Maintenance of mammalian cell lines and cell transfection
As in previous studies (e.g. [24–26], experiments were performed on HEK293 cells stably expressing WT hERG  or transiently transfected with hERG mutant cDNAs. The hERG stable-line was generously provided by Professor Craig January . HEK293 cells (ECACC, Porton Down, UK) were transiently transfected with cDNA plasmids using Lipofectamine 2000 (Invitrogen, Paisley, UK) according to the manufacturer's instructions. Expression plasmid encoding CD8 was also added (in pIRES, donated by Dr I Baró, University of Nantes, France) as a marker for successful transfection. Cells were passaged using enzyme free cell dissociation solution (Millipore, Watford, UK) and plated onto sterilized shards of glass coverslips in 40-mm petri dishes containing a modification of Dulbecco minimum essential medium with Glutamax-1 (DMEM; Invitrogen, Paisley, UK). This was supplemented with 10% fetal bovine serum (Gibco, Gloucester, UK), 50 µg/mL gentamycin (Invitrogen, Paisley, UK), and 400 µg/mL geneticin (G418, Invitrogen, Paisley, UK) for the WT hERG-expressing line. For experiments utilizing cells transiently transfected with hERG mutants, recordings were performed 12-72h after transfection. Successfully transfected cells (positive to CD8) were identified using Dynabeads® (Invitrogen, Paisley, UK) [25;26]. The N588K, S624A, Y652A, M651A and F557L mutations to hERG1a have all been used in prior studies from our laboratory (e.g. [24–26;28];, as has hERG1a/1b co-expression [25;30]. Zebrafish ERG (zERG) was synthesised and provided in pcDNA3.1 by Genscript (Leiden, Netherlands; NCBI Reference Sequence: NM_212837.1).
The F656V mutation was made to hERG1a as described in . F656T was made using QuikChange (Agilent) mutagenesis using conditions described in  and the following primer sequences:
forward − 5’-GTATGCTAGCATCACCGGCAACGTGTCG-3’
reverse − 5’- CGACACGTTGCCGGTGATGCTAGCATAC-3’
Experiments were performed on hERG1a current (IhERG1a) except for the data in Fig. 2B and C, which were conducted using co-expressed hERG1a and 1b channels (IhERG1a/1b), and supplementary Fig. 2 which shows data from zERG (IzERG).
Recordings were made as described previously [25;26]. In brief, electrophysiological recordings were made at least 24 hrs after transfection, using an Axopatch 200B amplifier (Molecular Devices) with a CV-4/100 headstage and data acquisition via a Digidata 1320 interface (Molecular Devices). Glass shards with plated HEK293 cells were placed in the recording chamber of an inverted microscope (Nikon Diaphot, USA). The extracellular superfusate was a standard Tyrode’s solution containing (in mM): 140 NaCl, 4 KCl, 2.5 CaCl2, 1 MgCl2, 10 glucose, and 5 HEPES (titrated to pH 7.4 with NaOH) [24–26]. Patch pipettes (AM-systems Inc, USA) had resistances of 2–4 MΩ and were filled with a solution containing (in mM): 130 KCl, 1 MgCl2, 5 EGTA, 5 MgATP and 10 HEPES (titrated to pH 7.2 with KOH) [24–26]. Series resistance was typically compensated by 60–80%. Currents were filtered at 1–5 kHz depending on the voltage protocol used and were digitized at 10 kHz. All measurements were made at room temperature (21 ± 1oC). Room temperature was employed to facilitate accurate evaluation of phenanthrene effects because in pilot experiments at 37oC, stability of phenanthrene-containing superfusate appeared variable. It also facilitated comparison of phenanthrene blocking potency between human and zebrafish ERG data in this study and with prior recordings from fish myocytes conducted at room temperature [11–13].
Evaluation of the effects of 3 and 30 µM phenanthrene on hERG channel trafficking was performed using a LI-COR based In-Cell/On-Cell Western assay, as described previously  see online supplement for more details).
Phenanthrene was obtained from Merck (Sigma-Aldrich) and dissolved in DMSO to give stock solutions of 20, 30 and 50 mM. Experimental solutions with phenanthrene concentrations shown in the Results contained a maximum of 0.1% DMSO.
Docking of phenanthrene to hERG was initially performed using the cryo-EM derived structure for hERG  (PDB code 5VA2), as described previously [26;29;34]. A model closely related to the cryo-EM structure was obtained from a short molecular dynamics (MD) simulation in which the F656 side chain of one of the four hERG subunits was found to reorient towards the pore – this subunit was then replicated around all four pore subunits to produce a model with all four F656 side chains facing the pore; this structure was used for the docking data shown in Fig. 6. The phenanthrene structure was converted from SMILE representation (obtained from PubChem database) to a 3D structure, hydrogens were added and the molecule energy minimised.
Phenanthrene was docked in the hERG structure and in the related model using GOLD (GOLD version 5.6; Cambridge Crystallographic Data Centre, Cambridge, UK). The central pore cavity was initially chosen as a binding site where a radius of 10 angstrom extended from the centre of the cavity and in a level with a middle point between the canonical aromatic residues F656 and Y652. The side chains of these aromatic residues were set to be freely flexible during docking simulations. Rotamer sampling was maximally set to 300,000 generations. Dockings were scored by Goldscore and rescored by Chemscore. Two hundred docking repeats were made in each case and the low-energy-score poses were selected and inspected. Phenanthrene was also docked within a side pocket under the selectivity filter in the open pore F656-rotated hERG model. This binding pocket was centred above the β-carbon of Y652 and encompassed a volume having a radius of 7 angstrom. Within this selection, the side chains for the following residues from subunit A of the tetrameric channel were permitted to rotate freely: F557, L622, T623, S624, L650, M651, Y652, I655. F656 side chains from adjacent subunits (A and B) of the channel were also allowed to rotate freely. Similar setting and parameters were used as above where also 200 docking repeats for each drug were generated and low energy poses identified.
Results were presented using PyMOL Molecular Graphics System, Version 2.0 Schrödinger, LLC.
Data presentation and analysis
Data are presented as mean ± SEM of the number of independent experiments indicated (n). Statistical comparisons were made using a Student t test, one- or two-way analysis of variance (ANOVA) followed by a Bonferroni or Dunnett’s post-test, as appropriate. P values < 0.05 were considered to be statistically significant.
Fractional block of hERG current (IhERG) by the different phenanthrene (Phen) concentrations studied was determined using the equation:
Fractional block = 1−((IhERG −Phen)/IhERG −Control) (1)
Where “Fractional block” refers to the degree of inhibition of hERG current by a given concentration of phenanthrene. IhERG- Phen and IhERG−Control represent current amplitudes in the presence and absence of phenanthrene.
Concentration-response data were fitted by a standard Hill equation of the form:
Fractional block = 1/(1+(IC50/[Phen]) h) (2)
Where IC50 is [Phen] producing half-maximal inhibition of the IhERG tail and h is the Hill coefficient for the fit.
Half maximal activation voltages for IhERG were obtained from current-voltage (I-V) relations from IhERG tails measured at -40 mV in the absence or presence of phenanthrene following voltage commands to different test potentials, using the following Boltzmann equation:
I = Imax/1 + exp((V0.5-Vm)/k) (3)
where I = IhERG tail amplitude following test potential Vm, Imax is the maximal IhERG tail observed during the protocol, V0.5 is the half maximal activation voltage of IhERG and k is the slope factor describing IhERG activation.
Voltage-dependent activation curves were constructed by calculating activation variables at 2 mV intervals between − 80 mV and + 40 mV. Values for V0.5 and k were derived from experimental fits to I-V data using Eq. (3) were inserted into the following equation:
Activation parameter = 1/1 + exp((V0.5-Vm)/k) (4)
where the ‘activation parameter’ at test potential Vm lies between 0 and 1 and V0.5 and k have the meanings described above for Eq. 3.