The importance of global histone methylation levels has first been published by Seligson et al. and has not yet been investigated for H4K20 in the view of prostate carcinogenesis and its prognosis [13
]. Compared to normal prostate tissue, the present study displays an aberrant pattern of H4K20 modifications during PCA progression with a general hypomethylation of H4K20me1 and H4K20me2 in mPCA and CRPC.
While H4K20me2 is shown to sufficiently differentiate between different stages of PCA except for normal prostate tissue versus localized PCA, H4K20me1 sufficiently differentiates CRPC from other stages of PCA development. These findings suggest that modifications in histone methylation may not be applied primarily as tumour markers, rather than help to distinguish the dynamic process of cancer progress and risk assessment. Analogous findings have been reported by Schneider et al. [9
] regarding the carcinogenesis of bladder cancer, with decreasing levels of H4K20me1-3 from normal urothelium over non-muscle invasive bladder cancer, muscle-invasive bladder cancer to metastatic bladder cancer. The authors concluded that H4K20me1-3 levels help identifying patients with poor prognosis after radical cystectomy
Recent studies indicate H4K20me3 to be strongly decreased during disease progression of squamous cell cancer of the lungs and, thus being an important marker for therapeutic approaches [14
]. To date, it remains unanswered whether hypomethylation is cause or consequence of carcinogenesis, although recent studies suggest that dynamics in H4K20 methylation contribute to deregulations in cell cycle control and oncogenic transformation [15
] which is conducive to correlations of H4K20 methylation levels with adverse clinical-pathological parameters: Interestingly, H4K20me1 was positively correlated to lymph node metastases (p
= 0.002) which is consistent with its expression profile in mPCA. Similar results are demonstrated for H4K20me2, which shows a positive correlation to the categorized Gleason sum.
Our regression analysis did not allow the prediction of PSA recurrence after radical prostatectomy via H4K20 methylation levels. In contrast, Zhou et al. reported that H4K20me3 levels, accompanied by preoperative PSA levels, allowed risk stratification for PSA recurrence after radical prostatectomy [16
]. It seems that other histone modifications are more important prognostic markers in patients with PCA: H3K4me1 [12
] and H3K4me2 [10
] were significant predictors of PSA recurrence following radical prostatectomy. In comparison, H3K27 methylation is correlated with many adverse clinical-pathological parameters (i.e. Gleason Score, pathological stage) but not with PSA recurrence [17
]. It has to be noted that follow-up information was only available for 77 patients in our study, and also established predictive parameters like pathological stage/grade failed to correlate with PCA recurrence indicating insufficient statistical power.
Furthermore, Ellinger et al. were able to show the acteylated Histone H3 (H3Ac) and di-methylated H3 lysine 9 (H3K9me2) to be valuable discriminators between non-malignant prostate tissue and PCA with a high sensitivity (> 78%) and specificity (> 91%) [12
]. Although recent studies suggest the PSA value and the PSADT to be associated with a risk development of metastatic disease in CRPC patients [18
] there is no clear surrogate so far which can determine disease progression or progression after first-line chemotherapy in these patients.
Though, H4K20me3 does not allow significant differentiations between various stages of PCA it shows strong methylation levels in mPCA and CRPC, compared to H4K20me1/2 (see Figure ). These observations might accompany recent findings by Vertino et al. [19
], who describe the tri-methylation of H4K20 via SUV420H2 to play a crucial role in silencing tumor suppressor genes, such as TMS1 (target of methylation-induced silencing), which encodes for caspase recruitment domains, contributing to the promotion of apoptosis in PCA cells [20
] and has been reported to be prognostic for PCA carcinogenesis [21
]. Although this study does not address effects of H4K20 methylation changes on gene expression, it offers perspectives for further investigations on which mechanisms contribute to hypomethylated states in different stages of carcinogenesis. To date there are no demethylases detected [22
] which are accountable for the significant reduction of H4K20me1/2 methylated states in mPCA and CRPC. The expression level of H4K20 methylases is yet no explored in PCA; however, loss of the H4K20 methylase Suv4-20h2 were associated with lower levels of H4K20me3 in a rodent-model of hepatocarcinogenesis [7
] and breast cancer cell lines [8