, 34). Analyses of the AIP sequence also reveal three regions that exhibit homology with the tetratricopeptide repeat (TPR), a motif found in several proteins that mediate a variety of functions, including the targeting of steroid receptors to the cell nucleus (35, 36). This finding suggests that AIP might participate in the movement of AhR from cytoplasm to nucleus.
Characterization of the AIP-AhR InteractionTo examine the
functional nature of the AIP-AhR interaction and to determine whether
AIP can also interact with Arnt, we used a Gal4-responsive, yeast
two-hybrid lacZ reporter system in which testing proteins
were fused to the DNA binding (DBD) or activation (AD) domains of Gal4.
Fig. 3A shows a representative result of such
studies. These results reveal that neither AIP nor AhR, by itself,
activates the lacZ reporter gene; however, the two proteins interact to produce an increase in lacZ gene expression
comparable with that produced by the AhR-Arnt interaction. Since Arnt
protein contains a constitutive transactivation activity toward its
carboxyl terminus (25), we used a Gal4-ArntNT fusion in which Gal4 DBD was fused with the amino-terminal region of Arnt to test the
interaction of AIP with Arnt. Our results indicate that AIP does not
interact productively with Arnt's amino-terminal region, which
contains its bHLH and PAS domains. Together, these observations imply
that AIP exhibits selectivity toward AhR (compared with Arnt) and that the AIP-AhR interaction is capable of producing a biological
response.
Immunoprecipitation experiments using T7-tagged AIP reveal that AhR, but not Arnt, coprecipitates with AIP (Fig. 3B). We obtained similar results using GST-tagged AIP (data not shown). These findings are consistent with our observations in yeast; they imply that AIP preferentially interacts with AhR and not with Arnt, even though AhR and Arnt have similar modular organizations. The basis for the preferential interaction remains to be determined.
To study the AIP-AhR interaction under more physiological conditions,
and to determine the effect of TCDD upon the interaction, we performed
immunoprecipitation experiments using wild-type, AhR-defective, and
Arnt-defective cells, in which we expressed a AIP-T7.tag fusion protein
(AIP-T7) that permits immunochemical studies with a monoclonal antibody
against the T7.tag sequence at the amino terminus of the fusion
protein. Fig. 4A shows that AIP-T7 was
overexpressed in wild-type, Arnt-D, and AhR-D hepatoma cells as
compared with endogenous native AIP, and the expression levels were
comparable in these cell lines. Immunoprecipitation studies (Fig.
4B) reveal that, in extracts from uninduced wild-type cells,
AIP coprecipitates with AhR and hsp90. Because unliganded AhR and hsp90
are cytoplasmic proteins (Fig. 5, Ref. 37), our findings
imply that AIP, hsp90, and AhR exist as a complex in the cytoplasm of
uninduced cells. In contrast, we find that little hsp90 and virtually
no AhR coprecipitates with AIP in TCDD-induced wild-type cells. We
obtained similar results using uninduced and TCDD-induced wild-type
cells that express AIP in normal amounts; these findings verify that
AIP interacts with AhR under physiological conditions (Fig.
4C). Our observations imply that TCDD binding by AhR is
associated with disruption of the AhR·AIP·hsp90 complex. Immunoprecipitation experiments using extracts from Arnt-defective cells reveal that AIP, hsp90, and AhR coprecipitate even in extracts from TCDD-treated cells (Fig. 4B). These findings indicate
that liganded AhR can maintain an interaction with AIP and hsp90 and that the presence of Arnt is necessary for disruption of the
three-protein complex. Studies in AhR-defective cells reveal that AIP
coprecipitates with hsp90 even in the absence of detectable AhR (Fig.
4B). These findings imply that AIP interacts with hsp90, as
well as with AhR.
We determined AIP's subcellular location by immunoblotting (Western) analyses of cytosolic and nuclear extracts from wild-type cells. Our findings (Fig. 5) reveal that AIP is present in the cytosol, but not in the nucleus, in both uninduced and TCDD-induced cells. In contrast, AhR is cytosolic in uninduced cells but is nuclear in TCDD-induced cells, as expected. These findings imply that AIP is cytoplasmic. Its location in the cytoplasm imposes a constraint upon the possible mechanisms by which AIP could influence AhR-mediated signaling.
Effect of AIP Overexpression on AhR-mediated CYP1A1 Induction by TCDDTo analyze the functional importance of AIP, we asked
whether overexpression of AIP influenced the response of the
CYP1A1 gene to TCDD. Our findings reveal that induction of
CYP1A1 mRNA is increased 2-3-fold in wild-type cells
that overexpress AIP (Fig. 6). Thus, AIP has a positive
influence on AhR-mediated signaling. Dose-response studies (Fig.
6A) reveal that AIP has no detectable effect on the
sensitivity of the CYP1A1 induction mechanism to TCDD. These
findings imply that AIP does not alter the affinity of AhR for TCDD.
Instead, AIP increases the extent of CYP1A1 induction at
each time point studied (Fig. 6B). Therefore, our data imply that AIP increases the amount of AhR that is transcriptionally active.
We have cloned and characterized a novel 37-kDa protein that interacts with AhR in the cytoplasm of the cell. We have designated the protein as AhR-interacting protein (AIP). Our studies suggest that AIP may act in concert with hsp90 and may contribute molecular chaperone and/or nuclear targeting activities to AhR-mediated signaling. In doing so, AIP produces a positive effect on TCDD-inducible CYP1A1 gene expression.
Mouse AIP exhibits amino acid sequence homology with FKBP12, a prototype of the FKBP family (30), and FKBP59, an immunophilin that interacts with unliganded steroid receptors (31). The FKBP family is implicated in several biological processes, such as protein folding, steroid receptor signaling and nuclear targeting, heat shock responses, and drug-induced immunosuppression (33, 34). Studies of steroid receptor-binding immunophilins reveal that FKBP59 is a component of the multi-protein complex containing unliganded steroid receptor as well as hsp90, hsp70, and an acidic protein, p23 (34, 38). FKBP59, together with other proteins in the complex, may act as a molecular chaperone that maintains the steroid receptor in a ligand-responsive configuration; the peptidylprolyl cis-trans-isomerase activity of FKBP59 may be important in this regard (39). Like the glucocorticoid receptor, AhR is a ligand-activated transcription factor that resides in the cell cytoplasm complexed with hsp90. By analogy with glucocorticoid receptor, the cytoplasmic AhR complex may also contain an immunophilin-like component. Previous immunochemical studies reveal that FKBP59 and CyP-40, two steroid receptor-binding immunophilins, are not associated with AhR (40). Our findings suggest that AIP might contribute an FKBP-like function(s) to AhR-mediated signaling.
The AIP protein contains three regions that exhibit homology with the tetratricopeptide repeat (TPR) motif, a degenerate conserved sequence of 34 amino acid residues (35). The TPR motif was first noted in CDC23, CDC16, and NUC2, three cell cycle genes required for completion of mitosis in yeast, and in SSN6 and SKI3, two genes required for mRNA synthesis in yeast (41). Subsequent studies have revealed that TPR exists in multiple copies in a diverse group of proteins that function in mitosis, transcription, RNA splicing, protein import, serine/threonine phosphorylation, and neurogenesis (35). Structural analyses of the TPR motif suggest that it may serve as an interface for protein-protein interactions (35, 42). Mutational analyses of some TPR domains indicate that the motif is essential for function (42, 43). It is notable that the two steroid receptor-binding immunophilins FKBP59 and CyP-40 each contain three TPR motifs, as does AIP (44, 45). Biochemical studies reveal that the TPR regions of FKBP59 are required for binding to hsp90 (46). Others have proposed that FKBP59 or CyP-40 interacts with TPR acceptor(s) in the glucocorticoid receptor·hsp90 complex and thereby directs the multi-protein complex to appropriate cellular targets (36). By analogy, we hypothesize that AIP contributes to the nuclear targeting of AhR through interactions between its TPR motifs and other proteins. Identification of the proteins that interact with AIP's TPR motifs may shed light on the molecular basis for the nuclear targeting and transport of liganded AhR.
AhR does not coprecipitate with AIP in TCDD-induced wild-type hepatoma cells (Fig. 4B). This observation could indicate that AIP dissociates from AhR following ligand binding. However, in Arnt-defective cells, AIP, AhR, and hsp90 coprecipitate in the presence of TCDD (Fig. 4B); this finding implies that AIP (as well as hsp90) remains associated with liganded AhR. Therefore, we infer that AIP has an effect on AhR-mediated signaling that occurs subsequent to ligand binding, and we hypothesize that AIP helps target the liganded AhR to the nucleus. Our findings also indicate that Arnt is required for the disruption of the AhR·AIP·hsp90 complex. Given the nuclear location of Arnt (37), we envision that Arnt binds liganded AhR upon its arrival in the nucleus, thereby stabilizing its nuclear localization.
Overexpression of AIP in mouse hepatoma cells enhances the response of the CYP1A1 gene to TCDD. These findings impute functional importance to the AIP-AhR interaction in AhR-mediated signaling. There are several possible mechanisms by which AIP may increase CYP1A1 induction. First, AIP might facilitate transport of the liganded AhR to the nucleus; thus, overexpression of AIP would enhance the nuclear accumulation of liganded AhR. Second, AIP might act as a molecular chaperone that maintains AhR in a ligand-responsive configuration; thus, overexpression of AIP would increase the amount of functional AhR in the cytoplasm. Third, AIP might interfere with AhR degradation; thus, overexpression of AIP might increase the overall cellular content of AhR. Each scenario would likely produce a positive effect on the induction of CYP1A1 gene expression in response to TCDD. Testing these hypotheses will provide new insights into the mechanism by which AhR transduces the signals of TCDD and other ligands.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) U85489[GenBank].
We thank Dr. G. P. Nolan for providing plasmid pMFG, M. L. Tuggle for secretarial assistance, and I. Kent Reed for a critical review of the manuscript.