A LANA peptide inhibits tumor growth by inducing CHD4 protein cleavage and triggers cell death

Abstract

Kaposi's sarcoma-associated herpesvirus (KSHV) establishes a latent infection, and viral genes are poised to be transcribed in the latent chromatin. In the poised chromatins, KSHV latency-associated nuclear antigen (LANA) interacts with cellular chromodomain-helicase-DNA-binding protein 4 (CHD4) and inhibits viral promoter activation. CHD4 is known to regulate cell differentiation by preventing enhancers from activating promoters. Here, we identified a putative CHD4 inhibitor peptide (VGN73) from the LANA sequence corresponding to the LANA-CHD4 interaction surface. The VGN73 interacts with CHD4 at its PHD domain with a dissociation constant (K_D) of 14 nM. Pre-treatment with VGN73 enhanced monocyte differentiation into macrophages and globally altered the repertoire of activated genes in U937 cells. Furthermore, the introduction of the peptide into the cancer cells induced caspase-mediated CHD4 cleavage, triggered cell death, and inhibited tumor growth in a xenograft mouse model. The VGN73 may facilitate cell differentiation therapy.

Keywords: CHD4; KSHV; LANA; apoptosis; autophagy; cancer; cell differentiation; leukemia; monocyte; peptide.

Figure 1.. Identification of CHD4 interacting peptide from KSHV LANA protein.

(A) LANA protein domain is depicted and non-structured regions (intrinsically unstructured regions) were predicted by the web interface (https://iupred2a.elte.hu/) and plotted below the diagram. The previously-identified CHD4 binding region (870–1042 amino acid) was enlarged. (B) Positions of the amino acid sequence are shown. (C) One hundred nanomolar of purified Flag tagged CHD4 protein was incubated with 1 μg of biotinylated peptides. Peptide bound to Flag-CHD4 was visualized by immunoblotting with anti-Flag antibody. Number 1–10 on top corresponds to the listed peptides. (D) Conserved amino acids were revealed by alignment with other gamma herpesvirus homologs. Conserved amino acids were depicted in bold case. (E) Based on consensus motifs in LANA protein sequence, VGN73 was designed. The phenylalanine was substituted with histidine, which is conserved among other gamma-herpesvirus homologues (shown in red). The cell penetrate peptide (TAT-sequence) was included at N-terminus. The mutant peptide (negative control) was made by changing three conserved amino acids (underlined) among other gamma-herpesvirus homologues (shown as bold in Figure 1D). d-R: D amino acid arginine, (ORN) ornithine, ( beta alanine were included to increase peptide stability. (F) Previously revealed crystal structure of LANA C-terminal domain was used as a reference to visualize the position of the CHD4 binding domain. Each monomer of LANA C-terminal domain was visualized in gray (left-side of dimer) and pink (right-side), respectively. The protein backbone and side-chains of the #9–10 peptide sequence (WKFAVIFWGNDPYGLKK) on the LANA monomer (gray; left-side) was highlighted with green (WKFAV----PYGLKK) and blue (IFWGND). The surface models with or without transparency were colored with the analogous to cartoon model. See also Figure S1.

Figure 2.. VGN 73 binding with CHD4 and LANA.

(A) Prepared CHD4 deletion proteins and protein domains are shown in schematic diagram. Increasing concentrations of deletion CHD4 proteins or full-length CHD4 (shown as 1–1912) was incubated in an ELISA plate coated with 1 μM of VGN73 in triplicate. Peptide binding measured as OD values at 450 nm are shown. (B) BLI sensorgram showing the binding of full-length Flag-CHD4 protein to N-biotinylated VGN73, mutant, or control TAT immobilized on a biosensor tip. Steady-state analysis of the interaction between VGN73 or mutant peptide and CHD4 protein in solution at various concentrations is shown. (C) Cy5-conjugated VGN73 and mutant peptide were used to track subcellular localization (15 min or 60 min) in BC3 cells. Nuclei were visualized by Hoechst 33342 staining (scale bar, 20 μm). (D) BC3 cells were incubated with biotinylated VGN73, biotinylated mutant peptide or biotinylated TAT peptide (10 μM of each) for 2 hours. 2.5% of the input reaction before peptide-pull down were used as control and beads alone was used to reveal non-specific binding of the beads. CHD4 protein bands were confirmed by immunoblotting with specific antibodies, and densitometry values were expressed as fold change compared to control values (normalized to 100%). See also Figure S2, Table S1 and Table S2.

Figure 3.. Effect of VGN73 on CHD4 and LANA in tissue culture.

(A) BC3 cells were treated with VGN73 (18 μM), mutant peptide (18 μM) or vehicle control for 24 hours followed by staining with the indicated antibodies (scale bar, 20 μm). (B) Cell volume, CHD4 and LANA signal intensity were compared by Wilcoxon signed-rank test (The number of cells measured; ctrl=89, mut=53, and VGN73=55). Data are represented as mean ± SD. **p < 0.01, ***p < 0.001 versus the VGN73 group. (C) BC3 and BCBL-1 cells were treated with VGN73 at their respective IC50 concentrations either in complete medium or low serum condition for 24 hours. CHD4 and LANA levels of total cell lysates were analyzed by immunoblotting with anti-CHD4 and anti-LANA antibody, respectively. Densitometry values of full-length CHD4 protein were measured with ImageJ and expressed as fold change compared with vehicle control values normalized to 100%. (D) BC3 cells were pretreated with the pan-caspase inhibitor, Z-VAD-FMK (50 μM) or the proteasome inhibitor, bortezomib (25 nM) for 1 hour, followed by treatment with VGN73 (18 μM) for 8 hours. CHD4 protein levels and cleavage were analyzed by immunoblotting with anti-CHD4 antibody. Densitometry values of full-length CHD4 protein were measured with ImageJ and expressed as fold change compared with vehicle control values normalized to 100%. See also Figure S3.

Figure 4.. VGN73 inhibits cell growth of various cancer cell types and induces apoptosis and autophagy in BC3 cells.

(A) MTT assays were performed with the indicated cell lines in triplicate treated with various VGN73 and mutant peptide concentrations for 24 hours. The OD₅₇₀ of mock-treated samples were set as 100%, and OD₅₇₀ from cells treated with 10% SDS was set as 0%. Mean percentage viability ± SD was calculated for each treatment (n = 3 samples/treatment). (B) BC3 cells were incubated with VGN73 (18 μM) for 0, 1, 4 and 24 hours, followed by measurements of percentage apoptosis induction (Annexin V staining) by flow cytometry. Median percentage apoptosis was calculated for each treatment time (n = 3 samples/treatment time) (i). Western blot analysis was performed to detect cleaved caspase 3. BC3 cells were treated with VGN73, mutant peptide (18 μM) or vehicle control for 24 hours. Densitometry values of cleaved caspase 3 protein were measured with ImageJ and expressed as fold change compared with vehicle control values normalized to 100% (ii). (C) BC3 cells were incubated with 18 μM VGN73. 50 nM rapamycin and starvation for 24 hours were used as positive controls. Flow cytometry analyses (Autophagy red staining) were performed. Median percentage apoptosis was calculated for each treatment time (n = 3 samples/treatment) (i). Western blot analysis was performed to detect LC3II and LC3I. BC3 cells were pretreated with lysosome inhibitor, chloroquine (7.5 μM) for 1 hour. Densitometry values of LC3I and II protein were measured with ImageJ and LC3II/LC3I ratios expressed as fold change compared with untreated sample values normalized to 100% (ii). (D) BC3 and BCBL-1 were pretreated with Z-VAD-FMK (40 μM) for 1 hour and then incubated with VGN73 (18 or 13 μM, respectively) for 8 hours. Cell viability was determined by MTT assay (n = 5 samples/treatment). b-d *p < 0.05, **p < 0.01, using by Wilcoxon signed-rank test. See also Figure S4.

Figure 5.. Direct target identification with SLAM-seq.

Vehicle control (PBS), VGN73 or mutant peptide at each concentration of IC50 were added to BC3, BCBL-1 or Raji cells 30 mins prior to incubation with 4sU, and RNA was labeled for 1.5 hours in the presence of each peptide. Each treatment was performed in duplicate. (A) Venn diagram shows commonly up-regulated gene sets by VGN73 incubation comparing with vehicle among BC3, BCBL-1, and Raji cells. (B) Peak scores of promoter region were compared between upregulated gene and non-upregulated gene in SLAM-seq data and depicted with violin plots. (C) To confirm the occupancies of CHD4 on promoter regions of upregulated gene, CUT&RUN-qPCR was performed with CHD4 antibody in BCBL-1 cells. Fold enrichment over IgG control was shown. Enrichment was also normalized with 10% of the input sample extracted before CUT& RUN. Data are represented as mean ± SD. *; p < 0.05, Promoter region of non-treated samples was compared with negative control genomic locus (NC) by Wilcoxon signed-rank test, and each promoter region of VGN73 treated samples was individually compared with its of non-treated samples for statistical testing. (D) The relative RNA levels were measured in CHD4 knockdown cells (shCHD4) compared to the negative control knockdown cells (shLuc). β-actin was used as an internal standard to normalize gene expression. The student unpaired t-test was used for significance calculation. Data are represented as mean ± SD. *p < 0.05; **p < 0.01. See also Figure S5.

Figure 6.. VGN73 enhances U937 cells differentiation.

(A) U937 cells were pretreated with VGN73 (4 μM) for 24 hours, followed by stimulated with PMA (10 ng/mL) in starved condition for 48 hours. And then, cells were stained with the indicated antibodies (scale bar, 20 μm). (B) U937 cells were pretreated with vehicle, mutant peptide or VGN73 (4 μM) for 24 hours, followed by stimulated with PMA (20 ng/mL) in starved condition for 24 hours. Box plot showed relative levels of gene expression involved in macrophage differentiation and inflammatory between each treatment for determined by RT-qPCR. β-actin was used as an internal standard to normalize viral gene expression. Data are represented as mean ± SD. *; p < 0.05. (C) U937 cells stably knockdown for CHD4 (shCHD4) and negative control knockdown cells (shLuc) were left unstimulated (−) or stimulated with PMA (+) (20 ng/mL) for 24 hours. Relative levels of gene expression involved in macrophage differentiation were determined by RT-qPCR. 18S was used as an internal standard to normalize viral gene expression. Data are represented as mean ± SD. (D) U937 cells were pretreated with vehicle, mutant peptide or VGN73 (4 μM) for 24 hours, followed by stimulated with PMA (10 ng/mL) in starved condition for 48 hours. Cells treated with vehicle alone was used as control. Flow cytometric analysis for the expression of the indicated markers was performed and depicted in the density plots. See also Figure S6.

Figure 7.. VGN73 inhibits BCBL-1 cell growth in xenograft mice.

BCBL-1 cells (5 × 10⁶ cells) were injected intraperitoneally (i.p.) into 23–30 week-old female NRG mice. After 2 days, 18 mice were randomly assigned to 3 treatment groups (6 mice per group): Vehicle control (PBS), mutant peptide (10 mg/kg), and VGN73 (10 mg/kg). Treatment drugs were administrated i.p. every other day. (A) The tumors at day 14 and 19 were measured by in vivo bioluminescence imaging (n = 6 mice per group), and intensities (photons/s) were compared between each group. (B) Mice were weighed every other day for 18 days and body weight increments were compared among each group on days 14 and 18 (n = 6 mice per group). (C) PEL cell counts in ascites fluid were measured on day 19 by flow cytometry. (n = 6 mice per group) (D) Cells were stained with the indicated antibodies (scale bar, 20 μm). (E) CD38 and CD19 gene expression was quantified by RT-qPCR (n = 5 mice per group). 18S ribosomal RNA was used as an internal standard to normalize viral gene expression. (B-C and E) Data are represented as mean ± SD.*p < 0.05, **p < 0.01, using by Wilcoxon signed-rank test. See also Figure S7.