The Q-rich/PST domain of the AHR regulates both ligand-induced nuclear transport and nucleocytoplasmic shuttling

Abstract

The aryl hydrocarbon receptor (AHR) shuttles continuously between cytoplasm and nucleus, unless ligand-binding triggers association with the AHR nuclear translocator (ARNT) and subsequent binding to cognate DNA motifs. We have now identified Val 647 as mandatory residue for export from the nucleus and AHR-function. This residue prevents inactivation of the receptor as a consequence of nuclear sequestration via constitutive import. Concomitantly mutants lacking this residue are exclusively localised in the nucleus. Although ligands accelerate nuclear import transiently, stable nuclear transition depends on a motif adjacent to Val 647 that comprises residues 650-661. Together, this defined region within the Q-rich domain regulates intracellular trafficking of the AHR in context of both nucleocytoplasmic shuttling and receptor activation. Nuclear export therefore depends on the previously characterised N-terminal NES and the newly identified motif that includes V647. Nucleocytoplasmic distribution of full-length human AHR is further affected by a section of the PST domain that shows sequence similarities with nuclear export signals. In concert, these motifs maintain a predominant cytoplasmic compartmentalisation, receptive for ligand binding.

Figure 1. Constitutive nucleocytoplasmic shuttling of the AHR is accelerated by ligands.

(a) Fluorescence images demonstrating the translocation of the full-length EYFP-AHR into the nucleus in HepG2 cells after treatment with 10 μM β-naphthoflavone (βNF), 100 μM kynurenine (Kyn) and 40 nM leptomycin B (LMB) for 15 min (scalebar = 20 μm). The selection of these concentrations was based on a dose-response analysis (Supplementary Fig. S2b). Graphs in the diagrams visualise the shift of nuclear staining in relation to total fluorescence of the analysed cells over 15 min. All three compounds triggered a comparable transition of EYFP-AHR into the nucleus, as indicated by comparable slopes of the graphs. Each bar (right diagram) represents the mean of at least 10 analysed cells +/− S.E.M. ***p < 0.001. (b) βNF accelerates the basal nucleocytoplasmic shuttling. Cells were treated for 10 min with 40 nM LMB, co-treated for another 10 min with 10 μM βNF (left side). Nuclear transition of EYFP-AHR was recorded and analysed as described above. Nucleocytoplasmic shuttling continues in the presence of ligand. Cells were treated with 10 μM βNF for 10 min, then co-treated with 40 nM LMB and analysed as described above (middle). Slopes of the recorded graphs were separately determined for single and combined treatments as indicated (right side). Each bar represents the mean of at least 10 cells +/− S.E.M. (c) Thermal shift assay with purified recombinant human AHR (protein purification is summarised in Supplementary Fig. S4). Shown is the relative thermal stability of AHR in the presence of LMB, Kyn and βNF at different concentrations. Compared to unliganded AHR, βNF and Kyn increased thermal protein stability, while LMB apparently has a destabilising effect. A decrease of stability was noted using 150 μM Kyn, possibly triggered by precipitation. Shown are the means of three biological replicates +/− S.E.M. (d) Time-dependent transcriptional activation of CYP1A1 and CYP1B1 in HepG2 cells. Induction was only seen after treatment with 10 μM βNF but not with 40 nM LMB. Displayed values represent relative inductions of transcripts normalised to the solvent control. Values shown are means of three biological replicates +/− S.E.M.

Figure 2. V647 determines the compartmentalisation of the AHR: expression of fluorescent AHR deletion mutants in HepG2 cells.

Deletion mutants were derived from full-length human pEYFP-AHR-C1 (AHR⁸⁴⁸). Mutants are named according to truncations sites, as defined by the last included residue. These sites are marked on the drafted full-length protein. Representative images that reflect the typical compartmentalisation are shown (scalebar = 20 μm). Mutants truncated after amino acid 391 (AHR^∆391), 509 (AHR^∆509), 640 (AHR^∆640) 644 (AHR^∆644) and 646 (AHR^∆646) show an exclusive nuclear staining, whereas the full-length protein (AHR⁸⁴⁸) is predominantly located in the cytosol. AHR^∆647 is nearly exclusively detected in the cytoplasm. Inclusion of the Q-rich domain does increase nuclear association (AHR^∆723). This is balanced by a motif localised between Pro 728 and Leu 744 within in the PST domain. AHR^∆647 and full-length AHR⁸⁴⁸ show a similar predominantly cytoplasmic localisation. Replacing of residues M645, Q646 and V647 (AHR^{M645A,Q646A,V647A}), or V647 only (AHR^V647A) by alanines led to an exclusive nuclear staining, whereas mutant AHR^Q646A showed wild-type compartmentalisation (lower panel left). On the other side, replacement of V647 with isoleucine (AHR^V647I;∆647) did not affect the cytoplasmic staining pattern (upper panel right).

Figure 3. Compartmentalisation of EYFP-AHR mutants in transfected cells.

A total of at least 300 positive cells that were found in randomly selected optical fields were analysed and classified after 24 h according to the defined staining patterns. Data represent the mean +/− S.D. out of three independent transfections. Insert: Staining patterns have been defined according to shown examples. N >> C exclusively nuclear; N > C predominantly nuclear; N = C equal distribution; N < C predominantly cytoplasmic.

Figure 4. Ligand-induced nuclear association of the AHR depends on the Q-rich domain.

(a) Translocation of EYFP-AHR^∆647 into the nucleus in HepG2 cells after treatment with 100 μM Kyn or 10 μM βNF for 15 min (left). Translocation of AHR^∆647 after exposure to 40 nM LMB for 10 min, followed by direct addition of 100 μM Kyn or 10 μM βNF for another 10 min (middle). Representative images of treated cells are shown for the indicated time points (scalebar = 20 μm). Nuclear transition was recorded and analysed as described in Fig. 1a. Slopes of the linear transition graphs have been separately determined for single and combined treatments (right). Each bar represents the mean +/− S.E.M. of 6 analysed cells. (b) Snapshots of cells imaged from transfected populations that were treated for 1 h with Kyn or left untreated. In response to ligand, AHR^∆661 and full-length AHR⁸⁴⁸ showed a nearly exclusive nuclear staining pattern, while AHR^∆650 remained predominantly cytoplasmic as in non-treated cells. Similar effects were observed after application of βNF. (c) Residues 648–661 are required for ligand-induced nuclear accumulation of the AHR. Cells expressing AHR^∆647, AHR^∆650, AHR^∆661 or AHR⁸⁴⁸ (full-length) were treated with 100 μM Kyn or 10 μM βNF. Both, AHR^∆661 and AHR⁸⁴⁸ showed significantly higher nuclear translocation rates than AHR^∆647 (two-way ANOVA, **p < 0.01, ***p < 0.001, ****p < 0.0001). No such significant differences were observed between AHR^∆647 and AHR^∆650. Values depicted represent the mean +/− S.E.M of at least 5 cells.

Figure 5. Nuclear export and intracellular trafficking of the human AHR are regulated by defined motifs.

(a) The nuclear localisation signal (NLS) within the N-terminal domain triggers continuous basal import (green arrows) into the nucleus (shuttling). Contrary to this autonomous import mechanism, function of the adjacent nuclear export signal (NES) depends on C-terminal motifs, especially the mandatory residue V647. (b) Ligands (marked with L) accelerate import, while continued export (blue arrow) counteracts nuclear sequestration of the AHR, thus maintaining a predominant cytoplasmic fraction that is receptive for interactions with ligands. Notably, mutants that lack parts of the C-terminal domain (AHR^Δ647 and AHR^Δ650) do not efficiently accumulate in the nucleus, although nuclear transfer is accelerated by ligands. (c) Export of the AHR continues in the presence of ligands. Activation of the AHR might involve several passages of receptor molecules that need to engage in further associations with nuclear components during limited time intervals. Stable associations of the AHR with the nucleus likely require a defined section of the Q-rich domain (green, Pro 661 is indicated). However, it is as yet completely unknown how this motif stabilises nuclear compartmentalisation or whether it promotes interactions of the transactivation domain with transcription factors. (d) The N-terminal NES and the V647 motif facilitate an efficient nuclear export of AHR^Δ647, leading to a nearly exclusive cytoplasmic pattern. On the other side, the full-length AHR contains an additional motif within the PST domain to maintain a predominantly cytoplasmic compartmentalisation.