Cytoplasmic Dynein Antagonists with Improved Potency and Isoform Selectivity

Cytoplasmic dyneins 1 and 2 are related members of the AAA+ superfamily (ATPases associated with diverse cellular activities) that function as the predominant minus-end-directed microtubule motors in eukaryotic cells. Dynein 1 controls mitotic spindle assembly, organelle movement, axonal transport, and other cytosolic, microtubule-guided processes, whereas dynein 2 mediates retrograde trafficking within motile and primary cilia. Small-molecule inhibitors are important tools for investigating motor protein-dependent mechanisms, and ciliobrevins were recently discovered as the first dynein-specific chemical antagonists. Here, we demonstrate that ciliobrevins directly target the heavy chains of both dynein isoforms and explore the structure–activity landscape of these inhibitors in vitro and in cells. In addition to identifying chemical motifs that are essential for dynein blockade, we have discovered analogs with increased potency and dynein 2 selectivity. These antagonists effectively disrupt Hedgehog signaling, intraflagellar transport, and ciliogenesis, making them useful probes of these and other cytoplasmic dynein 2-dependent cellular processes.

. C7-benzyl ether ciliobrevins reversibly disrupt IFT88 trafficking. IFT88 distributions along the ciliary axoneme in response to treatment with analog 43 for 1 hour and then compound washout for 4 hours. Representative micrographs of an NIH-3T3 cell-derived line are shown with staining for IFT88, ARL13B (primary cilia), γ-tubulin (basal body), and DNA. Scale bar: 2 µm. Each axoneme was divided into 21 bins, and the IFT88 signal intensity within each bin was normalized to the total ciliary signal using Matlab R2014A (Mathworks). Data represent the average IFT88 signal intensities for 65 cilia ± s.e.m.  Figure S4. Comparison of the nucleotide-binding sites in the cytoplasmic dynein 2 heavy chain. Stereoviews of the nucleotide-binding sites in DYNC2H1, visualized from their most solvent-accessible faces (PDB ID: ARH7). Space-filling (left) and surface (right) renderings are shown, and the linker region N-terminal to the AAA1 domain has been omitted for clarity. Adenine-interacting residues that are conserved between DYNC1H1 and DYNC2H1 are depicted in blue, variable residues in green, and nucleotides in red.

MATERIALS AND METHODS -SYNTHETIC PROCEDURES
Ciliobrevins A (1) and D (2) and analog 25 were synthesized as previously described. 1 which was recovered by vacuum filtration and washed sequentially with ethanol, water, ethanol, and diethyl ether. The solid was then dried to yield the substituted 2-(4-oxo-3,4dihydroquinazolin-2-yl) acetonitrile. Synthetic yields ranged from 30-50%.

2-(2,4-dichlorobenzoyl)-3,3-bis(methylthio)acrylonitrile.
To a stirred solution of 3-(2,4dichlorophenyl)-3-oxopropanenitrile (1.06 g, 4.95 mmol) in THF (15 mL) at 0 °C was added dry sodium hydride (0.24 g, 10 mmol). The suspension was stirred at 0 °C for 1 h, at which point carbon disulfide (0.403 g, 5.29 mmol) was added, and the reaction was then stirred at room temperature for another 2 h. The resulting red solution was cooled to 0 °C, iodomethane (1.56 g, 11.0 mmol) was added, and the mixture was stirred at room temperature for 18 h. The solvent was then removed in vacuo, and the residue was diluted in ether and washed with brine. The aqueous layer was extracted twice with ether, and the combined organic layers were washed twice with 5% sodium thiosulfate, and then brine. The organic layers were dried over MgSO 4 , filtered and concentrated to give the desired acrylonitrile as a yellow powder (1.08 g, 68%). This product was used in the next step without further purification or characterization.

1-(2,4-dichlorophenyl)-2-methyl-3,3-bis(methylthio)prop-2-en-1-one.
To a stirred solution of 1-(2,4-dichlorophenyl)propan-1-one (1.02 g, 5.02 mmol) in THF (15 ml) at 0 °C was added dry sodium hydride (0.24 g, 10 mmol). The suspension was stirred at 0 °C for 1 h, at which point carbon disulfide (0.403 g, 5.29 mmol) was added, and the reaction was then stirred at room temperature for another 2 h. The resulting red solution was cooled to 0 °C and iodomethane (1.56 g, 11.0 mmol) was added, and the mixture was stirred at room temperature for 18 h. The solvent was then removed in vacuo, and the residue was diluted in ether and washed with brine.
The aqueous layer was extracted twice with ether, and the combined organic layers were washed twice with 5% sodium thiosulfate, and then brine. The organic layers were dried over MgSO 4 , filtered and concentrated to give the desired propenone, which was used in the next step without further purification or characterization.

Generation of stable DYNC1H1-and DYNC2H1-expressing cells
Flp Bradford reagent, and the purified heavy chains were snap-frozen in assay buffer containing 1% (v/v) glycerol using liquid nitrogen and stored at -80 °C.

Expression and purification of 6xHis-DYNC1H1 and GFP-DYNC2H1 motor domains
The motor domains of DYNC1H1 (amino acids 1320-4647) and DYNC2H1 (amino acids 1091-4307) were expressed in insect cells and purified as described below. In the case of the DYNC1H1 motor, an N-terminal 6x-His tag was used for affinity-based purification. Construct preparation was performed as follows: A human DYNC1H1 cDNA clone (pF1KA0325) was obtained from Kazusa DNA Research Institute, and the coding region that encompasses the motor domain (residues Q1320 -V4647) was amplified and fused with a hexahistidine (His 6 )-tag by PCR using Phusion High-Fidelity DNA Polymerase (NEB). The assembled construct was The DYNC2H1 motor domain was prepared as described previously. 4 Briefly, the protein was expressed with N-terminal protein A tag and GFP tags separated by a tobacco etch virus (TEV) protease site and purified using IgG-Sepharose. The protein A tag was then proteolytically removed, and gel filtration of the product revealed a monodisperse peak at an elution volume of 12.6 mL, consistent with the expected molecular weight.

DYNC1H1 and DYNC2H1 ATPase assays
The containing trace γ-32 P ATP (6000 Ci/mmol, 10 mCi/mL, Perkin Elmer) and allowed to proceed at room temperature for a time predetermined to lie within the linear range of the assay. For the "low ATP" conditions, final concentrations of 1 nM dynein motor and 25 nM ATP were used, and the reaction was allowed to proceed for 10 minutes. For the "high ATP" conditions, final concentrations of 30 nM dynein and 100 μM ATP were used, and the reaction was allowed to proceed for 30 minutes. Reactions were quenched by the addition of 100 mM EDTA, and 2 μL of each reaction was spotted onto PEI-cellulose thin layer chromatography plates (Millipore).
The plates were developed in a glass chamber with a freshly prepared solution of 150 mM formic acid and 150 mM LiCl, dried, exposed to a storage phosphor tray, and scanned on a Typhoon imaging system (GE Healthcare Life Sciences). The fraction of γ-phosphate hydrolyzed in each condition was quantified using ImageJ and normalized to a DMSO control.

Hh signaling assays
Shh-LIGHT2 cells, 5 an NIH-3T3 cell line stably integrated with Gli-dependent firefly luciferase and thymidine kinase promoter-driven Renilla luciferase reporters, were cultured in DMEM containing 10% calf serum (CS), 100 U/mL penicillin, 0.1 mg/mL streptomycin, 1 mM sodium pyruvate, 400 µg/mL G418, and 150 µg/mL zeocin. The cells were seeded at 35,000 cells/well in a 96-well plate and treated with individual ciliobrevin analogs or an equivalent amount of DMSO vehicle (1%, v/v) in DMEM containing 0.5% CS and 10% ShhNconditioned medium. 6 After 30 h, the cells were lysed, and their firefly and Renilla luciferase activities were measured using a Dual Luciferase Reporter kit (Promega) and a Veritas microplate luminometer. Dose-response data were curve-fitted with a variable slope, sigmoidal dose-response algorithm using Prism software (GraphPad) to obtain IC 50 values.

Ciliogenesis assays
Shh-EGFP cells, 7  To quantify primary cilia lengths, the minimum threshold intensity for cilia staining was first established by manual inspection, and ImageJ software was used to quantify the total pixel area of ARL13B-positive pixels equal to or greater than the minimum threshold intensity. The Quantitative image analyses were conducted using MatLab R2014A (Mathworks).
ARL13B staining was used to mask and track the cilia and the γ-tubulin staining was used to orient the cilia from base to tip. The IFT88 signal along each axoneme was analyzed by dividing the total length of the cilium (as measured from the ARL13B staining) in 21 bins, each consisting of a 2-pixel-radius circle. Overlap correction was used to make sure that the summed fluorescence intensities over the 21 bins did not exceed the total ciliary fluorescence signal. The IFT88 signal within each bin was then normalized to the total ciliary signal to determine the fraction of ciliary IFT88 protein localized to each position along the axoneme. IFT88 signals from 70-120 cilia were analyzed from 5 fields of view to obtain traces for each compound, and the experiments were performed in duplicate.
To assess the effects of ciliobrevins on IFT88 movement in real time, murine inner medullary collecting duct (IMCD3) cells stably expressing mNeonGreen-IFT88 were imaged as previously described. 8 The cells were seeded on 25-mm coverslips and serum-starved for 24 h to induce ciliation. Imaging was conducted in phenol red-free media and on a DeltaVision system (Applied Precision) equipped with a PlanApo 603/1.49 NA internal reflection microscopy (TIRF) oil-immersion objective (Olympus). Images were captured with an sCMOS camera (Applied Precision) at 2 Hz, and line scan kymographs were generated by ImageJ. Velocities and frequencies of mNeonGreen-IFT88 foci movements were then quantified from the kymographs.

Mitotic spindle morphology assays
Shh-EGFP cells were seeded onto poly-D-lysine-coated, 12-mm glass coverslips in 24well plates at a density of 60,000 cells/well and cultured for overnight in the DMEM/10% CS Quantitative image analyses were conducted using MatLab R2014A (Mathworks). Golgi dispersal was analyzed by scoring specific morphologies as described in Figure S2. ARL13B staining was used to mask and track individual cilia, and GLI2 staining was used to designate the distal tip of each cilium. Each axenome was divided in 21 bins as described above, and the absolute fluorescent GLI2 signal at the ciliary tip (bin 19; maximum GLI2 signal) was averaged over 100-300 cilia analyzed from 4-6 fields of view. The GLI2 signal for each compound was then normalized to that observed in cells treated with the Hh pathway activator SAG. Three independent experiments were analyzed.