Diminished PDE4D7 Expression is Associated with Treatment-Resistant, Lethal Prostate Cancer and Manifests with Impaired Androgen Response, Neuroendocrine Differentiation and Alterations in DNA Repair

doi:10.21203/rs.3.rs-2611935/v1

Download PDF

Article

Diminished PDE4D7 Expression is Associated with Treatment-Resistant, Lethal Prostate Cancer and Manifests with Impaired Androgen Response, Neuroendocrine Differentiation and Alterations in DNA Repair

https://doi.org/10.21203/rs.3.rs-2611935/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Androgen signalling remains the seminal therapeutic approach for management of advanced prostate cancer. However, most tumours eventually shift towards an aggressive phenotype, characterised by androgen-independence and treatment resistance. The cyclic adenosine monophosphate (cAMP) pathway plays a crucial role in regulating various cellular processes, with the phosphodiesterase PDE4D7 being a vital modulator of cAMP signalling in prostate cancer cells. Our study provides evidence that loss of PDE4D7 expression represents a pivotal switch driving the transition from an androgen-sensitive state to hormone unresponsiveness and neuroendocrine differentiation. Additionally, we demonstrate that PDE4D7 loss results affects DNA repair pathways, conferring resistance to poly ADP ribose polymerase (PARP) inhibitors. Reinstating PDE4D7 expression sensitises prostate cancer cells to anti-androgens, DNA damage response inhibitors, and cytotoxic therapies. These findings provide significant insight into the regulatory role of PDE4D7 in the development of lethal prostate cancer and the potential of its modulation as a novel therapeutic strategy.

Biological sciences/Cancer/Urological cancer/Prostate cancer

Health sciences/Oncology/Cancer/Urological cancer/Prostate cancer

Biological sciences/Cancer/Tumour biomarkers

In 2020, approx. 1.4 million men were newly diagnosed with prostate cancer (PCa) worldwide leading to 375,000 cancer induced deaths [1]. In general, the majority of PCa patients are at low to intermediate risk of developing progressive disease and to die from their tumour [2]. However, 17–31% of patients are diagnosed with clinically high-risk disease with an elevated probability of death from PCa [3–4].

In 2012, the U.S. Preventive Services Task Force (USPSTF) recommended against the continuously wide-spread PSA screening in the U.S. to avoid over-detection of low-risk disease [5]. While in the following years the number of newly diagnosed patients declined across all age groups by an estimated annual percent change (APC) of close to 20% [6] the incidence of patients with locally advanced or metastatic disease at presentation increased [7, 8]. This in turn may promote rising PCa mortality [9].

Despite the shift to higher grade and advanced stage disease at diagnosis, the identified tumours are still heterogeneous concerning their biology and progressive characteristics. The most frequent molecular alterations in advanced PCa involve signaling in androgen receptor (AR), WNT, PI3K, DNA repair, and cell cycle [10, 11], next to mutations in additional genes of uncertain clinical significance as well as changes during disease progression in gene expression and epigenetics [12, 13]. Consequently, it is clinically relevant to accurately risk stratify these patients to apply the most appropriate therapeutic strategy [14].

Cyclic AMP (cAMP) can modulate AR activity via promotion of phosphorylation of the AR protein [15] or AR co-regulatory proteins [16] by protein kinase A (PKA). Alternatively, cAMP can activate cAMP response element-binding protein (CREB) through PKA phosphorylation in the presence of androgen and as such bind to promoter regions of androgen-regulated genes to enhance their transcriptional activation [17].

Phosphodiesterases (PDEs) are a superfamily of enzymes [18] which play a fundamental role in hydrolysing cAMP, leading to its degradation and termination of cell signal transduction in a spatially compartmentalised manner [19].

Previously, we reported that the expression of PDE4D7 is increased in TMPRSS2-ERG positive prostate tumors [20] and is inversely correlated to risk of biochemical recurrence (BCR) after radical prostatectomy. The PDE4D7 score independently adds value to clinical prognostic variables like ISUP Gleason grade or risk scores like CAPRA or CAPRA-S to predict post-surgical disease progression and poor patient outcome [21, 22]. Furthermore, we showed that PDE4D7 transcription is associated with androgen resistance and might be involved in the regulation of proliferation of PCa cells [23]. The patient cohorts used for these previous studies represented patients across the entire risk spectrum ranging from clinically (very) low to high-risk disease and post-surgical BCR was used as a surrogate endpoint for longitudinal patient outcomes [21, 22].

Here, we present evidence that the PDE4D7 score predicts survival outcomes in clinically high-risk PCa patients after post-surgical PSA recurrence followed by salvage radiation with or without addition of androgen deprivation therapy (ADT). We replicate in a cellular model the clinical observation that diminished PDE4D7 expression induces resistance to standard of care clinically used treatments and we suggest the manipulation of PDE4D7 expression or activity as a novel therapeutic option for patients who are at high risk to eventually succumb to their disease.

Patient cohorts and tissue samples

In total, 367 prostate cancer patients operated on between 1994–2011 were included into this study. Of those, 363 patients had at least one adverse feature in pathology (PSM: positive surgical margins; SVI: seminal vesicle invasion; EPE: extra-prostatic extension; LNI: lymph node invasion; Gleason 4 component) and/or were classified into post-treatment intermediate or high risk of disease progression according to the clinical risk metrics CAPRA-S [24] and/or the EAU-BCR risk model [25] while 4 patients demonstrated a post-operatively rising PSA in the absence of any of the stated adverse features; all 367 patients experienced biochemical recurrence during post-treatment follow-up and were stratified to salvage radiation therapy (SRT) of the pelvic bed with a total administered radiation dose between 50–72 Gy over multiple fractions. Of the 367 patients 188 also received anti-androgen therapy at different time points (before, during or after SRT) (Supplementary Fig. 1).

Two small biopsy punches (~ 1x2 mm) of a representative resected tumor area (the index lesion) of 367 patients were collected for downstream nucleic acid extraction. The local Institutional Review Board approved the collection of patient tissue for clinical research.

Generation of stable PDE4D7 modified LNCaP prostate cancer cells

- Stable shRNA mediated PDE4D7 knockdown LNCaP

LNCaP clone FGC (ATCC® CRL1740™) were cultured in RPMI 1640medium (Gibco™), supplemented with 10% (v/v) foetal bovine serum, 1% (v/v) _L-Glutamine and 1% (v/v) penicillin/streptomycin, and stored within a 37°C incubator with 5% CO₂.

The transfer plasmid of the lentivirus transfection system was designed to stably express an shRNA RNA hairpin targeting the first coding exon of the PDE4D7 transcript. As a control, a scrambled nucleotide sequence of the PDE4D7 targeting shRNA was cloned into the transfer plasmid of the lentivirus. LNCaP clone FGC cells were infected with lentiviruses at MOI = 10 in the presence of 10 µg/ml polybrene. Cells were maintained in puromycin-supplemented medium prior to selection and clonal expansion of stably transduced cells (Supplementary Materials). RNA extraction and qPCR verification of gene knockdown were performed as described below.

- Inducible PDE4D7 expression in PDE4D7 knockdown LNCaP P1 cells

PDE4D7 (GenBank ID AF536976) sequence was first subcloned into lentivirus vectors with CMV (LVR-1001-pLV-CMV-PGK-Puro) and then into Tet-On inducible vector pLV-tet-on-3G-P2A-Puro,

LNCaP-shRNA P1 cells were maintained in RPMI1640 medium supplemented with 10% (v/v) foetal bovine serum, 1% (v/v) _L-Glutamine and 1% (v/v) penicillin/streptomycin, and infected with Tet-On inducible lentivirus at MOI = 5–10 in the presence of 8 µg/ml polybrene. Cells were maintained in puromycin-supplemented medium prior to selection and clonal expansion of stably transduced cells (Supplementary Materials). Cells transduced with Tet-On inducible lentivirus have been tested for the induction of PDE4D7 expression in the presence of 1 µg/ml Doxycycline. Expression of PDE4D7 has been determined by RT-PCR.

Molecular Analysis

For downstream molecular analysis the following cell lines/clones were used: LNCaP FCG wildtype (WT); LNCaP transfected with scrambled shRNA control (clone SC2); LNCaP transfected with PDE4D7 targeting shRNA (clones P1, w5.2 and e6.3), and LNCaP clone P1 transfected with Tet-On inducible PDE4D7 vector (clone P1-TET4D7).

RT-qPCR Analysis

- Patient sample biopsies:

To account for potential tumour heterogeneity the two tissue punches of the RP cohort were combined before nucleic acid extraction. A potential difference in tumour cellularity was addressed by normalisation of the PDE4D transcript expression to four reference genes. All used molecular laboratory methods including RNA extraction, quantitative real-time PCR (RT-qPCR), oligonucleotide primers/probes and statistical analysis were done as previously described [21].

Cell lines: RNA was extracted from LNCaP cell lines using the RNeasy Mini Kit (QIAGEN) and one-step qRT-PCR was performed with the Superscript™ III Platinum™ One-Step qRT-PCR Kit (Thermo Fisher Scientific) as per manufacturers’ protocols. Experiments were carried out using the Step-One™ Real-Time PCR System (Thermo Fisher Scientific) at the following parameters: One cycle of 50^oC for 30 minutes and 95^oC for 5 minutes, followed by 45 cycles of 95^oC for 15 seconds and 60^oC for 45 seconds. Primers and probes were purchased from Integrated DNA Technologies (Supplementary Materials), with 400nM primer/200nM probe and 1ng/µl RNA per reaction. Ct values were quantified using the 2^−∆∆Ct method and statistical analysis was performed on ∆∆Ct values.

Calculation of Clinical and Genomic Risk Scores

- Calculation of clinical risk score CAPRA-S

The post-surgical CAPRA-S risk score and its corresponding low (1), intermediate (2), high-risk (3) categories were calculated as described earlier [24]. The EAU-BCR score and its two categories of low and high post-surgical risk of disease progression was determined as published [25].

- Calculation of the PDE4D7 Score and PDE4D7 Score Categories

Generation of normalized PDE4D7 transcript expression was performed by subtracting the RT-qPCR Cq of the PDE4D7 transcript from the averaged RT-qPCR Cq of the reference genes and transformed to the PDE4D7 score and its related PDE4D7 score classes [21]. In multivariable analysis Cox and logistic regression analysis for various available biological and treatment related outcomes the PDE4D7 score was used as a continuous variable. For Kaplan-Meier survival analysis we transformed the normalized PDE4D7 transcript expression values to a percentile distribution (pPDE4D7 score) which allowed more efficient comparison with other genomic scores as all scores were distributed between 0–1 (see also below). Two cut-offs for the pPDE4D7 score were defined by AUROC (Area under the ROC curve) analysis with 5-year prostate cancer specific death after start of SRT as the dependent variable and the pPDE4D7 score as the independent variable. One cut-off (pPDE4D7 > 0.2) was defined as the point in the AUROC with maximum sensitivity and specificity. The second cut-off was defined as the point in the AUROC at max sensitivity (pPDE4D7 > 0.87). Consequently, the pPDE4D7 score classes stratified patients into three sub-cohorts (pPDE4D7 scores > 0.87: ‘high’; pPDE4D7 scores > 0.2 and < = 0.87: ‘intermediate’, and pPDE4D7 scores < = 0.2: ‘low’).

- Calculation of the Cell Cycle Progression and the Genomic Prostate Score

To derive the risk scores for two commercially available prognostic tests which uses the CCP (Cell Cycle progression) 31 gene signature and GPS (Genomic Prostate Score) 12 gene signature we inferred the mxCCP and mxGPS scores from RNAseq expression data of the respective signature and reference genes based on previously published formulas to calculate the respective risk score [26]. The derived mxCCP and mxGPS scores were converted to percentile distributions (mxpCCP and mxpGPS, respectively). Where indicated we either used the mxCCP and mxGPS (AUROC analysis, Cox regression) or the mxpCCP and mxpGPS (Kaplan-Meier survival) for use in down-stream statistical analysis.

RNA sequencing

100 ng of total RNA was used as input to remove ribosomal RNA using Ribo-Zero Gold (Human/Mouse/Rat) rRNA Removal Kit (Illumina Inc.) according to the instructions of the manufacturer. For library construction, we used the total of the depleted RNA as input into the Scriptseq V2 RNA-Seq Library Preparation Kit (Epicentre/Illumina Inc.). Prepared RNAseq libraries were sequenced using NextSeq 500 sequencing system (paired-end; 2 x 75 bp read length; approx. 80 million total reads per sample for human tissue and approx. 20 million reads per samples for the sequenced cell lines). Details on RNAseq data processing and TPM calculation can be found in Supplementary Materials.

Gene Set Enrichment Analysis (GSEA)

Gene Set Enrichment Analysis (GSEA) was performed as described here (https://www.gsea-msigdb.org/gsea/index.jsp). As input for the gene set enrichment analysis the feature count matrices of the gene expression as derived from RNAseq were used.

Whole Genome Sequencing (WGS)

Approximately 100 ng gDNA per sample was randomly fragmented at 350 bp by ultrasonication. Libraries were constructed following the TruSeq DNA Nano protocol (Illumina). In brief, sheared gDNA was end-prepped, dA-tailed, and enriched for fragments of approx. 350 bp through size selection with AMPure XP beads (Beckman Coulter). Indexing PCR for a total of 8 cycles was performed for the size-selected fragments, and the products were purified with AMPure XP beads. The quality of the final libraries was checked on the LabChip GX Nucleic Acid Analyzer (PerkinElmer). The concentration was determined by performing qPCR on the samples using the 2X KAPA HiFi HotStart ReadyMix (Kapa Biosystems) as mastermix and a dilution of PhiX Sequencing Control v3 as standard. High coverage sequencing was performed on the Illumina Novaseq 6000 system with the S4 flowcell and PE150 configuration. Details on WGS data processing can be found in Supplementary Materials.

Plasmid DNA transfection

LNCaP WT and stable-PDE4D7 knockdown LNCaP clone P1 were transiently transfected with pcDNA3.1-PDE4D7-VSV plasmid DNA construct using Lipofectamine LTX with Plus Reagent (Thermo Fisher Scientific) according to manufacturer’s protocol. Cells were incubated for 24h post-transfection prior to further experiments.

xCELLigence Real-Time Cell Analysis

Real-time cell growth was analysed using the xCELLigence RTCA system (Agilent), which measures real-time cell adhesion across gold microelectrode biosensors on 96 well electronic plates within a 37°C incubator with 5% CO2. Cellular adhesion influences electrical impedance which is converted into the cell index (CI) via RTCA software. Initial background measurements were recorded with 100 µL culture medium, followed by additional 100 µL LNCaP cell suspension (10,000 cells/wells) for measurement of CI over time. Approximately 24h after seeding, cells were treated with specified concentrations of drugs. Negative controls containing either 0.1% DMSO or untreated cells were also included. CI was transformed to normalised CI via the RTCA software to timepoint of treatment.

Western Blotting

Cells were lysed in 3T3 lysis buffer (50 mM NaCl, 50 mM NaF, 30 mM sodium pyrophosphate, 25 mM HEPES, 2.5 mM EDTA, 10% (v/v) glycerol, 1% (v/v) Triton X-100; pH 7.5) supplemented with cOmplete™ EDTA-free protease inhibitor cocktail (Sigma-Aldrich). Lysed cells were rotated end-over-end for 30 minutes at 4°C, then centrifuged for 10 minutes at 13,000 RPM at 4oC. Supernatants were collected and protein concentration determined via Bradford assay. Equal concentrations of protein samples were separated via SDS–polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto nitrocellulose membranes for western blot. Membranes were blocked in 5% (w/v) milk powder in TBS-T for 1h prior to primary and secondary antibody incubations. Primary antibodies used: anti-PDE4D5 1:5000 (Baillie lab), anti-PDE4D7 1:500 (Baillie lab), anti-PDE4D9 1:5000 (Baillie lab), anti-Pan-PDE4D 1:5000 (Baillie lab), anti-VSV 1:5000 (Abcam, #ab1874), anti-GAPDH 1:5000 (Abcam, #ab8245), anti-AR 1:1000 (Cell Signal, #D6F11), anti-PSMA 1:1000 (Invitrogen, #1H8H5), anti-E-cadherin 1:500 (Cell Signalling, #3195S), anti-TMPRSS2 1:1000 (Abcam #ab109131), anti-NKX3.1 1:1000 (Cell Signalling, #92998), anti-PSA 1:1000 (Abcam, #ab76113), anti-AR-V7 (Cell Signalling, #68492), anti-GR (Cell Signalling, #12041T). Secondary antibodies used: IRDye® 800CW Donkey-anti-Rabbit IgG 1:5000 (LI-COR, #926-32213), IRDye® 680RD Donkey-anti-Goat IgG 1:5000 (LI-COR, #926-68074), IRDye 800CW Goat-anti-Human IgG 1:5000 (LI-COR, #926-32232), IRDye® 680RD Donkey-anti-Mouse IgG 1:5000 (LI-COR, #925-68072).

Immunofluorescence

LNCaP cells were seeded onto 13mm glass coverslips 24 hours prior to transfection as previously described. Cells were fixed with 4% (v/v) paraformaldehyde for 15 minutes, blocked and permeabilised with 10% donkey serum, 0.5% BSA and 0.2% Triton-X in PBS. Coverslips were incubated overnight with anti-PDE4D7 antibody 1:200 (Baillie Lab), followed by secondary antibody incubation with Alexa-Fluor 488 Donkey-anti-Goat 1:500 (Thermo Fisher Scientific, #A-11055). Coverslips were mounted onto slides with Duolink In Situ Mounting Medium with DAPI (Sigma-Aldrich) and imaged via the ZEISS LSM 880 laser scanning microscope with 63x oil immersion objective.

Data analysis and statistics

Multivariate Cox regression and Kaplan Meier analyses were applied to correlate the PDE4D7 score or the clinical metrics (CAPRA-S score, EAU-BCR risk score) and its categories to prostate cancer specific (PCSM) or all-cause mortality (ACM) in the total patient cohort (n = 367) (Supplementary Fig. 1). For statistical analysis the software package MedCalc (MedCalc Software v2.20 BVBA, Ostend, Belgium) was used using statistical tests as indicated in the main text of the manuscript and/or in the figure legends. For LNCaP data, values are presented as mean ± standard error of mean (S.E.M) or standard deviation (SD) for N ≥ 3, unless otherwise stated. Statistical analyses were performed either via one-way ANOVA + Dunnett’s multiple comparison, two-way ANOVA + Šidák’s multiple comparison, or via unpaired t-test with assumed Gaussian distribution. P ≤ 0.05 was deemed significant (* = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001, **** = p ≤ 0.0001). All statistical analyses and models were performed via GraphPad Prism 9.

The PDE4D7 score is associated with survival outcomes in high-risk prostate cancer patients

All subjects enrolled in this study experienced post-surgical PSA relapse and were subsequently stratified to salvage radiation therapy (SRT). Pathological analysis of surgical specimens revealed pT3a or higher disease in nearly 80% of the 367 patients (Table 1). Extra-prostate extension and positive surgical resection margins were identified in 62.4% and 72.2% of cases, respectively. Based on the CAPRA-S risk score [24], most patients (85%) were classified as intermediate-risk (41.1%; CAPRA-S scores 2–5) or high-risk (44.1%; CAPRA-S scores > 5) for post-surgical disease progression. Over half of the patients (188/367) received ADT at some point during the study period. During a median follow-up of 103 months post-prostatectomy, 11.7% of subjects died from prostate cancer and 17.4% died from any cause (Table 1).

Table 1

**Aggregated summary of the characteristics of the studied patient cohort.** Demographics of the radical prostatectomy (RP) patient cohort including the N = 367 patients eligible for statistical data analysis. For patient age and preoperative PSA the min and max values in the cohort are shown; median and IQR are shown in parentheses. Post-surgical pathology is given (Note: extracapsular extension was derived from pathology stage information). The outcome category illustrates the cumulative events in terms of recurrence to clinical metastases (CR), salvage radiation therapy (SRT) or androgen deprivation therapy (ADT) after surgery. Mortality is shown as prostate cancer specific mortality (PCSM) as well as overall all-cause mortality (ACM); (N/A = not available).
# Patients	N = 367
Age range [years]	45.3–79.2 (median: 62.5; IQR: 8.1)
pre-operative PSA [ng/ml]	0.3–168 (median: 10.0; IQR: 9.3)
pT Stage	N	%
pT2	78	21.3%
pT3a	131	35.7%
pT3b	103	28.1%
pT3c	44	12.0%
pT4	11	3.0%
ISUP Gleason
Grade Group 1	36	9.8%
Grade Group 2	147	40.1%
Grade Group 3	109	29.7%
Grade Group 4	19	5.2%
Grade Group 5	55	15.0%
Not available	1	0.3%
CAPRA-S Score
CAPRA-S < = 2	39	10.6%
CAPRA-S 3–5	151	41.1%
CAPRA-S > = 6	162	44.1%
Not available	15	4.1%
EAU-BCR Risk Score
EAU-BCR Risk 0	185	50.4%
EAU-BCR Risk 1	163	44.4%
Not available	19	5.2%
Surgical Margin Status
Negative	100	27.2%
Positive	265	72.2%
Not available	2	0.5%
Extra-Prostatic Extension
Absent	135	36.8%
Present	229	62.4%
Not available	3	0.8%
Seminal Vesicle Invasion
Absent	244	66.5%
Present	123	33.5%
Lymph Node Invasion
Absent	334	91.0%
Present	33	9.0%
Clinical Metastasis
No event	182	49.6%
Event	128	34.9%
Unknown	57	15.5%
Salvage Radiation
Not administered	0	0%
Administered	367	100%
Androgen Deprivation Therapy
Not administered	162	44.1%
Administered	188	51.2%
Unknown	17	4.6%
Prostate Cancer-Specific Mortality
No event	324	88.3%
Event	43	11.7%
All-Cause Mortality
No event	303	82.6%
Event	64	17.4%
Median Follow-Up [post-RP]
	103 months

Multivariate Cox regression analysis was conducted to assess the association between the PDE4D7 score and prostate cancer-specific mortality (PCSM) after SRT initiation, adjusting for pathological prognostic variables (i.e., ISUP Gleason grade group, pathological pT stage, status of extra-prostate extension and surgical margins, seminal vesicle involvement, and lymph node invasion). As previously reported for biochemical recurrence as a clinical endpoint, we observed a strong inverse association between the PDE4D7 score and PCSM after SRT initiation (HR = 0.37; 95% CI 0.23–0.58; p < 0.0001) in the multivariable model that included clinical prognostic parameters (Table 2). To determine whether the PDE4D7 score provided independent prognostic information in the context of the genomic GPS or CCP signatures, the respective RNAseq derived score (mxGPS, mxCCP) was included in the multivariable model. Interestingly, only the mxGPS signature demonstrated a significant association with PCSM in multivariable analysis (HR = 1.3; 95% CI 1.0-1.6; p = 0.03). In both analyses, the PDE4D7 score remained a significant variable (p = 0.01 and p = 0.007; Supplementary Tables 1 and 2, respectively). The PDE4D7 score was also tested in multivariable analysis for PCSM in combination with the clinical prognostic models CAPRA-S and EAU-BCR risk score [24, 25]. In both models, the PDE4D7 score remained the most significant predictor for PCSM after SRT (HR = 0.36, p < 0.0001; HR = 0.31, p < 0.0001, respectively; Supplementary Table 3).

Table 2

**Multivariable Cox regression analysis of the PDE4D7 score with the clinical prognostic parameters in a N = 342 clinical high risk prostate cancer patient cohort.** The tested endpoint was time to prostate cancer specific mortality (N = 42 events). The PDE4D7 score and the ISUP Gleason grade group were used as continuous variables; the pathology stage pT was used as categorical variable with pT2 as reference. Extra-Prostatic Extension (EPE), Seminal Vesicle Invasion (SVI), Surgical Margin Status (SMS), and Lymph Node Invasion (LNI) were used as binary variables (0=’no’; 1=’yes’). The Hazard Ratio (HR), 95% confidence interval (CI) of the HR, and the p-values are indicated.
N = 342 Patients	Multivariable
Variable	HR	95% CI	P
PDE4D7 score	0.37	0.23 to 0.58	< 0.0001
ISUP pGleason Grade Group	1.8	1.4 to 2.4	< 0.0001
pT_stage="pT3a"	0.45	0.14 to 1.4	0.19
pT_stage="pT3b"	0.1	0.01 to 0.9	0.04
pT_stage="pT3c"	0.01	0 to 0.28	0.006
pT_stage="pT4"	0.36	0.04 to 3.3	0.36
Extra-Prostatic Extension	1.6	0.3 to 5.9	0.63
Seminal Vesicle Invasion	17.1	1.4 to 205	0.02
Surgical Margin Status	0.45	0.23 to 0.93	0.03
Lymph Node Invasion	2.2	1.0 to 4.8	0.04

The Kaplan-Meier (KM) analysis of pPDE4D7 percentile scores stratified patients into three sub-cohorts with significantly different survival rates (HR intermed-high: NA; HR low-intermed: 4.9; 95% CI 2.1–11.3; logrank p < 0.0001; Fig. 1A). Patients in the high pPDE4D7 score class had 100% 10-year survival rate after SRT, while those in the intermediate and low score classes had a 10-year post-SRT survival rates of approximately 90% and below 50%, respectively (Fig. 1A). All-cause mortality exhibited similar trends in this patient cohort (Fig. 1B). The EAU-BCR Risk model stratified patients into low- and high-risk groups, with the low-risk group having a significantly better survival rate at 10 years post-radiation (HR = 3.7; 95% CI 2.0-6.8; logrank p < 0.0001; Fig. 1C). The survival rate in the low-risk group was greater than 90%, while it was approximately 72% in the high-risk group at 10 years post SRT. The CAPRA-S model did not show significant stratification between risk groups (not shown).

To assess the predictive ability of the PDE4D7 score for 5-year PCSM after SRT, we conducted receiver operating characteristic (ROC) curve analysis and calculated the corresponding area under the curve (AUC). We compared the PDE4D7 score to individual clinical parameters and clinical models, as well as combination models incorporating the PDE4D7 score with the EAU-BCR risk score and additional clinical variables (full risk model: PDE4D7 score, ISUP pGleason grade group, pT Stage, extra-prostatic extension and surgical margin status, SVI, and LNI). The individual clinical parameters and clinical models demonstrated AUCs below 0.7, while the PDE4D7 & clinical combination models exhibited AUCs of 0.81 (PDE4D7_EAU-BCR) and 0.88 (full risk model) (p < 0.0001). These results suggest that the PDE4D7 & clinical combination models have superior predictive ability for 5-year PCSM compared to the individual clinical parameters and models (Supplementary Table 4).

Finally, we applied KM survival analysis to investigate whether pPDE4D7 percentile score classes could stratify patients receiving anti-androgens in addition to SRT into distinct subgroups with varying risks of PCSM or ACM. Our results revealed a statistically significant difference in survival outcomes between pPDE4D7 score classes (HR intermed-high: NA; HR low-intermed: 4.3; 95% CI 1.9–9.6; logrank p < 0.0001; Fig. 1D). Specifically, patients in the low pPDE4D7 score class had a median survival of 89.6 months and an overall cancer-specific survival rate of 30% post-initiation of ADT, while those in the intermediate and high pPDE4D7 score classes had not reached median survival. A similar pattern was observed for ACM, with the overall survival rate dropping below 20% for the lowest pPDE4D7 class (Fig. 1E). In contrast, none of the other studied risk classification models (mxpGPS, EAU-BCR risk, CAPRA-S) exhibited the ability to stratify patients into different risk classes with significant performance in this setting (Supplementary Figs. 2–4).

Our data demonstrate that the PDE4D7 score is a significant predictor of disease specific (PCSM) and overall (ACM) mortality in patients with recurrent and progressive PCa following primary surgical resection. Patients with higher PDE4D7 scores had a more favourable long-term survival rate, while those with lower scores exhibited a poorer prognosis in terms of survival following post-surgical treatments. These findings suggest that the PDE4D7 score may be a useful predictor of patient outcomes in this patient population.

PDE4D7 knockdown is associated with loss of androgen signaling, enrichment of the mesenchymal epithelial transition, and neuroendocrine prostate cancer.

To further investigate the biological significance of PDE4D7 in PCa development and progression, we generated a derivative of the LNCaP (clone FCG) PCa cell line with selective knockdown of PDE4D7 expression using a lentivirus-mediated shRNA strategy (Supplementary Fig. 5A). Among several PDE4D-knockdown LNCaP clones, LNCaP_P1 cells exhibited the most significant downregulation of PDE4D7 mRNA expression, as confirmed by RT-qPCR (Supplementary Fig. 6). Western blot analysis and confocal microscopy further confirmed the reduction in PDE4D7 protein expression in LNCaP_P1 cells compared to the wildtype (LNCaP_WT) and a scrambled shRNA control (LNCaP_SC2) (Figs. 2A, 2B, and 2C). These findings demonstrate the successful generation of a PDE4D7 knockdown model in LNCaP cells.

We performed next-generation sequencing (NGS) RNA sequencing of the two control cell lines (LNCaP_WT and LNCaP_SC2) as well as three PDE4D7 knockdown clones (LNCaP_P1, LNCaP_w5.2 and LNCaP_w6.3) and used the corresponding gene expression count tables as input for gene set enrichment analysis (GSEA) of the 50 GSEA hallmark pathways (gsea-msigdb.org). For pathway analysis in phenotypes WT and SC2 vs. the PDE4D7 knockdown cell lines, we identified the AR response pathway as the most significantly depleted gene set in the shRNA knockdown cell lines (false discovery rate (FDR) q-value < 0.001; normalised enrichment score − 2.1; Figs. 2D, 2E, Supplementary Table 5, respectively). To confirm that RNAseq data was reflected at the protein level, western blotting for selected PCa-related proteins was performed, with significant downregulation observed in the P1 cell line compared to LNCaP WT (Fig. 2F). Importantly, the scrambled shRNA cell line SC2 did not show notable difference in expression of these genes (Supplementary Fig. 7). Amongst the most downregulated genes in the three PDE4D7 knockdown cell lines were the known AR-regulated kallikrein-related peptidases KLK2 and KLK3, the transmembrane serine protease 2 (TMPRSS2), and the transcription factor NK3 homeobox-1 (NKX3-1). Also, the AR itself – although not part of this hallmark AR response pathway – is strongly downregulated in the PDE4D7 knockdown cell line P1 (Fig. 2F). The described findings indicate that the selective depletion of the PDE4D7 transcript expression leads to a cellular phenotype which represents androgen-insensitive proliferation.

In contrast, we identified multiple hallmark pathways significantly enriched in the PDE4D7 knockdown LNCaP derivatives. Of those, the EMT pathway demonstrated the highest level of enrichment (FDR q-value < 0.001; normalised enrichment score 1.9; Figs. 2G, 2H, Supplementary Table 6, respectively). Other hallmark pathways with FDR q-values < 0.1 that may further contribute to the progressive phenotype of the PDE4D7-knockdown cells, and are potentially relevant in a clinical setting, are those representing WNT beta catenin and hedgehog signaling. Both pathways were previously linked to progression of PCa after the disease developed into hormone resistance [27, 28].

Neuroendocrine differentiation (NED) in PCa is a hallmark of aggressive, AR-independent and treatment-resistant disease [29]. It can occur in both primary and metastatic PCa [30]. NED is characterized by decreased expression of the AR and increased expression of neuroendocrine markers [31]. Given the importance of NED in aggressive PCa, we evaluated a range of genes altered in NEPC (Supplementary Table 7) for expression in our LNCaP models. This analysis revealed that many genes with previously reported induced expression levels in NEPC (Neuroendocrine Prostate Cancer) were up-regulated in the PDE4D7 knockdown models including classical NED markers like ENO2, SYP, AURKA, or NCAM1 (Fig. 2I).

To explore the potential clinical relevance of these findings, we also analysed expression of hallmark AR response genes and key markers for NED in RNAseq data of 533 human patient samples described previously by us [32] and found that all tested AR response genes were down-regulated along with reduced PDE4D7 expression in these patient tumours, while expression of the AR itself did not change (Supplementary Fig. 8). Similarly, we found several markers reported in NED with altered expression between low vs high PDE4D7 tumours (Supplementary Fig. 9) and the change of regulation was according to what was reported previously in NEPC [29, 33].

PDE4D7-knockdown in LNCaP cells alters phenotype and confers resistance to AR antagonist

Knockdown of PDE4D7 in LNCaP cell line P1 resulted in increased growth characteristics as confirmed by real-time growth assays comparing P1 LNCaP cells to WT and SC2 LNCaP cells (Fig. 3A). Importantly, no significant difference in growth characteristics were found between the WT and SC2 LNCaP cells.

Enzalutamide, an AR antagonist used to treat advanced PCa [34], was utilised to investigate the effects on growth of PDE4D7-knockdown PCa cells. Dose-response experiments using real-time growth assays showed a significant decrease in WT LNCaP proliferation at all concentrations tested (0.1–30 µM), with the most profound effects at 1 µM and above (Fig. 3B). In contrast, P1 LNCaP cells showed minimal reductions in proliferation upon treatment with enzalutamide, and even exhibited increased growth at higher concentrations (Fig. 3C). Re-expression of PDE4D7 in LNCaP P1 cells via a Tet-On inducible vector (Supplementary Fig. 5B) rescued the growth phenotype and restored sensitivity of P1 LNCaP cells to enzalutamide (Fig. 3D, 3E). This data strongly highlights that low PDE4D7 expression in PCa cells confers resistance to anti-androgens such as enzalutamide.

MR-L2, a PDE4 longform activator that binds allosterically to the PDE4 enzyme in order to mimic PKA phosphorylation in the UCR1 domain [35], was tested to rescue the enhanced growth phenotype caused by PDE4D7 knockdown in LNCaPs. Results showed inhibition of growth in P1 cells, but no effect on proliferation in WT LNCaPs (Fig. 3F, 3G). This data suggests that activation of PDE4 in PDE4D7-knockdown LNCaPs has an inhibitory effect on their proliferative potential and highlights again the importance of PDE4D7 in driving PCa.

PDE4D7 specific knockdown in LNCaPs is associated with expression changes in DNA damage repair genes

Multiple studies have conclusively demonstrated that genomic alterations in DNA damage repair (DDR) genes are prevalent in both primary and metastatic PCa tissue [36], leading to the clinical utilisation of PARP inhibitors (PARPi) in patients with DDR deficiencies [37]. However, the emergence of resistance mechanisms has been observed in patients receiving PARPi treatment [38]. To investigate the potential impact of PDE4D7 knockdown on DNA repair mechanisms, we used the REACTOME homology-directed repair (HDR) of DNA double strand breaks gene set (reactome.org: R-HSA-5693538) and expanded it with additional genes from other DDR pathways previously reported to be altered in PCa (Supplementary Table 8) [36]. Our analysis across the LNCaP WT and SC2 cell lines, as well as their PDE4D7 knockdown derivatives, revealed alterations in the expression of a subset of DDR genes in the knockdown cells, particularly among DDR genes that are commonly mutated in PCa.. Most of these genes are significantly enhanced in expression in the PDE4D7 knockdown cells (Fig. 4A), while we did not identify significant differences in single nucleotide variants between the parent and knockdown cell line (Supplementary Figs. 10A, 10B, respectively).

To further evaluate the correlation between low PDE4D7 expression and potential resistance to PARPi treatment, we re-analysed RNAseq data from 533 human clinical samples [32], focusing on the expression of PCa-relevant DDR genes. Our analysis revealed a most significant increase in BRCA2 expression with decreasing PDE4D7 expression in patient tumours (p < 0.0001) along with a range of other DDR-related genes (Fig. 4B; Supplementary Fig. 11). BRCA1, ATM and BRIP1 were also upregulated but at non-significant levels (not shown). Based on these findings, we posit that the altered expression of DDR genes with decreasing transcription of PDE4D7 may contribute to resistance to DDR targeting therapies, and that cell lines with low expression of PDE4D7 following selective knockdown may exhibit resistance to PARPi compounds such as Olaparib.

PDE4D7 knockdown in LNCaPs confers resistance to PARP-inhibition and is rescued upon reintroduction of PDE4D7

Olaparib is an FDA-approved PARPi that is effective against LNCaP cells [39]. Ceralasertib (AZD6738) is an ataxia-telangiectasia mutated and Rad3-related (ATR) inhibitor, which has been suggested to enhance synergistic effects in combination with Olaparib in metastatic castration resistant prostate cancer (CRPC) [40]. In light of this, we aimed to determine whether PDE4D7 knockdown affects LNCaP cells response to either of these treatments. Using real-time growth analysis, LNCaP WT proliferation decreased in a dose-dependent manner with Olaparib (Fig. 5A), whereas the effect on PDE4D7-knockdown LNCaP P1 was minimal (Fig. 5B). Evaluation of growth rate via slope analysis yielded similar data with significant reductions in WT growth at 3, 10 and 30 µM Olaparib, but with unexpected reduction in slope at low concentrations (0.3 and 1 µM) in cells with PDE4D7 knockdown. It is notable, however, that cells with reduced PDE4D7 expression were resistant to Olaparib at 10 and 30 µM (Fig. 5B). Next, we sought to determine whether we could rescue the phenotype through transient re-introduction of PDE4D7 into LNCaP P1. As with the TET-On induction of PDE4D7, transiently transfected PDE4D7 into P1 cells (P1-PDE4D7⁺) revealed significant downregulation in growth (Fig. 5C). Furthermore, P1-PDE4D7⁺ cells were re-sensitised to 10 µM Olaparib treatment (Fig. 5D), with a statistically significant downregulation in growth compared to both Olaparib-treated P1 cells, and DMSO-treated P1-PDE4D7⁺ cells. Additionally, proliferation of P1-PDE4D7⁺ cells decreased further with Olaparib treatment in comparison to DMSO (p = 0.057). This data re-confirms the role of PDE4D7 in blunting of PCa cell growth and promoting susceptibility to clinically relevant therapeutics.

PDE4D7-knockdown in LNCaP cells does not affect response to ATR inhibition yet may limit docetaxel treatment response

Docetaxel is a DNA-damage inducing drug which is clinically used in advanced PCa treatment. Compared to DMSO, Docetaxel significantly reduced growth of both WT and P1 LNCaPs (Supplementary Figs. 12A, 12B, respectively]. We then investigated how PDE4D7-reintroduction into P1 cells affected their response to this. As with Olaparib, P1-PDE4D7⁺ cells were more sensitive to Docetaxel treatment than P1 (Supplementary Fig. 12C), suggesting that a paucity of PDE4D7 promotes the ability to repair associated DNA damage and protects against induction of cell death. This is in line with the increased expression of several key DDR genes in PDE4D7-knockdown LNCaPs (Fig. 4A).

Interestingly, whilst PDE4D7-knockdown conferred resistance of LNCaP cells to Olaparib at 10–30 µM, these cells were extremely sensitive to ATR inhibition via Ceralasertib at the same concentrations. Growth curves for WT and P1 LNCaP cells treated with increasing concentrations of Ceralasertib showed similar trends in sensitivity to the compound (Figs. 5E, 5F, respectively). In both cases, the most significant growth inhibition occurred at 10–30 µM treatment. These results highlight that PDE4D7 knockdown confers resistance to PARP-inhibition via Olaparib, however these cells are highly susceptible to ATR-inhibition via Ceralasertib.

Cyclic AMP dynamics in response to PDE4D7 knockdown and effect of PDE4 activation

Using FRET microscopy [41], we investigated how disease-related PDE4D7 knockdown affects localized cAMP signaling in response to drug treatments or cellular conditions in PCa CUTie cAMP probes targeting the plasma membrane (AKAP-79) or cytosol (CYTO-pDUAL) were used [42]. Rolipram downregulated the FRET ratio in the membrane compartment of WT LNCaPs, indicating increased cAMP concentration [Fig. 6B]. Forskolin/IBMX treatment did not further affect these cells, indicating that PDE4 inhibition maximized cAMP concentration. P1 LNCaPs showed no significant difference in FRET ratio upon rolipram treatment, but forskolin/IBMX addition slightly increased intracellular cAMP [Fig. 6D]. PDE4D7 is substantially membrane-located [Fig. 4C], leading to increased cAMP at this locale in WT cells. PDE4D7 knockdown elevated basal cAMP levels, leading to no upregulation in P1 cells. No observable difference was seen in WT and P1 LNCaPs upon rolipram treatment in the cytosol, but forskolin/IBMX addition decreased FRET ratio/increased cAMP [Figs. 6A and 6B]. This data stipulates the importance of PDE4D7 as a key regulator of cAMP concentration in the membrane domain of prostate cancer cells.

PDE4D7 score predicts PSA relapse, mortality, and therapy response

Patients with PCa rely on treatment stratification in order to receive the most efficacious therapy for their specific disease presentation. SRT remains a commonly utilised treatment modality for PSA recurrent disease, however, reported 5-year biochemical failure rates range between 25–70% [43]. Further disease progression after SRT failure is driven by AR signaling [44], hence ADT has remained a mainstay treatment for advanced disease for several decades [45]. While initially responsive to androgen depletion, ADT may select for cellular clones that evade the selection pressure via androgen-independent survival mechanisms, inevitably leading to the development of CRPC with resistance to all forms of ADT [46].

Therefore, it is imperative to identify emerging treatment resistance early in order to potentially select alternative treatment regimens that may improve survival outcomes. This may include combination therapies that have been tested in metastatic hormone-sensitive clinical trials, such as ADT in conjunction with second generation anti-androgens or docetaxel [47–49]. While these combination therapies may be valuable in the future for treating patients at highest risk of disease progression with current standard therapies, the adverse effects must be carefully considered, and these treatments should only be prescribed to patients who stand to benefit.

Prognostic genomic biomarkers that objectively assess the risk of death from advanced PCa may aid in the selection of patients for treatment de-escalation (those at virtually no risk of death from the disease) versus treatment intensification for those with poor prognosis. Clinical trials such as NCT03371719, NCT04513717 and NCT04484818 are currently evaluating the utility of molecular stratification and prognostic tools in determining the optimal treatment approach for high-risk/recurrent PCa [50–52] while PREDICT-RT trial (NCT04513717) is examines the feasibility of less intense treatment for patients at low genomic risk. These trials will provide valuable insights into the use of prognostic tools to maximize benefit for patients.

The present study demonstrates that individuals with low expression levels of PDE4D7 exhibit a substantially elevated risk of mortality from PCa. Alterations in PDE expression result in dysregulation of the cAMP/PKA signaling pathway, which has been extensively linked to prostate carcinogenesis and therapy resistance. cAMP binding activates the guanosine exchange factor Epac, promoting activation of the Ras family proteins Rap1 and Rap2, which in turn regulate a variety of cellular processes, including migration, focal adhesion, proliferation and cell survival [53]. PKA-mediated phosphorylation of the transcription factor CREB has been shown to regulate the expression of oncogenes such as c-Jun and cyclin D1 [54]. Additionally, cAMP/PKA phosphorylates downstream protein kinases such as Raf and GSK3, which are involved in the regulation of cell survival [55, 56]. The focal adhesion kinase FAK, another substrate of PKA, promotes the formation of cancer metastases [57], while PKA-mediated inactivation of CAMKK2 affects cellular energy homeostasis and metabolism [58]. Furthermore, cAMP/PKA phosphorylates the AR-binding heat shock protein HSP90, thereby releasing the AR from the complex and facilitating nuclear translocation, leading to activation of AR-target genes [59].

PDE4 is one of the most highly expressed PDE families in the human prostate [60], a finding that is consistent with our previous studies [20]. Several studies have highlighted the relevance of PDE4D in the development and progression of PCa. Baca et al. proposed that PCa progresses through clonal and punctuated events, rather than gradual or catastrophic evolution. They identified the loss of PDE4D as an early, clonal event in cancer development [61]. Additionally, a study that employed whole genome sequencing of 112 primary and metastatic PCa samples identified PDE4D as a potential driver gene of the disease [62]. The PDE4D gene was often impacted in the studied tumour samples by chromosomal rearrangements and regions of loss of heterozygosity, particularly in the part encoding the long PDE4D transcripts, which includes the isoform PDE4D7. The study by Wedge and colleagues [62] has demonstrated distinct genomic alterations in ETS fusion positive and negative PCa tumours. Specifically, they found that homozygous deletions on chromosome 5 are less prevalent and occur later in progression in ETS fusion positive tumours, compared to ETS fusion negative tumours where these deletions are more common and occur earlier in the disease course. Additionally, our previous study [32] reported that the TMPRSS2-ERG gene fusion status has an impact on progression-free survival as determined by PDE4D7 expression levels. Intriguingly, our data here shows that low expression of PDE4D7 is associated with strongly elevated risk of death in ETS fusion negative tumours (Supplementary Figs. 13A and 13B). These findings suggest that loss of PDE4D7 may represent an early, clonal event in ETS fusion negative prostate cancer and is associated with the development of a progressive phenotype. Together, these data provide significant evidence that reduced expression of the PDE4D gene, particularly the isoform PDE4D7, is involved in the development and progression of PCa, as well as inversely associated with progression-free and disease-specific survival in advanced PCa after prostate-specific antigen relapse.

Therapy resistance in PCa may be attributed to the intrinsic biology of the tumour or may develop as a result of therapeutic interventions. One potential mechanism of resistance is the development of a neuroendocrine phenotype via neuroendocrine differentiation (NED) in primary adenocarcinoma, which is the most common histopathology of primary PCa. This phenomenon has been observed to occur both in treatment-naive tumours and in tumours that have undergone ADT [63]. In untreated tumours, cells that stain positively for neuroendocrine markers are referred to as NE-like cells, as they still genetically resemble prostate adenocarcinoma cells [64]. However, it is well established that PCa cells in culture can undergo NED in vitro through exposure to various stimuli, such as increase in cAMP and/or PKA activation [65]. In treated tumours, it is believed that the differentiation from adenocarcinoma to a neuroendocrine phenotype is driven by selection pressure induced by treatment, as is supported by in vitro [66] and in vivo data [67] which associated with reduced AR activity or expression [68]. It has been postulated that differentiated NE-like tumour cells, whether induced by treatment or by endogenous stimuli, may progress to small cell prostate carcinoma (SCPC) through the accumulation of additional genetic alterations, such as loss of RB1, MYCN, and AURKA amplification [63]. These cells are characterised by rapid proliferation, loss of AR and PSA expression, enhanced transcription of neuroendocrine markers, and strong resistance to further treatment interventions, resulting in a major treatment challenge due to a lack of effective therapeutic options [69].

Our findings demonstrate that targeted disruption of the PDE4D7 transcript results in the emergence of a phenotype that mimics that of SCPC. Notably, the AR and its associated response genes are downregulated, while cells exhibit accelerated proliferation and resistance to a broad range of therapeutic compounds. These transcriptomic alterations are partially replicated in patient samples with low PDE4D7 expression. We postulate that the reduction in PDE4D7 transcription results in a pre-differentiation state to neuroendocrine prostate cancer (NEPC), which can subsequently progress to SCPC under conditions of hormone deprivation. This could contribute to treatment resistance and poor prognosis, as demonstrated in patients with diminished PDE4D7 expression in their primary tumours. Based on these findings, we propose that PDE4D7 may serve as a biomarker for early identification of NEPC. Additionally, we suggest close monitoring of patients with low PDE4D7 expression who are treated with potent anti-androgens, as these tumours may rapidly progress to a state where further treatment options become limited.

PDE4D7 expression influences the DNA damage response in PCa

An important aspect of treatment resistance that we investigated is the alteration of DDR pathways. The targeting of DDR has become increasingly relevant in recent years due to the discovery of deleterious mutations in genes involved in various DDR pathways, which often results in loss of function of the corresponding DNA repair proteins. In particular, the inhibition of PARP1/2 activity has become a cornerstone of treatment for advanced cancers with DDR deficiencies and has been extensively studied in the literature [70]. However, it has also been reported that increase in the gene expression level of DDR genes may lead to resistance to cisplatin chemotherapy [71], while the overexpression of BRCA1, BRCA2, RAD51, and RPA1 has been observed in hypopharyngeal and nasopharyngeal carcinoma cells resistant to radiotherapy [72].

We present evidence that the selective suppression of PDE4D7 in LNCaP cells results in the transcriptional induction of a range of genes involved in DNA repair, which is replicated to some extent in clinical patient samples with low PDE4D7 expression. This may explain the resistance of the PDE4D7 knockdown cell line to the PARP inhibitor, Olaparib. Interestingly, the PARPi resistant PDE4D7 knockdown LNCaP cell line showed substantial response to inhibition of the DNA repair upstream key regulator ATR which might provide a potential future therapeutic option for these otherwise difficult-to-treat tumours.

It is noteworthy that a recent publication showed that oncogene-driven activation of cAMP signalling in pituitary somatotroph adenomas leads to the induction of double-strand break DNA damage. The DNA damage effect may be further augmented by the reported downregulation of the PDE4D gene in these tumours [73]. Furthermore, somatotroph tumours are characterized by widespread chromosomal copy number abnormalities, which may relate to the observation of extensive copy number variations (CNV) in the PDE4D7 knockdown cell lines compared to the LNCaP control cell lines (Supplementary Figs. 14A-D). Interestingly, we recently reported a protein-protein interaction between PDE4D7 and the helicase, DHX9. This enzyme is involved in multiple functional processes, including the resolution of R-loop DNA-RNA hybrids, which are an important source of replication stress and genome instability [74]. The downregulation of PDE4D7 transcription may impact the capacity of DHX9 to resolve these hybrid structures in the DNA, which may in turn be the result of cAMP pathway-driven DNA damage in prostate cancer cells.

PDE4D7 re-expression or activation as a therapeutic avenue for PCa

We have shown through re-expression of PDE4D7 in the knockdown LNCaP cells via either transient DNA transfection or an inducible Tet-On system, that we can not only restore a reduced growth phenotype, but also that these cells can be re-sensitised to treatments such as Enzalutamide and Olaparib, highlighting the potential for manipulating PDE4D7 gene expression as a novel therapeutic avenue for PCa independent of AR-targeting treatments.

Additionally, using a PDE4 long-form activating compound, MR-L2, which mimics PKA phosphorylation within the UCR1 domain [35], we can re-activate the residual PDE4D7 in PCa cells and slow down proliferation. Ultimately, our findings support the feasibility of either re-activation of PDE4D7 via compounds, or enhanced expression of PDE4D7 transcript such as via gene therapy, as a new treatment type for high-risk PCa patients to hinder tumour growth and re-sensitise cells to standard clinical treatments, overcoming therapeutic resistance in the treatment of advanced PCa.

This study suggests that decreased PDE4D7 expression is associated with a more aggressive form of PCa characterized by increased cell growth and resistance to standard treatments. Knockdown of PDE4D7 in LNCaP cells led to reduced androgen signaling, increased EMT, neuroendocrine differentiation, and DNA repair pathway alterations, partially replicating clinical findings. Our insights suggest that PDE4D7 may act as a key switch inducing changes mimicking PCa molecular biology after extensive anti-androgen treatment. Tumors with low PDE4D7 expression may be at high risk for progressing towards NEPC and therapeutic resistance, suggesting caution in using potent androgen inhibitors. Targeting PDE4D7 or its downstream pathways may maintain cAMP signaling during androgen deprivation therapy and may provide a novel strategy for extending the period of androgen response in advanced prostate cancers.

PCa – prostate cancer

RP – radical prostatectomy

BCR – biochemical recurrence

PCSM – prostate cancer specific mortality

ACM – all cause mortality

SRT – salvage radiation therapy

ADT – androgen deprivation therapy

CAPRA – Cancer of the Prostate Risk Assessment

NED – neuroendocrine differentiation

NEPC – neuroendocrine prostate cancer

PARPi – PARP inhibitor

DDR – DNA damage response

HDR – homology directed repair

HR – hazard ratio

CI – confidence interval

ROC – Receiver Operating Characteristics

AUC – area under the curve

CRPC – castration resistant prostate cancer

SCPC – small cell prostate carcinoma

Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68(6):394–424. https://doi.org/10.3322/caac.21492
Fletcher, S.A., von Landenberg, N., Cole, A.P. et al. Contemporary national trends in prostate cancer risk profile at diagnosis. Prostate Cancer Prostatic Dis 2020, 23, 81–87. https://doi.org/10.1038/s41391-019-0157-y
van den Bergh R, Gandaglia G, Derya Tilki D, Borgmann H et al. Trends in Radical Prostatectomy Risk Group Distribution in a European Multicenter Analysis of 28 572 Patients: Towards Tailored Treatment. Eur Urol Focus 2019, 5(2), 171-178. https://doi: 10.1016/j.euf.2017.07.003
Chierigo F, Borghesi M, Würnschimmel C, Rocco Simone Flammia RS et al. Contemporary Pathological Stage Distribution After Radical Prostatectomy in North American High-Risk Prostate Cancer Patients. Clinical Genitourinary Cancer 2022, 20(5), e380-e389. https://doi.org/10.1016/j.clgc.2022.04.005
Moyer VA. Screening for Prostate Cancer: U.S. Preventive Services Task Force Recommendation Statement. Ann Intern Med 2012, 157(2), 120-134. https://doi.org/10.7326/0003-4819-157-2-201207170-00459
Herget KA, Patel DP, Hanson HA, Sweeney C et al. Recent decline in prostate cancer incidence in the United States, by age, stage, and Gleason score. Cancer Med. 2016;5(1):136–41. https://doi.org/10.1002/cam4.549
Leyh-Bannurah, SR., Karakiewicz, P.I., Pompe, R.S. et al. Inverse stage migration patterns in North American patients undergoing local prostate cancer treatment: a contemporary population-based update in light of the 2012 USPSTF recommendations. World J Urol 2019, 37, 469–479. https://doi.org/10.1007/s00345-018-2396-2
Santino S. Butler SS, Muralidhar V , Zhao SG, Sanford NN et al. Prostate cancer incidence across stage, NCCN risk groups, and age before and after USPSTF Grade D recommendations against prostate-specific antigen screening in 2012. Cancer 2020, 126(4), 717-724. https://doi.org/10.1002/cncr.32604
Desai MM, Cacciamani GE, Gill K, et al. Trends in Incidence of Metastatic Prostate Cancer in the US. JAMA Netw Open 2022, 5(3), e222246. https://doi.org/10.1001/jamanetworkopen.2022.2246
Robinson D, van Allen EM, Wu YM, Schultz N et al. Integrative clinical genomics of advanced prostate cancer. Cell 2015, 161, 1215–1228. https://doi.org/10.1016/j.cell.2015.05.001
Abida W, Cyrta J, Heller G, Prandi D et al. Genomic correlates of clinical outcome in advanced prostate cancer. Proc Natl Acad Sci USA 2019, 116, 11428–11436. https://doi.org/10.1073/pnas.1902651116
Aggarwal R, Huang J, Alumkal JJ, Zhang L et al. Clinical and genomic characterization of treatment- emergent small- cell neuroendocrine prostate cancer: a multi- institutional prospective study. J Clin Oncol 2018, 36, 2492–2503. https://doi.org/10.1200/JCO.2017.77.6880
Beltran H, Prandi D, Mosquera JM, Benelli M et al. Divergent clonal evolution of castration- resistant neuroendocrine prostate cancer. Nat Med 22 (2016), 298–305. https://doi.org/10.1038/nm.4045
Ku, SY., Gleave, M.E. & Beltran, H. Towards precision oncology in advanced prostate cancer. Nat Rev Urol (2019), 16, 645–654. https://doi.org/10.1038/s41585-019-0237-8
Sadar MD. Androgen-independent Induction of Prostate-specific Antigen Gene Expression via Cross-talk between the Androgen Receptor and Protein Kinase A Signal Transduction Pathways. J. Biol. Chem 1999, 274, 7777–7783. https://doi.org/10.1074/jbc.274.12.7777
Dagar M, Singh JP, Dagar G, Tyagi RK et al. Phosphorylation of HSP90 by protein kinase A is essential for the nuclear translocation of androgen receptor. J Biol Chem 2019, 294(22), 8699-8710. https://doi.org/10.1074/jbc.RA119.007420
Kim J, Jia L, Stallcup MR, Coetzee GA. The role of protein kinase A pathway and cAMP responsive element-binding protein in androgen receptor-mediated transcription at the prostate-specific antigen locus. J Mol Endocrinol. 2005, 34(1), 107-118. https://doi.org/10.1677/jme.1.01701
Francis SH, Blount MA, Corbin JD. Mammalian Cyclic Nucleotide Phosphodiesterases: Molecular Mechanisms and Physiological Functions. Physiol Rev 2011, 91(2), 651-690. https://doi.org/10.1152/physrev.00030.2010
Baillie GS. Compartmentalized signalling: Spatial regulation of cAMP by the action of compartmentalized phosphodiesterases. FEBS J. 2009, 276(7), 1790–9. https://doi.org/10.1111/j.1742-4658.2009.06926.x
Böttcher R, Henderson DJP, Dulla K, Van Strijp D, Waanders LF, Tevz G, et al. Human phosphodiesterase 4D7 (PDE4D7) expression is increased in TMPRSS2-ERG-positive primary prostate cancer and independently adds to a reduced risk of post-surgical disease progression. Br J Cancer. 2015;113(10):1502-1511. https://doi.org/doi: 10.1038/bjc.2015.335
Alves de Inda M, van Strijp D, den Biezen-Timmermans E, van Brussel A, Wrobel J, van Zon H, et al. Validation of Cyclic Adenosine Monophosphate Phosphodiesterase-4D7 for its Independent Contribution to Risk Stratification in a Prostate Cancer Patient Cohort with Longitudinal Biological Outcomes. Eur Urol Focus. 2018, 4(3), 376-384. https://doi.org/10.1016/j.euf.2017.05.010
Van Strijp D, De Witz C, Vos PC, Den Biezen-Timmermans E, Van Brussel A, Wrobel J, et al. The Prognostic PDE4D7 Score in a Diagnostic Biopsy Prostate Cancer Patient Cohort with Longitudinal Biological Outcomes. Prostate Cancer. 2018, 2018:5821616. https://doi.org/10.1155/2018/5821616
Henderson DJP, Byrne A, Dulla K, Jenster G, Hoffmann R, Baillie GS, et al. The cAMP phosphodiesterase-4D7 (PDE4D7) is downregulated in androgen-independent prostate cancer cells and mediates proliferation by compartmentalising cAMP at the plasma membrane of VCaP prostate cancer cells. Br J Cancer 2014 Mar, 110(5), 1278-1287. https://doi.org/10.1038/bjc.2014.22
Cooperberg MR, Hilton JF, Carroll PR. The CAPRA-S score: a straight-forward tool for improved predicion outcomes after radical prostatectomy. Cancer 2011, 117(22), 5039-5046. https://doi.org/10.1002/cncr.26169
Tilki D, Preisser F, Graefen M, Huland H, Pompe RS. External Validation of the European Association of Urology Biochemical Recurrence Risk Groups to Predict Metastasis and Mortality After Radical Prostatectomy in a European Cohort. Eur Urol 2019, 75(6), 896-900. https://doi.org/10.1016/j.eururo.2019.03.016
Salami SS, Hovelson DH, Kaplan JB, Mathieu R, Udager AM, Curci NE, et al. Transcriptomic heterogeneity in multifocal prostate cancer. JCI insight. 2018;3(21):1–13. https://doi.org/10.1172/jci.insight.123468
Wang C, Chen Q, Xu H. Wnt/β-catenin signal transduction pathway in prostate cancer and associated drug resistance. Discov Oncol 2021, 12(1), 2-12. https://doi.org/10.1007/s12672-021-00433-6
Shaw G, Price AM, Ktori E, Bisson I, Purkis PE, McFaul S, et al. Hedgehog Signalling in Androgen Independent Prostate Cancer. Eur Urol 2008, 54(6), 1333-1343. https://doi.org/10.1016/j.eururo.2008.01.070
Merkens, L., Sailer, V., Lessel, D. et al. Aggressive variants of prostate cancer: underlying mechanisms of neuroendocrine transdifferentiation. J Exp Clin Cancer Res 41, 46 (2022). https://doi.org/10.1186/s13046-022-02255-y
Epstein JI, Amin MB, Beltran H, Lotan TL, Mosquera JM, Reuter VE, Robinson BD, Troncoso P, Rubin MA. Proposed morphologic classification of prostate cancer with neuroendocrine differentiation. Am J Surg Pathol. 2014 Jun;38(6):756-67. https://doi.org/10.1097/PAS.0000000000000208
Fine SW. Neuroendocrine tumors of the prostate. Mod Pathol. 2018 Jan;31(S1):S122-132. https://doi.org/10.1038/modpathol.2017.164.
Van Strijp D, De Witz C, Heitkötter B, Huss S, Bögemann M, Baillie GS, et al. The association of the long prostate cancer expressed PDE4D transcripts to poor patient outcome depends on the tumour’s TMPRSS2-ERG fusion status. Prostate Cancer. 2019;2019. https://doi.org/10.1155/2019/8107807
Xie Y, Ning S, Hu J. Molecular mechanisms of neuroendocrine differentiation in prostate cancer progression. J Cancer Res Clin Oncol. 2022 Jul;148(7):1813-1823. https://doi.org/10.1007/s00432-022-04061-7
Menon MP, and Higano CS. Enzalutamide, a second generation androgen receptor antagonist: development and clinical applications in prostate cancer. Curr Oncol Rep 2013, 15(2), 69-75. https://doi.org/10.1007/s11912-013-0293-9
Omar F, Findlay JE, Carfray G, Allcock RW, Jiang Z, Moore C, et al. Small-molecule allosteric activators of PDE4 long form cyclic AMP phosphodiesterases. Proc Natl Acad Sci USA 2019, 116(27), 13320-13329. https://doi.org/10.1073/pnas.1822113116
Lozano R, Castro E, Aragón IM, Cendón Y, Cattrini C, López-Casas PP, et al. Genetic aberrations in DNA repair pathways: a cornerstone of precision oncology in prostate cancer. Br J Cancer 2021, 124(3), 552-563. https://doi.org/10.1038/s41416-020-01114-x
Congregado B, Rivero I, Osmán I, Sáez C, López RM. PARP Inhibitors: A New Horizon for Patients with Prostate Cancer. Biomedicines 2022, 10(6), 1416. https://doi.org/10.3390/biomedicines10061416
Rose M, Burgess JT, O’Byrne K, Richard DJ, Bolderson E. PARP Inhibitors: Clinical Relevance, Mechanisms of Action and Tumor Resistance. Front Cell Dev Biol 2020, 8 :564601. https://doi.org/10.3389/fcell.2020.564601
Feiersinger GE, Trattnig K, Leitner PD, Guggenberger F, Oberhuber A, Peer S, et al. Olaparib is effective in combination with, and as maintenance therapy after, first-line endocrine therapy in prostate cancer cells. Mol Oncol. 2018, 12(4), 561-576. https://doi.org/10.1002/1878-0261.12185
Lloyd RL, Wijnhoven PWG, Ramos-Montoya A, Wilson Z, Illuzzi G, Falenta K, et al. Combined PARP and ATR inhibition potentiates genome instability and cell death in ATM-deficient cancer cells. Oncogene 2020, 39(25):4869-4883. https://doi.org/10.1038/s41388-020-1328-y
Deal J, Pleshinger DJ, Johnson SC, Leavesley SJ, Rich TC. Milestones in the development and implementation of FRET-based sensors of intracellular signals: A biological perspective of the history of FRET. Cell Signal 2020, 75:109769. https://doi.org/10.1016/j.cellsig.2020.109769
Surdo NC, Berrera M, Koschinski A, Brescia M, Machado MR, Carr C, et al. FRET biosensor uncovers cAMP nano-domains at b-adrenergic targets that dictate precise tuning of cardiac contractility. Nat Publ Gr [Internet]. 2017;8:1–14. https://doi.org/10.1038/ncomms15031
Ohri N, Dicker AP, Trabulsi EJ, Showalter TN. Can early implementation of salvage radiotherapy for prostate cancer improve the therapeutic ratio? A systematic review and regression meta-analysis with radiobiological modelling. Eur J Cancer 2012, 48, 837-844. https://doi.org/110.1016/j. ejca.2011.08.013
Jacob A, Raj R, Allison DB, Myint, ZW. Androgen Receptor Signaling in Prostate Cancer and Therapeutic Strategies. Cancers 2021, 13, 5417. https://doi.org/10.3390/cancers13215417
Huggins C & Hodges CV. Studies on prostatic cancer. I. The effect of castration, of estrogen and of androgen injection on serum phosphatases in metastatic carcinoma of the prostate. CA Cancer J Clin 1972, 22, 232–240. https://doi.org/10.3322/canjclin.22.4.232
Quigley DA, Dang HX, Zhao SG, Lloyd P, Aggarwal R, Alumkal JJ, et al. Genomic hallmarks and structural variation in metastatic prostate cancer. Cell 2018, 174, 758-769. https://doi.org/10.1016/j.cell.2018.06.039. Epub 2018 Jul 19
Kyriakopoulos, CE, Chen, YH, Carducci, MA, Liu, G., Jarrard, DF et al. Chemohormonal Therapy in Metastatic Hormone-Sensitive Prostate Cancer: Long-Term Survival Analysis of the Randomized Phase III E3805 CHAARTED Trial. Journal of Clinical Oncology y, 36(11), 1080–1087. https://doi.org/10.1200/JCO.2017.75.3657
James ND, Sydes MR, Clarke NW, Mason MD et al. Addition of docetaxel, zoledronic acid, or both to first-line long-term hormone therapy in prostate cancer (STAMPEDE): survival results from an adaptive, multiarm, multistage, platform randomised controlled trial. Lancet. 2016, 387(10024), 1163-77. https://doi.org/10.1016/S0140-6736(15)01037-5
Fizazi K, Tran N, Fein L, Matsubara N et al. Abiraterone acetate plus prednisone in patients with newly diagnosed high-risk metastatic castration-sensitive prostate cancer (LATITUDE): final overall survival analysis of a randomised, double-blind, phase 3 trial. Lancet Oncol. 2019 May, 20(5), 686-700. https://doi.org/10.1016/S1470-2045(19)30082-8.
NCT03371719: BIOMARKER TRIAL of APALUTAMIDE and RADIATION for RECURRENT PROSTATE CANCER (BALANCE); https://clinicaltrials.gov/ct2/show/NCT03371719
NCT04513717: Two Studies for Patients With High Risk Prostate Cancer Testing Less Intense Treatment for Patients With a Low Gene Risk Score and Testing a More Intense Treatment for Patients With a High Gene Risk Score, The PREDICT-RT Trial; https://clinicaltrials.gov/ct2/show/NCT04513717
NCT04484818: Testing the Addition of Darolutamide to Hormonal Therapy (Androgen Deprivation Therapy [ADT]) After Surgery for Men With High-Risk Prostate Cancer, The ERADICATE Study; https://clinicaltrials.gov/ct2/show/NCT04484818
Formoso K, Lezoualc’h F, Mialet-Perez J. Role of EPAC1 signalosomes in cell fate: friends or foes? Cells. 2020;9:E1954. https://doi.org/10.3390/cells9091954
Steven A, Friedrich M, Jank P, Heimer N, Budczies J, Denkert C, et al. What turns CREB on? And off? And why does it matter? Cell Mol Life Sci. 2020;77:4049–67. https://doi.org/10.1007/s00018-020-03525-8
Häfner S, Adler HS, Mischak H, Janosch P, Heidecker G, Wolfman A, et al. Mechanism of inhibition of Raf-1 by protein kinase A. Mol Cell Biol. 1994;14:6696–703. https://doi.org/10.1128/mcb.14.10.6696-6703.
Jensen J, Brennesvik EO, Lai YC, Shepherd PR. GSK-3beta regulation in skeletal muscles by adrenaline and insulin: evidence that PKA and PKB regulate different pools of GSK-3. Cell Signal. 2007;19:204–10. https://doi.org/10.1016/j.cellsig.2006.06.006
Cheng Y, Gao XH, Li XJ, Cao QH, Zhao DD, Zhou JR, et al. Depression promotes prostate cancer invasion and metastasis via a sympathetic cAMP-FAK signaling pathway. Oncogene. 2018;37:2953–66. https://doi.org/10.1038/s41388-018-0177-4
Langendorf CG, O’Brien MT, Ngoei KRW, McAloon LM, Dhagat U, Hoque A, et al. CaMKK2 is inactivated by cAMP–PKA signaling and 14–3-3 adaptor proteins. J Biol Chem. 2020. https://doi.org/10.1074/jbc.RA120.01375 6.
Dagar M, Singh JP, Dagar G, Tyagi RK, Bagchi G. Phosphorylation of HSP90 by protein kinase A is essential for the nuclear translocation of androgen receptor. J Biol Chem. 2019;294:8699–710. https://doi.org/10.1074/jbc.RA119.007420
Uckert S, Oelke M, Stief CG, Andersson KE, Jonas U, Hedlund P. Immunohistochemical distribution of cAMP- and cGMP-phosphodiesterase (PDE) isoenzymes in the human prostate. Eur Urol. 2006 Apr;49(4):740-5. https://doi.org/10.1016/j.eururo.2005.12.050
Baca SC, Prandi D, Lawrence MS, Mosquera JM et al. Punctuated evolution of prostate cancer genomes. Cell. 2013 Apr 25;153(3):666-77. https://doi.org/10.1016/j.cell.2013.03.021. PMID: 23622249
Wedge, D.C., Gundem, G., Mitchell, T. et al. Sequencing of prostate cancers identifies new cancer genes, routes of progression and drug targets. Nat Genet 50, 682–692 (2018). https://doi.org/10.1038/s41588-018-0086-z
Terry S, Beltran H. The many faces of neuroendocrine differentiation in prostate cancer progression. Front Oncol. 2014 Mar 25;4:60. https://doi.org/10.3389/fonc.2014.00060. PMID: 24724054
Sauer CG, Roemer A, Grobholz R. Genetic analysis of neuroendocrine tumor cells in prostatic carcinoma. Prostate (2006) 66:227–34. https://doi.org/10.1002/pros.20338
Cox ME, Deeble PD, Bissonette EA, Parsons SJ. Activated 3',5'-cyclic AMP-dependent protein kinase is sufficient to induce neuroendocrine-like differentiation of the LNCaP prostate tumor cell line. J Biol Chem. 2000 May 5;275(18):13812-8. https://doi.org/10.1074/jbc.275.18.13812. PMID: 10788503
Shen R, Dorai T, Szaboles M, Katz AE. Transdifferentiation of cultured human prostate cancer cells to a neuroendocrine cell phenotype in a hormone-depleted medium. Urol Oncol (1997) 3:67–75. https://doi.org/10.1016/S1078-1439(97)00039-2
Hirano D, Okada Y, Minei S, Takimoto Y, et al. Neuroendocrine differentiation in hormone refractory prostate cancer following androgen deprivation therapy. Eur Urol (2004) 45:586–92. https://doi.org/10.1016/j.eururo.2003.11.032
Wright ME, Tsai MJ, Aebersold R. Androgen receptor represses the neuroendocrine trans-differentiation process in prostate cancer cells. Mol Endocrinol (2003) 17:1726–1737. https://doi.org/10.1210/me.2003-0031
Merkens, L, Sailer, V, Lessel, D et al. Aggressive variants of prostate cancer: underlying mechanisms of neuroendocrine transdifferentiation. J Exp Clin Cancer Res (2022), 41, 46. https://doi.org/10.1186/s13046-022-02255-y
Jurkovicova, D.; Neophytou, C.M.; Gašparović, A.Č.; Gonçalves, A.C. DNA Damage Response in Cancer Therapy and Resistance: Challenges and Opportunities. Int J Mol Sci 2022, 23, 14672. https://doi.org/10.3390/ijms232314672
Wang, L.-E.; Yin, M.; Dong, Q.; Stewart, D.J.; Merriman, K.W.; Amos, C.I.; Spitz, M.R.; Wei, Q. DNA Repair Capacity in Peripheral Lymphocytes Predicts Survival of Patients with Non–Small-Cell Lung Cancer Treated with First-Line Platinum-Based Chemotherapy. J. Clin. Oncol. 2011, 29, 4121–4128. https://doi.org/10.1200/JCO.2010.34.3616
Wang, Z.; Zuo, W.; Zeng, Q.; Li, Y.; Lu, T.; Bu, Y.; Hu, G. The Homologous Recombination Repair Pathway is Associated with Resistance to Radiotherapy in Nasopharyngeal Carcinoma. Int. J. Biol. Sci. 2020, 16, 408–419. https://doi.org/10.7150/ijbs.37302
Ben-Shlomo A, Deng N, Ding E, Yamamoto M, Mamelak A, Chesnokova V, Labadzhyan A, Melmed S. DNA damage and growth hormone hypersecretion in pituitary somatotroph adenomas. J Clin Invest. 2020, 130(11), 5738-5755. https://doi.org/10.1172/JCI138540
Cristini A, Groh M, Kristiansen MS, Gromak N. RNA/DNA Hybrid Interactome Identifies DXH9 as a Molecular Player in Transcriptional Termination and R-Loop-Associated DNA Damage. Cell Rep 2018, 23(6), 1891-1905. https://doi.org/10.1016/j.celrep.2018.04.025

Yes there is potential Competing Interest. R.H. is an employee of Philips Research. R.H. and G.S.B hold patent rights relevant to the published work. The other authors do not declare any conflict of interest.

SupplementaryMaterialsDiminishedPDE4D7isAssociatedwithTreatmentResistantLethalPCa.docx
Supplementary Materials

Download PDF

Version 1

posted

You are reading this latest preprint version

Diminished PDE4D7 Expression is Associated with Treatment-Resistant, Lethal Prostate Cancer and Manifests with Impaired Androgen Response, Neuroendocrine Differentiation and Alterations in DNA Repair

Status:

Version 1

Abstract

Figures

Introduction

Methods

Results

Discussion

Conclusions

Key To Abbreviations

References

Additional Declarations

Supplementary Files

Status:

Version 1