LSD1 promotes prostate cancer reprogramming by repressing TP53 signaling independently of its demethylase function

Lysine-specific demethylase 1 (LSD1) is a histone demethylase that promotes stemness and cell survival in cancers such as prostate cancer. Most prostate malignancies are adenocarcinomas with luminal differentiation. However, some tumors undergo cellular reprogramming to a more lethal subset termed neuroendocrine prostate cancer (NEPC) with neuronal differentiation. The frequency of NEPC is increasing since the widespread use of potent androgen receptor signaling inhibitors. Currently, there are no effective treatments for NEPC. We previously determined that LSD1 promotes survival of prostate adenocarcinoma tumors. However, the role of LSD1 in NEPC is unknown. Here, we determined that LSD1 is highly upregulated in NEPC versus adenocarcinoma patient tumors. LSD1 suppression with RNAi or allosteric LSD1 inhibitors — but not catalytic inhibitors — reduced NEPC cell survival. RNA-Seq analysis revealed that LSD1 represses pathways linked to luminal differentiation, and TP53 was the top reactivated pathway. We confirmed that LSD1 suppressed the TP53 pathway by reducing TP53 occupancy at target genes while LSD1’s catalytic function was dispensable for this effect. Mechanistically, LSD1 inhibition disrupted LSD1-HDAC interactions, increasing histone acetylation at TP53 targets. Finally, LSD1 inhibition suppressed NEPC tumor growth in vivo. These findings suggest that blocking LSD1’s noncatalytic function may be a promising treatment strategy for NEPC.

) (A) MR42D cells were treated with DMSO vehicle or 600 nM SP2509 for 24 hours. ChIP was performed with anti-TP53 antibodies. qPCR was performed to amplify promoter regions of TP53 targets (CDKN1A, MDM2) or a negative control region (UNTR4), n=3. (B) MR42D cells were treated with DMSO vehicle or 600 nM SP2509 for 24 hours. Expression of TP53 target genes was analyzed by RT-qPCR. All data are reported as the mean ± SD. For statistical analysis, unpaired two-tailed Student's t-tests were performed, and P values are indicated. Figure 7) (A) Tumor volume of LASCPC-01 inducible shLSD1 xenografts in mice treated with control or dox diets was measured at harvest and is presented as a bar graph. The data are reported as the mean ± SEM. For statistical analysis, an unpaired Welch's t-test was performed, and the P value is shown. (B) Western blot analysis of LASCPC-01 inducible shLSD1 xenografts harvested at endpoint using antibodies specific for LSD1 or p21 (CDKN1A). Beta-actin served as a loading control (left). The densitometry analysis of LSD1 and p21 bands are presented as bar plots (right). The data are reported as the mean ± SEM. For statistical analysis, unpaired twotailed Welch's t-tests were performed, and P values are indicated. (C) Body weights of mice during the study was measured as a function of time and plotted. The data are reported as the mean ± SEM. For statistical analysis, a mixed-effects model two-way ANOVA was performed, and the P value is indicated. Gene Symbol Forward Primer Sequence 5'-3' Table 3: Primers used for ChIP-qPCR assays.

Viability and apoptosis assays
All cell viability measurements were determined using the CellTiter-Glo 2.0 (CTG) assay

Dose-response experiments
For dose-response experiments, indicated cells were treated in biological triplicate for 72 h with a 7-point, 5-fold dilution series from 10 M of the indicated drugs in DMSO. Cell viability was assessed using the CTG assay. Dose-response was normalized to the vehicle-treated growth rate and fitted with a logistic curve as previously described (1).

LSD1 inhibition assay
The LSD1 inhibition assay was performed using the LSD1 Inhibitor Screening Assay Kit (Cayman Chemical, cat# 700120) using the manufacturer's protocol. LSD1 activity was measured in the presence of DMSO vehicle or the LSD1 inhibitors GSK2870552, GSK-LSD1, SP2509, and SP2577.

Cell cycle analysis
For cell cycle analysis, cells were harvested in hypotonic propidium iodide buffer (0.1% Sodium citrate, 0.1% Triton-X100, 100 µg/mL RNase A, and 50 µg/mL propidium iodide) and incubated at room temperature for 5 minutes. The cells were analyzed using a BD LSRFortessa Cell Analyzer (BD Biosciences), and data were analyzed using FlowJo version 10.8.1 (BD Biosciences) to obtain the percentage of cells in each phase of the cell cycle. The data was used to plot graphs using GraphPad Prism version 9.4.1.

Plasmid Transfection:
For overexpression experiments, plasmids were transfected using Lipofectamine 3000 reagent (Thermo Fisher cat# L3000008) per manufacturer's recommendations. Cells were harvested 48 hours post-transfection and processed for downstream analyses. LSD1 expression constructs have been described previously (2). TP53 constructs were a kind gift from Basant Kumar Thakur, University Hospital Essen, Essen, Germany (3).

Transient knockdown experiments
Transient knockdowns were performed using siRNA oligonucleotides or shRNA constructs described previously (2). The siRNA oligonucleotides were transfected with DharmaFECT 3 (GE Dharmacon) transfection reagent for 96 h. Cells used for RNA and protein harvest were seeded and transfected in 6-well plates. Cells used in viability assays were seeded and transfected in 96well plates. Cell viability was measured at time 0 and endpoint using the CTG assay; these values were used to calculate the relative growth rate as described previously (4).

Western blotting
Western blotting experiments were performed by running protein lysates on SDS-PAGE gels (Thermo Fisher Scientific cat# NP0335BOX) and transferring them onto PVDF membranes as described previously (4). Blots were probed with indicated antibodies and imaged using a Chemidoc MP imaging system (Bio-Rad). Densitometry analysis was performed using NIH Image J (5).

Co-immunoprecipitation
For co-immunoprecipitation experiments, cell lysates were prepared in RIPA buffer (Thermo Fisher Scientific cat # 89900) with protease inhibitors (Thermo Fisher Scientific cat# A32965).
The cell lysates were centrifuged at 16,000 × g for 20 min at 4°C. The supernatants were collected and precleared with protein G-conjugated Dynabeads (Thermo Fisher Scientific cat# 10004D) for

In vivo anti-tumor activity of LSD1 suppression
All the studies were performed in 6-8 week-old male athymic homozygous nude-Foxn1nu mice (Jackson Laboratories cat# 002019). All cells for implantation were prepared in 1:

RNA-seq
Total RNA was extracted with RNeasy Plus Mini Kit as above. Library preparation and sequencing methods were performed as described previously (4)

Master regulator and pathway analysis
RNA-seq data from cell lines treated with vehicle or SP2509 were used to perform pathway analysis and to evaluate differential transcription factor activities. Differential gene expression analysis between experimental groups was first performed using DESeq2 (13). Gene expression differences were considered significant with Padj < 0.05. The Wald test statistic results from regulons (the regulatory network) used in this study were curated from four databases as previously described (22). ARG10 (23) and NEPC Up (24) signatures were derived from previously reported datasets.

TP53 signature analysis
The TP53 signature was described previously by Chipidza et al. (25). The TP53-WT Centroid genes of the TP53 signature were used to build a pseudo-regulon of TP53. Forty-one upregulated and 145 down-regulated genes served as positive and negative target genes of the pseudo-regulon of TP53, respectively. To measure the TP53 regulon activity in each sample, we used the VIPER algorithm (21) implemented in the VIPER R package (version 1.26.0). A log1P transformed TPM gene expression matrix and a regulatory network were used as inputs for VIPER analysis. The viper function was employed to calculate the regulon activities of the TP53 signature and other transcription factors on different datasets. The regulatory network used in VIPER analysis was the same as described above.