Tyrosine kinase gene rearrangements in epithelial malignancies

Abstract

Chromosomal rearrangements that lead to oncogenic kinase activation are observed in many epithelial cancers. These cancers express activated fusion kinases that drive the initiation and progression of malignancy, and often have a considerable response to small-molecule kinase inhibitors, which validates these fusion kinases as 'druggable' targets. In this Review, we examine the aetiologic, pathogenic and clinical features that are associated with cancers harbouring oncogenic fusion kinases, including anaplastic lymphoma kinase (ALK), ROS1 and RET. We discuss the clinical outcomes with targeted therapies and explore strategies to discover additional kinases that are activated by chromosomal rearrangements in solid tumours.

Figure 1. Molecular aetiology and types of chromosomal rearrangements

a | Several steps are thought to be required for the formation of a pathogenic fusion gene, symbolically shown here as A–B. First, double-strand breaks (DSBs) are initiated by cell-extrinsic mechanisms such as ionizing radiation or by various different cell-intrinsic mechanisms. Second, the ends of the broken DNA have to be brought into close proximity. This juxtaposition can occur after the formation of DSBs (as shown), or before the initiation of DSBs. Third, the DNA ends are aberrantly repaired, probably by alternative non-homologous end-joining (NHEJ). DNA junctions frequently show short stretches of homology that are referred to as microhomology and are indicated in the figure by the asterisk. In the final step, expression of the fusion gene confers a growth and/or survival advantage, which enables clonal selection and expansion. b | The boxes in the figure provide a summary of the different types of chromosomal rearrangements that lead to oncogenic tyrosine kinase fusions. Interchromosomal rearrangements, principally reciprocal translocations, are the most common type (left). However, intrachromosomal rearrangements (right), which include paracentric inversions, intrachromosomal deletion or tandem duplication, are also observed. All known anaplastic lymphoma kinase (ALK) and ROS1 fusions are shown here by the type of chromosomal rearrangement. Of note, complex rearrangements (not shown in the figure) may also lead to tyrosine kinase fusion genes. C2orf44, chromosome 2 open reading frame 44; CCDC6, coiled-coil domain-containing 6; CSR, class switch recombination; EML4, echinoderm microtubule-associated protein-like 4; EZR, villin 2; FIG, fused in glioblastoma; KDELR2, KDEL endoplasmic reticulum protein retention receptor 2; KIF5B, kinesin family member 5B; KLC1, kinesin light chain 1; LRIG3, leucine rich repeats and immunoglobulin-like domains 3; SDC4, syndecan 4; SLC34A2, solute carrier family 34 member 2; STRN, striatin; TFG, TRK-fused gene; TPM, tropomysin; VCL, vinculin; VDJ, variable diverse joining.

Figure 2. Genomic organization of tyrosine kinase rearrangements

Breakpoints (grey arrows), exons (drawn to scale) and introns are depicted. The numbering of the exons corresponds to coding, translated exons. Given the space limitations, sequences that correspond to untranslated regions are not depicted unless relevant (dashes). Regions that encode the transmembrane and kinase domains (as designated in UniProt) are shaded in blue and green, respectively, and were mapped using MapBack. The exon that encodes the GXGXXG motif is indicated with an asterisk. The isoform depicted for each kinase corresponds to the canonical isoform that is designated in UniProt unless a non-canonical isoform (for example, neurotrophic tyrosine kinase receptor type 3 (NTRK3)) or multiple isoforms (for example, fibroblast growth factor receptor 2 (FGFR2) and FGFR3 (alternative exon 7, isoform number in superscript)) were specifically indicated in the primary reference. ALK, anaplastic lymphoma kinase; PDGFRA, platelet-derived growth factor receptor-a.