B-Raf activation loop phosphorylation revisited

The Ras/Raf/MEK/ERK pathway plays a crucial role in physiological processes and its aberrant activity drives various tumor entities.¹ The three RAF isoforms (A-Raf, B-Raf and Raf-1) are major ERK pathway regulators. Commensurate with its role as a signaling hub, B-Raf, the most potent MEK activator, represents the most frequently mutated protein kinase in cancer. Knock-out/knock-in approaches delivered critical insights into the unique and overlapping functions of Raf isoforms. If Braf deficiency is introduced via the germline, Braf^−/− mice display a lethal placental phenotype and, if deficiency is limited to the nervous system, a myelination defect restricts viability to the first 5 postnatal weeks (see Refs. ^{2, 3} for discussion and references).

Tight regulation of B-Raf activity is achieved by the incompletely understood protein-protein interaction and (de)phosphorylation events guiding B-Raf through its activation cycle.^1,4 In its inactive state, B-Raf resides in an auto-inhibited conformation in the cytoplasm that is stabilized by 14-3-3 proteins binding to phospho-S365 in the N-terminal regulatory moiety and the C-terminus of the kinase (Fig. 1). Upon interaction with Ras, 14-3-3 is displaced from phospho-S365, thereby exposing the kinase domain of B-Raf, which now forms either homo- or heterodimers.¹ In addition, Ras promotes the in cis auto-phosphorylation of T599 and S602 in the evolutionary conserved TVKS-motif of the B-Raf activation loop (AL).^1,5 These phosphorylations induce conformational changes in the kinase domain, leading to the alignment of the so-called C- and R-spine residues, ATP uptake and ultimately MEK phosphorylation.¹ These conformational changes expose, the dimer interface (DIF), a set of amino acid residues forming a contact zone essential for the allosteric activation of the receiver by the activator protomer.^1,4 The relevance of TVKS-motif phosphorylation was demonstrated by experiments showing that substitution of T599/S602 by alanine (AVKA) and phospho-mimetic (EVKD) residues impairs Ras-induced activity and confers transforming properties to B-Raf, respectively.^3,5 The dominance of V600 substitutions in tumor-associated BRAF mutations further underscores that AL phosphorylation mimicking mutations induce conformational changes that cut the B-Raf activation cycle short. Indeed, B-Raf^V600E signals independently of RAS, 14-3-3 binding, critical phosphorylation sites and, although it forms particularly stable dimers in its normal state, an intact DIF.^3,4 Thus, various lines of evidence suggest that AL phosphorylation induces conformational changes in the B-Raf kinase domain, promoting both dimerization and kinase activity (Fig. 1) and that, once this conformation is stabilized by V600E mutation-specific effects,^1,4 AL phosphorylation becomes redundant.

Figure 1.

Upper sketch: Model of the B-Raf activation cycle. Negative feedback phosphorylations disrupt the B-Raf/Raf-1 hetero-dimer, leading to re-formation of the closed B-Raf monomer.¹ The regulatory moiety consists of the Ras-binding domain, the cysteine-rich domain and the conserved region 2 encompassing S365. Lower sketch: Interplay between AL phosphorylation, dimerization and transactivation. Top: Dimerization of B-Raf with Raf-1 induces AL phosphorylation (red arrow) followed by conformational changes promoting DIF formation and Raf-1 transactivation. Middle: The R509H mutation prevents DIF formation and Raf-1 transactivation. Bottom: The AVKA mutation in B-Raf precludes the conformational change leading to DIF formation and Raf-1 transactivation.

In addition to point mutations like BRAF^V600E, chromosomal rearrangements represent alternative tumor-associated BRAF alterations. The resulting fusion oncoproteins lack the N-terminal regulatory moiety, but expose an intact kinase domain with increased homo-dimerization potential.^1,6 Hence, they are regarded as constitutively active, although little is known about their regulatory requirements, e.g. whether they require AL phosphorylation.

To further investigate the relevance of the TVKS-motif for B-Raf signaling in development and physiology, we generated a conditional knock-in allele allowing the production a B-Raf mutant with alanine substitutions of T599 and S602.³ Although this Braf^AVKA allele produces a kinase with significantly impaired activity, mice homozygous for this allele were surprisingly viable, fertile and had a normal life span. Nevertheless, Braf^AVKA mice presented with mild abnormalities in the haematopoietic system, a distinct facial morphology, reduced MEK/ERK phosphorylation in the brain and slight gait abnormalities. Thus, genetic impairment of AL phosphorylation does not phenocopy the lethality of Braf^−/− mice, further supporting a scaffolding role for B-Raf. This concept is supported by (pre)clinical observations showing that drug-bound or kinase-dead B-Raf provokes paradoxical ERK-pathway activation, a phenomenon that underlies therapy resistance, inhibitor promoted secondary neoplasms and restricts the application of clinically approved B-Raf inhibitors to BRAF^V600E/K mutant tumors.^1,7 Interestingly, however, B-Raf^AVKA did not provoke paradoxical ERK activation like the kinase-dead B-Raf^D594A mutant,^3,4,7 suggesting that the inability of B-Raf^AVKA to undergo AL phosphorylation precludes a proper DIF conformation in which it can trans-activate another protomer, e.g., Raf-1 (Fig. 1). Consequently, Raf‐1, which still interacts with B‐Raf^AVKA, cannot be trans-activated, ultimately resulting in impaired MEK activation (Fig. 1). Interestingly, this phenotype mimics the DIF mutant B-Raf^R509H that also interacts with Raf-1, but fails to trans-activate its dimerization partner.⁴ These related phenotypes further illustrate the tight relation between AL phosphorylation and dimerization, as also proposed by structural studies (see Refs. ^{1, 3} for discussion and references).

Given the mild phenotype of Braf^AVKA mice, a pharmacological strategy mimicking the effects of the AVKA mutation might allow quenching B-Raf activity without inducing debilitating side-effects, incl. paradoxical ERK activation. Therefore, we analyzed the requirement for the TVKS-motif in B-Raf oncoproteins. As expected from structural studies showing that B-Raf^V600E is locked in an active conformation¹, alanine substitution of T599/S602 did not affect its signaling potential. In contrast, the AVKA mutation reduced the transformation potential of tumor-associated B-Raf oncoproteins such as the dimer promoting B-Raf^E586K mutant or the FAM131B-B-Raf fusion protein. These findings have two implications. Firstly, they highlight a strategy to target tumors with non-V600E mutants that might resist V600E/K selective drugs, as it has been observed for the vemurafenib-insensitive fusion proteins.^3,6 Secondly, our findings illustrate that B-Raf fusions and oncogenic non-V600E point mutants are not as constitutively active as previously thought, but still retain a certain level of regulation that could be exploited pharmacologically.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.