NACC2-NTRK2 promotes ligand-independent kinase and downstream activation

NACC2 consists of 587 residues (UniprotQ96BF6) encoded by 5 exons, whereas NTRK2 contains 822 residues (UniprotQ16620) encoded by 21 exons. NACC2-NTRK2 contains the NACC2 BTB dimerization domain and partial BEN domain, the NTRK2 transmembrane helix, and the NTRK2 tyrosine kinase domain (Fig 1A). The normal activation of the NTRK2 kinase relies on dimerization, which is induced by BDNF ligand binding to the extracellular domain. To examine the transforming activity of the NTRK2 kinase domain, in the context of the fusion protein NACC2-NTRK2, focus-forming assays were preformed in NIH3T3 cells. Transformation assays represent one of the original assays used to identify and characterize oncogenes [20–23]. Transfected NIH3T3 cells expressing an oncogene are able to overgrow the contact-inhibited monolayer of the cells and establish foci. The expression and efficiency of the transfections are determined by plating an aliquot of the transfected cells into selection media containing Geneticin. The colonies that form are then counted and used to quantitate the focus-forming ability of each clone. We included BCR-FGFR1 fusion protein, previously described, as a positive control [24]. Using this assay, expression of wild-type NACC2, NTRK2, and NACC2-NTRK2(KD) showed no focus-forming activity, whereas the NACC2-NTRK2 fusion showed high transformation activity, indicating that an active NTRK2 kinase domain is essential for the cell transforming ability of NACC2-NTRK2 fusion proteins (Fig 1B and 1C).

Next, we examined the downstream signaling activity that results from NACC2-NTRK2 activation. NTRK2, NTRK2(KD), NACC2-NTRK2, and NACC2-NTRK2(KD) were transfected into HEK293T cells, previously utilized to characterize downstream signaling stimulated by the oncogenic fusion protein FGFR3-TACC3 [25]. NTRK2(KD) was included as a negative control. We analyzed activation of MAPK, PLCγ1, and JAK/STAT pathways, shown previously to be activated in response to NTRK2 signaling [26]. Cells were collected and lysed in RIPA, and then analyzed by SDS-PAGE and immunoblotting with appropriate antibodies. Activation of downstream pathways was only observed in lysates expressing NACC2-NTRK2, as shown by immunoblotting for P-NTRK2, P-MAPK, P-PLCγ1, and P-STAT3 (Fig 1D, lane 4, 1^st, 3^rd, 5^th, and 7^th panels). In contrast, no activation of these pathways was observed in cells expressing NTRK2, NTRK2(KD), of NACC2-NTRK2(KD). Importantly, the immunoblot for total NTRK2 (Fig 1D, 2^nd panel) shows approximately equivalent expression of NTRK2 and the NACC2-NTRK2 fusion protein, both wild type and KD. Together, these data demonstrate that a functional tyrosine kinase domain in NACC2-NTRK2 is required for activation of downstream signaling pathways and transformation of NIH3T3 cells.

The NACC2 BTB domain is responsible for multimerization

After confirming the oncogenic activation of the NACC2-NTRK2 fusion protein, we tried to elucidate the function of the N-terminal fusion partner NACC2. In the fusion, the entire BTB domain (residue 20–120), part of the BEN domain (residue 351–418), and a large disordered region between these two domains are retained. We endeavored to characterize the relationship between the BTB domain, protein multimerization, and activation of the associated tyrosine kinase domain. We also wished to identify important regulatory residues whose disruption would result in inactivation of the fusion. Currently, there exists no detailed structure-function analysis of the NACC2 BTB domain. However, the BTB domain is an evolutionarily conserved domain retained by various protein families in different species [27]. We performed multisequence alignment of the NACC2 BTB domain with four other BTB domains that have all been structurally determined: NACC1, PLZF, BCL6, and BACH1 (Fig 2A). Multisequence alignment indicated that several conserved motifs that have been identified in BTB proteins such as PLZF and BCL6, also exist in NACC2. These motifs include a charged pocket motif that consists of a negatively charged aspartic acid (D) together with a positively charged arginine (R) or lysine (K), and a monomer core motif that consists of a tyrosine residue. These two motifs exist in all five BTB domain proteins presented (Fig 2A). Using Chimera, we compared the PLZF BTB domain structure (PDB ID: 1BUO) with the NACC2 BTB domain predicted by AlphaFold (AF-Q9DCM7-F1) [17,28]. The PLZF BTB domain and NACC2 BTB domain are highly similar in both structure and sequence (Fig 2A and 2B); thus, due to the conservation of BTB domains, mutations in conserved residues that disrupt multimerization in PLZF and BCL6 [17] could potentially disrupt multimerization of the NACC2 BTB domain (Fig 2C). Based on this comparison, we introduced charged pocket (D31N R45Q) and monomer core (Y86A) mutations into the BTB domain of the NACC2-NTRK2 fusion protein and examined their effects on multimerization.

The BTB domain mutants were expressed in HEK293T cells, collected, and lysed in E1A lysis buffer to stabilize protein-protein interactions [19,30]. Samples were prepared in non-reducing sample buffer and resolved by SDS-PAGE followed by immunoblotting. Using this protocol, high molecular weight multimeric complexes were observed (Fig 2D, left panel, lane 3). The approximate molecular weight of the observed multimeric complexes suggests that NACC2-NTRK2 primarily forms a quaternary structure. Incorporation of either the D31N/R45Q or Y86A mutations resulted in a significant reduction in homo-multimer formation (Fig 2D, left panel, lanes 4–5). Quantitation of the ratio of the multimer band compared to the monomer band (Fig 2E) indicated a significant reduction from 44% for the wild-type NACC2-NTRK2 fusion to 18% due to the charged pocket mutations D31N/R45Q, and a reduction from 44% to 12% due to the monomer core mutation Y86A. Surprisingly, the data indicate that all of the NACC2-NTRK2 fusion proteins multimerize more efficiently than NACC2 alone, which exhibited a multimer/monomer ratio of only 7%.

When these same samples were prepared with sample buffer containing β-mercaptoethanol to reduce disulfide bonds, multimeric bands were reduced to monomeric proteins (Fig 2D, right panel). This suggests the existence of covalent interactions stabilizing NACC2-NTRK2 multimerization.

In addition, we employed an assay for hetero-multimerization. The BTB domain mutants were expressed either alone or in combination with wild-type NACC2 in HEK293T cells (Fig 3A). Lysates were immunoprecipitated using antibodies against NTRK2, after which these immune complexes were analyzed by SDS-PAGE and immunoblotting using antibodies against NACC2. We postulated that the BTB domain present in NACC2-NTRK2 would undergo multimerization with the BTB domain of wild-type NACC2. As anticipated, NACC2-NTRK2 was observed to bind NACC2 (Fig 3A, bottom panel, lane 4), while incorporation of either the D31N/R45Q or Y86A mutants resulted in diminished binding of NACC2 (Fig 3A, bottom panel, lanes 5 and 6). Using the bottom panel ofFig 3A together with similar replicates, quantitation of the ratio NACC2 relative to the NACC2-NTRK2 fusion with which it was co-immunoprecipitated was determined (Fig 3B). This revealed a significant reduction from 72% for the wild-type NACC2-NTRK2 fusion to 45% due to the charged pocket mutations D31N/R45Q, and a reduction from 72% to 15% due to the monomer core mutation Y86A. These changes were statistically significant when compared in a two-tailed paired t test. These data indicate that introduction of either the charged pocket mutations or the monomer core mutation into NACC2-NTRK2 resulted in a significant reduction in its ability to bind wild type NACC2 in this hetero-multimerization assay.

Taken together, the assays examining multimerization of NACC2-NTRK2 and hetero-multimerization with NACC2 both demonstrate the importance of the BTB domain in the formation of these complexes. Both assays also demonstrate the ability of mutations within the BTB domain, such as the charged pocket (D31N/R45Q) and monomer core (Y86A) mutations, to disrupt multimerization.

NACC2-NTRK2 fusion activity requires BTB domain multimerization

The extracellular ligand BDNF induces dimerization of NTRK2, resulting in auto-phosphorylation of the NTRK2 kinase domain. In an oncogenic fusion protein such as NACC2-NTRK2, ligand-mediated activation is impossible due to deletion of the extracellular ligand-binding domain. This necessitates dimerization by another mechanism, usually provided by the N-terminal fusion partner [31,32]. We next examined the importance of a functional BTB domain in NACC2-NTRK2 for activation of downstream signaling pathways, as indicated by phosphorylation of MAPK, PLCγ1, or STAT3. In addition to the charged pocket (D31N/R45Q) and monomer core (Y86A) mutations discussed above, we also included an additional mutation, ΔBTB, which deletes the entire BTB domain. Results show that the specific BTB point mutations D31N/R45Q and Y86A were able to reduce PLCγ1 and STAT-3 activation to the same extent as deletion of the entire BTB domain itself (Fig 3C, 5^th and 7^th panels). All three BTB domain mutations were observed to partially reduce receptor autophosphorylation, as detected by immunoblotting for phospho-NTRK2 (Fig 3C, top panel). Somewhat surprisingly, the BTB domain mutations did not appear to significantly reduce MAPK activation, which remained similar to the wild-type NACC2-NTRK2 fusion (Fig 3C, 3^rd panel). Importantly, the immunoblot for total NTRK2 (Fig 3C, 2^nd panel) shows approximately equivalent expression of the NACC2-NTRK2 fusion proteins examined here.

Using multiple replicates of immunoblots, the relative changes were quantitated as shown inFig 3D. For all four phospho-proteins examined, P-NTRK2, P-MAPK, P-PLCγ1, and P-STAT3, the difference between wild-type NACC2-NTRK2 versus the KD derivative was statistically significant. For P-MAPK activation, the differences between wild-type NACC2-NTRK2 and the mutants D31N/R45Q, Y86A, and ΔBTB were not significant, suggesting that robust MAPK activation can be achieved by a NACC2-NTRK2 mutant which is inactive in other assays, but which retains an intact tyrosine protein kinase domain. However, for the activation of P-PLCγ1 and P-STAT3, the loss of activity exhibited by the D31N/R45Q, Y86A, and ΔBTB mutants was statistically significant in every case as shown inFig 3D.

When examined for biological activity using NIH3T3 cell transformation, all three BTB domain mutants were completely inactive in their ability to produce foci of transformed cells (Fig 3E).

Collectively, these data demonstrate that disruption of oligomerization mediated by the BTB domain, either by specific point mutation or by deletion, results in inactivation of the NACC2-NTRK2 fusion protein as evidenced by decreased activation of downstream PLCγ1 and STAT-3 signaling pathways and by loss of biological activity in a morphological transformation assay. The continued activation of MAPK signaling activity by these same BTB domain mutations suggests that activation of this pathway alone is insufficient to support the biological activity of NACC2-NTRK2.

NACC2 residues 121–418 stabilize NACC2-NTRK2 multimerization via covalent interactions

To investigate the possible role of the retained portion of NACC2 downstream of the BTB domain, which includes the disordered region (residues 121–350) and the beginning of the BEN domain (residues 351–418), we constructed a shorter NACC2:NTRK2 fusion protein that contains only the BTB domain, residues 1–120, referred to as BTB-NTRK2 (Fig 4A). We examined activation of downstream signaling pathways, protein multimerization, and biological activity for BTB-NTRK2 in comparison with NACC2-NTRK2. This fusion protein results in similar levels of NTRK2 autophosphorylation, as well as similar levels of MAPK activation (Fig 4B, 1^st and 3^rd panels). However, phosphorylation of PLCγ1 and STAT3 by BTB-NTRK2 was diminished relative to NTRK2-NACC2 (Fig 4B, 5^th and 7^th panels). Importantly, the immunoblot for total NTRK2 (Fig 4B, 2nd panel) shows approximately equivalent expression of NACC2-NTRK2 in comparison with BTB-NTRK2. Quantitation of replicate immunoblots (Fig 4C) revealed that BTB-NTRK2 exhibited significantly decreased phosphorylation of PLCγ1 and STAT3, although no significant change was observed in P-NTRK2 or P-MAPK.

We next examined the protein multimerization of BTB-NTRK2. As described above, samples were prepared in non-reducing sample buffer and resolved by SDS-PAGE to examine multimerization. As before, NACC2-NTRK2 exhibited significant multimerization (Fig 4D, lane 2); in contrast, BTB-NTRK2 did not exhibit detectable multimerization (Fig 4D, lane 3). This assay demonstrates that the NACC2-NTRK2 multimer is partially resistant to SDS denaturation under non-reducing conditions, whereas a putative BTB-NTRK2 multimer was not detectable under these conditions.

In the hetero-multimerization assay, BTB-NTRK2 was examined for its ability to bind wild-type NACC2 when co-expressed. Cell lysates were immunoprecipitated using antibodies against NTRK2, after which they were analyzed by SDS-PAGE and immunoblotting with antibodies against NACC2. Binding of NACC2 was observed with both NACC2-NTRK2 and with BTB-NTRK2 (Fig 4E, panel 3, lanes 4 and 5), and the location of the co-immunoprecipitated NACC2 is indicated by the red arrowheads.

When biological activity was examined by NIH3T3 transformation assay, BTB-NTRK2 exhibited only 25% transformation activity relative to NACC2-NTRK2 (Fig 4F). Considered together, the results of the multimerization assay and the hetero-oligomerization assay, together with the transformation assay, indicate that the BTB domain alone is sufficient for oncogenic activation, but residues 121–418 apparently contribute to the covalent stabilization and greater activation of NACC2-NTRK2 compared with BTB-NTRK2.

The NTRK2 transmembrane helix contains a putative degron

The NACC2-NTRK2 fusion utilized thus far throughout the present work, with exon 4 of NACC2 joined to exon 13 of NTRK2, is known to cause pilocytic astrocytoma. Going forward, we will refer to this fusion as NACC2-NTRK2(ex4:ex13) (Fig 5A, top). However, an isoform of NACC2-NTRK2 exhibiting a different breakpoint, NACC2-NTRK2(ex4:ex15) (Fig 5A, bottom), was observed in a case of pediatric glioblastoma [10]. We wished to investigate possible differences between these two NACC2-NTRK2 fusions that might explain the more severe pediatric glioblastoma resulting from the exon4:exon15 fusion. The most notable difference between these two fusions is the presence or absence of the TM helix. During the normal translation of NTRK2, an N-terminal signal peptide directs co-translational membrane insertion of the nascent polypeptide, a process which is then arrested upon membrane insertion of the TM domain to produce a Type I transmembrane protein. Lacking an N-terminal signal peptide, due to the substitution of the N-terminal domain by NACC2, either fusion of NACC2-NTRK2 is incapable of membrane insertion; consequently, the normal function of the TM helix as a membrane anchor is abrogated.

To approach this experimentally, we first deleted residues 431–454 of NACC2-NTRK2(ex4:ex13) to create NACC2-NTRK2-ΔTM (Fig 5A, middle); this deletes residues that encode the TM helix. This was first assayed for its downstream signaling activity and transformation ability in comparison with NACC2-NTRK2(ex4:ex13). Results show that NACC2-NTRK2-ΔTM increases NTRK2 autophosphorylation, as well as increasing PLCγ1 and STAT3 phosphorylation (Fig 5B, 1^st, 5^th, and 7^th panels). Importantly, the immunoblot for total NTRK2 (Fig 5B, 2nd panel) shows approximately equivalent expression of the mutants examined here: NACC2-NTRK2(ex4:ex15), NACC2-NTRK2(KD), and NACC2-NTRK2-ΔTM. Quantitation of replicate immunoblots (Fig 5C) revealed that NACC2-NTRK2-ΔTM exhibited significantly greater phosphorylation of P-NTRK2, P-PLCγ1 and P-STAT3, although no significant change was observed in P-MAPK. When assayed for biological activity using an NIH3T3 transformation assay, NACC2-NTRK2-ΔTM exhibited approximately 125% greater transformation activity compared with NACC2-NTRK2(ex4:ex15) (Fig 5D). A Student’s t-test of these data returned a p-value of 0.09, indicating the significance of this difference.

Next, cycloheximide chase experiments were utilized to compare the stability of NACC2-NTRK2-ΔTM and NACC2-NTRK2(ex4:ex13). Because cycloheximide blocks protein translation, the amount of protein remaining at different times reflects the half-life of the protein. HEK293T cells expressing the fusion proteins were treated with 300 ng/ml of cycloheximide for 0, 5, 10, and 20 hours before preparing cell lysates, analysis by SDS-PAGE, and immunoblotting. After addition of cycloheximide, the expression level of NACC2-NTRK2(ex4:ex13) decreased over time (Fig 5E, left lanes). In marked contrast, the observed amount of NACC2-NTRK2-ΔTM remained stable over the course of the experiment with little or no apparent degradation (Fig 5E, right lanes). When this was quantitated (Fig 5F), the half-life of NACC2-NTRK2 was determined to be approximately 18 hours, whereas the greater stability of NACC2-NTRK2-ΔTM precluded calculation of its half-life over the time course of this experiment.

As presented inFig 5A, we also constructed a more extensive deletion in which all of exons 13 and 14 of NTRK2 were removed, to create the NACC2-NTRK2(ex4:ex15) fusion. When the stability of NACC2-NTRK2(ex4:ex15) was examined, the half-life was determined to be approximately 42 hours, more than twice that of the more commonly observed NACC2-NTRK2(ex4:ex13) fusion (Fig 5F).

These data suggest that the occurrence of the altered breakpoint exhibited by NACC2-NTRK2(ex4:ex15), resulting in greater stability of the fusion protein, may contribute to the more severe clinical outcome of pediatric glioblastoma [10]. Removal of the transmembrane helix, either in NACC2-NTRK2-ΔTM or in NACC2-NTRK2(ex4:ex15), slows the proteasomal degradation of the fusion protein. These observations suggest that the transmembrane helix may contain a signal for degradation, known as a degron. Inspection reveals the presence of the D-box motif RxxL, characterized as a degradation signal for cyclin B1 [33]. In NTRK2, this sequence occurs as the sequence REHL at the very beginning of the TM helix, where it would normally be partly membrane embedded. The increased oncogenicity as measured by focus assays of NACC2-NTRK2(ex4:ex15) may result from deletion of this degron, leading to a longer half-life and greater ability to activate downstream signaling pathways.

Characterization of NACC2-NTRK2 activation

Our experiments show that multimerization of the N-terminal BTB domain in the NACC2-NTRK2 fusion protein induces constitutive activation of the NTRK2-derived tyrosine kinase. Our downstream signaling analysis demonstrates that NACC2-NTRK2 kinase activation triggers the phosphorylation of signaling proteins in the ERK/MAPK, PLCγ1, and JAK/STAT pathways, and results in the morphological transformation of NIH3T3 cells (Fig 1B and 1D). The complete lack of activity displayed by the kinase-dead mutant, NACC2-NTRK2(K572R), demonstrates the importance of a functional NTRK2 kinase domain. Mutations which disrupt BTB-mediated multimerization, either in the charged pocket of the BTB domain or in the monomer core, exhibit diminished downstream activation, as does the complete deletion of the BTB domain (Fig 3C).

These mutants are also unable to transform NIH3T3 cells, indicating that the oncogenic activity of NACC2-NTRK2 relies not only on the constitutive activation of the tyrosine kinase domain but also requires an intact BTB domain to drive multimerization. Importantly, the diminution of P-PLCγ1 and P-STAT3 activation exhibited by the D31N/R45Q, Y86A, and ΔBTB mutants (Fig 3C and 3D) correlated with the loss of biological activity (Fig 3E).

Deletions which remove the transmembrane helix result in increased fusion protein stability

The first case of NACC2-NTRK2(ex4:ex13) fusion was identified in pilocytic astrocytoma [9]. In 2021, a case report described a different NACC2-NTRK2 fusion with a distinct breakpoint, NACC2-NTRK2(ex4:ex15), in pediatric glioblastoma [10]. The transmembrane helix of NTRK2 emerges as a key structural motif missing from this novel fusion. Thus we constructed two different mutants to examine this more closely: NACC2-NTRK2-ΔTM which deletes the transmembrane helix, and NACC2-NTRK2(ex4:ex15) which deletes exons 13 and 14 entirely (Fig 5A). In cycloheximide chase experiments (Fig 5F), both mutants exhibit a significantly longer half-life than NACC2-NTRK2(ex4:ex13) which has a half-life of approximately 18 hours. The variant described in glioblastoma, NACC2-NTRK2(ex4:ex15), exhibited a half-life of approximately 42 hours. The NACC2-NTRK2-ΔTM variant was even more stable and did not decrease over the 20 hours of the experiment. Further experiments showed that NACC2-NTRK2-ΔTM resulted in increased downstream activation and a higher level of NIH3T3 cell transformation (Fig 5B and 5D). These data indicate that the NTRK2 transmembrane helix regulates NACC2-NTRK2 degradation, suggesting that the transmembrane helix contains a potential degron sequence [36]. By removing the transmembrane helix, the fusion protein becomes resistant to proteasomal degradation and results in more persistent kinase activation and eventually leads to higher transformation activity.

BTB domain as a druggable target

Treatment of NTRK fusion-positive cancers currently relies on TKIs that target the NTRK tyrosine kinase activity. In this study, we demonstrate that disrupting the normal functioning of the NACC2 BTB domain could be an alternative therapeutic approach.

The BTB domain is an evolutionary conserved domain that can be found in various zinc finger proteins. The structure and function of BTB domains have been well characterized [37], but its role in NTRK fusion proteins has not been investigated. Here we demonstrate that the NACC2-NTRK2-induced signaling activity and cell transformation can be inhibited by disruption of the BTB domain of NACC2. As shown by the focus assay results of NACC2(D31N/R45Q)-NTRK2, NACC2(Y86A)-NTRK2, and NACC2(ΔBTB)-NTRK2, disruption of the BTB domain charged pocket and monomer core abrogates the cell transformation activity (Fig 3E). In our multimerization study, the NACC2(D31N/R45Q)-NTRK2 charged pocket mutation, as well as the NACC2(Y86A)-NTRK2 monomer core mutation, show reduced homo- and hetero-multimerization in addition to reduced downstream activities (Figs2D and3A). In another BTB domain containing fusion protein, PLZF/RAR-alpha that causes AML, similar mutations result in inactive fusion-positive cells [19]. Thus, a novel inhibitor that targets these conserved motifs may be beneficial for multiple BTB domain protein associated cancers.

Precedent of BTB domain inhibitors can be found in another BTB domain containing oncogene, B-cell lymphoma 6 (BCL6). Two types of inhibitors for this protein have been developed. BI-3802 inhibitor lead to BCL6 dimer degradation [38], and BI-3812 prevents BCL6 dimer formation [39]. Despite limitations in bioavailability,in vitro analysis proved their efficacy [40]. Thus, the BTB domain is a druggable domain and a novel inhibitor targeting the NACC2-NTRK2 BTB domain charged pocket or monomer core can be envisioned as an alternative treatment method to overcome possible TKI resistance.

The study of a shorter form of the fusion protein BTB-NTRK2 reveals another interesting aspect of NACC2 in the fusion protein. In the homo-multimerization experiment, protein samples were prepared in non-reducing sample buffer and resolved in SDS-PAGE. The NACC2-NTRK2 fusion and BTB domain point mutants were still able to multimerize whereas the homo-multimerization of BTB-NTRK2 was not observed (Fig 4D). However, the coimmunoprecipitation experiment proved that BTB-NTRK2 was able to interact with the BTB domain of native NACC2 as does NACC2-NTRK2 (Fig 4E, 3^rd panel). Also, BTB-NTRK2 focus assay results showed that the BTB domain alone, in BTB-NTRK2, is able to mediate downstream activation and transform NIH3T3 cells, although at a reduced level (Fig 4F). These data together indicate that NACC2 residues in the disordered region (residues 121–350), possibly together with the residual BEN domain (residues 351–418), provide structural support for NACC2-mediated multimerization through covalent interactions, making it resistant to SDS denaturation. Since this multimer is sensitive to β-mercaptoethanol reduction (Fig 2D), this establishes the existence of one or more intermolecular disulfide bonds stabilizing the multimer. This retained portion of NACC2 from residues 351–418 includes a total of seven Cys residues: four in the disordered region, and three in the partial BEN domain. It will be interesting in future experiments to determine which of these seven residues are responsible for this covalent disulfide bonding. In summary, the BTB domain itself is responsible for self-association–a precondition for multimerization–but other sequences, contributed by NACC2 and which lie downstream of the BTB domain, provide a major enhancement to structural stability and biological activity.

These data also suggest, for NTRK2 fusion proteins, that domains other than the multimerization domain in the N-terminal fusion partner may contribute important regulatory functions. For example, in the QKI-NTRK2 fusion there exist three domains from QKI: the Qua1 domain responsible for protein dimerization, the KH domain that mediates RNA recognition, and the Qua2 domain that interacts with RNA [41]. As a homodimerization domain, the Qua1 domain fulfils a similar function in NTRK2 activation as the BTB domain; however, the other two domains, KH and Qua2, may contribute additional regulatory functions to the QKI-NTRK2 fusion. Thus, it will be interesting to investigate the roles of protein domains within the N-terminal fusion partners above and beyond multimerization.

Through this study, we briefly characterized the functions of every domain in the NACC2-NTRK2. The BTB domain (1–120) is responsible for fusion protein multimerization and the residues 121–418 are responsible for stabilization of the multimer. In the NTRK2-derived portion, the kinase domain is responsible for the oncogenic activation of the fusion, and lastly, removal of the transmembrane helix results in greater stability of the fusion protein and higher-grade activation. Point mutations in the NACC2 BTB domain abrogate the activity of the NACC2-NTRK2 fusion protein. Thus, patients may benefit from a combined treatment of TKI inhibition and BTB domain inhibition. These findings suggest that characterizations of oncogenic fusion proteins are important for the discovery of novel therapeutic methods and facilitate the development of novel treatments in the battle against fusion-positive cancers.

DNA constructs

Genes for NACC2 (SC319661) and NTRK2 (RG221838) were purchased from Origene (Rockville, MD, USA). NTRK2 1770G was found to be missing in RG221838; this was repaired by site-directed mutagenesis. To subclone both genes into pCDNA3 vector, an XbaI restriction site was introduced at the 3’ ends of both NACC2 and NTRK2 by quick-change PCR mutagenesis. NACC2 and NTRK2 were subcloned into pCDNA3 vector using upstream EcoRI sites and the introduced XbaI sites. A ClaI restriction site was introduced at residue V417 of NACC2 and Y397 of NTRK2. NACC2 and NTRK2 were joined together following ClaI and XbaI double digestion and ligation. Point mutations (D31N/R45Q, Y86A, K572R) were introduced by site-directed PCR mutagenesis. Domain deletions (ΔBTB, Δ121–418, ΔTM, and (ex4:ex15)) were introduced by PCR site-directed mutagenesis [42]. For NIH3T3 transformation assays, DNA constructs were subcloned into the murine retroviral vector pLXSN [43] using EcoRI and XbaI restriction sites. All NACC2-NTRK2 genes expressed in pcDNA3 clones are identical to those expressed in pLXSN clones, and all maintain an identical sequence surrounding the initiation codon ATG. The only difference is that in pcDNA3 vectors, genes are expressed from a Human Cytomegalovirus (CMV) immediate-early promoter for high-level expression [44], whereas in pLXSN vectors, genes are expressed from the Moloney Murine Leukemia Virus (MoMLV) LTR promoter for stable long-term expression [43].

Antibodies and reagents

Antibodies were purchased as following: NACC2 (Bethyl A304-991A), P-NTRK (Cell Signaling 4621S), NTRK2 (Invitrogen PA5-86241), P-MAPK (Cell Signaling 4370), MAPK (Cell Signaling 9102), P-PLCγ1 (Cell Signaling 2821), PLCγ1 (Santa Cruz sc-81), P-STAT3 (Cell Signaling 9145), STAT3 (Cell Signaling 9139). Enhanced Chemiluminescence (ECL) reagents, horseradish peroxidase (HRP) anti-mouse (NA931V) and HRP anti-rabbit (NA934V) secondary antibodies were purchased from Cytiva.

Other reagents were purchased as following: G-418 Sulfate (Fisher Scientific), Lipofectamine 2000 (Invitrogen), Protein A-Sepharose (Sigma P3391), and Cycloheximide (Fisher J66901.03).

Transfection and immunoblotting

HEK293T cells were grown in 1% Pen/Strep and 10% FBS and seeded at 10⁶ cells per 10 cm plate. 5 μg of pCDNA3 DNA constructs were transfected using CaCl₂ transfection as described previously [25]. Cells were incubated in 3% CO2 for 14–18 hours, recovered in 10% CO2 for 6–8 hours, and then starved with 0% FBS media for 14–16 hours. Cells were lysed in RIPA buffer [50 mM Tris HCl pH 8.0, 150 mM NaCl, 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM NaF] with 1 mM sodium orthovanadate, 1 mM PMSF and 10 μg/mL aprotinin. Lysate concentrations were determined by Lowry Assay. Samples were resolved by SDS-PAGE and transferred to Immobilon-P PVDF membranes (Millipore, Burlington, MA, USA). Membranes were blocked with either 5% BSA in 0.1% TBS-T or 5% nonfat milk in 0.1%TBS-T.

Immunoprecipitation and multimerization assay

DNA constructs were transfected into HEK293T cells using CaCl₂ transfection as above and collected with E1A buffer [250 mM NaCl, 50 mM HEPES, 5 mM EDTA and 0.1% NP-40] [30]. Total protein concentrations were measured by Lowry Assay. For multimerization assay, protein samples were prepared with either reducing sample buffer (50mM Tris-Cl, 10% β-mercaptoethanol, 2% SDS, 0.1% bromophenol blue, 10% glycerol) or non-reducing sample buffer (4% SDS,10 mM NaPO4 pH7.0, 20% glycerol, 0.08% Bromophenol blue). Samples were resolved by SDS-PAGE. For immunoprecipitation, 300 μg of total protein was diluted to 1mL with E1A wash buffer [125 mM NaCl, 50 mM HEPES, 0.2% NP-40 and 5 mM EDTA]. Diluted lysates were precleared with Protein A-Sepharose for 3 hours rocking at 4°C, and then incubated with primary antibody overnight at 4°C. Protein A-Sepharose was added to recover immune complexes. Beads were washed with E1A wash buffer 5 times followed by addition of 30 μL of reducing sample buffer. Samples were resolved by SDS-PAGE and immunoblotted as above.

Focus assay

NIH3T3 cells were maintained in DMEM with 10% calf serum and 1% penicillin/streptomycin in 10% CO₂ at 37 degC Focus assays were performed basically as described [45]. NIH3T3 cells were plated at a density of 4 x 10⁵ cells on 60mm plates 24 hours before transfection. Cells were transfected by Lipofectamine 2000 Reagent with 10 μg of pLXSN DNA constructs. Approximately 24 hours after transfection, cells were re-fed with DMEM + 10% calf serum. After an additional 24 hours, cells were split 1:12 onto duplicate 100-mm plates later containing 2.5% calf serum for the focus forming assays, and 1:240 onto duplicate 100-mm plates containing 10% calf serum and Geneticin (G418, 0.5 mg/mL), in order to measure the expression and efficiency of the transfections. After 14 days, focus assay plates were fixed in methanol, stained with Geimsa stain, and scored. Geneticin-resistant colonies were also fixed with methanol after 14 days and scored. Numbers of foci were normalized to the number of G418-resistant colonies, and quantitated relative to a positive control–/+ standard error of the mean (SEM).

Cycloheximide chase

HEK293T cells were seeded at 10⁶ cells per 10 cm plate and transfected with 2–5 μg of pcDNA3 DNA constructs. After recovery at 10% CO₂ for 6–8 hours, cell were treated with 300 μg/ml of cycloheximide in ethanol for 0, 5,10 and 20 hours. Cells were harvested and lysed in RIPA.

• 1.Amatu A, Sartore-Bianchi A, Siena S. NTRK gene fusions as novel targets of cancer therapy across multiple tumour types. ESMO Open. 2016;1(2):e000023. Epub 2016/11/16. doi: 10.1136/esmoopen-2015-000023 ; PubMed Central PMCID: PMC5070277. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 2.Solomon JP, Benayed R, Hechtman JF, Ladanyi M. Identifying patients with NTRK fusion cancer. Ann Oncol. 2019;30(Suppl_8):viii16–viii22. Epub 2019/11/19. doi: 10.1093/annonc/mdz384 ; PubMed Central PMCID: PMC6859817. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 3.Chao MV. Neurotrophins and their receptors: a convergence point for many signalling pathways. Nat Rev Neurosci. 2003;4(4):299–309. Epub 2003/04/03. doi: 10.1038/nrn1078 . [DOI] [PubMed] [Google Scholar]
• 4.Cunningham ME, Greene LA. A function-structure model for NGF-activated TRK. EMBO J. 1998;17(24):7282–93. Epub 1998/12/19. doi: 10.1093/emboj/17.24.7282 ; PubMed Central PMCID: PMC1171074. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 5.Amatu A, Sartore-Bianchi A, Bencardino K, Pizzutilo EG, Tosi F, Siena S. Tropomyosin receptor kinase (TRK) biology and the role of NTRK gene fusions in cancer.Ann Oncol.2019;30(Suppl_8):viii5–viii15. Epub 2019/11/19. doi: 10.1093/annonc/mdz383 ; PubMed Central PMCID: PMC6859819. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 6.Kim JH, Hong JH, Choi YL, Lee JA, Seo MK, Lee MS, et al. NTRK oncogenic fusions are exclusively associated with the serrated neoplasia pathway in the colorectum and begin to occur in sessile serrated lesions. J Pathol. 2021;255(4):399–411. Epub 2021/08/18. doi: 10.1002/path.5779 . [DOI] [PubMed] [Google Scholar]
• 7.Vaishnavi A, Le AT, Doebele RC. TRKing down an old oncogene in a new era of targeted therapy. Cancer discovery. 2015;5(1):25–34. Epub 2014/12/21. doi: 10.1158/2159-8290.CD-14-0765 ; PubMed Central PMCID: PMC4293234. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 8.Jiang T, Wang G, Liu Y, Feng L, Wang M, Liu J, et al. Development of small-molecule tropomyosin receptor kinase (TRK) inhibitors for NTRK fusion cancers.Acta Pharm Sin B.2021;11(2):355–72. Epub 2021/03/02. doi: 10.1016/j.apsb.2020.05.004 ; PubMed Central PMCID: PMC7893124. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 9.Jones DT, Hutter B, Jager N, Korshunov A, Kool M, Warnatz HJ, et al. Recurrent somatic alterations of FGFR1 and NTRK2 in pilocytic astrocytoma. Nat Genet. 2013;45(8):927–32. doi: 10.1038/ng.2682 . [DOI] [PMC free article] [PubMed] [Google Scholar]
• 10.Britton HM, Levine AB, Shen Y, Mungall K, Serrano J, Snuderl M, et al. NTRK2 Fusion driven pediatric glioblastoma: Identification of oncogenic Drivers via integrative Genome and transcriptome profiling. Clin Case Rep. 2021;9(3):1472–7. Epub 2021/03/27. doi: 10.1002/ccr3.3804 ; PubMed Central PMCID: PMC7981675. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 11.Basta J, Rauchman M. The nucleosome remodeling and deacetylase complex in development and disease. Transl Res. 2015;165(1):36–47. Epub 2014/06/01. doi: 10.1016/j.trsl.2014.05.003 ; PubMed Central PMCID: PMC4793962. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 12.Xuan C, Wang Q, Han X, Duan Y, Li L, Shi L, et al. RBB, a novel transcription repressor, represses the transcription of HDM2 oncogene. Oncogene. 2013;32(32):3711–21. Epub 2012/08/29. doi: 10.1038/onc.2012.386 . [DOI] [PubMed] [Google Scholar]
• 13.Abhiman S, Iyer LM, Aravind L. BEN: a novel domain in chromatin factors and DNA viral proteins. Bioinformatics. 2008;24(4):458–61. Epub 2008/01/22. doi: 10.1093/bioinformatics/btn007 ; PubMed Central PMCID: PMC2477736. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 14.Melnick A, Carlile G, Ahmad KF, Kiang CL, Corcoran C, Bardwell V, et al. Critical residues within the BTB domain of PLZF and Bcl-6 modulate interaction with corepressors. Mol Cell Biol. 2002;22(6):1804–18. Epub 2002/02/28. doi: 10.1128/MCB.22.6.1804-1818.2002 ; PubMed Central PMCID: PMC135611. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 15.Albagli O, Dhordain P, Deweindt C, Lecocq G, Leprince D. The BTB/POZ domain: a new protein-protein interaction motif common to DNA- and actin-binding proteins. Cell Growth Differ. 1995;6(9):1193–8. Epub 1995/09/01. . [PubMed] [Google Scholar]
• 16.Collins T, Stone JR, Williams AJ. All in the family: the BTB/POZ, KRAB, and SCAN domains. Mol Cell Biol. 2001;21(11):3609–15. Epub 2001/05/08. doi: 10.1128/MCB.21.11.3609-3615.2001 ; PubMed Central PMCID: PMC86980. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 17.Ahmad KF, Engel CK, Prive GG. Crystal structure of the BTB domain from PLZF. Proc Natl Acad Sci U S A. 1998;95(21):12123–8. Epub 1998/10/15. doi: 10.1073/pnas.95.21.12123 ; PubMed Central PMCID: PMC22795. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 18.Melnick A, Ahmad KF, Arai S, Polinger A, Ball H, Borden KL, et al. In-depth mutational analysis of the promyelocytic leukemia zinc finger BTB/POZ domain reveals motifs and residues required for biological and transcriptional functions. Mol Cell Biol. 2000;20(17):6550–67. Epub 2000/08/11. doi: 10.1128/MCB.20.17.6550-6567.2000 ; PubMed Central PMCID: PMC86130. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 19.Puccetti E, Zheng X, Brambilla D, Seshire A, Beissert T, Boehrer S, et al. The integrity of the charged pocket in the BTB/POZ domain is essential for the phenotype induced by the leukemia-associated t(11;17) fusion protein PLZF/RARalpha.Cancer Res. 2005;65(14):6080–8. Epub 2005/07/19. doi: 10.1158/0008-5472.CAN-04-3631 . [DOI] [PubMed] [Google Scholar]
• 20.Shih C, Shilo BZ, Goldfarb MP, Dannenberg A, Weinberg RA. Passage of phenotypes of chemically transformed cells via transfection of DNA and chromatin. Proc Natl Acad Sci U S A. 1979;76(11):5714–8. PubMed Central PMCID: PMC411720. doi: 10.1073/pnas.76.11.5714 [DOI] [PMC free article] [PubMed] [Google Scholar]
• 21.Padua RA, Carter G, Hughes D, Gow J, Farr C, Oscier D, et al. RAS mutations in myelodysplasia detected by amplification, oligonucleotide hybridization, and transformation. Leukemia. 1988;2(8):503–10. . [PubMed] [Google Scholar]
• 22.Bafna S, Singh AP, Moniaux N, Eudy JD, Meza JL, Batra SK. MUC4, a multifunctional transmembrane glycoprotein, induces oncogenic transformation of NIH3T3 mouse fibroblast cells. Cancer Res. 2008;68(22):9231–8. doi: 10.1158/0008-5472.CAN-08-3135 ; PubMed Central PMCID: PMC2610629. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 23.Sun M, Wang G, Paciga JE, Feldman RI, Yuan ZQ, Ma XL, et al. AKT1/PKBalpha kinase is frequently elevated in human cancers and its constitutive activation is required for oncogenic transformation in NIH3T3 cells. Am J Pathol. 2001;159(2):431–7. doi: 10.1016/s0002-9440(10)61714-2 ; PubMed Central PMCID: PMC1850562. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 24.Peiris MN, Meyer AN, Nelson KN, Bisom-Rapp EW, Donoghue DJ. Oncogenic fusion protein BCR-FGFR1 requires BCR-mediated oligomerization and chaperonin Hsp90 for activation. Haematologica. 2019. Epub 2019/08/24. doi: 10.3324/haematol.2019.220871 . [DOI] [PMC free article] [PubMed] [Google Scholar]
• 25.Nelson KN, Meyer AN, Siari A, Campos AR, Motamedchaboki K, Donoghue DJ. Oncogenic Gene Fusion FGFR3-TACC3 Is Regulated by Tyrosine Phosphorylation. Mol Cancer Res. 2016;14(5):458–69. doi: 10.1158/1541-7786.MCR-15-0497 . [DOI] [PubMed] [Google Scholar]
• 26.Khotskaya YB, Holla VR, Farago AF, Mills Shaw KR, Meric-Bernstam F, Hong DS. Targeting TRK family proteins in cancer. Pharmacol Ther. 2017;173:58–66. Epub 2017/02/09. doi: 10.1016/j.pharmthera.2017.02.006 . [DOI] [PubMed] [Google Scholar]
• 27.Zollman S, Godt D, Prive GG, Couderc JL, Laski FA. The BTB domain, found primarily in zinc finger proteins, defines an evolutionarily conserved family that includes several developmentally regulated genes in Drosophila. Proc Natl Acad Sci U S A. 1994;91(22):10717–21. Epub 1994/10/25. doi: 10.1073/pnas.91.22.10717 ; PubMed Central PMCID: PMC45093. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 28.Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O, et al. Highly accurate protein structure prediction with AlphaFold. Nature. 2021;596(7873):583–9. Epub 2021/07/16. doi: 10.1038/s41586-021-03819-2 ; PubMed Central PMCID: PMC8371605. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 29.Simossis VA, Heringa J. PRALINE: a multiple sequence alignment toolbox that integrates homology-extended and secondary structure information. Nucleic Acids Res. 2005;33(Web Server issue):W289–94. Epub 2005/06/28. doi: 10.1093/nar/gki390 ; PubMed Central PMCID: PMC1160151. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 30.Chan CH, Lee SW, Li CF, Wang J, Yang WL, Wu CY, et al. Deciphering the transcriptional complex critical for RhoA gene expression and cancer metastasis. Nat Cell Biol. 2010;12(5):457–67. Epub 2010/04/13. doi: 10.1038/ncb2047 ; PubMed Central PMCID: PMC3855841. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 31.Gatalica Z, Xiu J, Swensen J, Vranic S. Molecular characterization of cancers with NTRK gene fusions. Mod Pathol. 2019;32(1):147–53. Epub 2018/09/02. doi: 10.1038/s41379-018-0118-3 . [DOI] [PubMed] [Google Scholar]
• 32.Aepala MR, Peiris MN, Jiang Z, Yang W, Meyer AN, Donoghue DJ. Nefarious NTRK oncogenic fusions in pediatric sarcomas: Too many to Trk. Cytokine Growth Factor Rev. 2022;68:93–106. Epub 2022/09/25. doi: 10.1016/j.cytogfr.2022.08.003 . [DOI] [PubMed] [Google Scholar]
• 33.Zur A, Brandeis M. Timing of APC/C substrate degradation is determined by fzy/fzr specificity of destruction boxes. EMBO J. 2002;21(17):4500–10. Epub 2002/08/29. doi: 10.1093/emboj/cdf452 ; PubMed Central PMCID: PMC126191. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 34.Cocco E, Scaltriti M, Drilon A. NTRK fusion-positive cancers and TRK inhibitor therapy.Nat Rev Clin Oncol. 2018;15(12):731–47. Epub 2018/10/20. doi: 10.1038/s41571-018-0113-0 ; PubMed Central PMCID: PMC6419506. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 35.Somwar R, Hofmann NE, Smith B, Odintsov I, Vojnic M, Linkov I, et al. NTRK kinase domain mutations in cancer variably impact sensitivity to type I and type II inhibitors. Commun Biol. 2020;3(1):776. Epub 2020/12/18. doi: 10.1038/s42003-020-01508-w ; PubMed Central PMCID: PMC7745027. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 36.Fortmann KT, Lewis RD, Ngo KA, Fagerlund R, Hoffmann A. A Regulated, Ubiquitin-Independent Degron in IkappaBalpha. J Mol Biol. 2015;427(17):2748–56. Epub 2015/07/21. doi: 10.1016/j.jmb.2015.07.008 ; PubMed Central PMCID: PMC4685248. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 37.Stogios PJ, Downs GS, Jauhal JJ, Nandra SK, Prive GG. Sequence and structural analysis of BTB domain proteins. Genome Biol. 2005;6(10):R82. Epub 2005/10/07. doi: 10.1186/gb-2005-6-10-r82 ; PubMed Central PMCID: PMC1257465. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 38.Slabicki M, Yoon H, Koeppel J, Nitsch L, Roy Burman SS, Di Genua C, et al. Small-molecule-induced polymerization triggers degradation of BCL6. Nature. 2020;588(7836):164–8. Epub 2020/11/20. doi: 10.1038/s41586-020-2925-1 ; PubMed Central PMCID: PMC7816212. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 39.Nitsch L, Jensen P, Yoon H, Koeppel J, Burman SSR, Fischer ES, et al. BTB(BCL6) dimers as building blocks for reversible drug-induced protein oligomerization. Cell Rep Methods. 2022;2(4):100193. Epub 2022/05/03. doi: 10.1016/j.crmeth.2022.100193 ; PubMed Central PMCID: PMC9046236. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 40.Kerres N, Steurer S, Schlager S, Bader G, Berger H, Caligiuri M, et al. Chemically Induced Degradation of the Oncogenic Transcription Factor BCL6. Cell reports. 2017;20(12):2860–75. Epub 2017/09/21. doi: 10.1016/j.celrep.2017.08.081 . [DOI] [PubMed] [Google Scholar]
• 41.Aberg K, Saetre P, Jareborg N, Jazin E. Human QKI, a potential regulator of mRNA expression of human oligodendrocyte-related genes involved in schizophrenia. Proc Natl Acad Sci U S A. 2006;103(19):7482–7. Epub 2006/04/28. doi: 10.1073/pnas.0601213103 ; PubMed Central PMCID: PMC1464365. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 42.Liu H, Naismith JH. An efficient one-step site-directed deletion, insertion, single and multiple-site plasmid mutagenesis protocol. BMC Biotechnol. 2008;8:91. Epub 2008/12/06. doi: 10.1186/1472-6750-8-91 ; PubMed Central PMCID: PMC2629768. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 43.Miller AD, Rosman GJ. Improved retroviral vectors for gene transfer and expression. Biotechniques. 1989;7(9):980–2, 4–6, 9–90. Epub 1989/10/01. ; PubMed Central PMCID: PMC1360503. [PMC free article] [PubMed] [Google Scholar]
• 44.Pasleau F, Tocci MJ, Leung F, Kopchick JJ. Growth hormone gene expression in eukaryotic cells directed by the Rous sarcoma virus long terminal repeat or cytomegalovirus immediate-early promoter. Gene. 1985;38(1–3):227–32. Epub 1985/01/01. doi: 10.1016/0378-1119(85)90221-5 . [DOI] [PubMed] [Google Scholar]
• 45.Nelson KN, Meyer AN, Wang CG, Donoghue DJ. Oncogenic driver FGFR3-TACC3 is dependent on membrane trafficking and ERK signaling.Oncotarget.2018;9(76):34306–19. Epub 2018/10/23. doi: 10.18632/oncotarget.26142 ; PubMed Central PMCID: PMC6188140. [DOI] [PMC free article] [PubMed] [Google Scholar]

Supplementary Materials

Data Availability Statement

All relevant data are within the manuscript or deposited at the Zenodo data repository as DOI10.5281/zenodo.10416608.