We performed the first gene expression profiling experiment with rhizomes of Cimicifuga racemosa
(black cohosh) to identify molecular effects in the human breast cancer cell line MCF-7. In initial experiments analyzing cell proliferation we observed growth inhibition in response to treatment with a lipophilic extract of black cohosh, the major cycloartane-type triterpene glycoside actein and a cycloartane aglycon mixture. IC50
values were comparable to previously reported results [22
Effects of a black cohosh extract on gene expression in MCF-7 cells were determined by means of Affymetrix GeneChip® Human Genome U133 Plus 2.0 arrays, enabling almost complete analysis of the transcriptome. After treatment of MCF-7 cells for 24 h with the black cohosh extract at the IC50 concentration (15 μg/ml) changes of the expression levels of 431 genes were detected using highly stringent selection criteria. Random permutation of gene expression profiles showed a mean of 22 to 32 regulated probe sets in our data set with our selection criteria indicating a false positive rate of about 5%.
Comparing the expression patterns, action of black cohosh is opposite to estradiol and more similar to tamoxifen regarding proliferation and cell survival. However, the effect of black cohosh treatment appears more complex, as significantly more genes were regulated than either with E2 or tamoxifen. The genes regulated by black cohosh did not include well-known estrogen-regulated genes, with the exception of ESR1 (the estrogen recptor α gene) and VEGF which both could be regulated via hypoxia response. In contrast, a wider range of cellular pathways and targets were affected by black cohosh but not E2 or tamoxifen. Hence, action of black cohosh in MCF-7 cells seems to be neither estrogenic nor antiestrogenic but rather multifacetted. Because MCF-7 is a model for determining estrogenic or antiestrogenic effects, we can conclude that the benefical effects of black cohosh in alleviating postmenopausal complaints might be rather due to central nervous action via dopamine, serotonin or μ opioidic receptors.
Among all regulated genes, those related to the functional categories of proliferation (cell cycle and DNA replication) and stress response (including members of different functional categories) were found to be highly significantly overrepresented, and these groups appear to be interconnected. Given that many genes were regulated at a lower level and, therefore, excluded by the stringent filter setting, a statistically highly significant accumulation of genes in these functional categories emphasizes the significance of our results.
The cell cycle inhibition observed in the proliferation experiments appears to be due to an arrest at both G1/S- and G2/M-transition points, as indicated by an accumulation of regulated genes. This finding is corroborated by recent results of a flow cytometric analysis [24
]. Several genes involved in apoptosis appeared to be regulated, most genes in a pro-apoptotic manner. Some genes are related to apoptotic as well as anti-apoptotic events. These genes are linked to different cellular stress response mechanisms. Altogether, the expression levels of approximatively 40 genes related to stress response were affected by black cohosh. Given the fact that results of stress response pathways are contradictory, pro- and anti-apoptotic gene expression may not be surprising. The main purpose of stress response is an adaptation of the cells to stress factors, resulting in cell survival, angiogenesis and promoted tumor growth. However, when stress reaches a certain threshold, the protecting pathways become saturated and cells undergo apoptosis [49
]. This may explain the seemingly contradictory effects observed in our experiment. Altogether, pro-apoptotic signalling seems to outweigh. Stress response regulation has already been previously reported for other xenobiotics such as the cancer chemoprotective phytochemical indole-3-carbinol (I3C) and its physiological condensation product diindolylmethane (DIM), whose antitumor activity has been widely investigated [42
]. Its influence on gene expression in different tumor cells – among others in breast cancer cells such as MCF-7 – shows a certain similarity to our results with black cohosh. However, the mechanism of induced stress response by I3C is still unknown.
In association with unfolded protein response various transcripts of protein turnover were affected by black cohosh. Regulation of several transcripts whose products are involved in ubiquitinylation may be due to augmented degradation of malformed proteins. Increased mRNA levels of not less than eight different aminoacyl-tRNA synthetases (ARSs) are not only linked to protein synthesis. Some ARSs are rather multifunctional proteins involved in different cellular processes [61
]. For example, the secretion of tyrosyl-tRNA synthetase (YARS) is linked to apoptotic events [39
] and tryptophanyl-tRNA synthetase (WARS) has been shown to possess angiostatic and proliferation-inhibitory activity [61
As response to exposure to black cohosh extract a group of transcripts coding for enzymes with oxidoreductase activity was upregulated ("response to xenobiotics"). A strong upregulation was observed for CYP1A1
, and, to a somewhat lesser extent, of CYP1B1
. Apart from well-known involvement in xenobiotic metabolism, thereby mediating toxic and tumorigenic effects of several chemicals, these two oxidoreductases are involved in the metabolism of 17β-estradiol [62
]. CYP1A1 metabolizes E2 to non-carcinogenic 2-hydroxy-E2 whereas CYP1B1 is responsible for the formation of carcinogenic 4-hydroxy-E2. The two enzymes are not always expressed at the same level in tissues. An increased production of 2-hydroxy-E2 relative to 4-hydroxy-E2, due to a higher expression level of CYP1A1 than CYP1B1, has been suggested to be contributing to the antitumor activity of indol-3-carbinol and, therefore, being of clinical importance [62
]. As described, we also observed significantly stronger induction of CYP1A1 transcripts than CYP1B1 with black cohosh treatment.
is known as the
classical target of the aryl hydrocarbon receptor (AhR). Interestingly, the receptor has also been upregulated with our experiment. The AhR, upon binding of a ligand, forms a heterodimeric complex with ARNT (aryl hydrocarbon receptor nuclear translocator) which induces transactivation of the CYPs and other target genes via binding to xenobiotic response elements (XREs) in their promoter regions [63
]. Classical AhR ligands and, therefore, CYP1A1 inducers are hydrophobic and planar or coplanar molecules of polycyclic structure. Although the cycloartane-type triterpenoids do not satisfy these structural requirements, gene expression of CYP1A1
was not only induced in our experiments by black cohosh extract but also by the cycloartane aglycons (5.0 ± 1.7 fold, p < 0.01 and 12.2 ± 3.4 fold, p < 0.01) and, to a lower but statistically significant extent, by actein (2.1 ± 0.8, p < 0.05 and 2.0 ± 0.2, p < 0.01). CYP1B1
was also differentially expressed by the triterpenoids. The upregulation of CYP1A1
expression by extract and purified cycloartanes cannot be explained by AhR binding. Indeed, black cohosh extract and compounds did not show any AhR activity in a reporter gene assay [64
] in rat hepatoma H4IIE cells [see Additional file 2
]. These findings are in contradiction with previous experiments showing weak AhR binding and transactivation activity of black cohosh extracts. However, these experiments were carried out with high concentrations (up to 200 μg/ml) and did not include single compounds [20
]. The upregulation of CYP1A1
may be linked to activation of genes involved with hypoxic stress response, although the consequences of upregulated stress response appear somewhat contradictory. AhR and hypoxia-induced signal transduction are connected via ARNT, and AhR and HIF1α compete in binding to ARNT. Hypoxic conditions reportedly inhibited AhR-dependent gene expression [65
], on the other hand, a connection between CYP1A1
induction and hypoxia-induced gene regulation in Caco2 cells has been found [66
Besides CYP1A1 and CYP1B1, 11 other selected mRNAs representing all regulated functional categories were investigated by real-time RT-PCR to verify the microarray results and to investigate the effects of the cycloartane-type triterpenoids. Briefly, all genes were regulated by the triterpenoids in the same direction and to a fairly comparable extent than with the extract. Hence, we provide here first evidence that cycloartane glycosides as well as their aglycons, which are most likely formed in the intestine prior to absorption, are putative active principles in black cohosh.