Serine and one-carbon metabolisms bring new therapeutic venues in prostate cancer

Serine and one-carbon unit metabolisms are essential biochemical pathways implicated in fundamental cellular functions such as proliferation, biosynthesis of important anabolic precursors and in general for the availability of methyl groups. These two distinct but interacting pathways are now becoming crucial in cancer, the de novo cytosolic serine pathway and the mitochondrial one-carbon metabolism. Apart from their role in physiological conditions, such as epithelial proliferation, the serine metabolism alterations are associated to several highly neoplastic proliferative pathologies. Accordingly, prostate cancer shows a deep rearrangement of its metabolism, driven by the dependency from the androgenic stimulus. Several new experimental evidence describes the role of a few of the enzymes involved in the serine metabolism in prostate cancer pathogenesis. The aim of this study is to analyze gene and protein expression data publicly available from large cancer specimens dataset, in order to further dissect the potential role of the abovementioned metabolism in the complex reshaping of the anabolic environment in this kind of neoplasm. The data suggest a potential role as biomarkers as well as in cancer therapy for the genes (and enzymes) belonging to the one-carbon metabolism in the context of prostatic cancer.

In some lung, ovarian and breast cancers SHMT1 acts as an oncoprotein, promoting tumor progression. The role of SHMT1 in hepatocellular carcinomas is still unclear, although some recent studies revealed a significantly decreased expression of SHMT1. This reduction is possibly correlated to adverse clinical features and poor prognosis of these cancer patients.
In several types of cancer an accentuated deregulation of SHMT2 is observed and is often associated with tumor progression. For example, in vitro studies of hepatocellular carcinoma cells show that inhibition of SHMT2 significantly decreases tumor progression while its overexpression is insufficient to induce tumor transformation, thus suggesting that SHMT2 is involved in tumorigenesis but it is not able to start malignant transformation alone. Anyhow, both the enzymes are somehow de-regulated by many modalities in the context of cancer [43][44][45].

Serine metabolism in prostate cancer
In the pathogenesis of prostate cancer, the initial pathogenetic event might be considered the dependency from the androgen stimulation [46][47][48][49][50], but many more molecular alterations have been found associated to its progression [51][52][53][54][55] as well as to the development of specific clinical features of the disease such as bone metastasis [56] or studied as potential therapeutic targets [57]. The microbiome has also been shown to impact on the pathogenesis of this kind of cancer [58][59][60][61][62], also due to development of algorithms allowing to interpret its role in the context of a disease [63][64][65][66][67]. Anyhow, specific attention is focused on epigenetic remodeling. DNA methylation has been associated to prostate cancer pathogenesis under different aspects such as in response to radiotherapy [68] or in the tracking of circulating tumor DNA [69] where androgen receptor sequences are hypomethylated in the blood stream and correlate with a poorer prognosis, suggesting also a possible role in liquid biopsies [70][71][72][73][74][75] and the overcome of the classical histology based biopsies [76,77]. The methylation of the DNA strongly depends on the availability of methyl groups, by the activity of the enzymes involved in the 1C unit metabolism. An alteration of this metabolic pathway has therefore been explored, beneath partially, in the context of prostatic cancer.
The enzymes involved in the one-carbon metabolism have been evaluated in a prostate cancer population through genomic sequencing defining at least two methionine synthase gene (MTR) polymorphisms (rs2837281 T>G and rs1131450 G>A) associated to a poorer prognosis in the Han Chinese population [78]. Moreover, folate depletion has been shown to induce transcriptional reprogramming and to reshape DNA damage response and the DNA methylation pattern in a human xenograft model of prostate cancer following androgen withdrawal [79][80][81][82].
Anyhow, the one-carbon metabolism has also been implied in the pathogenesis of the highly resistant to treatment neuroendocrine histotype of prostatic cancer (NEPC) [83][84][85][86] since the de-regulation of protein kinase C (characteristic in NEPC) upregulates serine biosynthesis and SAM intracellular levels which allow the epigenetic reshaping of this peculiar type of cancer [87][88][89].
Prostate cancer has also been shown to depend on the activity of the methyladenosine phosphorylase enzyme (MTAP), the rate-limiting enzyme of the methionine cycle, which is used by prostatic cells to overcome the loss of carbon units in the secretion of polyamines. The dual inhibition on polyamines secretion and of MTAP has proved to induce apoptosis in ex-vivo explants [90].
Finally, the central component of the mitochondrial one-carbon metabolism MTHFD2 is upregulated in prostate cancer from at least two transcription factors ATF4 and c-Myc, which have already been implicated in this disease and its silencing inhibits prostate cancer cells growth as well as prostatosphere formation [91,92].
Consistently with this preliminary evidence of an alteration of some of the one-carbon metabolism component in prostate cancer at different stages of the disease, we performed a bioinformatic analysis on publicly available gene expression and clinical data from the TCGA (Firehose Legacy) [93][94][95], comprehensive of 499 samples of prostate cancer specimens. The database has been explored using the cBioportal online platform [96,97] and the GEPIA web server [98].

3
(except for a few cases in which the gene is deleted). Dehydrofolate reductase (DHFR) [100] is the second most altered gene in the 9% of all cases. Differently form PSPH the DHFR gene does not have a clear pattern of alteration, showing gene overexpression in almost half of the cases in which the alteration is present, followed by downregulation. SHMT1 and SHMT2 [25,101,102] enzymes show altered expression in 5% of the cases, comparably to the other genes which are essentially involved in the folate cycle. From this analysis it is evident that in prostate cancer samples in which there is an alteration in one of the genes related to the one-carbon metabolism, it is quite probable that other genes from the pathway are altered so that it is unlikely that a single gene is the only one whose expression is perturbed. This might suggest that the reshape of the serine metabolism in prostate cancer is a pervasive metabolic event that is in place to adapt the transformed cell to gain selective advantage.
As a matter of fact, all together these genes are altered in nearly 40% of cancer specimens and most of the alterations of the genes involved in serine and glycine metabolism are due to increased expression of the relative mRNA rather that due to mutation. This evidence might still suggest that the serine and glycine metabolism is more an adaptive event, rather than a driver genetic aberration of the cancer itself (Fig. 2B). As shown on Fig. 2A, there is a tendency to find more than one gene from the one-carbon metabolism altered in a single prostate cancer. This evidence can be further dissected showing that many genes have a statistically significative co-occurrence trend: methylene-tetrahydro-folate reductase MTHFR) is altered in samples in which also the MTHFD2 gene is altered (p < 0.001, q = 0.003), while PSPH overexpression is associated to MTHFD1 and DHFR (p < 0.001, q = 0.003 for both), which is also the case for PSAT1 and MTHFD2L (p < 0.001, q = 0.003), or with MTHFD1 (p < 0.001, q = 0.004) (Fig. 2C).
Consistently with the role played by the one-carbon metabolism in providing carbon units which are partially used to synthesize S-adenosyl-methionine, and its effect on the epigenetic remodeling of the chromatin, it is quite interesting to observe that in the serine metabolism altered group of patients (comprising all the patients showing one alteration in at least one of the genes shown in Fig. 1A) there is a relatively higher alteration (considering any kind of event, form mRNA over-or down-expression, to mutations, etc.) in many genes which are considered cancer drivers in prostate cancer, such as obscurin (OBSCN) [37], titin (TTN) [103], the ETS transcription factor (ERG) [104] and many mucins (MUC5B, MUC6, MUC2) [105,106] (Fig. 2D). Alteration of the serine metabolism might therefore show promising therapeutic approaches if its role in epigenetic remodeling of prostate cancer is confirmed.
Serine metabolism reshaping in prostate cancer might also plays a rather pervasive rearrangement of cellular metabolism and expression profile and might be associated to changes of the expression of many more genes which have been clearly associated to tumor genesis and progression such as the apoptotic protein BCL2L11 [107][108][109][110][111] (more expressed in the group of patients harboring a perturbation of the serine metabolism, p < 0.001 and q < 0.001) and well as the chromatin remodeler BAP1 [112] (less expressed in the altered groups, p < 0.001 q = 0.003) (Fig. 2E).

Mutational status and copy number alterations of one-carbon metabolism enzymes in prostate cancer patients
As said initially, genes in the serine metabolism are rarely mutated in prostate cancer. The most reported mutations are for the PSPH gene, with a mutation in its HAD domain, in DHFR, in PSAT1 and in the FTHFS domain of the MTHFD1 gene (Fig. 3A).
Since the genes involved in the serine metabolism are mostly subject to expression alteration, an analysis on the putative copy-number alteration form GISTIC shows a relatively increasing cumulative alteration in cancer specimens with higher Gleason scores (> 7), except for Gleason 10 cases which are just less abundant among all the cases and show quite a substantial number of samples harboring copy-number modifications (Fig. 3B).
Some of the genes involved in the serine metabolism are differentially expressed between prostate cancer tissue and the surrounding normal prostate epithelium as in the case of PSAT1, overexpressed in the tumor (p < 0.005) (Fig. 4A). PSAT1 expression also shows a strong direct positive correlation with the expression of MTHFD2 gene with Spearman coefficient = 0.48 (p < 0.001) and Pearson coefficient = 0.54 (p < 0.001) (Fig. 4B).

One-carbon and serine metabolism impact on prostate cancer progression
The progression of prostate cancer involves several molecular mechanisms, underlying a complex regulation of autophagy [113][114][115][116][117], micro-RNA [118][119][120][121][122], p53 [123][124][125][126], Bcl-2 [127,128] family and interleukins [31,32,129,130] able to also impinge on the above-mentioned metabolisms. Therefore, the analysis followed considering correlation to clinical data, the analysis of the disease-free survival (DFS) of patients stratified according to the expression levels of some of the key genes involved in the serine metabolisms fails in showing statistically significant differences among the groups for PSAT1 (Fig. 4C) as well for SHMT1, SHMT2 and MTH-FD1L where, anyhow, high expression shows a tendency to a worse disease course with higher number of progressions (Figs. 4D-F). Anyhow, grouping all the possible alterations in the genes involved in the serine metabolism stratifying patients between the altered and the unaltered group, there is a strong difference in terms of progression free survival with a better prognosis for the unaltered group (p < 0.001) (Fig. 4G), this might be regarded as the development of a potential prediction algorithm as already done in many other clinical contexts [131][132][133][134].
Apart from a possible role of the enzymes involved in the 1C metabolism as potential biomarker of progression in prostate cancer (specifically as a molecular signature) this metabolism might also represent a therapeutic option to be explored in this clinical context. The overall management of this cancer depends on a combined approach relying on surgery, radiotherapy, hormonal therapy or chemotherapy, depending on the status of the disease [135]. When 1 3 chemotherapy is necessary, often in the case of tumors which show resistance to androgen deprivation therapy or inhibition of androgen synthesis, the use of microtubules synthesis inhibitors is the standard of care with drugs such as docetaxel or cabazitaxel [136]. Anyhow, resistance to these treatments, as well as resistance to targeted therapies (PARP inhibitors) can occur and potential alternative strategies are required. The inhibition of one-carbon metabolism, with an anti-folate therapy, might represent a significative option to be explored in a clinical context, and some data on the use of methotrexate, a known anti-folate drug, show efficacy in reducing prostate cancer progression in combination with luteinizing hormone releasing hormone, in cell lines end mouse xenograft models [137,138]. The effect of the inhibition of the one carbon metabolism, should anyway be regarded with caution, since its impact on the immune system, inducing immunosuppression, might play a contradictory result promoting cancer progression [139].

Conclusion
Despite being and extremely preliminary evaluation of the possible role of serine and one-carbon metabolism in prostate cancer, the present review shows that a pervasive alteration of these biochemical pathways is abundantly evoked during prostatic cancer pathogenesis and evolution. Anyhow, the analysis is quite partial due to the heterogeneity of the cohort of patients and the evaluation of the expression of genes at an RNA level rather than at a protein level. To address this specific issue, a multi-omics approach [140] might be beneficial for the dissection of complex metabolic scenarios like the one described. Of course, the generation of omics data and further analysis should rely on the developing machine learning algorithms [141,142], allowing to detect molecular signatures among cancer specimens through the analysis of big data repositories. Moreover, alteration at the levels of such a complex metabolism might also be exploited in order to develop biomarker for prediction or monitoring of the progression of the disease [77,143,144], as it is happening in the context of other types of cancer [145,146].

Competing interests
The authors declare no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.