Evaluation of phosphatidylinositol-4-kinase IIIα as a hepatitis C virus drug target

Abstract

Phosphatidylinositol-4-kinase IIIα (PI4KIIIα) is an essential host cell factor for hepatitis C virus (HCV) replication. An N-terminally truncated 130-kDa form was used to reconstitute an in vitro biochemical lipid kinase assay that was optimized for small-molecule compound screening and identified potent and specific inhibitors. Cell culture studies with PI4KIIIα inhibitors demonstrated that the kinase activity was essential for HCV RNA replication. Two PI4KIIIα inhibitors were used to select cell lines harboring HCV replicon mutants with a 20-fold loss in sensitivity to the compounds. Reverse genetic mapping isolated an NS4B-NS5A segment that rescued HCV RNA replication in PIK4IIIα-deficient cells. HCV RNA replication occurs on specialized membranous webs, and this study with PIK4IIIα inhibitor-resistant mutants provides a genetic link between NS4B/NS5A functions and PI4-phosphate lipid metabolism. A comprehensive assessment of PI4KIIIα as a drug target included its evaluation for pharmacologic intervention in vivo through conditional transgenic murine lines that mimic target-specific inhibition in adult mice. Homozygotes that induce a knockout of the kinase domain or knock in a single amino acid substitution, kinase-defective PI4KIIIα, displayed a lethal phenotype with a fairly widespread mucosal epithelial degeneration of the gastrointestinal tract. This essential host physiologic role raises doubt about the pursuit of PI4KIIIα inhibitors for treatment of chronic HCV infection.

Fig 1

Schemes for the synthesis of compound A and compound B as described in U.S. patent 2007/0238746 A1 (16) and U.S. patent 2007/0238730 A1 (17).

Fig 2

(A, B) Schematics of activity assays developed to identify inhibitors of PI4KIIIα: fluorescent polarization (FP) assay to detect the formation of PI4P by PI4KIIIα (A) and Kinase-Glo assay to detect the consumption of ATP by PI4KIIIα (B). (C, D) Correlation of PI4KIIIα in vitro potency (IC₅₀) with HCV replication inhibition (replicon EC₅₀) obtained with chemotype 1 (slope of the observed correlation, 0.43; R² = 0.49; P < 0.05; 57 compounds tested) (C) and with chemotypes 2 (squares) and 3 (diamonds) (slope of the observed correlation, 0.99; R² = 0.75; P < 0.05; 51 compounds tested) (D).

Fig 3

Resistance study using PI4KIIIα inhibitors. (A) Genetic mapping of the region involved in resistance. Fifteen amino acid changes (red arrows) were identified in the HCV replicon sequence isolated from the clonal line resistant to compound A. These were distributed throughout the nonstructural region with 3 changes in NS2 (L33P, I86T, and F103S), 3 in NS3 (G262S, M470I, and N556S), 1 NS4A (T2A), 1 NS4B (S258P), 6 in NS5A (R70S [G70 in Con-1], S107T, G267E, D358N, L419P, and G421R), and 1 in NS5B (I585V). The FseI-PacI replicon-resistant restriction fragment I (RRRF-I) and the SrfI/MluI replicon-resistant restriction fragment II (RRRF-II) were obtained from the RT-PCR DNA product of the clonal line resistant to compound A. (B) Luciferase levels obtained after transient transfection of the chimeric replicon RNA in HuH-7.5 cells (left panel) and in the stable PI4KA knockdown cell line (HuH-7.5-shPIK4A; right panel). Luciferase levels were determined at 4, 72, and 96 h posttransfection. Values are corrected for transfection efficiency by measuring the signal at 4 h and expressing all luciferase levels as a percentage of this value. Luciferase levels of the baseline replicon (R3-derived replicon) with the FseI/PacI region of the S22.3 parent replicon are shown in blue (closed diamonds); with the RRRF-I region of the compound A-resistant clone, in red (open diamonds); and with the RRRF-II region of the Cpd A-resistant clone, in green (closed triangles). As a negative control, a replication-incompetent replicon (NS5A deletion of residues 55 to 60) was used (purple, open squares).

Fig 4

Stability of PI4KA RNA transcript suppression observed in a stable HuH-7.5 shRNA PI4KA knockdown clone. The levels of PI4KA in the control HuH-7.5 cell line are represented by open diamonds, and the levels of PI4KA in the HuH-7.5 PI4KA stable knockdown clone that expresses an shRNA targeting the PI4KA gene are represented by open squares.

Fig 5

Generation of Pi4ka conditional KO and KI mice. (A, B) Targeting strategies used to generate Pi4ka conditional KO (A) and KI (B) mice. (C) Southern blot analysis of KO homologous recombinant (HR) ES cell clones using a BamHI digest and probe A. W, wild-type signal; Targ, targeted signal. (D) PCR analysis of heterozygous conditional (cond) KO and wild-type (W) mice. (E) Southern blot analysis of KI homologous recombinant ES cell clones using a BsrGI (B) and PvuI (P) digest and probe B. Clone 1 is a homologous recombinant ES cell clone derived from the transfection with the first targeting vector. Clone 2 is derived from clone 1 and has undergone the recombination-mediated cassette exchange with the second targeting vector. W, wild-type signal; Targ 1 and Targ 2, targeted signals. (F) PCR analysis of heterozygous conditional (cond) KI mice. W, wild-type sample; C, internal control.

Fig 6

Gross phenotype and histopathology analysis of induced homozygous conditional Pi4ka KO mice. (A) Distended GI tract gross phenotype observed at necropsy. (B) Histopathology analysis of stomach lesions: ulceration of nonglandular mucosa (left panel) and parietal cell degeneration of glandular mucosa (right panel). (C) Histopathology analysis of small intestine lesions: mucosal epithelial degeneration of duodenal mucosa. (D) Histopathology analysis of large intestine lesions: mucosal epithelial degeneration of cecal mucosa.

Fig 7

Histopathology analysis of induced homozygous conditional Pi4ka KI mice: stomach lesions showing degeneration of glandular mucosa (A), small intestine lesions showing mucosal epithelial degeneration of duodenal mucosa (B), large intestine lesions showing mucosal epithelial degeneration of cecal mucosa (C), and small and large intestine lesions showing focal atypical hyperplasia of crypt (D).