Abstract

   Prostate cancer (PCa) is the most common malignancy in men in developed
   countries. Prostate-specific antigen (PSA) remains the most widely used
   serum marker for prostate cancer. Here, we reported that the expression
   of phosphoglucomutase-like protein 5 (PGM5) is significantly lower in
   prostate cancer tissue. The low expression of PGM5 and its related gene
   signature were found to be linked to poor clinical outcome and high
   Gleason score. In vitro assays showed that overexpression of PGM5
   significantly repressed proliferation and migration of prostate cancer
   cells. GO and pathway analyses showed the enrichment of genes in
   regulation of cell growth and migration, and pathways related in
   cancer. Our additional results showed that the downregulation of PGM5
   is closely related to DNA methylation. Taken together, our findings
   provide the first evidence that PGM5 expression is associated with
   prostate cancer progression. These results also highlight a preclinical
   rationale that PGM5 represents a prognostic marker and a promising
   target for new therapeutic strategies in prostate cancer.

Supplementary Information

   The online version contains supplementary material available at
   10.1007/s12672-022-00525-x.

   Keywords: Prostate cancer, PGM5, Prognostic marker, Bioinformatics
   analysis

Introduction

   In developed countries, prostate cancer is still the leading cause of
   cancer-related mortality in men [[36]1]. Prostate cancer has been
   recognized as a common diagnostic disease due to the development of
   diagnostic methods, especially the introduction of PSA testing [[37]2,
   [38]3]. However, PSA test often causes misdiagnosis or overdiagnosis,
   indicating that its accuracy in predicting the pathological and
   clinical outcome remains limited [[39]4]. Moreover, most of the
   patients have indolent tumors that can be safely followed without
   treatment, while others with an aggressive disease course need
   immediate therapeutic intervention [[40]5–[41]7]. It is still difficult
   for urologists to distinguish such patients in clinic. Therefore, there
   is still an acute need for new biomarkers to help urologists improve
   the detection and prediction of prostate cancer.

   Phosphoglucomutase-like protein 5 (also known as aciculin) is an enzyme
   encoded by the PGM5 gene. Gene functional studies show that PGM5 is
   similar to PGM1 but lacks enzymatic activity [[42]8]. PGM5 is tightly
   associated with the actin cytoskeleton and has primarily been
   investigated as an adhesion protein. It also functions as a
   cytoskeletal component of cell–matrix and cell–cell contacts in muscle
   and non-muscle cells [[43]8–[44]10]. Recent studies showed that PGM5
   has fundamental additional functions in sarcomeric development,
   stability and remodeling [[45]11]. The expression and function of PGM5
   in cancer, especially in prostate cancer is still largely unknown.

   In the present study, we did in silico analyses based on the RNA
   sequencing (RNA-Seq) data from various authoritative database online to
   identify some new potential diagnostic and prognostic markers in
   prostate cancer. We finally found in multiple databases that PGM5 is a
   gene that is down-regulated with prostate cancer development. We
   further used immunohistochemistry to evaluate the expression of PGM5
   protein in a wide spectrum of prostate tissues to determine its
   expression in situ, and evaluated the prognostic value of PGM5 in
   prostate cancer.

Materials and methods

Microarray and RNA-Seq data Information

   We obtained prostate cancer and normal tissue gene expression profiles
   of [46]GSE35988, [47]GSE21032, GDS2546 and GDS1439 from the NCBI-GEO
   microarray database. We also downloaded the RNA-Seq dataset of prostate
   cancer with corresponding clinical profiles from The Cancer Genome
   Atlas (TCGA) database. The corresponding information related to
   patients, including copy number and PGM5 methylation, was obtained from
   the cBioPortal for Cancer Genomics website
   ([48]http://www.cbioportal.org/index.do) [[49]12]. We also digged out
   the relationships between PGM5 TCGA expression, DNA methylation and
   clinical data directly through MEXPRESS website
   ([50]https://mexpress.be/) (MEXPRESS: visualizing expression, DNA
   methylation and clinical TCGA data).

Publically available gene expression data sets

   In this article, we used RNA and protein expression data of PGM5 in
   different human tissues and prostate cancer tissues through The Human
   Protein Atlas portal ([51]http://www.proteinatlas.org/) [[52]13]. All
   data are available directly online.

Patient information

   The cancer and para-cancer tissue samples from a group of 71 patients
   with prostate cancer were collected in this study. All the tissues were
   fixed in formalin after surgical removal. The Ethics Committee of
   Nanjing Medical University approved the research and written informed
   consents were obtained from the participants. Patients with a previous
   history of malignant tumors were excluded from this study. We showed
   the clinicopathological information of the tissues collected in
   Additional file [53]1: Table S1.

Tissue microarrays (TMAs) and immunohistochemistry (IHC)

   We collected the prostate cancer and para-cancer tissues to construct
   TMAs as previously described [[54]14]. IHC was performed as described
   elsewhere [[55]15, [56]16]. In brief, slides were deparaffinized and
   heated in citrate buffer pH6 for antigenic retrieval. The primary
   antibody was PGM5 (Novus). Immunohistochemistry was performed using the
   streptavidin–biotin-peroxidase method with diaminobenzidine as the
   chromogen (KitLSAB, Dakocytomotion, Glostrup, Denmark). Negative
   controls were obtained after omission of the primary antibody or
   incubation with an irrelevant antibody.

   PGM5 staining was simultaneously examined by two independent observers
   (including one pathologist), and a consensus score was reached for each
   core. A positive reaction for PGM5 was scored in four grade categories
   depending on the intensity of the staining, i.e., 0, 1, 2 and 3, and
   the percentage of PGM5-positive cells was also scored in four groups: 0
   (0%), 1 (1 to 33%), 2 (34 to 66%) and 3 (67 to 100%). In cases with
   discrepancies between duplicated cores, the higher score of the two
   tissues was taken as the final score. The sum of the intensity and
   percentage scores was used as the final staining score. The staining
   pattern was defined as follows: 0, negative; 1 to 2, weak; 3 to 4,
   moderate; and 5 to 6, strong [[57]17].

Gene set enrichment analysis (GSEA)

   We analyzed differences in biological function annotations of gene mRNA
   expression levels between PGM5 up- and down-regulated patients by GSEA
   v3.0. The results indicated the effect of PGM5 expression on various
   biological function gene sets in prostate cancer patients. The number
   of permutations was set at 10. In this analysis, we used the following
   values as cutoff: nominal P-value cutoff of 0.05; false discovery rate
   (FDR q-val) < 0.25.

Cell culture

   We obtained LNCaP, PC-3 and DU145 cells from ATCC (Bethesda, USA). All
   cells were maintained in RPMI-1640 supplemented with 10% FBS and
   antibiotics (0.1 mg/ml streptomycin and 100 units/ml penicillin) as
   previous [[58]18]. All cell lines used in our study were authenticated
   by STR profiling and tested for mycoplasma contamination.

Constructs and construction of stable cell lines

   We used pPB-CAG-EBNXN vector (Sanger Institute) to deal with constructs
   and pPB-CAG-ires-Pac was generated as previously described [[59]18,
   [60]19]. We ligatied full length PGM5 into the multiple-cloning sites
   (MCS) of pPB-CAG-ires-Pac to generate pPB-CAG-PGM5-ires-Pac. Control
   and PGM5 overexpression stable cells were obtained as previously
   described and all stable cell lines were selected and identified by
   western blotting [[61]18].

Antibodies and immunoblotting

   In western blotting, cells were lysed in 1 × SDS loading buffer (50 mM
   Tris–HCl pH6.8, 10% glycerol, 2% SDS, 0.05% bromophenol blue and 1%
   2-mercaptoethanol). Antibodies were listed as follows: anti-PGM5
   antibody (NBP2-62654, Novus Biologicals) and anti-ACTIN (ab8227,
   Abcam).

   We did immunoblot as previously described [[62]18]. In brief, all
   proteins were separated by SDS–PAGE and were transferred to
   polyvinylidene difluoride membranes (Millipore). HRP-labeled secondary
   antibodies and enhanced chemiluminescence system was used for signal
   detection. Protein was visualized using ChemiDoc XRS chemiluminescence
   detection and imaging system (Bio-Rad Laboratories) or KODAK film
   machine.

Cell proliferation (MTS) assay

   Cell proliferation (MTS) assay was performed as described previously
   [[63]20]. In brief, all cells were seeded at 4000 cells/well (0.1 ml)
   in 96-well plates and were incubated overnight at 37 °C for 5–6 days.
   In each indicated time point, cells were incubated with 20 μl of
   CellTiter 96 AQueous One solution reagent MTS (Promega) in 100 μl of
   RPMI-1640 for one hour. Cell number was then estimated using a
   microtiter plate reader (Bio-tek).

Identification of differentially expressed genes (DEGs)

   EdgeR for examining differential expression of RNA-Seq count data as
   previously described [[64]21, [65]22]. DEGs were identified with the
   following criterion: |fold change (FC)|≥ 2; the P-value and FDR < 0.05.
   The DEGs were used for further bioinformatics analysis. A heat map of
   top 100 genes and volcano plot of the DEGs were drawn.

Cell migration assays

   We used Transwell (Corning) system in 24-well tissue culture plates to
   deal with cell migration assays as described previously [[66]18].

Functional annotation and pathway enrichment analysis

   We used the database for annotation, visualization and integrated
   discovery (DAVID) website to undergo GO analysis and KEGG analysis
   [[67]23], while specifying a P-value < 0.05 for statistical
   significance.

Online ChIP-seq analysis

   Online ChIP-seq data of different Histone methylation in prostate
   cancer cells were obtained by UCSC browser from Cistrome Data Browser
   online [[68]24].

Statistical analysis

   The results are presented as the mean values ± SEM. The association
   between PGM5 staining and the clinicopathologic parameters of the
   prostate cancer patients were evaluated by a χ2 test. Differences
   between groups were estimated using Student’s t test. For analysis of
   the association between PGM5 expression and biochemical recurrence, we
   used Kaplan–Meier method and log-rank test by Graphpad. Receiver
   operating characteristic (ROC) curve analysis was also dealt using
   Graphpad. Areas under the curves (AUCs) were used to evaluate the
   diagnostic value of each marker combination. An AUC greater than 0.9
   indicated excellent diagnostic efficacy. An AUC between 0.7 and 0.9
   indicated good diagnostic efficacy. FDR in edgeR and GSEA were adjusted
   for multiple testing with the Benjamini–Hochberg procedure to control
   FDR, respectively [[69]25, [70]26]. A value of P < 0.05 was considered
   statistically significant. All the statistical analyses were conducted
   with Graphpad and R 3.3.0.

Results

Identification of differentially expressed genes (DEGs)

   To identify the DEGs, we downloaded and analyzed the prostate cancer
   and normal tissue gene expression profiles of [71]GSE21032, GDS2546 and
   GDS1439 from the NCBI-GEO microarray database. Based on the in silico
   analysis, using P < 0.05 and fold change (FC) ≥ 2.0 or ≤ 0.5 as
   criteria. After integrated bioinformatical analysis, only one gene,
   PGM5, was identified from the three profile datasets above (Fig. [72]1
   A). We also showed the epxression patterns of PGM5 in the three
   datasets in Fig. [73]1 B-D, respectively. To confirm PGM5 expression,
   we analyzed PGM5 expression in another dataset, [74]GSE35988. The same
   result showed PGM5 downregulation in [75]GSE35988 (Fig. [76]1 E). To
   further investigate the mRNA expression of PGM5, we explored The Human
   Protein Atlas portal and found relatively high expression of PGM5 in
   normal prostate tissue (Fig. [77]1 F). Thus, our findings above
   indicated that PGM5 expressed much lower in prostate cancer tissue,
   which makes it a potential biomarker for prostate cancer development.

Fig. 1.

   [78]Fig. 1
   [79]Open in a new tab

   Associations between PGM5 expression and prostate cancer progress.A
   Identification of only one common DEGs from the 3 cohort profile data
   sets ([80]GSE21032, GDS1439, and GDS2546) using an online tool
   Calculate and draw custom Venn diagrams
   ([81]http://bioinformatics.psb.ugent.be/webtools/Venn/). B–D PGM5 mRNA
   level was downregulated in primary and metastatic prostate cancer in
   datasets GDS1439, GDS2546 and [82]GSE21032. E Data from [83]GSE35988
   confirmed the low PGM5 mRNA level in primary and metastatic prostate
   cancer. F Data from The Human Protein Atlas portal show PGM5 expression
   in different tissues

Clinical impact of PGM5 expression on prostate cancer progression

   We next determined whether PGM5 expression influences prostate cancer
   progression. We first investigated the relationship between PGM5
   expression and the clinic pathological characteristics in the TCGA
   prostate cancer dataset using cBioPortal for Cancer Genomics [[84]20].
   Results showed that the downregulation of PGM5 appeared to
   significantly correlate with higher Gleason score, but not the tumor
   stage (Fig. [85]2 A and Table [86]1). More importantly, we found that
   patients with a lower PGM5 level in tumor tissues showed much poorer
   prognosis (Fig. [87]2 B). All these results revealed that the low level
   of PGM5 positively correlates with advanced stage, increased
   metastasis, and poor prognosis in PCa patients.

Fig. 2.

   [88]Fig. 2
   [89]Open in a new tab

   PGM5 expression is associated with prostate cancer progress severity
   and prognosis. A Correlation between PGM5 expression and Gleason score.
   B Kaplan–Meier survival curves for prostate cancer patients stratified
   by PGM5 expression. C ROC curves of PGM5 expression in prostate cancer
   patients. D Representative immunohistochemistry of PGM5 on benign
   prostatic epithelia and prostate cancer tissues

Table 1.

   Clinical Characteristics of PCa Patients and PGM5 Expression in TCGA
   Characteristics Expression of PGM5 in PCa  P Value
                   Low High
   Age, years
     ≤ 65          232          119           0.3472
     > 65          86            54
   Gleason score
     < 7           22            23            0.009
    7              152           92
     > 7           144           58
   Tumor stage
     ≤ T2a          7            6            0.4279
    T2b             8            2
     ≥ T2c         299          162
   N stage
    N0             222          120           0.0796
    N1             58            19
   M stage
    M0             288          161           0.1961
    M1              3            0
   [90]Open in a new tab

   P < 0.05 was considered statistically significant. Chi-square test was
   used

   To analyze the value of PGM5 in the diagnosis and disease prognosis of
   prostate cancer, we performed ROC curve analysis for PGM5 in healthy
   individuals and in PCa patients, or in primary cancer and in metastatic
   cancer. Results in Fig. [91]2 C showed that in different data sets,
   PGM5 was useful for prostate cancer diagnosis in patients. Moreover,
   the expression level of PGM5 was also useful for distinguish metastatic
   patients from primary patients. All the AUCs were greater than 0.9.

   In addition to the RNA level, we also detected PGM5 expression in
   protein level. We examined PGM5 protein expression by IHC in prostate
   cancer and matched para-cancer tissues. As shown in Fig. [92]2 D,
   immunoreactivity of the PGM5 protein was observed in cell cytoplasm.
   After analyzing PGM5 expression pattern in prostate cancer tissues, we
   found that in cancer tissues, 4 cases were strong (5–6 points), 19
   cases were moderate (3–4 points), 37 cases were weak (1–2 points), and
   11 cases were negative (0 point), while in para-cancer tissues, we
   found that 33 cases were strong, 24 cases were moderate, 14 cases were
   weak, and 0 case was negative. Data from HPA showed the similar result
   (Additional file [93]2: Figure S1A). Our data indicated that prostate
   cancer tissues express much lower PGM5 in compared with benign tissues.

PGM5 could suppress prostate cancer proliferation and migration

   To further investigate the role of PGM5 in prostate cancer progression,
   we overexpressed PGM5 in LNCaP, DU145 and PC-3 prostate cancer cell
   lines (Additional file [94]2: Figure S1B). We determined the effect of
   PGM5 on prostate cancer cell proliferation and migration. In vitro
   assays showed that up-regulation of PGM5 repressed cell proliferation
   and migration (Figs. [95]3 A-D). These results above suggested the
   tumor suppressor role of PGM5, and also confirmed the prognostic role
   of PGM5 in prostate cancer.

Fig. 3.

   [96]Fig. 3
   [97]Open in a new tab

   PGM5 suppresses prostate cancer cell proliferation and migration. A
   Cell proliferation was measured at the indicated time points in LNCaP
   cells (** P < 0.01, *** P < 0.001, N = 3). B Cell proliferation was
   measured at the indicated time points in PC-3 cells (** P < 0.01, ***
   P < 0.001, N = 3). C Cell proliferation was measured at the indicated
   time points in DU145 cells (** P < 0.01, *** P < 0.001, N = 3). D
   Transwell assay analyses of the indicated cell lines (N = 3)

Functional enrichment analyses of PGM5 correlated genes

   To further explore the mechanisms of PGM5 in prostate cancer and find
   out some testable hypotheses, we did the correlation analysis using the
   datasets from NCBI-GEO above. Data of [98]GSE35988, [99]GSE21032 and
   GDS1439 were recruited for the analysis. Correlation coefficient was
   calculated and the cutoff was 0.4/-0.4. After integrated
   bioinformatical analysis, total of 934 consistently correlated
   expressed genes were identified from the three profile datasets,
   including 635 positively correlated genes and 299 negatively correlated
   genes (Fig. [100]4 A, B, Additional file [101]3: Table S2).

Fig. 4.

   [102]Fig. 4
   [103]Open in a new tab

   DAVID enrichment results of correlated genes of PGM5. A Bioinformatic
   analysis showed 635 positively correlated genes in [104]GSE21032,
   [105]GSE35988 and GDS1439. B Bioinformatic analysis showed 299
   negatively correlated genes in the three datasets above. C The GO
   enrichment terms of positively correlated genes. D The KEGG pathway
   analysis of positively correlated genes

   In order to analyze the correlated genes at the functional level, we
   submitted the two groups of genes separately online for further
   analyses with DAVID. GO and KEGG analyses showed enrichment of
   positively correlated genes, while negatively correlated genes showed
   no significant enrichment. We showed the results of GO analysis of
   positively correlated genes in Fig. [106]4 C. These results indicated
   that most of the positively correlated genes were significantly
   enriched in cell migration, proliferation and differentiation. We
   showed the results of signaling pathway enrichment in Fig. [107]4 D.
   KEGG pathway analysis showed that positive correlation genes had common
   pathways related in cancer development.

   To further investigate the effect of PGM5 downregulation on the
   development of prostate cancer, we analyzed the effects of PGM5
   downregulation on various gene sets by the GSEA. Both data from
   [108]GSE35988 and TCGA datasets were used to confirm the results. GSEA
   results of [109]GSE35988 and TCGA dataset were shown in Additional file
   [110]4: Figure S2, respectively. This suggested that PGM5
   downregulation may affect disease progression and prognosis in prostate
   cancer patients by influencing cell cycle, DNA repair and multiple
   oncologic pathways, which could be further explored and confirmed in
   future analysis.

PGM5 promoter methylation correlates with PGM5 downregulation

   Considering the important role of PGM5 downregulation in prostate
   cancer disease progression, we further analyzed the mechanisms of
   decreased PGM5 expression and conducted preliminary investigation. DNA
   methylation, especially promoter methylation, plays a key role in
   transcriptional regulation and always contributes to silencing gene
   expression. To test the hypothesis that the methylation of DNA could
   play a role in regulating PGM5 expression in prostate cancer, we
   explored the correlation between PGM5 methylation and PGM5 expression
   in TCGA dataset by cBioPortal for Cancer Genomics. Our database
   analysis showed that high PGM5 methylation level is correlated with low
   PGM5 expression level in prostate tumors and with poor prognosis in
   prostate cancer patients (Fig. [111]5 A, B). What’s more, we confirmed
   the result in another online data visualization tool, MEXPRESS
   ([112]https://mexpress.be). We found a much lower level of methylation
   in CpG island in normal tissues than that of prostate cancer tissues,
   which also suggests that DNA methylation is likely to be one of the
   important mechanisms for downregulation of PGM5 expression (Additional
   file [113]5: Figure S3A).

Fig. 5.

   [114]Fig. 5
   [115]Open in a new tab

   PGM5 mRNA expression correlated with PGM5 promoter methylation level. A
   PGM5 mRNA expression is inversely correlated with PGM5 methylation
   level in prostate samples from TCGA dataset. B Kaplan–Meier survival
   curves for prostate cancer patients stratified by PGM5 methylation
   level. C PGM5 mRNA expression is not correlated with PGM5 copy number
   in prostate samples from TCGA dataset

   Further more, we also detected other potential mechanisms for PGM5
   downregulation. We first examined whether DNA copy number alteration
   influenced PGM5 expression. Data from TCGA indicated that DNA copy
   number alteration did not corelate to PGM5 mRNA expression level, thus
   ruling out this possibility (Fig. [116]5 C). We next investigated
   whether histone methylation caused a decrease in PGM5 expression. To
   test this hypothesis, we explored the public ChIP-seq (chromatin
   immunoprecipitation assays with sequencing) data of histone methylation
   markers in prostate cancer cell lines [[117]36]. We noticed no binding
   events of either active or repressive histone methylation marks (e.g.
   H3K4me3 and H3K27me3) in the DNA region of PGM5 in LNCaP-abl or LNcaP
   cell lines (Additional file [118]5: Figure S3B), suggesting that
   histone methylation does not primarily affect the expression of PGM5.

Discussion

   In the present study, our results demonstrate that PGM5 expression is
   decreased in primary and metastatic prostate cancer by bioinformatic
   analyses and immunohistochemical assessment, which indicate a novel
   role for PGM5 as an important diagnostic and prognostic marker in
   prostate cancer.

   Current markers for diagnosis and prognosis of PCa, such as PSA test,
   tumor stage and Gleason score, have been used as the main methods to
   diagnose or predict prostate cancer in the past few decades [[119]27].
   However, such methods are limited in their accuracy and prognostic
   ability [[120]28]. A typical example is that the single test of PSA
   also caused a high rate of misdiagnosis or overdiagnosis [[121]4].
   Therefore, the development of additional prognostic markers is urgent.
   Our results provide a new sight into discovering a new diagnostic and
   prognostic marker in addition to PSA. We summarized the gene expression
   data of several authoritative prostate cancer clinical sequencing
   databases and found PGM5 downregulation in tumor tissues in all
   datasets. Further analysis indicated the tumor suppressor role of PGM5
   in prostate cancer, which was also confirmed by in vitro assays and IHC
   in tumor tissues from our patients.

   Radical prostatectomy is still the best treatment for clinically
   localized prostate cancer [[122]29]. Recurrence of the disease in men
   after radical prostatectomy suggests that undetected disease may have
   spread beyond the prostate gland before surgery [[123]30, [124]31].
   Increase the detection rate of this part of the patients will assist in
   designing a more effective therapeutic strategy, including aggressive
   treatment. In this study, our results indicate that patients with lower
   PGM5 expression are more prone to poor prognosis, and may benefit from
   early intervention.

   We also digged out the main biologic pathways implicated in this
   process. Correlation analysis was performed in several clinical
   datasets to find out the primary genes that were consistently expressed
   in association with PGM5 expression. Through the analysis of these
   positively/ negatively corelated genes, we found that several biologic
   processes related to oncogenesis were enriched. Signaling pathway
   analysis also showed enrichment in pathways that have been proven to be
   implicated in cancer development. In addition to the findings above,
   GSEA analysis also found that PGM5 also implicated in DNA repair.
   Previous researches have shown that metastatic castration-resistant
   prostate cancer have genomic aberrations that interfere with DNA repair
   [[125]32, [126]33]. Poly (adenosine diphosphate [ADP]–ribose)
   polymerase (PARP) is involved in multiple aspects of DNA repair, and
   PARP inhibition has been shown to exert antitumor activity in sporadic
   cases of metastatic, castration-resistant prostate cancer by dealing
   with DNA-repair defects in tumor cells [[127]34]. All these results
   above indicate many testable hypotheses that these biology processes
   and pathways might implicate in prostate cancer initiation and
   progression. PGM5 may influence cell proliferation, migration and
   differentiation through mutiple signaling pathways, thus promoting or
   inhibiting cancer progression, which needs further validation via
   molecular biologic experiments.

   As for how PGM5 downregulated in prostate cancer initiation and
   development, our results suggest DNA methylation might be the potential
   mechanism. DNA methylation is an important process needed for normal
   development. It has been quite clear that the methylation of gene
   promoters often leads to repression of transcription [[128]35]. In
   prostate cancer, critical oncogenes or tumor suppressor genes were
   methylated or de-methylated in cancer progression by mutiple
   mechanisms. Changes in DNA methylation influence gene expression and
   have been regarded as potential therapeutic targets [[129]36, [130]37].
   Methylation of PGM5 might be a potential prostate cancer research and
   drug target, which also needs further molecular biological experiments
   to validate.

   Our study contained several limitations. We utilized a series of online
   RNA-seq data to explore the role of PGM5 in prostate cancer by
   bioinformatics analysis. Most of the results are preliminary, such as
   the mechanisms of PGM5 downregulation in prostate cancer. In the
   following research, we will use molecular biological experiments to
   validate our results.

Conclusions

   In conclusion, we found the expression of PGM5 is significantly lower
   in prostate cancer tissue. Lower expression of PGM5 was found to be
   linked to high Gleason score and poor clinical outcome. Moreover,
   downregulation of PGM5 is closely related to DNA methylation. Taken
   together, our findings provide the first evidence that PGM5 expression
   is associated with prostate cancer progression, which suggest PGM5 as a
   promising prognostic marker and therapeutic target in prostate cancer.

Supplementary Information

   [131]12672_2022_525_MOESM1_ESM.xlsx^ (18KB, xlsx)

   Additional file 1: Table S1. Clinical characteristics of prostate
   cancer patients.
   [132]12672_2022_525_MOESM2_ESM.tif^ (1.6MB, tif)

   Additional file 2: Figure S1. PGM5 expression in HPA portal and
   overexpression effect in prostate cancer cells. A IHC data from The
   Human Protein Atlas portal show PGM5 expression is lower in prostate
   cancer tissues than in normal tissues. B Western blot analyses of PGM5
   expression in the indicated cell lines with overexpression.
   [133]12672_2022_525_MOESM3_ESM.xlsx^ (14.3KB, xlsx)

   Additional file 3: Table S2. Correlated genes of PGM5 identified from 3
   datasets.
   [134]12672_2022_525_MOESM4_ESM.tif^ (3.5MB, tif)

   Additional file 4: Figure S2. GSEA results of PGM5 expression in
   prostate cancer patients in GSE35988 (A) and TCGA (B).
   [135]12672_2022_525_MOESM5_ESM.tif^ (1,018.6KB, tif)

   Additional file 5: Figure S3. PGM5 promoter methylation level in online
   datasets. A PGM5 promoter methylation in MEXPRESS
   ([136]https://mexpress.be). B Online ChIP-seq data of present Histone
   methylation marks in prostate cancer cells.

Acknowledgements