Challenges and Opportunities in the Crusade of BRAF Inhibitors: From 2002 to 2022

Abstract

Serine/threonine-protein kinase B-Raf (BRAF; RAF = rapidly accelerated fibrosarcoma) plays an important role in the mitogen-activated protein kinase (MAPK) signaling cascade. Somatic mutations in the BRAF gene were first discovered in 2002 by Davies et al., which was a major breakthrough in cancer research. Subsequently, three different classes of BRAF mutants have been discovered. This class includes class I monomeric mutants (BRAF^V600), class II BRAF homodimer mutants (non-V600), and class III BRAF heterodimers (non-V600). Cancers caused by these include melanoma, thyroid cancer, ovarian cancer, colorectal cancer, nonsmall cell lung cancer, and others. In this study, we have highlighted the major binding pockets in BRAF protein, their active and inactive conformations with inhibitors, and BRAF dimerization and its importance in paradoxical activation and BRAF mutation. We have discussed the first-, second-, and third-generation drugs approved by the Food and Drug Administration and drugs under clinical trials with all four different binding approaches with DFG-IN/OUT and αC-IN/OUT for BRAF protein. We have investigated particular aspects and difficulties with all three generations of inhibitors. Finally, this study has also covered recent developments in synthetic BRAF inhibitors (from their discovery in 2002 to 2022), their unique properties, and importance in inhibiting BRAF mutants.

1. Introduction

The rat sarcoma (RAS)/rapidly accelerated fibrosarcoma (RAF)/mitogen extracellular kinase (MEK)/extracellular signal-regulated kinase (ERK) or mitogen-activated protein kinase (MAPK) signaling pathway is a crucial intracellular signaling pathway,¹ responsible for intracellular signal transduction, namely, acute hormone responses, embryogenesis, cell differentiation, and apoptosis involved in the control of cell growth, proliferation, migration, and survival.² By acting downstream of receptor tyrosine kinases, the MAPK pathway regulates cell fate decisions. Under physiological conditions, MAPK activation occurs via extracellular binding of growth factors to receptor tyrosine kinases,³ at the cell surface, leading to phosphorylation and activation of RAS proteins by guanosine-5′-triphosphate (GTP)-bound.⁴ RAF, a downstream effector of RAS, is a serine-threonine-specific protein kinase that stimulates MEK, which in turn activates ERK.⁵ Activation of ERK results in a growth-promoting and transforming signal (Figure 1).^6,7

Figure 1.

BRAF signaling pathway. (green arrow) Normal BRAF pathway. (red arrow) Oncogenic BRAF pathway.

The RAF kinase family includes three members of the MAPK pathway: ARAF, BRAF, and CRAF. BRAF is the major activating kinase,⁸ and its activation occurs by binding to its RAS-binding domain.⁹ It encodes a cytoplasmic serine/threonine kinase and is a major regulator of the RAS pathway (RAS → RAF → MEK → ERK → MAP kinase).^10,11 BRAF is essential for regulation of a variety of cellular processes, including growth, proliferation, survival, and migration.^12,13

The discovery of BRAF mutations was a major breakthrough in cancer research. Somatic mutations in BRAF were first discovered in 2002 by Davies et al.¹⁴ BRAF is among the most frequently mutated oncogenes in human cancers.¹⁵ One of the most studied pathways, with over 90% mutations, is the RAS > RAF > MEK > ERK (MAPK) pathway.¹⁶ BRAF mutations occur to varying degrees in human cancers, including approximately 70–90% in melanoma, 5–20% in colon cancer,^17,18 30–50% in thyroid cancer,¹⁹ 5–30% in ovarian cancer,²⁰ 1–4% in nonsmall cell lung cancer, and 1–3% in hairy cell leukemia and Langerhans cell histiocytosis (Figure 2).²¹

Figure 2.

Global cancer statistics for new cases and deaths for 2020 and BRAF^V600E (%) mutation in different cancer types.

All BRAF mutations are located in the kinase domain, either in the glycine-rich loop encoded by exon 11 or in the activation sequence encoded by exon 15.^14,22 The BRAF gene has a variety of activating mutations that have been discovered and classified into three classes. The monomeric BRAF^V600 codon mutation is classified as class I. In the RAS/RAF/MEK/ERK pathway, this mutation occurs in an RAS-independent manner.²³ The BRAF^V600 gene mutation BRAF^V600E is the most common mutation 90%²⁴ at position 600 with a single point mutation (Val600 → Glu) in which the polar, hydrophilic glutamic acid replaces the hydrophobic valine.²⁵ This breaks the hydrophobic link between the A-loop and the P-loop, resulting in an active conformation.²⁶ The constitutive kinase activity of BRAF^V600E is 500 times higher than that of wild-type BRAF kinase.²⁷ Another type of mutation is BRAF^V600WT, with BRAF^V600K (10%) being the second most common mutation, where valine is replaced by lysine.²⁸ Rarer codon BRAF^V600 mutations are BRAF^V600R, BRAF^V600D (6–3%), and BRAF^{V600E2,V600M,V600G} (<1%).²⁹ The class II includes non-V600 mutants and is regulated by mutant BRAF homodimers without involvement of RAS. Here, L597 and K601 were found in the activation segment, whereas G464 and G469 were found in the glycine-rich region of BRAF. Class III involves non-V600 mutants and is regulated by mutant BRAF heterodimers involving the RAS mutant. In this, D287, V459L, G466 V, S467, G469, N581, D594, F555, and G596 were detected. In BRAF gene mutations, the classes II and III are very rare codon mutations.^23,30

Statistics of new cases and deaths in all cancers worldwide for 2020 and details of specific cancers affected by the BRAF^V600E mutation are shown in (Figure 2).

2. Current Insight BRAF Protein

2.1. BRAF’s Skeletal System and Binding Pocket

The BRAF protein consists of 766 amino acid sequences.^31,32 It has two lobes, the N-terminal and the C-terminal lobes. These two lobes consist of three conserved regions (CRs): CR1 (R150–290), CR2 (R360–375), and CR3 (R457–717). CR1 and CR2 are regulatory domains.³³ CR1 contains the RAS-binding domain (RBD) (R155–227) and the cysteine-rich domain (CRD) (R234–280).^34,35 CR1 interacts with both RAS and membrane phospholipids. CR1 acts as an autoinhibitor of the BRAF kinase domain (CR3) and ensures that signal transduction is controlled. CR2 is a serine/threonine-rich domain that serves as a flexible hinge between CR1 and CR3 and contains a binding site for the 14–3–3 protein.^35,36 CR3 is the catalytic protein kinase domain located near the C-terminus [Figure 3a]. The N-terminus of CR3 is involved in adenosine triphosphate (ATP) binding. It primarily comprises an antiparallel β-sheet structure, anchors, and a glycine-rich ATP-phosphate binding loop (P-loop) (R464–471) and an αC-helix (R492–504). Through the activation segment, the P-loop region is critical for stabilizing ATP binding and maintaining an inactive BRAF conformation. The C-terminal of CR3 contains a hinge region/adenine region (R530–535), catalytic loop (R574–581), DFG motif [Asp-Phe-Gly(R594–595–596)], and an activation section (R594–623). In the CR3 domain, the ATP-phosphate binding loop, DFG motif, and activation segment are the major active binding sites.³⁷⁻³⁹ The ATP pocket has several binding regions that interact with inhibitors. These are an adenine region (similar to the hinge region), a sugar region (ATP ribose pocket), type I and II hydrophobic regions, and a solvent-accessible region that is also part of the ATP region but is not occupied. At the C-terminus, the dimerization interface (DIF) (R504–511) is another binding site for the 14–3–3 protein.³⁶ The end of CR3 facilitates the binding of substrate proteins. This occurs through its catalytic loop, which facilitates the transfer of phosphate from ATP to BRAF substrates [Figure 3b,c].^32,40 The gatekeeper (catalytic cleft between the N- and C-termini) plays an important role by controlling the access of ligands to a hydrophobic pocket deep in the active site that is not in contact with ATP, which affects the binding and specificity of inhibitors.⁴¹

Figure 3.

BRAF protein structure and binding pocket. (a) Regulatory (CR1 & CR2) and kinase (CR3) domain between N and C-terminal in BRAF kinase protein. (b) BRAF protomer (PDB ID:6P7G) with different orientations. (c) BRAF dimer (PDB ID:2FB8). Different binding pockets (p-loop, αC-helix, DIF, DFG, catalytic loop, activation segment, and hinge region) are labeled in specific colors.

2.2. Importance of BRAF Dimerization and Its Challenges

Recent studies have shown that Raf dimerization is required for normal Ras-dependent activation of Raf kinase in various cellular activities.⁴² The active conformation of BRAF is facilitated by dimerization.⁴³ It was discovered that dimerization of BRAF kinases involves insights into a core cluster of residues known as the dimer interface (DIF) [(R504–511) in BRAF], which plays a critical role in dimer formation.⁸ DIF is located at the tail end of the αC-helix in the BRAF kinase protein. Many of the ATP-competitive inhibitors promote BRAF dimerization in an RAS-dependent manner. BRAF dimerization can be homo- or heterodimerized.⁴⁴ The 14–3–3 domain protein in DIF is responsible for BRAF dimerization. The two-protein structure of BRAF–BRAF in the 14–3–3 complex in DIF is responsible for homodimer, and BRAF-MEK1–14–3–3 complex is heterodimer. Two BRAF serine sites pS729 and Ps365 open monomers in the 14–3–3 protein complex, promote formation of active BRAF dimers, and link them.³⁶ Dimerization is enhanced by Ras and is subject to negative feedback regulation by ERK.⁴⁵

The structure of an active BRAF kinase in dimerization is that of a side-by-side dimer, in which only one partner needs to be catalytically active.⁴⁵ Drugs that inhibit BRAF kinase can bind to a protomer of the dimer complex (homodimers and heterodimers of BRAF) and inhibit its ATP-binding site; however, this binding enhances the activity of the protomer (in the absence of drug), leading to the induction of an active conformation.⁴⁶ Although recent BRAF dimer structures have shown that the 14–3–3 dimer can bind to two kinase domain protomers simultaneously, it is unclear whether both protomers are catalytically active.⁴⁷

The RBD:14–3–3 interface has a dual function, contributing first to RAF autoinhibition and second to the full spectrum of RAS-RBD interactions.⁴⁷ There are two types of autoinhibitory mechanisms that must be overcome in BRAF activation. The first is 14–3–3-assisted locking of the N- and C-terminal portions, and the second is destabilization of the inactive conformation of the kinase domain itself.⁴⁶ BRAF dimerization is also associated with multiple phosphorylation events. RAF dimers may have an allosteric function in facilitating activation loop autophosphorylation. The autophosphorylation of the P-loop of RAF is inextricably linked to the catalytic activity of RAF.⁴⁸ The dimer-promoting potential of BRAF inhibitors may be associated with some “paradoxical” activation. Paradoxical triggers may enhance concerted movements of the αC-helix and A-loop regions.⁴⁹ As mentioned earlier, the class II mutation is responsible for the homodimer, and the class III mutation is responsible for the heterodimer. Rajakulendran et al. suggested that BRAF mutations G450W, F488A, M489W, and Y538F at the dimerization interface inhibit dimer formation.⁴⁶ BRAF dimerization contributes to the pathogenic effect of disease-associated mutant Raf proteins and, when activated, exhibits similar activity to the constitutively active V600E mutant. It has also been shown to influence treatment response and disease progression in individuals receiving BRAF inhibitors.⁵⁰ Raf dimerization and its role in normal and oncogenic states in cell signaling is shown in Figure 4.

Figure 4.

Dimerization of Raf in cell signaling. Note: (A) In the normal RAS-dependent signaling pathway, Raf dimerization is required for Raf kinase activation and signaling to MEK for further cell proliferation and regulation. (B) In oncogenic states, it is required for MEK/ERK signaling, which is upregulated by Raf dimerization. These include: (1) mutant c-Raf protein with b-Raf in dimerization impaired kinase activity from normal to oncogenic, (2) RTKs and RasGTPases induced by mutation, (3) in the context of active Ras, treatment with ATP-competitive Raf inhibitors, and (4) Raf inhibitor resistance is mediated by self-homodimerizing BRAF^V600E splice variants.

2.3. Concept of Inactive and Active Conformation of BRAF

BRAF protein switches from inactive to active conformation and is also important for kinase activation. Several regulatory elements play key roles in this process. The inactive conformation of BRAF is characterized by the position DFG-OUT/αC-OUT and is also known as the closed conformation. The active conformation is characterized by the position DFG-IN/αC-IN and is also referred to as the open conformation.⁵¹⁻⁵³ In BRAF protein kinases, the conformation of the conserved DFG motif containing a catalytic [Asp-Phe-Gly (594–595–596)] residue is known to be critical for selective binding of many kinase inhibitors.⁵⁴

In the DFG-OUT conformation, BRAF shows a reverse orientation. The DFG residue [Asp (594)] oscillates and is displaced from the active site, occupying part of the ATP-binding site. The Mg²⁺ of the DFG motif is not bound to ATP, facilitating ATP c-phosphate transfer as the catalytic cleft of BRAF becomes inaccessible to ATP. The glycine-rich loop and activation segment are close to each other due to hydrophobic interactions.^55,56 The DFG-OUT conformation has created an ATP-binding site and an adjacent hydrophobic pocket. In this conformation, the DFG residue moves from the ATP-binding site to the hydrophobic pocket, which is known as allosteric movement extending outward from the ATP-binding site. In the DFG-OUT conformation, the inhibitors bind to the allosteric site (hydrophobic pocket, near the ATP-binding pocket).⁵⁴

In the DFG-IN conformation, the DFG residue [Asp(594)] is in the active site, and the catalytic cleft of BRAF becomes accessible to ATP, so that the Mg²⁺ of the DFG motif is bound to and occupies part of the ATP-binding site. Disruption of the hydrophobic interaction occurs between the glycine-rich loop and the activation segment (due to phosphorylation in the activation segment).^55,56 The DFG-IN conformation has no allosteric movement and no allosteric binding site. In this conformation, inhibitors or agonists form interactions with the hinge region by forming ∼1–3 hydrogen bonds and the adenine region (ATP binding site) by hydrophobic interaction and also interact with other parts of the ATP binding site.⁵⁴

The αC-helix near the ATP-binding site is called the αC-IN position. In this catalytic position, the Glu501 residue of the αC-helix causes the formation of a salt bridge with the conserved K483 residue. The αC-helix away from the ATP site is referred to as the αC-OUT position.^57,58

3. Current Insight into BRAF Inhibitors

BRAF binding modes are classified into four basic categories based on the conformation of the binding pocket of the DFG motif and the αC-helix.⁵⁹ These conformations classify inhibitors as follows and are shown in (Figure 5).

Figure 5.

On the basis of DFG (purple color) and αC (cyan color) movement, four different binding conformations are shown. (a) PDB ID: 4MNF (GDC-0879); type I inhibitors (αC-IN/DFG-IN) (b) PDB ID: 5HI2 (Sorafenib); type II inhibitors (αC-IN/DFG-OUT) (c) PDB ID: 4RZV (Vemurafenib); type I_1/2 or type III (αC-OUT/DFG-IN) (d) PDB ID: 5CSX (BI 882370); type_I/II or type IV (αC-OUT/DFG-OUT).

1. Type I inhibitors (αC-IN/DFG-IN) 2. Type II inhibitors (αC-IN/DFG-OUT) 3. Type I_1/2 or Type III (αC-OUT/DFG-IN) 4. Type_I/II or Type IV (αC-OUT/DFG-OUT).⁶⁰⁻⁶³

3.1. First-Generation BRAF Inhibitors: (αC-IN)

Originally, this class was designed to target the RAS (CRAF) mutant in the MAPK pathway before the BRAF mutation was discovered.^58,64 The first-generation inhibitors (Figure 6) are designed as ATP-competitive small molecules with αC-IN conformation. They are BRAF monomer inhibitors and bind in dimers to the active site of a protomer within an RAF. They are not selective enough to block mutant RAF dimers.^65,66

Figure 6.

First-generation BRAF inhibitors.

Sorafenib (formally known as BAY43–9006) is a “multitargeted” RAF kinase inhibitor that inhibits both BRAF and CRAF. It is a nonselective RAF kinase inhibitor. It was the first Food and Drug Administration (FDA)-approved oral drug in 2005, targeting monomeric BRAF tumors (wild-type and V600E mutant) but ineffective against dimeric mutant.⁶⁷⁻⁷¹ In addition to sorafenib, many small molecules have been developed as BRAF inhibitors, but most are still in the preclinical phase. L-779,450 is a selective BRAF kinase inhibitor, but it lacks relative therapeutic efficacy due to poor bioavailability.⁷²⁻⁷⁵ GW-5074 (GlaxoSmithKline (GSK)) is an RAF inhibitor with stronger (10-fold) inhibitory activity on RAF1 than on BRAF. A phase I study demonstrated the safety and antitumor efficacy of GW-5074 (MG–005) in combination with sorafenib, providing a unique method of antitumor activity that targets cancer cell necroptosis caused by mitochondrial dysfunction and differing from conventional Raf inhibitor therapies.⁷⁶⁻⁸⁰ Clinical studies have shown that GDC-0879 is a potent and selective BRAF^V600E inhibitor in melanoma and colorectal cancer, inhibiting the RAF/MEK/ERK pathway.⁸¹⁻⁸³ ZM-336372 is a pan-RAF inhibitor that was the first BRAF inhibitor investigated but simultaneously hyperactivates and inhibits CRAF, resulting in paradoxical CRAF activity without pathway activation.^78,84−86 SB-590885 is a triarylimidazole moiety that has greater potency and activity for BRAF^V600E mutant than for CRAF in RAF kinases.⁸⁷⁻⁸⁹

3.2. Second-Generation BRAF Inhibitors: (αC-OUT)

Following the identification of BRAF mutations in 2002, second-generation drugs were found by screening inhibitors for the BRAF^V600E48 mutant. The second generation (Figure 7) has an αC-out binding conformation with specific selectivity for the tumor driven by the BRAF monomer to target the BRAF^V600E mutant with a broad therapeutic window. As monotherapies or in combination with other targeted treatments, these inhibitors significantly improve clinical outcomes for patients with BRAF^V600 mutation-driven melanoma and some solid malignancies.^53,75,90

Figure 7.

Second-generation BRAF inhibitors.

Dabrafenib is an ATP-competitive inhibitor of BRAF kinases was first developed under the name GSK2118436 by GlaxoSmithKline (GSK). It suppresses BRAF^V600E more effectively than BRAF^V600WT. The Biopharmaceutics Categorization System (BCS) classifies dabrafenib as a class II.⁹¹⁻⁹⁵ Vemurafenib is the first drug approved by the FDA in 2011 for the treatment of patients with advanced BRAF exon 15 ^V600E mutation metastatic melanoma.⁹⁶⁻⁹⁹ In June 2018, the FDA approved encorafenib (LGX-818), a pyrazolo-pyrimidine-based phenylsulfonamide derivative, for the treatment of BRAF^V600E/K-mutated cancer. Encorafenib, a selective inhibitor of BRAF kinase, is being paired with binimetinib, an MEK inhibitor, to treat metastatic and advanced malignant melanoma. Compared with other second-generation BRAF inhibitors, encorafenib has a longer duration of action.^100−103 BI882370 inhibited the growth of human BRAF-mutated melanoma cells with 100-fold greater efficacy (1–10 nmol/L) than vemurafenib. BI-882370 is a potent and selective inhibitor of both the BRAF^V600E oncogenic mutant and BRAF wild-type.^104−106 XL-281 is an orally bioavailable RAF inhibitor that is generally well-tolerated. XL-281 has a lower rate of keratoacanthoma and squamous cell carcinoma (4%) than BRAF inhibitors such as dabrafenib (6–10%) and vemurafenib (18–26%).^107−109

3.3. Third-Generation BRAF Inhibitors: (αC-IN with Allosteric Binding Site)

Third-generation drugs mainly have the αC-IN/DFG-OUT conformation (Figure 8). They were developed to overcome the problems with BRAF dimerization and paradoxical activation. These inhibitors with αC-IN conformation have specific selectivity for tumors driven by BRAF monomers and RAS-dependent BRAF dimers. They have target specificity for BRAF monomers and RAF dimers with a narrow therapeutic window. These inhibitors with αC-OUT conformation have specific selectivity for tumor-driven BRAF monomers and RAS-independent BRAF dimers. They have target specificity for BRAF homodimers and RAF monomers with a broad therapeutic window. The DFG-OUT conformation exposes an additional hydrophobic binding site directly adjacent to the ATP-binding site, commonly referred to as the “allosteric site.” Some third-generation drugs are pan-RAF inhibitors that inhibit paradoxical activation and are also not activated during RAF dimerization.^{64,66,110,111}

Figure 8.

Third-generation BRAF inhibitors.

AZ-628 is more effective against the BRAF^V600E mutation. AZ-628 compared with sorafenib has shown excellent preclinical results in terms of catalytic inhibition of BRAF protomers and dimers.^112−115 MLN-2480 has a wider cerebral distribution. MLN-2480 is an oral, experimental phase I BRAF^V600E inhibitor with a delayed inactivation rate. The allosteric kinase inhibitor MLN-2480 is used to treat melanoma, colorectal, lung, and pancreatic cancer.^116−119 CCT-241161 and CCT-196969 are equally effective against BRAF^V600E, CRAF, and SFKs (Src family kinase). The development of BRAF^V600E skin tumors was inhibited by CCT-241161 and CCT-196969. The efficacy of CCT-241161 and CCT-196969 was observed in PDXs derived from patients exhibiting innate resistance to vemurafenib.^78,120−122 INU-152 only partially activates the paradoxical pathway in melanoma cells with mutated RAS.^120,123 LY-3009120 inhibits the isoforms of ARAF, BRAF, and CRAF with comparable affinity. LY-3009120 stimulates dimerization of BRAF-CRAF and causes paradoxical ERK activation.^{120,124−126} CEP-32496 is an orally bioavailable potent BRAF inhibitor. It is also known as agerafenib or RXDX-105. CEP-32492 shows high affinity against BRAF^V600E. Agerafenib (CEP-32496) inhibited activation of the ERK, MAPK pathway in neuroblastoma cells.^127−131

BAL–3833 (CCT-3833) works well in in KRAS-mutated tumors because RAF and SRC are important junctions in them. It inhibits both the monomeric and dimeric forms of BRAF.^{78,112,132,133} LSN-307453, which inhibits dimerization of all RAF isoforms, is currently in preclinical testing. LSN-3074753 is more active against all ten mutations than BRAF^V600E.^{120,134−136} BGB-283 (lifirafenib) inhibits the BRAF dimer with potent reversible inhibition of all RAFs as well as EGFR.^120,137,138 RAF-709 inhibits monomers and dimers alike.^120,139,140 PLX-8394 inhibits paradoxical formation of RAF dimers and is more potent than vemurafenib. It is a BRAF inhibitor that can be taken orally and does not lead to paradoxical activation of MAPK.^141−144 Compared to vemurafenib, PLX-7904 inhibits MAPK signaling for a longer period of time, resulting in greater blockade of proliferation and lower viability. In BRAF^V600E melanoma cells, PLX-7904 was found to effectively suppress RAF signaling, whereas it had no paradoxical effects on wild-type cells.^{120,145−147} RAF-265 is a dual inhibitor of BRAF and VEGFR2, preventing both osteoclast development and resorption. RAF-265 shows synergistic antitumor activity with ZSTK-474 in medullary thyroid cancer.^{117,147−149} BGB-659 is able to inhibit class I and class II BRAF mutations because it can bind both monomeric BRAFs and both protomers of an RAF dimer. BGB659 showed higher activity against BRAF^WT kinase.^{117,136,150,151}

3.4. Challenges with BRAF Inhibitors

3.4.1. Challenges with First- and Second-Generation BRAF Inhibitors

First-generation BRAF inhibitors (αC-IN) have a lack of selectivity. They are ineffective against malignancies caused by II and III BRAF mutants due to homo/heterodimerization, leading to a paradoxical outcome.^66,152

Second-generation αC-OUT BRAF inhibitors are extremely selective for cancers that rely on monomeric BRAF species, resulting in a broad treatment window. Resistance to first- and second-generation BRAF kinase inhibitors is mediated by BRAF dimerization. However, they have negative allostery and paradoxical activation and are ineffective in cancers based on dimer BRAF mutations. Second-generation BRAF inhibitors (αC-OUT/DFG-IN) exhibit paradoxical activation and allosteric transactivation of BRAF dimers. As a result of treatment with these inhibitors, healthy (BRAF^WT) cells may be stimulated, leading to secondary malignancies. These drugs are highly effective in BRAF^V600E tumors; however, non-V600 mutated malignancies are resistant.¹⁵³

Secondary skin lesions such as hyperkeratosis, keratoacanthomas, and squamous cell carcinomas may occur in patients on BRAF inhibitors. Secondary melanomas, gastric and colonic polyps, and recurrences of previous cancers have also been observed in patients on BRAF inhibitors.¹⁵⁴ Combinations of BRAF inhibitors and MEK inhibitors increase response rates and prolong progression-free survival (PFS) and overall survival (OS) in patients with BRAF^V600 mutated metastatic melanoma. Although long-term effects have been documented, many patients develop acquired resistance to these drugs.¹⁵³

3.4.2. Challenges with Third Generation BRAF Inhibitors

The ability of inhibitors to efficiently and persistently target both protomers is critical for effective inhibition of such dimers. The goal is to develop third-generation BRAF inhibitors that stabilize BRAF in the αC-IN conformation that can bind to both protomers in dimeric structures while inhibiting cellular CRAF inhibits both monomeric and dimeric forms of RAFs or does not allow dimer formation. The ability of both αC-OUT and αC-IN inhibitors to promote trans-activating dimerization of RAFs, as well as their general mechanism of inhibition, results in a paradoxical stimulation of ERK signaling. αC-IN inhibitors are more potent than αC-OUT inhibitors in suppressing dimeric RAF activity. αC-OUT and αC-IN inhibitors both stimulate dimerization; however, αC-IN inhibitors often do so more strongly because they greatly increase RAF binding to the active RAS. αC-IN inhibitors promote paradoxical activation more strongly than αC-OUT inhibitors.^39,66

Paradoxical Activation by BRAF Inhibitors

Oncogenic BRAF dimers show resistance to BRAF inhibitors and cause paradoxical activation. Paradoxical activation in dimerizing cells by BRAF inhibitors can be explained by two different mechanistic explanations (Figure 9). The first hypothesis states that BRAF is autoinhibited. Another proposed mechanism involves BRAF and CRAF conformational changes induced by physical binding of the RAF inhibitor that promote dimer formation between an uninhibited CRAF protomer and BRAF or CRAF bound to the inhibitor. The phenomenon of paradoxical activation is mostly unknown and hypothetical.^49,154

Figure 9.

Paradoxical activation. Note: Mechanism of autoinhibition: (1) In this case, inhibition of BRAF in the presence of a mutant or growth factor-activated RAS leads to abrogation of BRAF autoinhibition, so that it homodimerizes with BRAF and becomes hyperactivated. Conformational changes: (2, 3) At low doses, the drug binds only one RAF protomer and causes the other to transactivate. (4) At high doses, it binds to and inhibits both RAF dimers, effectively knocking down the signaling complex.

4. Recent Advancements

Following the discovery of BRAF mutations in 2002 and their importance in various cancers, many academic scientists/researchers started to work on it to solve the problem related to different BRAF mutations, dimerization, and paradoxical activation. In this context, various scaffold derivatives, such as pyrazine, imidazole, pyridine, pyrazole, pyrimidine, quinoxaline, etc., and their hybrids were synthesized. Their inhibitory activity against various cell lines such as A375, WM266–4, B16BL6, LOX-IMVI, SK-MEL-5 (melanoma), T-29RKO (colorectal wild type), A549 (lung adenocarcinoma), cancer cell lines, etc., and enzyme kinase assay BRAF^V600E and BRAF^WT were investigated, and their results were published. A summary of the various synthesized scaffolds and their BRAF inhibitory activity and binding conformation in relation to the existing difficulties (from the discovery of BRAF mutant in 2002 to 2022) is reported in Figure 10 and Table 1.

Figure 10.

Various synthesized compounds with parents’ scaffolds as BRAF mutant inhibitors. Note: (*) Cpd No.: Compound number; (**) Parent scaffold: Different derivatives of the parent scaffold were synthesized, but only the most potent compounds based on cell line or enzyme kinase activity were chosen.

Table 1. Different Scaffolds and Their Key Aspects in BRAF Mutant.

Compound No.	Parent scaffold*	BRAFV600E(Enzyme Kinase)/Cell line	IC50(μM)	Key points	Refs
1	Pyrazine	BRAFV600E	1.2	Pyrazine scaffold shows strong selectivity against BRAFV600E mutant compared with CRAF	(155)
2	Imidazo-pyridine	BRAFV600E	0.001	Demonstrated interaction with the DFG pocket	(156)
3	Pyrazolo-pyrimidine	WM266–4	0.30	Bound to the active conformation of BRAF. Fusion of the tropane moiety with the pyrimidine core led to novel, potent BRAF inhibitors that unfortunately did not show good efficacy in cells.	(157)
4	Pyrazolo-pyrimidine	A375	0.28	Introduction of groups interacting with the hinge region of the kinase in the 2-position of the scaffold with potent BRAF inhibitors	(158)
5	Pyrazolo-pyrimidine	BRAFV600E	0.5	Replacement of esters with amides was tolerated with a slight increase in enzyme potency	(159)
6	Amidoheteroaryls (quinoxaline-benzamide)	BRAFV600E	0.001	Association with DFG-out allosteric inhibitors of BRAF kinase	(160)
7	Imidazo-pyridinone	BRAFV600E	0.015	Inhibitors of the type II BRAF, in which the central phenyl ring is linked to the hinge binder and 1,3 substitutions bind to the allosteric pocket of BRAF	(161)
8	Indazolylpyrazolopyrimidine	A375	0.024	Type I BRAF inhibitors. Induced dimerization/activation mechanism. In BRAF mutant xenograft models, it shows good inin vivo good anticancer activity	(162)
9	Imidazole	BRAFV600E	0.3	Allosteric DFG-OUT inhibitors of BRAF kinase. The quinazoline moiety binds to the hinge region, the phenyl ring binds to the gatekeeper residue, and the imidazole moiety binds to the selectivity site.	(163)
10	Imidazopyridin -Imidazole	BRAF	0.142	Type II kinase inhibitors with DFG-OUT pocket conformation. Substituted phenyl groups as C-rings (adjacent to imidazole) resulted in a significant improvement in cellular potency	(164)
11	Pyrazoline	BRAFV600E	4.0	Chemical modification of the pyrazoline scaffold led to the development of a potent and selective BRAFV600E inhibitor. (2-Fluoro,6-hydroxy) substitution enhances cellular potency	(165)
12	Pyrimidin-4-yl-1H-imidazole	WM3629	0.62	Superior antiproliferative activities compared to sorafenib. Selective and potent CRAF inhibitor.	(166)
13	Benzothiazole	BRAFV600E	0.004	DFG-OUT pocket conformation in which substituted anilines replace Phe593, which swings out and interacts with aromatic systems in the hinge region and selectivity pocket.	(167)
14	Isoindoline-1,3-diones	BRAFV600E/WT	0.006	DFG-OUT pocket conformation. Structure-based drug design technology was used to develop both type I and type II BRAF inhibitors with unique chemical scaffolds and hinge binders.	(168)
15	Imidazolopyrazole	A375	0.27	The compound showed competitive antiproliferative activity to sorafenib. Potential and selective BRAFV600E and CRAF inhibitor	(169)
16	Quinazoline	BRAFV600E	0.07	DFG-OUT conformation	(170)
17	4,5-Dihydro-2H-pyrazole 2-hydroxyphenyl	BRAFV600E	0.22	Compound showed potent and strong anti-BRAFV600E activity	(171)
18	Benzylsulfonyl chalcone	BRAFV600E	0.17	More potent and selective BRAFV600E inhibitor compared to BRAFWT	(172)
19	Salicylamide moiety containing 3,5-diarylpyrazoline derivatives	BRAFV600E	0.16	Selective inhibition of proliferation of BRAFV600E and BRAFWT mutants	(173)
20	4,5-Dihydro-1H-pyrazole niacinamide	BRAFV600E	0.20	Compound has highest bioactivity compared to positive control sorafenib.	(174)
21	2,3-Dihydrobenzo [b][1,4]dioxin-containing 4,5-dihydro-1H-pyrazole	BRAFV600E	0.11	More potent than erlotinib and compound with selective BRAFV600E inhibitor activity	(175)
22	Imidazole	BRAFV600E	>10	The novel entity benzo[b]thiophen-3-yl)-2,2,2-trifluoroethanol interacts with the BPII pocket of BRAF, resulting in potent BRAF kinase inhibition in vitro and in cells.	(176)
23	5-Phenyl-1H-pyrazol	WM266.4	1.50	Compound showed high antiproliferative activity comparable to positive control vemurafenib	(177)
24	N-(thiophen-2-yl) benzamide	BRAFV600E	0.63	Developed compound proves to be a potent BRAFV600E kinase inhibitor	(178)
25	(1,3-Diphenyl-1H-pyrazol-4-yl) methyl benzoate	WM266.4	0.94	High antiproliferative effect comparable to positive control vemurafenib and highly selective BRAFV600E in vitro inhibitory activity	(179)
26	5-Phenyl-1H-pyrazole	WM266.4	2.63	Compound exhibits highly selective BRAFV600E in vitro inhibitory activity comparable to positive control vemurafenib	(180)
27	2-(1H-Imidazol-2-yl) pyridine	A375	2.93	The nitrogen of the imidazole ring forms a critical hydrogen bond with Cys531 in the hinge region, while the amide moiety forms two additional hydrogen bonds with Glu500 and Asp593 in the DFG region.	(181)
28	Benzyl 2-(1H-imidazole-1-yl) pyrimidine	A375	1.82	Conformation with two different BRAFV600E/V599E mutants and CRAF inhibitor activity	(182)
29	2-Amino-3-purinylpyridine	A375	1.839	DFG-OUT conformation with BRAFV600E inhibitory activity. The compounds decreased the growth of melanoma cells A375 (BRAFV600E) via the ERK pathway, paradoxical activation was not observed in ERK melanoma cells SKMEL-2 (BRAFWT)	(183)
30	Diaryl-pyrazol	BRAFV600E	0.15	Addition of ortho-hydroxyl to the 4,5-dihydro-1H-pyrazole scaffold enhanced antitumor activity, while expansion of the group at N-1 of pyrazoline was similarly beneficial. The compound is more potent than vemurafenib and erlotinib.	(184)
31	4,5-Dihydro-1H-pyrazole thiazole	WM266.4	0.12	The entire BRAFV600E complex was characterized as a receptor, and the sphere was selected based on the ligand binding position of SB–590885.	(185)
32	Pyrazole	BRAFV600E	0.63	Potent inhibitory activities against BRAFV600E and BRAFWT	(186)
33	Guanidinium-diaryl	T-29	0.9	DFG-OUT conformations. Sorafenib binds to the adenine region of the ATP-binding site while penetrating deeper into the hydrophobic pocket.	(187)
34	Benzo-α-pyrone containing piperazine	BRAFV600E	0.37	The drug induces cell apoptosis and marked DNA fragmentation. Cell cycle arrest in G0/G1 phase in melanoma cells. Compound tightly binds to the crystal structure of BRAFV600E at the active site.	(188)
35	Dioxin-containing triaryl pyrazoline	BRAFV600E	0.04	Selective for BRAFV600E against BRAFWT, CRAF and EGFR and low in toxicity	(189)
36	1,3,4-Triaryl pyrazole	A375	1.82	The compound was tested against BRAFV600E, BRAFWT, and C-RAF. Its type 2-kinase inhibitors bind to the inactive form of kinases (DFG-OUT).	(190)
37	1,3,4-Triaryl pyrazole	SK-MEL-5	1.54	The compound was tested against seven kinases: BRAFWT, BRAFV600E, RAF1, EGFR, P38a/MAPK14, ABL1, and ABL1 (T315l). Good inhibitory activity against mutant BRAFV600E and RAF1 kinases	(191)
38	Purine based cyclopropyl formamide fragment	BRAFV600E	3.49	Inhibits proliferation of SK-MEL-2 cell lines without paradoxical activation of ERK. DFG-OUT conformation inhibits all subtypes of Raf proteins. More potent than vemurafenib and sorafenib	(192)
39	Pyrimidine based cyclopropyl formamide fragment	BRAFV600E	12.6	The compound was tested against seven kinases: BRAFWT, BRAFV600E, A-RAF, C-RAF. DFG-OUT, BRAFV600E inhibitors inhibited A375 cell line proliferation via the ERK pathway without paradoxical activation of ERK in SK	(193)
40	N-(5-Phenyl-1H-pyrazol-3-yl) benzenesulfonamide	WM266.4	1.58	The binding mode of the developed compound is similar to that of the BRAFV600E inhibitor. It consists of a terminal aromatic group (phenyl or pyrazole) filling the allosteric pocket formed by the displacement of the DFG loop and an amide linker connecting an aryl group in the hydrophobic pocket to a hinge-binding heterocycle. The use of an amide linker would allow rapid and efficient screening of many possible hinge-binding groups.	(194)
41	2-(Aminopyrimidin-5-yl)-4-(morpholin-4-yl)-6-substituted triazine	A375	0.12	Compound was tested against PI3K(α,β,γ), mTOR BRAFV600E BRAFWT and CRAF kinase activity, proving to be a PI3Kα and BRAFV600E dual inhibitor.	(195)
42	Pyrazolo[3,4-d]pyrimidine	A375	1.74	DFG-OUT-type pan-RAF inhibitors. High inhibitory activities against both BRAFV600E and VEGFR-2 kinase, comparable to sorafenib act as dual BRAFV600E and VEGFR-2 inhibitor	(196)
43	Pyrazole	A375	0.96	Compound has potential as a BRAFV600E inhibitor compared to vemurafenib	(197)
44	Pyrazino[1,2-a]indole	A-549	0.8	Compound showed both BRAFV600E and antioxidant activity.	(198)
45	1,2,4-triazole	H-460	1.6	Compound showed potent BRAFV6OOE, EGFR and tubulin inhibitory activity	(199)
46	Pyrazole	A375	0.39	Type II BRAFV600E inhibitors in this hit compound showed hydrogen bonding with Cys532 in the hinge region, Asp594 in the DFG motif and Glu501 in the αC-helix. The compound showed no significant effect on phosphorylation of ERK and MEK and is more potent than vemurafenib	(200)
47	Indazole	A375	1.97	The compound does not induce paradoxical effect in normal cells. Shows good inhibitory BRAFV600E activities and antiproliferation activities.	(201)
48	Thioether and nicotinamide containing sorafenib analogs	B16BL6	19.75	DFG-OUT conformation: the synthesized compound hasBRAF, BRAFV600E and VEGFR-2-muti kinase inhibitor	(202)
49	Imidazo[2,1-b]thiazole	BRAFV600E	9.30	The developed compound was tested against leukemia, nonsmall cell lung cancer, colon, melanoma, ovarian, kidney, prostate, and breast cancer and obtained moderate results. The compound showed the greatest potential inhibitory effect against BRAFV600E compared to BRAFWT and CRAF.	(203)
50	N-phenyl-(2,4-dihydroxypyrimidine-5-sulfonamido) phenylurea	A549	0.67	The compound outperformed BRAF kinase inhibition and EGFR kinase inhibition	(44)
51	4-Amino-2-thiopyrimidines	BRAFWT	0.11	Dual VEGFR-2 and BRAF inhibitors. In the designed compound, the 4-amino-2-thiopyrimidine core scaffold is to be located in the central gate region of the inactive DFG-OUT conformations of both enzymes. The hydrophobic substituent at the 4-amino group occupies the hydrophobic posterior pocket on one side. The substituent on the sulfide group, on the other hand, expands to fit into the hinge region.	(204)
52	N-(3-(3-Alkyl-1H-pyrazol-5-yl) phenyl)-aryl amide	BRAFV600E	0.70	Superior selectivity over BRAFV600E and CRAF over BRAFWT kinases	(205)
53	3-Carbonyl-5-phenyl-1H-pyrazole	BRAFV600E	0.10	Paradoxical activation can be avoided by inhibiting BRAFV600E and CRAF more selectively than BRAFWT. In silico studies have shown that 3-carbonyl-5-phenyl-1H-pyrazole (3) has three hydrogen bonds near the hinge site. In addition, the pyrazole and middle phenyl rings of the compound interlock with Phe595 of the DFG motif.	(206)
54	Imidazo[2,1-b]oxazole	BRAFV600E	0.034	The developed compound was tested against leukemia, nonsmall cell lung cancer, CNS, colon, melanoma, ovarian, renal, prostate, and breast cancers and obtained moderate results. Most effective against BRAFV600E compared to BRAFWT and RAf1	(207)
55	Pyrrolo[2,3-b]pyridine	BRAFV600E	0.080	Compound with potent BRAFV600E inhibitor	(208)
56	Imidazole-sulphonamides	BRAFV600E	32.90	The developed compound was tested against leukemia, nonsmall cell lung cancer, CNS, colon, melanoma, ovarian, kidney, prostate and breast cancer, achieving moderate results. Compound with potent BRAFV600E inhibitor with active binding site	(209)
57	1-Aryl-3-[4-(pyridin-2-ylmethoxy)phenyl]urea	A549	2.39	Using the colorimetric MTT technique, compounds were tested in vitro for their antiproliferative activity against cancer cell lines A549, HCT-116, PC–3, and HL7702, a normal human liver cell line. Compound more effective than sorafenib	(210)
58	Quinoline/chalcone/1,2,4-triazole	BRAFV600E	1.1	The compound was tested against pancreatic (Panc-1), breast (MCF–7), colon (HT–29), and epithelial cell lines (A-549). The compound showed binding affinity to the (ATP) active site of BRAFV600E and was comparable to that of vemurafenib. They have potent inhibitory effects on BRAFV600E and EGFR.	(211)
59	7,8-Disubstituted-1,3-dimethyl-1H-purine-2,6(3H,7H)-dione	BRAFWT	115.1	Molecular docking in the active site of BRAFWT revealed that C-8 and C-7 substitutions for the N-7 benzyl and tricyclic derivatives, respectively, were responsible for binding interactions with the DFG loop amino acid Asp593 and the αC-helical amino acid Glu500. The compounds showed promising multikinase activity against PI3Kα, BRAFV600E and BRAFWT	(212)
60	2-Anilinoquinoline-based arylamides	BRAFV600E	0.888	The developed compound was tested against leukemia, nonsmall cell lung cancer, CNS, colon, melanoma, ovarian, renal, prostate and breast cancer and obtained moderate results. Inhibitory activity against BRAFWT, BRAFV600E (DFG-OUT) and C-RAF kinases	(213)
61	Imidazothiazole	BRAFV600E	0.021	The developed compound was tested against leukemia, nonsmall cell lung cancer, CNS, colon, melanoma, ovarian, kidney, prostate and breast cancer and achieved moderate results Compound with BRAFV600E and BRAFWT inhibitory activity.	(214)
62	Imidazol-5-yl-pyrimidine	LOX-IMVI	13.90	Association with interactions with RAS, ribose, hydrophobic pockets and hinge region, additional binding with allosteric pocket of BRAF kinase. With BRAFV600E/p38α inhibitors.	(215)
63	Imidazo[2,1-b]thiazole	BRAFV600E	0.07	Activity against mutant BRAF was higher than for CRAF and BRAFWT. The compounds are more potent than vemurafenib in terms of inhibition of MEK and ERK phosphorylation.	(216)
64	4-(3-Hydroxy anilino)-6-(1H-1,2,3-triazol-4-yl)quinazolines	BRAFV600E	0.051	The compound was tested against BRAFV600E, CRAF EGFR, EGFRT790M, VEGFR-2, and PDGFR-β. It interacts with the DFG-OUT conformation with and strongly inhibits BRAFV600E at low dose compared to CRAF, EGFRT790 M and VEGFR-2.	(217)
65	Imidazol-5-yl-pyridine	BRAFV600E	0.530	The designed compound has been tested against leukemia, nonsmall cell lung cancer, CNS, colon, melanoma, ovarian, renal, prostate and breast cancer and moderate results were obtained. Dual BRAFV600E/p38α inhibitors with high affinity in kinase pockets.	(218)

5. Conclusion

Currently, three types of BRAF mutations are reported, i.e., class I: monomeric mutants (BRAF^V600); class II: BRAF homodimer mutants (non-V600); and class III: non-V600 BRAF heterodimers. Currently, FDA-approved targeted therapy specifically targets BRAF^V600E mutant monomers but is insufficiently effective against non-V600E dimers. Third-generation pan-RAF inhibitors are particularly intriguing because they were designed as “paradox breakers” that do not cause paradoxical BRAF activation but, rather, perform activation in a specific manner that leads to the development of secondary cancers. The mobility of the DFG motif is critical for the selectivity of BRAF inhibitors. Many inhibitors are designed based on the αC-helix and DFG conformation; they are ineffective for current problems such as dimerization and paradoxical activation. Therefore, a structure-based design is strongly recommended to design BRAF inhibitors that exploit the properties of BRAF binding sites, conformational changes, and DFG rearrangement.

In recent years, researchers have synthesized various BRAF inhibitors, and after compiling the data, among these inhibitors, a few target the second and third classes of BRAF mutations as well as have problems like dimerization and paradoxical activation. They should design inhibitors that have binding affinity in the core active site [the nucleotide binding site (ADP or ATP), the DFG motif, the phospho-acceptor site (activation segment) adjacent to the DFG motif, and the helix] and an allosteric binding pocket. In cases of BRAF dimerization, new molecules should aim to bind to both protomers instead of one. These efforts could reduce the number of paradoxical breakers, which would provide new hope for a fourth-generation drug to overcome the current challenges of BRAF.

6. Future Perspective

Many questions regarding the BRAF protein and its regulation remain unanswered. The structure of BRAF kinase is very flexible; different ligands can induce different states of the kinase. The differences in structures may explain the selectivity of BRAF inhibitors for active and inactive BRAF. BRAF monomer, homodimer, and heterodimer protein structures have different mutations. Therefore, based on the above-mentioned mutations and the change of their conformations from active to inactive, further studies are needed.

Because of paradoxical activation, alternative methods of inhibiting Raf (BRAF) signaling using ATP-competitive inhibitors are needed. Little is known about the regulatory mechanisms controlling the formation of BRAF homodimers to heterodimers. In the long term, a better knowledge of RAF regulation would help to develop more effective and better tolerated therapies for BRAF-related malignancies.

Glossary

List of Abbreviations

ADP

Adenosine diphosphate

ATP

Adenosine triphosphate

BRAF

v-RAF murine sarcoma viral homologue B1

CR

Conserved regions

CRD

Cysteine-rich domain

DIF

Dimerization interface

EGFR

Epidermal growth factor receptors

ERK

Extracellular-signal regulated kinase

FDA

Food and drug administration

MAPK

Mitogen-activated protein kinase

MEK

Mitogen extracellular Kinase

RAF

Rapidly Accelerated Fibrosarcoma

RAS

Rat sarcoma

VEGFR-2

Vascular endothelial growth factor receptor-2

Author Contributions

Conceptualization: Pradeep Kumar; Data collection: Adarsh Kumar, Harshwardhan Singh; Writing the manuscript: Ankit Kumar Singh and Pankaj Sonawane; Sketching of figures and data interpretation: Vladislav Naumovich, Prateek Pathak; Writing, review, and final editing of the manuscript: Habibullah Khalilullah, Mariusz Jaremko, Abdul-Hamid Emwas, Amita Verma, Maria Grishina, and Pradeep Kumar.

The APC was funded by King Abdullah University of Science and Technology, Thuwal, Jeddah, Saudi Arabia.

The authors declare no competing financial interest.