Lycopene Oxidation, Uptake, and Activity in Human Prostate Cell Cultures - Carotenoids: Physical, Chemical, and Biological Functions and Properties - page 455

The fact that lycopene appears to have so many different bioactivities in prostate cancer cell
lines points to a common mechanism that might explain a variety of its effects. What are the can-
didates for a common modality of action and is it a chemically feasible mechanism for lycopene
action? There are four modalities that could be explored through cell culture studies: (1) the effects
of lycopene on shifting the aberrant DNA methylation pattern that is seen as prostate cells undergo
carcinogenic transformation, (2) the modulation of retinoid receptor signaling, (3) the modulation
of redox controlled cell signaling mechanisms, and (4) selective binding to catalytic and signaling
proteins with common structural motifs.

21.9.1 L YCOPENE AND G ENE M ETHYLATION
Cytosine is methylated to form 5-methylcytosine, which combines with guanine to form cytosine-
guanine dinucleotides (CpGs) that cluster in key regulatory regions called CpG islands. CpG methy-
lation is a key epigenetic regulatory mechanism. Hypermethylation tends to inactivate genes, and
promoter hypermethylation and the silencing of associated genes are widespread in prostate cancer
(Murphy et al. 2008). The identii cation of a signature gene methylation pattern that would dif-
ferentiate clinically signii cant prostate cancer from indolent cancer is a current research focus.
For example, the gene for glutathione-S-transferase pi (GSTP1) has been found to be methylated in
>75%-90% of prostate cancers but not methylated in normal epithelium (Lin et al. 2001, Woodson
et al. 2004). The gene for IGFBP-3 has also been found to be methylated in prostate cancer (Perry
et al. 2007) as is RAR , the retinoid receptor (Bastian et al. 2007, Woodson et al. 2004). Increased

DNA methylation of the connexin promoter region and the down-regulation of Connexin 43 expres-
sion are features of human lung tumors (Chen et al. 2003).
King-Batoon et al. have found that the physiological concentrations of lycopene (2

β

M) partially
demethylated the promoter for GSTP1 and restored its expression in the breast cancer cell line
MDA-MB-468, but RAR

μ

β was not demethylated. However, lycopene did induce the demethylation

of RAR in MCF10A i brocystic cells (King-Batoon et al. 2008). Genistein, in this study, was less

active compared to lycopene.
Is it chemically feasible to suppose that lycopene could directly modulate methylation/
demethylation? A family of DNA methyltransferases is responsible for methylation, and it has been
recently proposed that the DNA methyltransferases DNMT3A and DNMT3B are also responsible
for active demethylation through a very complicated mechanism that cycles every 100 min during
transcriptional cycling (Kangaspeska et al. 2008, Metivier et al. 2008). So how could lycopene
intervene in this process? Interestingly, Kangaspeska et al. (2008) used doxorubicin to reduce
methylation with the hypomethylation of the proximal pS2 promotor occurring about 45 min after
its introduction into MDA-MB-231, estrogen receptor-negative breast cancer cells. Doxorubicin is
thought to act by intercalating within the DNA structure and this interferes with the unwinding of
DNA for replication (Formari et al. 1994). It is possible that lycopene could act in the same manner
since all- trans lycopene has a l at, rod-like structure and 5

β

cis lycopene is almost as l at, and 55%
of the lycopene is found in the nuclear membrane and 26% in the nucleus matrix of prostate cancer
cells (Liu et al. 2006). Bathaie et al. found that the saffron carotenoids, crocetin, dimethylcrocetin,
and crocin bind to calf thymus DNA in the outside groove-binding pattern (Bathaie et al. 2007).
However, both the doxorubicin and the saffron carotenoids are highly oxygenated and water-soluble,
whereas lycopene is the most hydrophobic of the carotenoids. Furthermore, how could lycopene,
concentrated in the nuclear bi-membrane be in proximity to DNA that is about to undergo replica-
tion? Of interest is the observation that silenced genes, methylated and/or tightly wrapped around
histones, are more likely to be found peripherally, nearest to the nuclear membrane (Shaklai et al.
2007). Alternatively, lycopene or a lycopene oxidation product could be acting as a weak androgen
or estrogen antagonist or even agonist (Hirsch et al. 2007, Wertz et al. 2004). Estrogen stimulates
cell division and thus methylation/demethylation cycling.

′