Fig 2: CDK8 deficiency in MSCs suppresses osteoclastic activities, leading to high bone mass(A) Schematic diagram of generation of tissue-specific CDK8 knockout mice. Black arrows represent primer binding sites.(B) Deletion efficiency of CDK8 in the marrow-flushed bone of Prx1-Cre;CDK8fl/fl at the genomic DNA and mRNA levels (n = 4 independent replicates, ∗p < 0.05).(C and D) μCT (C) and von Kossa staining (D) and BV/TV measured by each method of femurs from CDK8fl/fl and Prx1-Cre;CDK8fl/fl mice (n = 6–7, ∗p < 0.05).(E) Toluidine blue staining (left) and N.Ob/T.Ar of femurs from CDK8fl/fl and Prx1-Cre;CDK8fl/fl mice (n = 6–7). Arrowheads indicate osteoblasts.(F) Calcein labeling (left) and BFR/BS of femurs from CDK8fl/fl and Prx1-Cre;CDK8fl/fl mice (n = 6–7).(G) TRAP staining (left) and Oc.S/BS of femurs from CDK8fl/fl and Prx1-Cre;CDK8fl/fl mice (n = 6–7, ∗∗p < 0.01). Arrowheads indicate TRAP-positive osteoclasts.(H) TRAP serum level in CDK8fl/fl and Prx1-Cre;CDK8fl/fl mice (n = 6–8, ∗p < 0.05).(I) Deletion efficiency of CDK8 in the marrow-flushed bone of Col1a1-Cre;CDK8fl/fl at the genomic DNA level and the mRNA level (n = 4, ∗p < 0.05).(J) μCT analysis and BV/TV measurement of femurs from CDK8fl/fl and Col1a1-Cre;CDK8fl/fl mice (n = 6–7).(K–O) BV/TV measured by von Kossa staining (K), N.Ob/T.Ar (L), BFR/BS (M), Oc.S/BS (N), and N.Oc/B. Pm of femur from CDK8fl/fl and Col1a1-Cre;CDK8fl/fl mice (n = 6–7) (O).All of the mice used in this study were female. Scale bars, 1 mm (C, D, J, and K), 50 μm (G), and 10 μm (E and F). μCT, micro-computed tomography; BV/TV, bone volume/tissue volume; N.Ob/T.Ar, number of osteoblasts/tissue area; BFR/BS, bone formation rate/bone surface; TRAP, tartrate-resistant acid phosphatase; Oc.S/BS, osteoclast surface/bone surface; N.Oc/B.Pm, number of osteoclasts/bone perimeter; N.S., not significant.

Fig 3: CDK8 deficiency in MSCs attenuates osteoclastogenesis through STAT1-RANKL axis(A) Protein levels of CDK8, p-STAT1Ser727, and STAT1 in MSCs infected with shCDK8. β-Actin served as a loading control.(B) BMMs prepared from WT mice were co-cultured with MSCs infected with shCDK8, followed by TRAP staining (n = 4 independent replicates, ∗∗∗p < 0.001).(C) mRNA levels of Tnfsf11 and Tnfrsf11b and Tnfsf11/Tnfrsf11b ratio in MSCs infected with shCDK8 (n = 4 independent replicates, ∗∗p < 0.01, ∗∗∗p < 0.001).(D) BMMs prepared from WT mice were co-cultured with MSCs infected with shCDK8 in combination with STAT1 expression vector, followed by TRAP staining (n = 5 independent replicates, ∗∗p < 0.01 [shCtrl + E.V. versus shCDK8 + E.V.], #p < 0.05 [shCDK8 + E.V. versus shCDK8 + STAT1]).(E and F) MSCs were infected with shCDK8 in combination with STAT1 expression vector, followed by determination of mRNA levels of Tnfsf11 (E) and Tnfrsf11b (F) (n = 4 independent replicates, ∗∗∗p < 0.001 [shCtrl + E.V. versus shCDK8 + E.V.], #p < 0.05, ###p < 0.001 [shCDK8 + E.V. versus shCDK8 + STAT1]).(G) mRNA levels of STAT1-targeted genes in MSCs were infected with shCDK8 (n = 4 independent replicates, ∗p < 0.05, ∗∗p < 0.01).(H) Tnfsf11 expression in CDK8low (n = 1,185) and CDK8high (n = 845) MSCs (∗p < 0.05).(I) mRNA levels of STAT1-targeted genes of CDK8low (n = 1,788) and CDK8high (n = 1,164) MSCs (∗p < 0.05, ∗∗∗p < 0.001).(J) MSCs were infected with shCDK8, followed by determination of CFU-Fs and CFU-Obs (n = 3 independent replicates). Scale bars, 800 μm (B and D) and 2 mm (J).BMM, bone marrow macrophage; CFU-F, colony-forming unit fibroblast; CFU-Ob, colony-forming unit osteoblast; shCtrl, shControl; E.V., empty vector.

Fig 4: Depletion of STAT1a or the STAT1 TAD induces growth arrest in JAK2-mutant AML cells.A. Diagram depicts location of various sgRNAs used to target STAT1. Dotted lines represent locations of the Cas9 cut site for each guide. Inset, part of the TAD (transactivation domain) DNA and protein sequence, with the CDK8 recognition motif highlighted in blue and serine 727 in red.B. CRISPR-Cas9-mediated knockout (KO) of STAT1 using sg1 and sg9n shows efficient depletion of STAT1 by immunoblot. STAT1-KO (mCherry) cells show resistance to CA compared to control AAVS1-KO (AmCyan) cells (mean ± s.e.m. ratio of n = 3 biological replicates). Representative growth curves shown for experiments repeated at least three times.C. The pie charts represent the percentage of frameshit versus in-frame mutations generated in pools of cells edited with guide sg5, sg4, sg11, or sg9n. SET-2 or UKE-1 cells were edited two independent times with each sgRNA and sequenced; one experiment is shown.D. Model for the role of STAT1 in proliferation. A proliferative state is maintained when STAT1a/ß are present or absent. Growth arrest is induced when the TAD of STAT1a is inhibited by CA or when STAT1a is knocked out.See also Figs. S3 and S4.

Fig 5: Constitutive Basal STAT1-S727 Phosphorylation in NK Cells Is Mediated by CDK8(A) Western blot of freshly isolated DX5+ splenocytes from wild-type animals shows constitutive basal phosphorylation on STAT1-S727 in vivo.(B and C) In vitro-expanded LAK cells from wild-type and Stat1-S727A mice were treated with or without (B) 100 U/ml IFN-β or (C) 5 ng/ml IL-12 for the given times, and the phosphorylation on STAT1-Y701 and STAT1-S727 was investigated by western blot. Stat1−/− LAK cells served as negative control and HSC-70 as loading control.(D) Wild-type LAK cells were treated with different kinase inhibitors targeting CDKs (roscovitine and flavopiridol) or MAPK p38 (BIRB 0796 and SB 203580), and the phosphorylation status of STAT1-S727 was determined in a western blot.(E) CDK8 knockdown in primary NK cells with two hairpins led to reduced STAT1-S727 phosphorylation. A random hairpin served as control.(F) Knockdown of CDK8 with five hairpins in an immortalized Ink4a−/− NK cell line is shown. A random hairpin served as control.(G) FACS-based cytotoxicity assay of immortalized NK cells transduced with a random hairpin or a hairpin against CDK8 (#3) against YAC-1 target cells is presented. Symbols represent mean, and error bars indicate SEM of triplicates (∗p < 0.05, ∗∗∗p < 0.001; unpaired t test).See also Figure S2.