NACC2-TRK2 in astrocytoma

Browsing NGS databases for TRK gene rearrangements can lead to more profound questions regarding their origin.

In our sister website we had only one example of an TRK2 gene rearrangement, QKI-TRK2. The COSMIC database was searched revealing one more TRK2 gene rearrangement. The NACC2 gene codes for the nucleus encumbens-associated protein 2 (UniProt-Q96BF6). TRK2 codes for the TrkB BDNF/NT-3 receptor tyrosine kinase. One of these patients was a seven year old boy with a brain stem astrocytoma. Both publications did not really concern themselves with individual Trk mutations. Perhaps a closer examination of the fruits of NGS would lead to a greater understanding of a category of Trk mutations that are “actionable targets.”

Reference Title Author Year Status COSMIC Pubmed
Recurrent somatic alterations… Jones DT et al 2013 Curated COSP32378 23817572
Pilocytic astrocytoma: pathology,… Collins VP et al 2015 Review COSP39334 25792358

I. Quick overview of the NACC2 and TRK2 genes

  • This information came from COSMIC
  • Chromosome diagrams of the NACC2 and TRK2 genes on chromosome 9 (Genecards) gives an idea of just how  far apart these two genes really are.
  • The NACC2 gene resides on the complement strand of DNA on chromosome 9. Note just how far these two genes are apart on chromosome 9.  Does the fact that one is on the sense and the other on the anti-sense strand mean that a simple inversion is all that is needed to create a gene fusion? Were no other genes deleted in this process?

cartoon showing fusion of NACC2 and TRK2 gene and protein

II From exons to protein domains

One of the beauties of  databases such as COSMIC, NCBI, and on line nucleotide translating programs is that it is easy to take information from COSMIC and determine the approximate protein domains involved in the fusions.

A. NACC2

The NACC2 transcript is easy because it has only five exons. As a proof of concept in reading sequences of the NCBI gene database, a stop translation codon was found at the end of exon5. The “ochre” stop codon of TAA is right above the dark green horizontal bar with the white arrow in it. A red box is also drawn around it. The end exon 4 was found to code for K418, a lysine, in the BEN domain  of the translated protein.

NACC2 transcript coding sequence with stop codons

B. TRK2

Exon counting of the TRK2 gene from the NCBI database is a bit more tricky due to the large number of splice variants and the sheer number of exons in the NTRK2 gene. The top panel is the whole gene in which individual exons are not visible. The next panel zooms in on sequential regions of exons. Finally, a region that appears to be exon 13 is sequenced at Expasy.org. This region codes for a region (388-398) that maps between the last immunoglobin like repeat and the tansmembrane α-helix of TrkB (365-431). It does not look like this exon codes for a well defined protein domain.

Exon counting of the TRK2 gene

III Domains of the fusion protein

A. The nucleus accumbens associated protein 2

The entire BTB dimerization domain is intact in the fusion protein we can only speculate that it is capable of forming dimers. Some but not all of the DNA associating BEN domain is missing. Close proximity to the transmembrane helix of TrkB might hinder what little DNA binding ability it might have.

Domains of the fusion protein NACC2

B. The kinase domain of TrkB

The full kinase domain is present in the fusion. The exon to protein domain diagram was modified from Luberg 2010. Some have speculated that any region to the N-terminus of the transmembrane helix of TrkB will end up on the extracellular side of the plasma membrane. If this were the case, it would allow for BTB domain dimerization in absence of BDNF and NT-3.

 

 

The questions left unanswered by the NGS data

  1. Is there there enough fusion protein to detect by immunohistochemistry with a TrkB antibody or even a pan- receptor tyrosine kinase antibody?
  2. Is this fusion protein active?
  3. Are all of the usual TrkB pathways active or just a few?
  4. Where does the fusion protein localize in the cell?

IV Does the NACC2-NTRK2 fusion make sense?

The nucleus accumbens is associated with reward and addiction pathways. It is assumed that the nucleus accumbens associated protein is expressed in the nucleus accumbens. TrkB is not only expressed in the nucleus accumbens but is considered a target of therapy for various mood disorders PubMed. The protein products of both genes are part of the standard operating of the nucleus accumbens. What is different in the fusion protein is the absence of the requirement for BDNF to activate TrkB kinase activity.

Cartoon of human brain showing reward and addiction pathways with nerve ending blow up showing various kinases

  1. Does having both genes actively transcribed make them more susceptible to rearrangement?
  2. Are these genes also transcribed in the glial cells of associated neurons?
  3. If so, could the brain stem astrocytomas have originated in the nucleus accumbens?

Important information

An open clinical trial testing an ALK, ROS1 and Trk inhibitor is accepting patients with TRK gene rearrangements. An FDA pre-screening test, Trailblaze Pharos, addresses a question raised in this blog. Does the gene rearrangement result in protein expression detectable by immunohistochemistry? The particular small molecule has been successfully used in patients and xenograph bearing nude mice to treat (1) a TRK1 fusion driven glial neuronal tumor of the brain, (2) neuroblastoma xenographs, (3) and  a neuroblastoma combination therapy.