As 7-azaindoles, meriolins represent structural hybrids of meridianins and variolins that have been isolated from marine invertebrates (Fig. 1A) [8, 46]. Like their parental natural compounds, meriolins potently inhibit a broad range of CDKs and appear to be even more active than variolins and meridianins in vitro and in vivo [1, 8, 10–14]. Thus, it has been shown that meriolins display a strong cytotoxic potential in nanomolar range in various tumor cell lines of different origin, such as colon cancer (HCT116, LS-174T), hepatoma (Huh7, F1), cervix carcinoma (HeLa), breast carcinoma (MCF-7), glioma (GBM, SW1088, U87), neuroblastoma (SH-SY5Y), leukemia (Jurkat, Molt-4), lymphoma (Ramos), and myeloma (KMS-11) [1, 7, 11, 19, 20, 47]. However, nontransformed cells as human foreskin fibroblasts appear to be more resilient [7]. In recent years, this substance class underwent derivatization and optimization in order to enhance their efficacy in cancer therapy. In this context, we have previously shown that 6 out of 13 newly synthesized meriolin derivatives displayed a pronounced cytotoxicity with meriolin 31 and 36 comprising the highest cytotoxic potential [1].
Here, we investigated the bioactivity of the novel derivative meriolin 16 and compared its cytotoxic potential to meriolin 31 and 36. Thus, we could show that meriolin 16 was extremely cytotoxic (IC50: 50 nM) in Ramos lymphoma cells as compared to meriolin 31 (IC50: 90 nM) and meriolin 36 (IC50: 170 nM) upon 24 h treatment. Meriolin 16 was even more potent than meriolin 3 (IC50: 70 nM) that has already been tested in preclinical trials [16]. Since meriolin 3 exhibits the highest cytotoxic potential amongst all meriolines described so far, this renders meriolin 16 as the most potent meriolin derivative to date.
The increased cytotoxicity of meriolin 16 compared to meriolin 31 and 36 might be attributed to the additional methoxy group at the aromatic pyridine ring of meriolin 16 (see Fig. 1A). The methoxy group was intentionally added in order to modify its pharmacokinetic properties and to enhance its biological activity. Besides affecting the polarity and solubility of a molecule, an additional methoxy group can increase potential reactivity by enhancing the electron density of the aromatic ring, thus rendering it more nucleophilic [48]. Since meriolin 3 also contains a methoxy group at the same position as meriolin 16 (Suppl. Figure 1B), the methoxy group might be of importance for the structure-activity relationship of these derivatives.
Most interestingly, a 5 min exposure to meriolin 16, 31 or 36 was sufficient to activate the endogenous suicide program in Ramos lymphoma cells after 24 h (Fig. 2). We further investigated the apoptosis signaling pathways activated by meriolins. Using cell lines deficient for caspase-8, -9 or Apaf-1, we could show that meriolin-induced apoptosis was independent of external death receptor signaling but required Apaf-1 and caspase-9 as central mediators of the mitochondrial apoptosis pathway (Fig. 4). Activation of the mitochondrial apoptosis pathway was further confirmed by the mitochondrial release of proapoptotic Smac, breakdown of the mitochondrial membrane potential (ΔΨm), processing of the dynamin-like GTPase OPA1 and subsequent mitochondrial fission (Fig. 4,5).
The question however remains, how meriolins activate the mitochondrial apoptosis pathway. We initially observed that meriolin 16 and 36 and other CDK inhibitors such as R547 induce the ATM-mediated phosphorylation of CHK at Thr68. In addition, meriolin 16 induced the ATM-mediated phosphorylation of γH2AX (at Ser139) and KAP1 (at Ser824). However, in contrast to DNA damaging agents as the topoisomerase II inhibitor etoposide, inhibition of caspases completely blocked the meriolin 16 induced phosphorylation of γH2AX and KAP1. Thus, these data indicate that meriolin-inflicted DNA damage is rather an apoptotic downstream event mediated by caspases. This effect might be attributed to the caspase-mediated degradation of the inhibitor ICAD during apoptosis that subsequently releases and activates the caspase-activated DNase (CAD) which then produces the characteristic internucleosomal DNA cleavage [49, 50]. Apoptotic internucleosomal DNA cleavage might in turn activate ATM and induce the phosphorylation of γH2AX and KAP1, which in case of meriolin 16, could be inhibited upon caspase-inhibition but not in case of direct DNA damage inflicted by etoposide. Thus, meriolin-induced activation of the mitochondrial apoptosis pathway is obviously not initiated upon DNA damage – as in case of radio- and chemotherapy.
Bettayeb et al. could demonstrate that meriolin 3 inhibits the kinase activity of various cyclin dependent kinases (such as CDK1–7 and CDK9) and prevents the phosphorylation of CDK1, CDK4, CDK9, retinoblastoma protein and RNA polymerase II. In addition, they observed that meriolin 3 induced the downregulation the antiapoptotic Bcl-2 protein Mcl-1 [7]. Consequently, they proposed that meriolin 3-mediated inhibition of CDK9 reduces RNA polymerase II activity, which in turn downregulates short-lived Mcl-1 proteins. However, the authors did not include caspase-inhibitors in their study to exclude a potential caspase-mediated degradation of Mcl-1, as previously shown [51, 52]. Usually, antiapoptotic Bcl-2 members (such as Bcl-2, Bcl-xL or Mcl-1) inhibit the activation of the mitochondrial apoptosis pathway by counteracting proapoptotic Bcl-2 members (such as Bax, Bak or BH3-only proteins). However, since we observed that meriolin 16, 31 and 36 were able to induce apoptosis in the presence of antiapoptotic Bcl-2 (Fig. 4C, D), it is rather unlikely that meriolins induce apoptosis via downregulation of antiapoptotic Bcl-2 proteins. The observation that meriolin 16, 31 and 36 induce apoptosis even in the presence of Bcl-2 rather suggests that they activate the mitochondrial apoptosis pathway in a direct way downstream of pro- and antiapoptotic Bcl-2 members.
Antiapoptotic members of the Bcl-2 family, such as Bcl-2, Bcl-xL, or Mcl-1, were first identified in leukemia and lymphoma and are frequently overexpressed in other types of neoplasia to prevent apoptotic cell death during tumorigenesis. Moreover, since the major mechanism of genotoxic radio- and chemotherapy is the p53-mediated activation of the mitochondrial apoptosis pathway, tumor cells gain therapy resistance by inactivating this route e.g. via overexpression of anti-apoptotic Bcl-2 proteins [33, 53]. Thus, the observation that meriolins are capable of directly activating the mitochondrial apoptosis pathway even in the presence of Bcl-2 renders them as valuable agents for the treatment of therapy-resistant tumors. This feature is corroborated by our observation that the tested meriolin derivatives were capable of inducing cell death in imatinib-resistant K562 and KCL22 CML cells as well as in cisplatin-resistant J82 urothelial carcinoma and 2102EP germ cell tumor cells (Fig. 7).
Beside cell death induction, it has also been shown that meriolins as multikinase inhibitors target kinases such as CDKs (e.g. CDK1-7 and CDK9) [7] and induce cell cycle arrest in G2/M-phase [11]. Thus, meriolins act as a double-edged sword, since on the one hand they inhibit CDKs (i.e. proliferation) and on the other hand they can also induce cell death (i.e. apoptosis). This renders meriolins as valuable anticancer drugs since they simultaneously target two Achilles' heels of the tumor – i.e. unlimited proliferation and inhibition of cell death [21, 22].