TRK Gene Rearrangements

The terms (1) gene fusion, (2) gene rearrangement, and (3) genomic alterations are sometimes used interchangeably to describe cancer driving genomic events for which there is an open clinical trial testing a specific Trk inhibitor. Gene transcripts rearrange themselves on the mRNA level creating protein products commonly called splice variants. This cutting and splicing occurs within a gene transcript. It can also occur between different genes. Perhaps a greater understanding of the natural splicing can give us a greater understanding of “unnatural splicing” that occurs on the DNA level to create Trk gene rearrangements.


Kristi Luberg and coworkers of the Department of Gene Technology, Tallinn University of Technology, in Estonia have published a detailed analysis of the TRK1 gene.


  • lacks exon 9, an 18 bp region coding for the juxta membrane region of full length TrkA.
  • is expressed in non-neuronal tissues.
  • binds to nerve growth factor with similar affinity  as TrkAII.
  • binds to  neurotrophin 3 (NT-3) with  weaker affinity  than TrkAII.


  • is the full-length, canonical isoform.
  • is mainly expressed in neuronal tissues.


  • lacks exons 6,7, and 9
  • is missing some of the immunoglobin like domains as well as several glycosylation sites.
  • does not bind NGF
  • exists in a constitutively active, autophosphorylated state.
  • is spliced in response to hypoxia.
  • resides in the endoplasmic reticulum and golgi rather than the plasma membrane.
  • can drive neovascularization in neuroblastoma cells (as reviewed by Luberg 2015).

TrkA borrows  regions from other genes

Prior to the study of Luberg 2015,  the TRK1 gene was known to overlap with the anti-sense direction of sarcoma protein homology 2 domain protein 2A (SH2D2A) and the insulin receptor-related protein (INSRR) genes.  This terminology of sense and antisense can be a bit confusing.

  • The “anti-sense” strand is used to transcribe the messenger RNA.
  • When transcribed, the mRNA has the same nucleotide sequence as the “sense” strand except thymine is replaced by uracil.
  • On chromosome 1 TRK1 occupies the position of 156,815,750 to 156,881,850 base pairs.
  • Also on chromosome 1, “complement strand” INSRR reads 5′ to 3′ from 156,858,920 to 156,840,873 base pairs.
  • Numbering on chromosome 1 starts at the telomer of the short p arm to the telomer of the long q arm.

Modern techniques explore the intricacies of gene overlap

Luberg and coworkers used 5′ RACE to discover new exons in the TrkA transcript that are also parts of introns of the SH2D2A gene.

The newly discovered exon A is located about 45,000 bp upstream of the previously characterized exon 1. Luberg and coworkers (2015) claim that transcription start sites of TRKA exon 1 and the INSRR gene are less than 2000 bp apart. The newly discovered exon D and a coding exon in INSRR  overlap by 80 bp.

Figure 1 a. The TRK1 region on chromosome 1 b. Traditional exons are labeled with numbers. Exons borrowed from neighboring genes are given letter codes

Exons from neighboring genes get spliced into the transcript but are not translated into protein (no fill).

Figure 1 continued c. Splice variants of Trk A Colored boxes correspond to b. protein domains of the Trk A kinase

The difference between canonical TrkA and TrkAα is how much of exon 1 gets translated into protein.  There seems to be an alternative transcription start site in exon 1.  TrkAγ and TrkAδ borrow some of their translated N-termini from other proteins but have the major TrkA domains.  TrkAε lacks the leucine rich extracellular domains but has the Ig repeats and the transmembrane region.  TrkAξ has fewer Ig repeats but has the transmembrane region.  TrkAκ has a translation start site in the middle of exon 12 that leaves the translated protein with only the kinase domain.

Some TrkA splice variants have higher levels of autophosphorylation than the canonical TrkAII.

  • The starred TrkA transcripts(Fig 1c) were loading into vectors with a V5-tag and transfected into  HEK293 cells.
  • Antiboides against the V5-tag were used to retrieve the TrkA splice variants by immunoprecipitation.
  • Antibody against phosphotyrosine was used to assess the degree of autophosphorylation (pY/V5 ratios).  Each experiment was repeated three times.

Figure 2 Autophosphorylation of TrkA splice variants a. Transfected TrkA transcripts also have the V5-tag to distinguish them from endogenous TrkA. Anti-phosphotyrosine antibody was used to detect autophosphorylation b. results from statistical analysis comparing the canonical version of TrkA with its splice variants.

For additional reading, a small molecule Trk treatment has been tested for its ability to inhibit cell growth in cultured cells over expressing some of these TrkA splice variants.


The TRK2 genomic neighborhood is no where near as interesting as that of TRK1.  Many of the neighboring genes are non-protein coding.    Alterations in splicing are interesting in that they have ramifications on the regulation of TrkB fusion proteins.  Various splice variants of exons 1-5 appear in in the 5-UTR in a manner that may  affect translation.

Figure 3 TrkB . a. The TRK2 genomic home on chromosome 9.   b. Borrowing from neighboring genes gives variation in the 5-UTR (non protein coding) region of he TrkB transcript.

One of the remarkable things to realize about the splice variants of TrkB that some of them do not contain kinase domains.


Figure 3 continued c. Splice variants of Trk B  b. Colored boxes correspond to d. protein domains of the d Trk B kinase

TrkB-T-Shc  contains no tryosine kinase domain but does contain an Shc domain that ends in exon 19.  Note that several micro RNAs have been shown to regulate the translation of this transcript (Wong 2014).  Luberg (2010) discussed developmental regulation of various TrkB transcripts.  One may  wonder how much this is due to micro RNA regulation.


We do not know of any extensive mapping of splice variants of TrkC.  Most of our information was obtained from UniProt and NCBI.

Figure 4 TrkC splicing a. The genomic neighborhood of TRK3 on chromosome 15 b. Domains of  translated TrkC.   Amino acid differences in splice variants from the canonical sequence of isoform 1 are also given.  c  Cartoon diagram of the five splice variants of TrkC

Is there any correlation between splicing and gene rearrangements?

Here are some spice sites from

Figure 5. Splice sites of Trk from UniProt


BCAN-TRK1, V418 first TrkA in fusion

LMNA-TRK1, V418 first TrkA in fusion

NFASC-TRK1, T400 first TrkA in fusion

TPM3-NTRK1, D399 first TrkA in fusion

All four of these fusions are close to a splice site.


QKI-TRK2 , P466 is the first amino acid of TrkB.

This fusion is on a splice site.


ETV6-TRK3, V530 is the first TrkC in the fusion

MYO5-TRK3, D411 is the first TrkC in the fusion

These two fusions are on splice sites.


Luberg K, Wong J, Weickert CS, Timmusk T.(2010) Human TrkB gene: novel alternative transcripts, protein isoforms and expression pattern in the prefrontal cerebral cortex during postnatal development. J Neurochem. 113(4):952-64.  PubMed

Luberg K, Park R, Aleksejeva E, Timmusk T.(2015) Novel transcripts reveal a complex structure of the human TRKA gene and imply the presence of multiple protein isoforms. BMC Neurosci. 2015 Nov 18;16:78.  PubMed

Wong J.(2014)Regulation of a TrkB Alternative Transcript by microRNAs. Dement Geriatr Cogn Dis Extra. 4(3):364-74. PubMed


Important information

We have presented some Trk mRNA transcript splicing variations, tied them to functional domains of these kinases, and presented similarities to splice sites of fusion proteins. Some of the studies presented hint the kinase domains might be active even if the fusion partners fail to dimerize. There is an open clinical trial accepting patients with solid tumors containing Trk gene rearrangements and the resulting fusion proteins. The Trk inhibitor in this study binds the kinase domain with  high affinity.