SCYL3-NTRK1, HEAT repeats with an active kinase

An exciting and perplexing new TrkA fusion: SCYL3-NTRK1.

Yes, we have yet another chromosome 1 intra-chromosome rearrangement with NTRK1. This time it is with a rather poorly understood protein whose gene name is SCYL3 and protein name is PACE1 (Protein-associating with the carboxyl-terminal domain of ezrin ). This one was reported by Massimo Milione of Istituto Nazionale dei Tumori in Milan, Italy and his colleagues. They used an RNA-based anchored multiplex polymerase chain reaction to amplify the mRNA. Next Generation Sequencing (NGS) was customized for the detection of rearrangements of selected tyrosine kinases and simultaneous identification of fusion partners. The goal was to identify patients with solid tumors that might be good candidates for a Trk treatment.

SCYL3-TRK1, overview of a new Trk Fusion

PACE1 is a member of a family of proteins that contain a domain that resembles a protein kinase but lacks the necessary residues for phospho-transferase. Replacement residues may allow the SCYL family members to serve different functions (Pelletier 2016).

  • The protein product of the SCYL1 gene is also known as the “protein with the N-terminal kinase like domain (NTKL).
  • The protein product of SCYL2 gene is also known as the coated vesicle associated kinase of 104kDa (Pelletier 2016). This observation influenced the interpretation of the “clod-like” distribution of the PACE1-TrkA fusion protein in the colon adenocarcinoma.
  • PACE1 , protein that associates with the C-terminus of ezrin, was identified on a fishing expedition to “catch” binding partners of ezrin (Sullivan 2003).

SCYL gene products are proposed to regulate intracellular Golgi trafficking.

Another chromosome 1 gene rearrangement

We have gone into the acrobatics of chromosome 1 gene rearrangements that result in Trk fusion proteins on another website. Milione and coworkers (2017) attributed this gene rearrangement to an inversion rather than a deletion.

Ensembl picture showing location of SCYL3 and TRK1 gene location on chromosome 1

Figure 1 Locations of SCYL3 and TRK1 genes on the q arm of chromosome 1.

SCYL3/PACE1 expression is tricky to assess

Two sources of data were used for proteomics data. Genecards uses mass spectrometry data whereas ProteinAtlas uses coverage and intensity of immunohistochemistry staining (Figure 2, panels A and B, respectively). Proteomics data does not show any appreciable expression of PACE1 in the colon (Figure 2A). IHC shows moderate expression in the colon and in the rectum.

figure A shows PACE1 expression in various tissues and figure B shows imunohistochemical expression in colon and rectum as marked with arrowsdata

Figure 2. PACE1 expression in normal tissue A. Genecards.org, based on proteinics data B. ProteinAtlas.com, based on immunohistochemical data

 

PACE1 is expressed in both colonic glandular cells as well as endothelial cells in two normal colon slices (Figure 3 B,C). One their website Protein Atlas claims that PACE1 cannot be found in most of the cancerous cells in the colon tumors. One might surmise that the PACE1 staining that seems to be present is in normal cells.

IHC stain showing PACE1-TrkA expression in various human tissues

Figure 3. The expression of A. the PACE1-TrkA fusion in a colon adenocarcinoma using TrkA antibody immunohistochemistry Milione 2017 B, C. in normal colonic tissue with ProteinAtlas codes and D. in a colonic adenocarcinoma with the ProteinAtlas code.

Protein domains in the  PACE1-TrkA fusion protein

The fusion contains the N-terminal myristoylation site of PACE1 and the four HEAT repeats, but lacks the ezrin biniding domain. The TrkA component of the  fusion contains most of the juxta membrane domain of TrkA complete with SQSTM1 and SHC1 binding domains. It should be remembered that the kinase domain of PACE1 is inactive.  Note that a fusion junction occurs at a splice site for isoform 2. Trk gene rearrangement at splice sites is not unusual.

Diagram showing protein domains of SCK3/PACE1 and TrkA.

Figure 4 domains of PACE1 and TrkA. Note that the fusion site occurs in the region of a splice variant of PACE1.

Can the function of SCYL protein products explain PACE1-TrkA localization in a tumor?

Milione and coworkers had a nice but brief discussion of the localization of the PACE1-TrkA in regards to the function of protein products of the SCYL  gene family.  Green boxes are regions of coiled coil. The tan box in SCYL1 is a proline rich region of unknown function.  Tannish-pink boxes are HEAT repeats.

Comparison of PACE1-TrkA protein domains relative to SCYL gene family

Figure 5. SCYL domains are from Pelletier 2016. The mTOR structure is from Yoshimura and Hirano (2016)

SCYL1

  • Coatomer (COPI)-coated vesicles facilitate membrane trafficking in secretory pathways.
  • ARF4 is a small GTPase.
  • SCYL1 links the two.
  • HEAT domains mediate  homo-oligomerization and tRNA binding (Pelletier 2016).
  • GORAB, 58k, and Arf4 are Golgi proteins.
  • RNA transferase is the export of tRNA from the nucleus to the cytosol.

SCYL2

  • Vpu is an HIV protein.
  • Clathrin is a coated vesicle protein

SCYL3

  • N-terminal myristoylation targets proteins to membranes.
  • The protein ezrin links the cell membrane to the actin cytoskeleton.

mTOR

  • mTOR is a serine/threonne protein kinase that links protein synthesis and cell division to nutrient availability, among other things.
  • mTOR oligomerization in response to amino acids is mediated by HEAT domains ( Takahara 2006)

The PACE1-TrkA fusion protein is expressed in cultured cells and is subjected to a Trk Treatment when the patient cannot.

The patient was a 61-year-old female diagnosed with an adenocarcinoma of the right colon, infiltrating the pancreas.

  • The patient’s cancer progressed on an immuno therapy directed against the epidermal growth factor receptor (EGFR). This agent goes by the name of FOLFOX and panitumumab.
  • The EGFR antibodytherapy was followed by FOLFIRI-aflibercept. This agent prevents tumor blood vessel growth by acting as a trap for vascular endothelial growth factor.
  • As part of a wide phase I screening program at  institutions of Milione and coworkers,  a deeper molecular characterization of the patient’s primary tumor was performed.
  • This evaluation was  for a  clinical trial testing a  TRK inhibitor.

In addition to the SCYL3-TRK1 rearrangement, results indicated that the tumor was wild type for

  • RAS
  • BRAF
  • EGFR

A high microsatellite instability (MSI-H) profile was also present.

In other words, it was deemed likely that the SCYL3-TRK1 fusion was the driver of her cancer. Unfortunately, the conditions of this patient rapidly deteriorate making it impossible to treat her with entrectinib. The gene for the PACE1-TrkA fusion protein was transfected into Ba/F3 cells for testing with entrectinib. Ba/F3 cells do not express endogenous TrkA. These cells also require interleukin 3 (IL3) for growth.  Without IL3, they die.  Milione and coworkers found that transfection with the PACE1-TrkA gene allowed growth in the absence of IL3.

The growth IC50 for TrkA inhibitors in PACE1-TrkA expressing cells.

The inhibitory concentration 50 (IC50) is the concentration needed to kill 50% of the cells. Recall that the Ba/F3 cells are not viable in the absence of IL3 unless they have an active PACE1-TgrkA fusion protein. Three TrkA inhibitors were tested: entrectinib (IC50= 1.5nM) , LOXO-101 (11.2nM) and the FDA approved criznotinib (IC50=160nM). The log scale in Figure 6A occludes just how much more effective entrectinib is than the current FDA approved small molecule criznotinib.

Graph and Bar chart shpwing inhibitory concentration 50 (IC50) needed to kill 50% of Ba/F3 cells testing Three TrkA inhibitors: entrectinib (IC50= 1.5nM) , LOXO-101 (11.2nM) and the FDA approved criznotinib (IC50=160nM). The log scale in Figure A shows that entrectinib is way more effective than the current FDA approved small molecule criznotinib.

Figure 6. The growth inhibitory concentration 50 of TrkA inhibitors A. From Milione 2017 B. The IC50 in bar graph form for easy visualization of relative potency.

The ability of two TrkA inhibitors to reduce downstream activation is compared

Three cell signaling pathways are activated by the Trk family of receptors. The ability of criznotinib and entrectinib to inhibit the PLCγ pathway were compared (Figure 7A) using Western blot analysis and antibodies as indicated (Figure 7B).

Cartoon on left side shows PLYCgama signalling and a western blot on right side shows levesl of pTRKA-Y490, Pan TrkA, pPLCgama-Y783, PLCy as related to alpha tubulin levels

Figure 7 TrkA inhibitors and PLCϒ signaling A. PLCϒ is one of three signaling pathway activated by Trk family signaling B. Western blot analysis was used to compare entrectinib and crizotinib.

  • p-TrkAY490. This antibody is specific for one of the TrkA autophosphorylation sites. Autophosphorylation is needed to active kinase activity of TrkA. Note that it is high in absence (0 nM) of either TrkA inhibitor. Also note that entrectinib is more effective than crizotinib in reducing TrkA auto activation.
  • panTrkA. This antibody recognizes TrkA without regard to phosphorylation status.
  • p-PLCγ-Y783. This antibody is specific for PLCγ that has been phosphorylated (activated) by TrkA or related receptor tyrosine kinase. Note that entrectinib is more effective than criznotinib in reducing phosphorylation/activation.
  • PLCγ. This is a general antibody that recognizes all PLCγ regardless of phosphorylation status.
  • Tubulin. This is a loading control to prove that the same amount of protein was loaded in each lane.

Entrectinib induced cell cycle arrest

A brief overview of the cell cycle is given in figure 8.  Propidium iodine is a fluorescent DNA stain used to visualize fixed cells in a procedure called flow cytometry.

Diagram showing how Flow cytometry is used to count cells

Figure 8 The cell cycle and checkpoints. Flow cytometry is used to count cells as they flow by a detector that measures the intensity of propidium iodine stained DNA.

Flow cytometry was used to compare two different concentrations of entrectinib in PACE1-TrkA transformed cells.

Flow cytometry graphs showing effect of two different concentrations of entrectinib in PACE1-TrkA transformed cells.

Figure 9. Entrectinib inhibits the cell cycle

Note that just 10 nM entrectinib increases the percentage of cells in the G0/G1 phase and decreases the percentage of cells in the G2/M mitosis phase.

Entrectinib induced apoptosis

The role of the Trk ligand nerve growth factor (NGF) deprivation in neuronal apoptosis has been addressed in a review (Franklin 2011).  From the Franklin review (Figure 10A) we know that

  1. NGF deprivation results in BAX (bcl-2 associated protein X) forming pores in the mitochondrial membrane that release  cytochrome C (CC).
  2. CC binds to the apoptosome activating caspase 9.
  3. Caspase 9 cleaves pro-caspase 3 and 7 to active enzymes that can degrade cellular protein.

The Franklin review got into an excellent discussion of the influence of reactive oxygen species in neuronal apoptosis.  A fine point is that bcl-2 is not only regulated by growth factor receptor phosphorylation, but it also prevents BAX pore formation (Figure 10B).

Fig A Cartoon showing intrinsic apoptotic pathways. Fig B shows Bax interaction with bcl-2 in mitochondra. Fig. C graph shows relative caspase activity at different concentrations of entrectinib.

Figure 10 Entrectinib induces apoptosis in cells transfected with the SCYL3-TRK1 gene A. The intrinsic apoptotic pathway in neurons B. a close up of the interaction BAX oligomers and heterodimer with bcl-2 in the mitochondria. C. Dose dependent activation of caspases 3 and 7 by entrectinib (Milione 2017).

Additional Reading

The clinical trial of the Trk inhibitor is also open for patients with solid tumors with ROS1 and ALK gene rearrangements.  ALK gene rearrangements have also been found in colorectal adenocarcinoma.

Entrectinib inhibits the growth of TPM3-NTRK1 transformed cells from an actual colorectal cancer patient.

A patient with colorectal cancer driven by a LMNA-NTRK1 gene fusion has been treated with the Trk inhibitor entrectinib.

References

Milione M, Ardini E, Christiansen J, Valtorta E, Veronese S, Bosotti R, Pellegrinelli A, Testi A,(2017) Identification and characterization of a novel SCYL3-NTRK1 rearrangement in a colorectal cancer patient. www.impactjournals.com/oncotarget/

Franklin JL. (2011) Redox regulation of the intrinsic pathway in neuronal apoptosis. Antioxid Redox Signal.14(8):1437-48.  Review. PubMed

Pelletier S.(2016)SCYL pseudokinases in neuronal function and survival. Neural Regen Res.11(1):42-4. PubMed

Sullivan A, Uff CR, Isacke CM, Thorne RF. (2003) PACE-1, a novel protein that interacts with the C-terminal domain of ezrin. Exp Cell Res. 284:224–238. PubMed

Takahara T, Hara K, Yonezawa K, Sorimachi H, Maeda T.(2006)Nutrient-dependent multimerization of the mammalian target of rapamycin through the N-terminal HEAT repeat region. J Biol Chem.281(39):28605-14. PubMed

Yoshimura SH, Hirano T (2016)HEAT repeats – versatile arrays of amphiphilic helices working in crowded environments? J Cell Sci. 129(21):3963-3970. PubMed

 

Leave a Reply