Cortistatin-A

Cortistatin-A : Inhibitor of CDK19, CDK8

Structure

Information

  • CDK19
  • CDK8
  • Inhibitor
  • 1-1000 nM, 100 nM provides full inhibition of CDK8 and CDK19 in cells

In Vitro Validations

Uniprot ID: Q9BWU1
Target Class: Kinase
Target SubClass: CMGC
Potency: IC50
Potency Value: 100 nM
Potency Assay: immunoblot showing capture of native CDK19 from MOLM-14 lysates
PDB ID for probe-target interaction (3D structure): 4CRL 4CRL
Target aliases:
Cyclin-dependent kinase 19, KIAA1028, CDK11, CDC2L ...

DOI Reference: 10.1038/nature14904

In Cell Validations

In Vivo Data

Off-Target Selectivity Assesments

Potency end-point : IC50
Potency assay (off target): In vitro kinase assay and Kinativ cell lysate displacement assay.
Probe Selectivity in Vitro:
No reproducible off-targets were validated up to 1 uM Cortistatin A. The only kinase affected in the 294 recombinant kinase panel (other than CDK8) was GSG2, but GSG2 did not bind Cortistatin A in the Kinativ assay.
I have extra information to add

SERP ratings and comments


SERP Ratings

In Cell Rating
In Model Organisms

SERP Comments:

This is a potent and selective probe for CDK8 and its paralogue, CDK19. It was screened against a panel of 387 kinases (KiNativ) and shown to be completely selective for CDK8.

(last updated: 24 Oct 2016 )

SERP Ratings

In Cell Rating
In Model Organisms

SERP Comments:

The cortistatins are a group of steroidal alkaloids first isolated from the marine sponge Corticium simplex. Cortistatin A is a highly potent and selective inhibitor of CDK8 and CDK19, the kinases that associate with Mediator complex. Of the 387 kinases evaluated, cortistatin A only inhibited CDK8 and CDK19. At 100X its CDK8 IC50 value, Cortistatin A is fully selective in MOLM-14 cell lysate for CDK8 and CDK19, and in vitro it only inhibits the CDK8–CCNC complex and GSG2 (latter disqualified as a cellular target of Cortistatin A). Cortistatin A has acceptable pharmacokinetic properties in mice for once-daily intraperitoneal dosing. It inhibits growth of acute myeloid leukemia cells and AML in in vivo mouse models. Identification of dominant drug-resistant alleles of CDK8 and CDK19 demonstrate that these kinases mediate the activity of cortistatin A in AML cells (disseminated human MV4 11 model & AML model using SET-2 cells). Cortistatin A caused selective and disproportionate up-regulation of super-enhancer-associated genes in AML cells which contributed to its anti-leukemic activity.

(last updated: 22 Dec 2016 )

SERP Comments:

Author’s note: Although Cortistatin A showed antiproliferative activity in HUVECs (IC50=1.8 nM) at a concentration below the Kd for Cortistatin A on CDK8 as reported in that publication (Angew Chem 2009, 48, 8952, http://onlinelibrary.wiley.com/doi/10.1002/anie.200904778/full), we reported a Kd for Cortistatin A for the CDK8/CCNC dimer (Kd=0.2 nM), which is lower than the IC50 reported in the HUVEC article. Thus, we believe these data are consistent with Cortistatin A acting on target against CDK8 (and CDK19) in HUVECs.

(last updated: 9 Mar 2017 )

SERP Ratings

In Cell Rating

SERP Comments:

Cortistatin A is a comprehensively characterized CDK8 inhibitor. Inhibitory activity was demonstrated in biochemical assays with three known substrates: CDK8 (autophosphorylation), STAT1-S727, and RNA polymerase CTD. Inhibitory activity was also demonstrated in cellular assays of IFN-gamma stimulated STAT1-S727 phosphorylation and TGF-beta stimulated SMAD2 T220 and SMAD3 T179 phosphorylation. Moreover, a protein/ligand co-crystal structure exists that clearly demonstrates a specific interaction with CDK8. To demonstrate selectivity, Cortistatin A was tested a panel of 294 recombinant kinases and found to only inhibit CDK8 and GSG2. Kinase selectivity was also demonstrated on a cellular level, evaluating the displacement of ATP-site binding probes from endogenous kinases by cortistatin A. The first method used was limited in scope and had a western blot readout of several CDKs and ROCKs 1 and 2. Only CDK8 and CDK19 showed any inhibition of probe binding. Additionally, a broader set of kinases (~220) was investigated using the KiNativ technology with a mass spec-based proteomic readout, and only CDK8/19 showed inhibition of probe binding by cortistatin A (0.1 and 1 uM). Antiproliferative activity in select cell lines was reported in 3 to 7 day assays, with varying activities depending on the cell line. To demonstrate specificity of inhibition of proliferation, a mutant form of CDK8 was identified (W105M) that retained activity but was not inhibited by cortistatin A (confirmed in vitro and in cells). Importantly, minimal inhibition of proliferation was observed in MOLM-14 cells transfected with these mutant forms of CDK8 and CDK19 relative to cells expressing the wild type proteins. An older publication reports antiproliferative activity data for cortistatin A against HUVECs (IC50 = 1.8 nM), (Angew Chem 2009, 48, 8952, http://onlinelibrary.wiley.com/doi/10.1002/anie.200904778/full). It is surprising that the number is is almost one order of magnitude lower than the observed Kd for CDK8 (17 nM) and CDK11 (10 nM) reported in the same publication. While the kinase selectivity profile does not suggest kinase off-target activity, off-target activities unrelated to kinases cannot be ruled out. No published data on screening in secondary pharmacology panels exist.

It is clear that Cortistatin A is an inhibitor of a small number of specific kinases and is well characterized as such (enzyme inhibition data for CDK8 and 11, ROCK, crystal structure determination bound to CDK8). It is also clear the compound has antiproliferative activity across a number of tumor lines, as well as in HUVECs. It has not been convincingly demonstrated that these two observations are connected, however, and I am inclined to believe that Cortistatin A has additional activities that are contributing to the observed cellular phenotypes. Thus, any data emerging from the use of this compound should be treated with caution.

When a cellular phenotype is one of growth inhibition or cell kill, the burden of proof for target specificity is much higher than simply activity in a kinase assay. Some limited evidence is presented in the form of a mutated CDK8 allele that is less reponsive to the compound in one cell line, but this alone is not sufficient to rule out other activities of the compound. The observation of enzyme activity in the same region as cell activity (or indeed greater cell activity) is always a flag for this - particularly where a compound is a type I, ATP-competitive inhibitor (as is Cortistatin A - based on the crystal structure). Potent cell activity with no or negative drop off from enzyme activity would suggest the cell effects are being driven by something other than, or in addition to, its noted enzyme effects. The chemical structure of Cortistatin A might also flag potential for other activities (steroidal natural product). To understand whether the reported antiproliferative effects are being driven by CDK8, one would like to see evidence of target engagement within cells, evidence of pathway modulation (such as inhibition of a phospho-substrate), evidence that target ablation (for example siRNA or CRISPR) phenocopies small molecule inhibition, and an identical phenotype observed with a completely orthogonal chemical series.

(last updated: 23 Jun 2017 )