Reduced PDE4D7 in Prostate Cancer Correlates With Genomic Downregulation Within the Upstream PDE4D Coding Region

Abstract

Aim:PDE4D7 expression is significantly associated with prostate cancer (PCa) progression, representing an attractive prognostic biomarker. We sought to determine whether other genes in the PDE4D coding region were associated. Patients & methods: RNA from biopsy punch samples of resected tumor tissue was analyzed via RNA sequencing. RT-qPCR was used to determine PDE4D7 score. Results: Numerous genomic sequences within the PDE4D coding region on Chr5q12 revealed similar mRNA expression profiles to PDE4D7. PART1 had a significantly similar expression pattern to PDE4D7 across samples, correlating with disease progression. However, many other genes also exhibited matched expression to PDE4D7, including miRNAs and lncRNAs. Conclusion: These novel PDE4D7-associated genes, many of which are previously uncharacterized in cancer, represent putative PCa biomarkers and could have mechanistic roles in PCa progression.

Plain Language Summary

Identification of biological molecules that can indicate the presence of a disease are known as biomarkers and are valuable in medical research. For prostate cancer, the PDE4D7 gene is helpful for determining how the disease is getting worse. To understand more, this study explored whether genes located near PDE4D7 are also connected to prostate cancer. The analysis of tumor biopsies revealed that many of these nearby genes exhibited similar expression levels to PDE4D7, indicating their association with disease development. These findings suggest that several of these nearby genes could serve as potential biomarkers for prostate cancer by giving important clues to understand and detect the disease better.

Keywords:

• biomarkers
• cancer
• diagnostic
• genomics
• phosphodiesterase
• prognostic
• prostate
• stratification

Given the substantial heterogeneity in prostate cancer (PCa) treatment responses to standard therapies vary, therefore a prognostic biomarker which could accurately indicate disease progression would influence personalized treatment decisions and be beneficial in identifying novel therapeutic targets [1]. Dysregulated cyclic AMP (cAMP) signaling is associated with many diseases including cancer, and is particularly implicated in the progression from localised androgen-sensitive (AS) PCa to aggressive castration-resistant PCa (CRPC) [2]. Viable cAMP signalling depends on the compartmentalization of signaling intermediates such as cAMP effector proteins and phosphodiesterases (PDEs) [3]. The PDE4D sub-family is germane in PCa, with the long isoform PDE4D7 being of unique importance [4]. Previously we identified that PDE4D7 expression inversely correlates with disease progression, with significant downregulation of the isoform observed between localised PCa and metastatic CRPC samples [1,4]. Diminished PDE4D7 transcript level also correlated with an increased risk of post-surgical biochemical relapse in a patient cohort, identifying its value in risk stratification [5,6]. Additionally, there is a positive association between PDE4D7 expression and presence of the TMPRSS2 and ETS transcription factor family member ERG (TMPRSS2-ERG) gene rearrangement which has been linked to PCa development, leading to the proposition of a combined ‘CAPRA & PDE4D7’ score [5,7]. Interestingly, combining analysis of other long PDE4D isoforms, PDE4D5 and PDE4D9, can be used to further improve the prognostic power of the combination risk model [8]. These studies have highlighted the potential of PDE4D7 not only as a prognostic biomarker in patient risk stratification to distinguish between insignificant and aggressive PCa tumors, but also the potential of pharmacological enhancement of PDE4D7 activity, representing novel therapeutic avenues for aggressive PCa.

A study by Wedge et al. (2018) investigating genomic rearrangements via whole genome sequencing of >100 PCa tissue samples identified losses at chr5:60–100 Mb in ETS-negative tumors, as well as homozygous deletions within chr5:55–59 Mb (region encoding PDE4D) in ETS-positive tumors [9]. Given that exons 1–3 of PDE4D7 are located between the region of chr5:59–60 Mb, presence of the TMPRSS2-ERG gene fusion may influence the variation observed in PDE4D isoform expression [6,8]. Given this, and the evident involvement of the PDE4D gene in PCa development and progression, we sought out to investigate the genomic profile of the PDE4D coding region on chromosome 5 in PCa tissue samples to ascertain differences in other genes within this area in correlation with PDE4D7 expression.

Methods

Two independent NGS transcriptomics expression data sets (n = 533 and n = 151) from PCa patient biopsy punch samples of surgically resected tumor tissue were used in this study. The local Institutional Review Boards approved the collection of patient tissue for clinical research. RNA sequencing of tissue extracted RNA and data processing was performed as previously described [8]. Generation of normalized PDE4D transcript expression was performed via RT-qPCR, by subtracting the Cq of respective transcripts from the averaged reference gene Cq, then transformed to their respective transcript scores. To investigate the expression of genomic elements on chromosome 5 in the region upstream and downstream of where exons 1–3 of PDE4D7 are located (approximately chr5:59,000,000 to chr5:60,500,000), we created a heatmap of the expression of respective transcripts together with the level of PDE4D5, PDE4D7 and PDE4D9 expression. PDE4D5, PDE4D7 and PDE4D9 scores, as well as the transcript per million expression values for the respective transcripts, were transferred into z-scores across all samples. As a result, all genes of the heatmap have a mean expression of 0 and a SD of 1. These z-score transformed expression values were used as input for the expression heatmaps.

Results

Our analysis of the data suggests a distinct correlation between PDE4D7 expression and the expression of a subset of lincRNAs, miRNAs and pseudogenes located between chr5:60,000,000 to chr5:60,500,000 [Figure 1B]. Importantly, all trends observed are correlated between both independent datasets [Figures 1A (i) & (ii)].

Figure 1. Genome expression on chromosome 5 within PDE4D coding region in prostate cancer patients in comparison to the PDE4D7 score.

(A) Heat map of gene expression in PCa patient samples from datasets 1 n = 533 (i) and 2 n = 151 (ii) ordered according to their PDE4D7 scores from high to low. Gene expression values of the shown transcripts located in the genomic region of the PDE4D gene on chromosome 5 were determined by NGS RNA sequencing. Transcript names highlighted in yellow show p-value <0.05 in expression difference between samples with high versus low PDE4D7 score. (B) Chromosome 5 genome alignment in PDE4D coding region between approx. 59,000,000–60,500,000 bp from UCSC genome browser (GRCh38/hg38 Human Dec 2013 assembly), accessed on 14 April 2022. PDE4D7 isoform annotated by arrow and orange text box. (C) Heat map of gene expression in PCa patient samples from datasets 1 (n = 533) ordered according to their TMPRSS2-ERG (T2-ERG) fusion status.

Specifically, RNA transcripts located within the first 3 exons of PDE4D7 are coincidentally downregulated, with AC034234.1, RNU6-806P, AC109486.1, SETP21, PART1 and the entire PDE4D gene exhibiting a similar overall expression pattern to PDE4D7 among samples. These genes showed a p-value < 0.05 in difference of expression between samples with high versus low PDE4D7 scores [Figures 1A (i) & (ii)], alongside AC092343.1, AC008833.1, NDUFB4P2 and MIR582. While AC109486.1 exhibits the similarity in expression, AC109486.2 and AC109486.3 show no expression in any sample, alongside AC016591.1. Interestingly, all samples with no expression are found in the region overlapping PART1 which exhibits similar expression to PDE4D7, revealing that this entire region is not deleted.

CAB39P1 showed mostly lack of expression across all samples, however a subset of samples exhibited increased expression irrelevant to PDE4D7 level. AC016642.1, RAB3C, RPL5P15 and AC008852.1 revealed similar lack of expression to CAB39P1 however with more anomalous samples showing altered expression. The most varied expression profiles were DEPDC1B, KRT8P31 and ELOVL7, as well as PDE4D5 and PDE4D9, reflecting no obvious correlation to PDE4D7. Of note is that TMPRSS2-ERG fusion negative (TMPRSS2-ERG-) samples show more pronounced downregulation of genomic elements than TMPRSS2-ERG fusion positive (TMPRSS2-ERG+) samples, similar as observed relative to PDE4D7 score [Figure 1C].

Discussion

Our data shows that PDE4D is the only protein coding gene exhibiting an identical expression pattern to PDE4D7. Böttcher et al. (2016) noted that long PDE4D isoforms are significantly downregulated during PCa progression. However, while PDE4D5 and PDE4D9 are known to be downregulated in both localized AS PCa and CRPC tissue, PDE4D7 is only diminished in CRPC, rendering it a more specific measure of disease progression [6].

Other RNAs mapped to the PDE4D coding region exhibiting PDE4D7-like downregulation are pseudogenes (SETP21, RNU6-806P), nucleotide sequences (AC034234.1 and AC109486.1) and long non-coding (lnc) RNAs (PART1). Aside from PART1, no function has yet been ascribed to these genes, however PART1 has been extensively researched with regards to cancer development and progression. PART1 is predominantly expressed in the prostate, relies on androgens for transcriptional regulation in PCa and is upregulated in PCa tissue [10,11]. Interestingly, while PART1 is regulated via androgen signaling, PDE4D7 transcription is not, despite the PDE4D7 5′UTR coding region overlapping with antisense PART1 [4]. Henderson et al. (2014) reported that PART1 and PDE4D7 exhibit positively correlated mRNA expression within PCa cell lines and xenografts, which our results further confirm [4]. However, we are the first to report the similarity in expression among various genes located within the PDE4D7 coding region of Chr5q12. Given the lack of functional characterization of many of these genes, and their correlation with PDE4D7 expression in PCa, further research is warranted to determine their functional significance in PCa progression.

miR-582, which is implicated in PCa metastasis, correlated with PDE4D7 expression however the association was less robust than that of PART1. Similar to PDE4D7, downregulated expression of miR-582-3p and miR-582-5p is associated with an advanced PCa phenotype, with diminished expression observed in bone metastatic PCa tissue, correlating with reduced bone metastasis-free survival [12]. It is notable that although miR-582 expression correlated with PDE4D7 pattern, there is high variability among samples, with many exhibiting no expression at all.

Correlation between PDE4D7 level and TMPRSS2-ERG fusion status has previously been identified, with upregulated PDE4D7 expression in positive TMPRSS2-ERG tumors (TMPRSS2-ERG+) and low-grade PCa phenotype [7]. This gene rearrangement is evident in approximately 50% PCa patients, leading to overexpression of the oncogene ERG and subsequently enhanced proliferation and many hallmarks of cancer [13,14]. Given that the PDE4D7 gene contains an ERG binding site, it was speculated that PDE4D7 may be regulated via ERG transcription [7]. Our data further supports this theory, with negative TMPRSS2-ERG tumor samples (TMPRSS2-ERG-) exhibiting downregulation of genes akin to those identified upon stratification by PDE4D7 score.

Conclusion

In summation, our data identifies the downregulation of an entire chromosomal region of Chr5q12 mapping to the PDE4D7 domain throughout PCa progression. This pinpoints not only driver genes but also passenger genes which may be a fundamental feature of carcinogenesis. PDE4D7 and PART1 have previously been proposed as separate biomarkers for PCa [1,11], however measuring these two genes in combination, or alongside others identified here with correlative expression, could further enhance the specificity of these biomarkers in prognostic approaches. In this era of precision medicine, biomarkers that can be measured to reflect an accurate state of disease, or are implicated in disease pathology, are vital for improving diagnostics and therapeutics.

Summary points

• PDE4D7 represents a promising biomarker for PCa progression, with inverse expression correlating with disease aggression.

• Next generation RNA sequencing identified many additional genetic sequences located in the PDE4D coding region with similar expression patterns to PDE4D7 in PCa tissue samples.

• Many of the biological functions of the PDE4D7-associated genes are currently uncharacterized, with these findings suggesting their involvement in PCa.

• These newly identified genes with correlating expression to PDE4D7 could be used as biomarkers to enhance prognostics when measured alongside PDE4D7.

Author contributions

C Gulliver, R Hoffmann analyzed the data. C Gulliver, GS Baillie, R Hoffmann wrote and reviewed the manuscript.

Ethical conduct of research

The authors state that they have obtained appropriate institutional review board approval or have followed the principles outlined in the Declaration of Helsinki for all human or animal experimental investigations.

Acknowledgments

We acknowledge support provided by the University of Glasgow MVLS Doctoral Training Programme award to C Gulliver.

Financial & competing interests disclosure

This project was supported by the framework of CTMM (The Center for Translational Molecular Medicine; The Netherlands), PCMM project (grant no. 03O-203). R Hoffmann is an employee of Philips Research. R Hoffmann and GS Baillie hold patent rights relevant to the published work. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.