Both dietary and environmental factors have been shown to induce epigenetic changes, i.e
. mechanisms that permit stable transmission of cellular traits without an alteration in DNA sequence or amount. Epigenetic deregulation frequently participates in tumorigenesis through inactivation of tumor suppressor genes.15
Prostate cancer tends to be multifocal, suggesting a field effect that might be influenced by epigenetic changes due to diet and oxidative stress.42
Dietary compounds have also been shown to reverse these epigenetic changes31,43
and to be effective agents in the prevention and treatment of prostate cancer.25,26
The plant Momordica charantia
(MC) or bitter melon grows in tropical Asia, where it is utilized both as a vegetable and medicinal herb. A 30kDa protein isolated from bitter melon seeds has demonstrated in vitro
and in vivo
anti-tumor activity in several tumor model systems.31,32,43
We have isolated a 30 kDa protein (MCP30) from bitter melon seeds. The purified fraction was analyzed by SDS-PAGE, and both the intensely silver-stained upper bands (~30 kDa) as well as the faint lower band (~22 kDa) were identified by mass spectrometry to contain 2 highly related ribosome-inactivating proteins (RIPs), α-momorcharin and β-momorcharin (MAP30). These 2 proteins have been reported to have similar biological activities, including N-glycosidase activity (cleaves the glycosidic bond between adenine and ribose in ribosomal RNA at a specific location34
) and have also been shown to act on nucleic acids (e.g
. intrinsic nuclease activities44
). Although the precise mechanism of action of the Type I family of single chain ribosome-inactivating proteins is not well delineated, they are known to enter the cell through endocytosis27,45
and can induce apoptosis in cancer cells.46
The precise mechanisms whereby single-chain RIPS selectively enter viral-infected and malignant cells, induce apoptosis and exert selective effects on malignant cells have not been well delineated. Saporin, a Type I RIP from Saponaria officinalis
(soapwort) induces caspase-dependent apoptosis in the human histiocytic lymphoma cell line U937 via
the mitochondrial (intrinsic) pathway and this apoptotic activity occurs before the onset of any significant inhibition of protein synthesis.46
Trichosanthin (TCS), another Type I RIP, is highly toxic to choriocarcinoma JAR cells while hepatoma H35 cells are relatively resistant. Differences in receptor-mediated endocytotic uptake and intracellular routing of TCS were found between the 2 cell lines and higher accumulation in the choriocarcinoma JAR cells may explain their higher sensitivity to TCS.47
We tested the effects of MCP30 on cell growth and apoptosis in a variety of human prostate cell lines in vitro and demonstrate that it selectively induces both cell cycle arrest and apoptosis in premalignant and malignant prostate cells. Furthermore, in vivo administration of MCP30 decreased PC3 human prostate cancer cell growth subcutaneously in nude mice and this effect was due primarily to the induction of apoptosis, with no significant differences in markers of proliferation or MVD between control and treated animal tumors.
The selective induction of apoptosis in neoplastic cells is also a hallmark of a class of anti-tumor compounds known as HDAC inhibitors. HDACs, which catalyze the removal of acetyl groups from the N-terminus of histones, lead to chromatin condensation and transcriptional repression.15
Altered expression of individual HDACs in tumor samples has been reported48
and several HDAC inhibitors are in clinical trials for cancer therapy. We determined the effects of MCP30 on HDAC1 in prostate-derived cell lines because this particular HDAC was previously shown to be overexpressed in human premalignant and malignant prostate lesions, with the highest increase in expression in hormone refractory prostate cancer.11
We confirmed that HDAC1 activity is increased in premalignant and malignant prostate cancer cell lines as compared to the non-neoplastic RWPE cell line. Furthermore, our data demonstrate that that the Type I RIPs contained in MCP30 inhibit HDAC1 expression levels and activity selectively in the neoplastic cell lines.
Although many studies demonstrate that both HDACi and Type I RIPs can selectively kill tumor cells,14,32,46
the molecular pathways underlying these observed effects remain to be elucidated. HDAC inhibitors have been reported increase acetylation of non-histone proteins and may induce apoptosis through tumor specific induction of tumor suppressor and pro-apoptotic genes. As has been reported for other compounds that inhibit HDAC,14
MCP30 selectively increased tumor suppressor protein expression, including PTEN, KLF6 and E-cadherin (data not shown). This was not just a generalized effect on protein expression, as MCP30 did not modulate total protein levels of Akt, β-catenin or the β-actin housekeeping gene. We also present evidence that MCP30 may restore normal PTEN signaling as demonstrated by decreased activity of Akt by dephosphorylation at Ser-473, increased Ser-9 phosphorylation of GSK-3β, inhibition of canonical Wnt signaling and decreased expression of Cyclin-D1 and c-Myc
in the neoplastic prostate cells. However, these events may be merely correlative and need to be further explored.
It was previously reported that 5-aza-2′-deoxycytidine, a DNA methyltransferase inhibitor, reactivates the transcription of PTEN in prostate cancer cells. More recently, the dietary compound genistein, was shown to induce PTEN expression in LNCaP and PC3 prostate cancer cells by inhibiting SIRT1, an HDAC that belongs to the atypical class III histone deacetylase family. Genistein activation involved demethylation and acetylation of H3-K9 at the PTEN promoter.49
Our data shows re-expression of PTEN mRNA and protein in PIN, LNCaP and PC3 cells which may result from the inhibitory effect of MCP30 on HDAC-1 levels and activity. Eighteen HDACs have been identified in humans and it is possible that MCP30, genistein and other dietary compounds modulate the expression and activity of multiple HDACs in a tissue-specific manner with resultant activation of a variety of tumor suppressor and pro-apoptotic genes.
A large number of structurally diverse HDACi have been purified from natural sources or synthetically developed and several of these are in clinical trials for cancer.15
Although most HDAC inhibitors are molecules of low molecular weight, maspin, a 42 kDa protein in the serine protease inhibitor family, was recently identified as an endogenous inhibitor of HDAC-1.50,51
To our knowledge, this is the first report that Type I ribosomal inactivating proteins derived from dietary bitter melon (Momordica charantia
) possess HDACi activity and can selectively induce apoptosis in premalignant and malignant prostate cells and inhibit human prostate cancer cell growth in vivo
. Our data imply that effective and non-toxic prostate cancer chemopreventive and therapeutic agents may be developed from plant-derived Type I ribosomal inactivating proteins.