The microarray analyses revealed different expression profiles in the brain sections from ovariectomized (OVX) or iCR-treated rats after ovariectomy (OVX + iCR) compared to intact and untreated rats (PRAE). 3682 features out of > 30,000 in the hippocampus and 2862 features in the hypothalamus were significantly up- or downregulated. Thereof, 1732 features belonged to both tissue types resulting in 4812 in total. Comparing OVX and OVX + iCR, 2210 genes in the hippocampus and 1759 in the hypothalamus were identified as differentially expressed, i.e., resembled iCR-effects in tissues of ovariectomized rats.
Three approaches were used to discover relevant pathways and target genes in menopausal transition and iCR treatment: ‘compensation’, ‘iCR exclusive effects’, and those of ‘general’ interest. 213 genes in the hippocampus and 349 genes in the hypothalamus, counted in the compensation approach, were identified by testing them for contrary fold change. The cluster with 941 genes in the hippocampus and 644 genes in the hypothalamus displayed candidates of an exclusive iCR effect, i.e., in genes not modulated by ovariectomy. GSEA analyses on 1543 genes that were not changed by OVX revealed that black cohosh affects pathways involved in system classes such as osmotic regulation, hormone-related activities, metabolism, and immune system. Next to the contribution of plenty of OR family genes in signalling pathways, which were already covered in previous analyses , some additional pathways also appeared to specifically contribute during the menopausal transition and the abolishment by iCR. Black cohosh compensated the effects of OVX in 518 genes, which were allocated by GSEA analysis to following the processes and systems, respectively: immune system, general signal transduction, aging, bone biology, body temperature, and hormone-related processes, particularly neuropeptide hormone activity and neuropeptide receptor binding.
In order to verify the results of microarray and functional analyses with a second technique, target genes were chosen for subsequent qPCR assays. In respect to the first two approaches mentioned above, the picked pathways of interest from the two lists GSEA “compensation” and GSEA “iCR exclusive” were screened for specific target genes that best present the respective pathways. In addition to the 17 target genes thus selected, five further target genes were chosen due to scientific discussions [10,11,12,13,14,15,16,17,18,19] even though the criteria for approach 1 and 2 were not met. The 22 target genes were only few of the many genes tested on the 30,423-feature microarray. Unfortunately, the available RNA was restricted, so no more target genes could be assayed by qPCR. Nevertheless, due to plenty of overlapping functions, the whole panel of pathways of interest was represented by these 22 target genes. Thus, the selected target genes were not a random selection. An aim of this study was to find genes possibly targeted by black cohosh and to investigate theories about its mode of action. The procedures were chosen to maximize the gain in knowledge about iCR’s influence on brain function.
The expression values of the residual RNA were not sufficient enough to measure technical replicates. Therefore, the two biological replicates were used as the basis of statistical studies. This procedure differed in the qPCR from the microarray analysis; therefore, the values may partially differ resulting in some source of error-proneness. Nevertheless, the averages of both samples and the inference of the results of both methods (microarray and qPCR) are decisive. A major limitation of this study is the low sample size; the interpretation should be tentative. Additional studies may confirm these results with more impact.
The concordance of the results was assessed in regard to the selection criteria: compensating or iCR exclusive effect on the hippocampus or hypothalamus, or general interest in the gene. Statistical analysis was not reasonable for this assessment; a graphical comparison sufficed. All in all, predominantly compensation at the hippocampus was confirmed by qPCR. Particularly GAL, CALCA, and HCRT showed good consistency. The results of the genes AVPR1A and PNOC were somewhat less consistent showing compensation in the microarray but only a trend of an OVX-induced increase of the expression value in comparison to PRAE in the qPCR. However, the iCR-induced decrease was still significant (Supplementary Fig. 1). Thus, if OVX increases the expression of these genes, iCR can compensate such effect in hippocampus. This effect might alternatively be interpreted as over-compensatory with residual uncertainty as to which interpretation holds true regarding these two genes in this tissue. The signal intensity of IL5 was very low; therefore, the missing concordance with the qPCR results was expectable. Apart from IL5, all of the selected target genes in the hippocampal compensation query were more or less confirmed.
The compensation at the hypothalamus and all iCR exclusive effects found in the microarray analysis were not confirmed by the qPCR in the tissue of primary focus, e.g., this concerns pathways regarding metabolism and pathways regarding the immune system (Table 4). Unexpectedly, the pattern in the PCR sometimes gave a mirror image of that in the microarray, meaning an effect was obvious but in the other direction, e.g., hypothalamus PLCB1, MAPK9, TRPV3. This finding cannot be adequately explained. Limited sample size, sample material, and signal intensities may contribute to these circumstances.
As a matter of completeness, the expression of the selected genes was also investigated by qPCR in the other tissue. Regarding AVPR1A, GAL, CALCA, HCRT, and PNOC, the pattern of effects or no effects on gene expression in the hypothalamus differed from the clear pattern in the hippocampus (Table 4). Interestingly, IGFBP1 in the hippocampus and PRL7A3 in the hypothalamus showed a compensation pattern in the qPCR of the other tissue with only a trend in the microarray, which did not meet the FC-threshold criteria.
The compensating effect of iCR was found by qPCR also for the genes ESR1, ESR2, and TAC3 in the hippocampus even though there was no such effect in the microarray. The patterns of TAC3 at the hypothalamus were similar in both methods; a compensational effect was observed. This gene was not filtered by the selection procedure of the microarray results, because the 1.5-fold change was not attained (fold change OVX vs. OVX + iCR = − 1.32). Furthermore, the KISS1 microarray pattern of the hypothalamic sample was confirmed. OVX and OVX + iCR showed higher expression values than PRAE; an OVX effect was shown. For OPRM1 the qPCR confirmed no relevant effect of OVX or iCR on OPRM1 expression in the hippocampus. In the hypothalamus, the microarray’s p value results for OPRM1 might indicate a compensation of the OVX-induced decrease by iCR. However, this OPRM1 result did not meet the FC-threshold-criteria, and in the qPCR, only a trend of an OVX-effect accompanied a significant iCR-effect.
Since the target region of the genes through the qPCR is larger (80–129 bp, mean 108.8 bp) than in the 60-mer microarray, the specificity of the technique is enlarged too. On the other hand, the sample number was higher for the microarray. Therefore, finding a balance between the results from both techniques is difficult. In general, the results of the two techniques match well , but this depends on filtering parameters etc. . The absolute Ct values or signal intensities should also be taken into account when assessing the concordance of the results. The most critical factors influencing the correlation between microarrays and qPCR are fold change and p value values of the microarray, but Ct values and up- vs. downregulation are also important . The match of the transcripts in both methods also plays a role in the agreement of the results . Even if the qPCR transcripts are as close as possible to the corresponding microarray features, there are, in principle, certain differences in the targeted gene region yielding different results.
Although microarray analyses revealed compensation by iCR in the hippocampus and qPCR even revealed overcompensation of some genes (AVPR1A, PNOC), this result should not be overestimated as the number of samples was too small to reach a final conclusion. In addition, statistical analyses solely for significance should be interpreted cautiously as the sample size was not appropriate for such a test. A combination of the thresholds for FC and p-value is more informative. This study gives hints to help identify the underlying mechanisms after OVX and the mode of action of iCR. Additional studies may be done to follow-up on this information.
However, system processes are not only regulated by changed gene expression. Posttranslational protein modifications and elevated depletion or stability of the gene product also contribute to cell regulation. In such cases, protein analyses should then also be performed.
The GSEA of hippocampal samples resulted in some compensated pathways/terms of major interest within the transformation process during menopause as follows: positive regulation of blood pressure, hormones, feeding behaviour, cytokine-cytokine receptor interaction, inflammatory response, response to insulin, orexigenic neuropeptide QRFP/P518, opioid receptor binding, cytokine activity, and positive regulation of vasoconstriction. The good concordance of the target gene analyses by PCR with the microarray results in the hippocampal compensation approach leads to the conclusion that the GSEA obtained pathways/terms are also confirmed. From these analyses, it cannot be determined whether there is one single receptor for iCR or in which order the downstream processes happen; only involvement is shown. All in all, some signalling paths in brain performance are altered due to ovariectomy and several of these changes are restored by iCR in the hippocampus.
Of course, ovariectomy also results in more widespread effects and is not completely identical to the processes that occur during menopausal transformation. Nevertheless, studies with this animal model have provided hints towards a better understanding of the CNS processes caused by oestrogen decline and of the mode of action of Cimicifuga racemosa.
The analysis of the mode of action for CR is very difficult because the multicomponent herbal preparation—the extract—performs many different individual actions. As previously hypothesized, there is a multi-target mode of action and the resulting effect is that CR restores imbalanced functions and mitigates menopausal symptoms by a multifaceted mechanism [23, 24].
Gaube et al.  investigated the effects of a lipophilic black cohosh rhizome extract, E2, and tamoxifen on the oestrogen receptor positive human breast cancer cell line MCF-7 using an Affymetrix GeneChip® Human Genome array. A wide range of cellular pathways and targets were affected by black cohosh. Most of the regulated genes could be clearly assigned to five larger groups of functionally related genes: apoptosis, proliferation, general growth, signalling and transport, and metabolism. Previously reported actions of black cohosh were confirmed by gene expression assigned by “contra cell proliferation” and “pro apoptosis”. The action of black cohosh in MCF-7 cells seems to be neither oestrogenic nor antiestrogenic but rather multifaceted. The analyses presented in this study do not accord with that of Gaube et al.  in detail, probably due to the different sample material and study design, in particular (in vitro/in vivo; cancer cell line/healthy primary samples; treatment only/treatment after OVX).
Beginning with 30,423 features in the microarray, a reduction to a few individual genes with a contribution to concise signalling pathways was achieved in the course of all the analyses. iCR can compensate OVX-induced changes/hypoestrogenism in the brain, particularly the hippocampus. This is reflected by changes in the expression profiles of genes that not only contribute to the formation of hot flushes or central thermoregulation but also to secondary effects such as blood pressure, metabolism, nociception, hormonal regulation, homeostasis, cognition, mood regulation, neuroendocrine modulation, peripheral thermoregulation through altered blood flow at the cutaneous microvascular level, regulation of sleep and arousal, learning, and memory. These relationships to processes in the menopause deserve further clarification.