Plexxikon’s discovery of PLX4032, a selective targeted therapeutic for metastatic melanoma

In our March 2, 2010 blog post, we focused on a Phase I clinical trial of Plexxikon/Roche’s PLX4032, in which clinical researchers led by Keith T. Flaherty achieved a dramatic breakthrough in treatment of metastatic melanoma. Now we shall discuss the discovery of the drug itself, PLX4032.

In 2002, a research consortium based at the Wellcome Trust Sanger Institute in the U.K. found B-Raf somatic missense mutations in 66% of malignant melanomas (as well as in a subset of other cancers). V600E (valine substituted by glutamic acid at position 600) accounted for 80% of these mutant forms of B-Raf. The V600E mutation causes destabilization of the inactive conformation of B-Raf kinase, shifting the equilibrium toward the catalytically active conformation.

B-Raf is a serine/threonine protein kinase that is a component of an intracellular pathway that mediates signals from growth factors. B-Raf is regulated by binding to Ras. In turn, B-Raf activates MEK (mitogen-activated protein kinase kinase), which activates ERK (extracellular signal-regulated kinase). Activated ERK goes on to upregulate transcriptional pathways that promote cellular proliferation and survival.

Growth factors → →Ras→ B-Raf→ MEK→ ERK→ →upregulation of cell proliferation and survival

Growth factor signaling via Ras also controls other signaling pathways that upregulate cell proliferation, notably the PI3K-Akt (phosphatidylinositol-3-OH kinase-Akt) pathway.

The Sanger researchers found evidence that cells carrying B-Raf(V600E) no longer require Ras function for proliferation. This would mean that melanoma cells carrying this mutation could proliferate independently of growth factor signaling, resulting in the runaway proliferation characteristic of the malignant phenotype.

These results suggested that B-Raf(V600E) would be a good target for novel kinase inhibitors to treat malignant melanoma. The first such kinase inhibitors to be developed, although they had inhibitory activities at low nanomolar concentrations against B-Raf (both wild-type and mutant), were not successful in the clinic, due to their inhibition of multiple nonspecific targets and/or their poor bioavailability. Plexxikon researchers therefore set out to discover inhibitors that are highly selective for B-Raf(V600E). The result was the discovery of PLX4032.

The discovery of PLX4720 (a tool compound or chemical probe related to PLX4032) by Plexxikon researchers and their academic colleagues, and its preclinical validation, is described in a 2008 publication, Tsai et al. Plexxikon used its proprietary “scaffold-based drug design” technology platform to discover PLX4720. Scaffold-based drug design involves synthesizing sets of low-molecular weight “scaffold-like’” compounds. These compounds interact (typically at low affinity) with many members of a protein family by targeting their conserved regions.

In the B-Raf study, the researchers identified protein kinase scaffolds by screening a select library of 20,000 150-350-dalton compounds for inhibition of a set of three structurally characterized protein kinases at a concentration of 200 micromolar (μM). Of this library, 238 compounds were selected on the basis of their inhibition of the kinases by at least 30% at the 200 μM concentration. Each of the compounds was cocrystallized with one if the three kinases, Pim-1. Using this method, the researchers found that 7-azaindole bound to the ATP-binding site of Pim-1 kinase. They further modified this compound by adding side chains on the 3 position of 7-azaindole, resulting in a “scaffold candidate” with increased affinity for the ATP binding site of PIm-1 and other kinases. The researchers further modified this scaffold, based on structural data from other kinases. Ultimately, they cocrystallized their modified compounds with wild-type B-Raf and B-Raf(V600E), and optimized the structure of their compounds to give a compound, PLX4720, with selectivity for B-Raf(V600E) and against wild-type B-Raf and other kinases. This process (including the relevant chemical and protein structures) is illustrated in Figure 1 of Tsai et al.

In biochemical assays, the researchers found that PLX4720 inhibited B-Raf(V600E) at low nanomolar concentrations, and was 10-fold more selective for B-Raf(V600E) than for wild-type B-Raf, and was even more selective for B-Raf(V600E) than for other kinases.

Surprisingly, in cellular assays, PLX4720 is over 100-fold (not 10-fold) more selective in inhibiting proliferation of tumor cell lines that bear B-Raf(V600E) as compared to those that bear wild-type B-Raf. Moreover, as first found by researchers at Pfizer and their academic collaborators, a specific inhibitor of MEK (Pfizer’s CI-1040) is also similarly selective for tumor cell lines bearing B-Raf(V600E). This suggests that the B-Raf-MEK-ERK pathway is critical for the proliferation of B-Raf(V600E) cells, but not for cells bearing wild-type B-Raf. [For example, tumor cells that bear wild-type B-Raf might use the PI3K-Akt pathway to upregulate pathways that control cell proliferation independent of ERK signaling, while tumor cells that bear B-Raf(V600E) cannot.]

The B-Raf-MEK-ERK pathway dependence of B-Raf(V600E) cells may in part be related to feedback inhibition of B-Raf (and other isoforms of Raf). Activated ERK can phosporylate wild-type Raf isoforms at specific inhibitory sites. This results in downregulation of signaling via the Raf-MEK-ERK pathway. However, in cells bearing B-Raf(V600E), this feedback inhibition is disabled, resulting in uncontrolled signaling.

The Plexxikon researchers (Tsai et al.) tested PLX4720 against tumor xenograft models. Oral administration of PLX4720 blocked tumor growth, and in 4 out of 9 cases caused tumor regressions, in mice with tumor xenografts bearing B-Raf(V600E). Treatment with PLX4720 was well tolerated, and showed no adverse effects. Growth of tumor xenografts bearing wild-type B-Raf was not affected by PLX4720. In mice with tumors bearing B-Raf(V600E), PLX4720 blocked B-Raf-MEK-ERK pathway signaling, as demonstrated by immunohistochemical assays.

The exquisite specificity of PLX4720/PLX4032 for B-Raf(V600E) as compared to wild-type B-Raf was made possible by Plexxikon’s structure-guided “scaffold-based drug design” technology. Other structure-guided drug design technologies, such as fragment-based lead design, as is carried out in several companies, might be used to design comparably specific drugs.

The discovery of PLX4720/PLX4032 is an example of the use of new-generation chemistry technologies (or the revival of the old, and now disused natural products chemistry approach), coupled with biology-driven drug discovery strategies, to discover promising new drugs. We have discussed this strategy in several articles on this blog. (For example, see here and here).

Despite the promising results seen in Phase I clinical trials of PLX4032, it must be emphasized that the establishment of the efficacy and safety of this compound awaits the completion of the ongoing Phase III trials. Moreover, despite the dramatic regressions and increased survival seen in the Phase I trials, all the patients apparently eventually suffered relapses. Dr. Flaherty, as discussed in our earlier blog post, sees the need for combination therapies to effectively combat metastatic melanoma. In early 2009, Dr. Flaherty and his colleague Keiran S Smalley published a mini-review on potential strategies for developing such combination therapies.