Fig 2: Effects of pevonedistat treatment on cell cycle progression, p21 and CDT1 accumulation. (a) Cell cycle analysis by double staining with (propidium iodide) PI and pH3 showing an accumulation of cells in the S or G2 phase of the cell cycle predominantly in sensitive cells (Mero-82, MSTO211H and ONE58) treated with 0.3-µM pevonedistat for 24 h compared to control (treated only with dimethylsulfoxide (DMSO)). We noted a drastically increased accumulation of cells containing >4N DNA after the treatment, predominantly in the sensitive cell lines, without increased accumulation of cells undergoing mitosis (pH3+, circles). (b) Cells were treated with an increasing concentration of pevonedistat for 24 h and subjected to Western blot analysis for the indicated proteins. Number above each line of the blots represents protein expressions (ratio densitometry of the bands normalized with β-actin) relative to no pevonedistat treatment. Arrows indicate neddylated CUL4A or CUL4B. (c) A panel of MPM cell lines treated with DMSO or 0.6-µM pevonedistat for 24 h shows an accumulation of CDT1 and p21 in both sensitive (underlined) and resistant cell lines. Number above each line of the blots represents protein expressions (ratio densitometry of the bands normalized with β-actin) relative to no pevonedistat treatment.

Fig 3: RIF1 does not control Cdt1 and Cdc6 RIF1 does not control the cell cycle‐specific localization of Cdc6 protein. Subcellular localization of Cdc6 protein was previously shown to be cell cycle regulated through its phosphorylation by CDK, which leads to export of Cdc6 protein from the nucleus to the cytoplasm 47. To test whether Rif1 removal affects Cdc6 regulation, subcellular localization of Cdc6 was analyzed by immunofluorescence microscopy in U2OS cells treated with either siCtrl or siRIF1. Abundance of cytoplasmic Cdc6 and DNA content was measured for more than 1,700 cells under each condition using CellProfiler software as follows. Adaptive thresholding was applied to the blue (DAPI‐stained nuclear DNA) and green (anti‐Cdc6 stain) channels to identify areas corresponding to individual nuclei and cells, respectively. The cytoplasmic area was then defined by excluding the nuclear area from the cell area. For each cell, the integrated intensity in the green (Cdc6) channel was then measured separately for the cytoplasmic area, the nuclear area, and the whole cell. Integrated intensity in the blue channel was measured for each nuclear area to obtain a cellular DNA content value. Cells at the edge of images were excluded. Scatter plots show cytoplasmic Cdc6 abundance (y‐axis; log10 scale) against DNA content (x‐axis), while the histograms (lower panels) show DNA content of the corresponding cells. Representative images are presented at the top. The scale bar indicates 30 μm.RIF1 has only a minor effect on Cdt1 stability. Degradation of Cdt1 protein is regulated in a similar way to ORC1: outside G1 phase, Cdt1 is phosphorylated and degraded 48. The effect of depletion followed by ectopic expression of RIF1 on Cdt1 abundance was tested by flow cytometry. Flp‐In T‐REx 293 cells with stable GFP, GFP‐RIF1, or GFP‐RIF1‐pp1bs constructs were depleted for endogenous RIF1 by siRNA, and 1 day later GFP, GFP‐RIF1, or GFP‐RIF1‐pp1bs was induced. After further 3 days, cells were prepared for flow cytometry and abundance of Cdt1 protein per cell was analyzed by flow cytometry.

Fig 4: Efficacy and mechanisms of pevonedistat in an in vivo model. (a) Scheme depicting the schedule for tumor implantation and pevonedistat treatment cycles for the mouse model. (b) Kaplan-Meier plots showing significant differences in the overall survival (days) between the vehicle and pevonedistat-treated mice bearing an ACC-Meso-1 tumor (left) and MSTO211H tumor (right). (c) Comparison of tumor growth (pH3+), apoptosis (cleaved caspase 3, Cl.C3), CDT1 intensity, p21 H-score, mouse macrophage recruitment (F4/80+ cells per tumor area) and mouse vessel formation (CD31+ group of cells with visible lumen per tumor area) between vehicle-treated (filled circles) and pevonedistat-treated mice (filled squares). # not significant; * p < 0.05, ** p < 0.001, (d) Representative images of immunofluorescent staining for the indicated proteins with a green signal. They were co-stained with DAPI (blue) and pan-cytokeratin (yellow or red). The DAPI signal was omitted from CDT1 staining for a better visualization of the CDT1 signal. (e) Immunohistochemical staining of the indicated markers and counterstained with hematoxylin (blue). Scale bars represent 100 µm.

Fig 5: Effect of pevonedistat treatment on the expression of CCL2. (a) Comparison of the baseline expression levels of CCL2 in four different MPM cell lines in vitro. Lower Ct is equivalent to a higher expression. (b) qPCR analysis using mouse and human-specific CCL2 primers (using human-specific β-actin and mouse-specific β-2-microglobulin as a reference gene) showing a significant reduction of human CCL2 in the ACC-Meso-1 tumor treated with pevonedistat. (c) Change in the expression of human CCL2 in vitro after pevonedistat treatment in MPM cell lines. (d) Scheme illustrating the rationale and findings of our current study. CUL4A and CUL4B are shown to be regulated by Merlin dysfunction; here, we sought to investigate the importance of CUL4A and CUL4B expressions in MPM progression and tested the efficacy of pevonedistat, a small molecule inhibiting cullins via protein neddylation. We observed that a high CUL4B expression was associated with the worst clinical outcome. The treatment of MPM cells with pevonedistat resulted in an accumulation of CDT1 and p21, known targets of CUL4 and CUL1, leading to cell cycle arrest. Reduced nuclear factor- κB (NF-κB) pathway resulting from CUL1 inhibition caused a reduced CCL2 expression in tumor cells and reduced macrophage infiltration in tumors. Via an unknown mechanism, a reduced macrophage infiltration may result in more tumor cell death. It is important to note that CUL4A and CUL4B are shown to regulate several ubiquitination targets; thus, these targets may be important for MPM progression, and this remains to be investigated. In addition to cullins, pevonedistat also can affect some other proteins that are also regulated by neddylation. # not significant; * p < 0.05, Ct, threshold cycle and ND, not done.