Trans-activating mutations of the pseudokinase ERBB3

doi:10.21203/rs.3.rs-2568092/v1

Download PDF

Article

Trans-activating mutations of the pseudokinase ERBB3

https://doi.org/10.21203/rs.3.rs-2568092/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 28 May, 2024

Read the published version in Oncogene →

You are reading this latest preprint version

Genetic changes in the ERBB family of receptor tyrosine kinases serve as oncogenic driver events and predictive biomarkers for ERBB inhibitor drugs. ERBB3 is a pseudokinase member of the family that, although lacking a fully active kinase domain, is well known for its potent signaling activity as a heterodimeric complex with ERBB2. Previous studies have identified few transforming ERBB3 mutations while the great majority of the hundreds of different somatic ERBB3 variants observed in different cancer types remain of unknown significance. Here, we describe an unbiased functional genetics screen of the transforming potential of thousands of ERBB3 mutations in parallel. The screen based on a previously described iSCREAM (in vitro screen of activating mutations) platform, and addressing ERBB3 pseudokinase signaling in a context of ERBB3/ERBB2 heterodimers, identified 18 hit mutations. Validation experiments in Ba/F3, NIH 3T3, and MCF10A cell backgrounds demonstrated the presence of both previously known and unknown transforming ERBB3 missense mutations functioning either as single variants or in cis as a pairwise combination. Drug sensitivity assays with trastuzumab, pertuzumab and neratinib indicated actionability of the transforming ERBB3 variants.

Biological sciences/Cancer/Cancer genetics

Biological sciences/Genetics/Mutation

Biological sciences/Cell biology/Cell signalling/Growth factor signalling

While tyrosine phosphorylation-mediated signaling is critical for dynamic signaling in multicellular organisms [1], a significant subset of the human protein kinome consist of “kinase-dead” enzymes, with no or very limited enzymatic activity. These pseudokinases have been estimated to comprise 10% of all human protein kinases based on substitutions of conserved amino acid residues with their catalytic sites [2, 3]. While harboring little kinase activity, pseudokinases can still promote signaling and carry out important noncatalytic functions such as serving as scaffolding or docking platforms, regulatory domains, or structural elements [4, 5]. Aberrant signaling by pseudokinases, such as JAK2, ROR1, and TRIB2, has also been demonstrated to be relevant for tumorigenesis in different cancer types [6–10].

The receptor tyrosine kinase (RTK) ERBB3 is another pseudokinase frequently attributed a role in cancer. ERBB3 (HER3) belongs to the ERBB family of RTKs, together with EGFR (ERBB1/HER1), ERBB2 (HER2), and ERBB4 (HER4). EGFR and ERBB2 are well-known human oncogenes and somatic missense mutations, small in-frame indels or amplifications in these genes are clinically used as biomarkers predicting response to EGFR- or ERBB2-targeted therapies [11–13]. While lacking significant kinase activity, ERBB3 can readily heterodimerize with other kinase competent ERBB receptors and thus promote intracellular signaling. In particular, ERBB3 has been proposed to be the preferred dimerization partner for ERBB2, and to critically contribute to ERBB2-mediated transformation in breast and other cancer types [14–16].

While several monoclonal antibodies targeting ERBB3 have been evaluated in clinical trials, including the antibody-drug conjugate patritumab deruxtecan (clinicaltrials.gov), no ERBB3-targeting therapies are currently clinically available. However, the ERBB2-targeting antibodies trastuzumab and pertuzumab block the formation of signaling-competent ERBB3/ERBB2 heterodimers, and this mechanism has been shown to be critical for the anti-tumor effect [14, 16, 17]. In addition, the ERBB2-targeting tyrosine kinase inhibitors (TKIs) such as neratinib and tucatinib may inhibit the signaling through ERBB3/ERBB2 heterodimers [18–22].

While hundreds of different somatic ERBB3 mutations have been reported in clinical samples representing a number of different cancer types, the functional significance of the variants is mostly unknown. The cBioPortal database [23, 24] currently lists over 800 different somatic ERBB3 mutations of which the great majority remain without functional annotation. In spite of the clear potential of better understanding the predictive value of the ERBB3 variants for personalizing ERBB-targeting therapies, functional screens to identify gain-of-function alterations in ERBB3 have been hampered by its natural enzymatic inactivity as a pseudokinase.

Here, we describe a functional genetics screen of gain-of-function ERBB3 variants using a platform based on the previously described iSCREAM [25, 26] pipeline. The approach leveraged expression of 8 055 randomly mutated ERBB3 missense or nonsense variants in Ba/F3 cells and the clonal outgrowth of cell populations expressing the activating variants. As none of the thousands of ERBB3 variants promoted Ba/F3 cell growth when expressed alone – as expected for a pseudokinase – the pipeline was modified with co-expression of ERBB2. The modified iSCREAM and subsequent functional and structural validation experiments resulted in identification of several ERBB3 variants that promoted cellular transformation as heterodimeric complexes with ERBB2. Drug sensitivity analyses confirmed the sensitivity of the activating mutations to ERBB2-targeting therapeutics demonstrating the potential value of the ERBB3 pseudokinase variants as new predictive biomarkers.

Cell culture

Ba/F3 (DSMZ), NIH 3T3 (ATCC), and Phoenix Ampho cells (a gift from Dr. Garry Nolan) were cultured as previously described [27]. MCF-10A (ATCC) cells were cultured in DMEM/F-12 (Lonza) supplemented with 10% FCS, 50 U/ml penicillin, and 50 U/ml streptomycin, 20 ng/ml EGF (Peprotech), 0.5 mg/ml hydrocortisone (Sigma, #H-0888), 100 ng/ml cholera toxin (Sigma, #C-8052), and 10 µg/ml insulin (Sigma, #I-1882). Cells were routinely tested for Mycoplasma infection using MycoAlert (Lonza).

Expression library of random ERBB3 mutants

Primer pair 5’- GGGGACAAGTTTGTACAAAAAAGCAGGCTTCACC ATGAGGGCGAACGACGCT -3’ and 5’- GGGGACCACTTTGTACAAGAAAGCTGGGTTTTACGTTCTCTGGGCATTAGCCTT -3’ was used to produce 4 093 bp amplicon from pcDNA3.1ERBB3-HA plasmid [28] containing wild-type ERBB3 cDNA insert flanked by aatB1 and aatB2 gateway recombination sites. The amplicon was cloned into the pDONR221 vector [25].

Error-prone PCR was used to generate the expression library encoding randomly mutated human ERBB3 variants [25]. Primer pair 5’-TTGATGCCTGGCAGTTCCCTA-3’ (binds 78 bp upstream of the attL1 site at the 5’-end of the ERBB3 insert) and 5’-ATCTTGTGCAATGTAACATCAGAGATT-3’ (binds 80 bp downstream of attL2 site at the 3’-end of the ERBB3 insert) was used to produce randomly mutated 4 439 bp ERBB3 cDNA amplicon from pDONR221-ERBB3. The PCR product was purified and cloned into a pBABE-puro-gateway vector [25].

Generation of cell lines with stable ERBB expression

Phoenix Ampho cells were transfected with pBABE-based retroviral expression vectors encoding the ERBB proteins [27]. Retroviral infection with supernatants from Phoenix Ampho cells and selection to produce cell lines with stable ERBB3 expression was carried out, as previously described [25,27]. Cells with stable expression of activator-incompetent ERBB2 V956R [29–31] were produced by selecting the cells with 500 µg/ml neomycin (Geneticin, Gibco) for 5 days. Selection pressure was subsequently maintained by the presence of 250 µg/ml neomycin in the culture medium. To generate cells with simultaneous expression of both the ERBB2 V956R and ERBB3 library, stable neomycin-selected cells expressing ERBB2 V956R infected with retroviruses produced from pBABE-neo-gateway-ERBB2 V956R plasmid were transduced with supernatants from Phoenix Ampho cells expressing the pBABE-puro-gateway-ERBB3 library. When generating cells with simultaneous expression of both ERBB2 V956R and an individual ERBB3 mutant, cells infected with a pBABE-puro-gateway-ERBB3 plasmid and selected with puromycin were transduced with supernatants from Phoenix Ampho cells expressing pBABE-neo-gateway-ERBB2 V956R.

Establishment of IL-3-independent Ba/F3 cell populations and cDNA sample preparation

Ba/F3 cells expressing the ERBB3 mutation library together with ERBB2 V956R were cultured in the presence of IL-3, in the absence of IL-3 but in the presence or absence of 20 ng/ml of the ERBB3 ligand NRG-1 (R&D Systems), or in the absence of both IL-3 and NRG-1 (with or without pretreating the cells first with NRG-1 but without IL-3 for 48 hours to promote ERBB expression). After 15–48 days, the established ligand-independent cell populations were further analyzed.

To amplify ERBB3 inserts for next-generation sequencing (NGS), genomic DNA was extracted from the Ba/F3 cells (NucleoSpin Tissue, Macherey Nagel) and a total of 24 µg of genomic DNA was used as a template for PCR amplification with the primers: 5’- GGGGACAAGTTTGTACAAAAAAGCAGGCTTCACC ATGAGGGCGAACGACGCT -3’ and 5’- GGGGACCACTTTGTACAAGAAAGCTGGGTTTTACGTTCTCTGGGCATTAGCCTT -3’. The original plasmid containing the cDNA expression library (1 ng) used for transduction was similarly PCR-amplified. A mixture of 1:1 of Phusion (Thermo Scientific) and Velocity (Bioline) high fidelity DNA polymerases was used for PCR.

Cell sorting

To quantify ERBB3/ERBB2 expression levels, 500 000 or 3 million Ba/F3 cells were washed three times with PBS and resuspended into 100 µl or 500 µl of flow cytometry staining buffer (FCSB; 3% FBS, 0.09% NaN₃in PBS). Non-specific binding was blocked with incubation in 1 µg/1 million cells of IgG (Invitrogen, #10500C) for 15 minutes. Human ERBB3/HER3 Alexa Fluor 647-conjugated Antibody (FAB3481R, R&D Systems) and Human ERBB2/HER2 Alexa Fluor 488-conjugated Antibody (FAB1129G, R&D Systems) were used. BD LSRFortessa Cell Analyzer (BD Biosciences) or Sony SH800 Cell Sorter (Sony Biotechnology) were used for cell sorting.

Analysis of Ba/F3 and MCF-10A cell growth

Ba/F3 cell growth was analyzed using MTT cell viability assays (CellTiter 96 nonradioactive cell proliferation assay; Promega), as previously described [27], in the presence or absence of 20 ng/ml of NRG-1. To analyze MCF-10A cell growth, 2 000 cells/well were seeded in quadruplicates into 96-well plate wells, cultured for 0, 2, 4, or 6 days in DMEM/F-12 medium deprived from EGF (Peprotech) and FBS, and supplemented with or without 50 ng/ml NRG-1 and analyzed with the MTT assay. Statistical analysis was performed with the unpaired two-sample t test between cells expressing wild-type or mutant ERBB3 variants. P < 0.05 was considered as significant. GraphPad Prism 8 was used for creating the dot plots or bar graphs.

Western blotting and antibodies

Cells were lysed in lysis buffer and analyzed by western blotting, as previously described [27,32], using the following primary antibodies: anti-ERBB2 (Thermo Scientific, Cat#MA-5-14057), anti-ERBB3 (Cell Signaling Technology, Cat#4754), anti-phospho-ERBB2 Tyr1248 (Cell Signaling Technology, Cat#2247), anti-phospho-ERBB3 Tyr1289 (Cell Signaling Technology, Cat#4791), anti-AKT (Cell Signaling Technology, Cat#2920), anti-phospho-AKT (Cell Signaling Technology, Cat#4060), anti-ERK (Cell Signaling Technology, Cat#9102), and anti-phospho-ERK (Cell Signaling Technology, Cat#9101). Loading was controlled using β-actin (Sigma Aldrich, Cat#A5441) antibody. Signals were detected using LI-COR Odyssey CLx and quantified by Image Studio Lite 5.

Focus formation analyses

Two hundred and fifty thousand NIH 3T3 cells in DMEM supplemented with 3% FBS were seeded into 6-well plates and cultured for 2 weeks. The cells were fixed with 100% methanol for 10 minutes and stained with 0.4% crystal violet in 25% methanol for 5 minutes. Wells were washed with milli-Q water five times and the plates were scanned (Epson scanner). Foci were analyzed from four independent experiments using the ImageJ plugin ColonyArea [33]. Statistical analysis was performed by unpaired two-sample t test comparing cells expressing wild-type or mutant ERBB3 variants. P < 0.05 was considered as significant. GraphPad Prism 8 was used for creating the dot plots.

Drug response assays

Ba/F3 cell responses to neratinib (Santa Cruz Biotechnology), pertuzumab (Genentech), or trastuzumab (Genentech) were analyzed as previously described [27]. Statistical analysis was performed by unpaired two-sample t test between cells expressing wild-type or mutant ERBB3 variants. P < 0.05 were considered as significant. GraphPad Prism 8 was used for creating the dot plots and to calculate IC50 values using the four-parameter inhibitor model.

Supplementary methods

Additional experimental procedures including plasmid constructs for individual mutations, identification of mutations enriched during IL3-depletion, analysis of DNA with long-read sequencing, and structural analysis of the ERBB3 variantsare available in the Supplementary information.

Somatic ERBB3 mutations in cancer

ERBB3 has a somatic mutation frequency of 2.1% in pan-cancer analysis with relatively high frequencies observed e.g. in urothelial (6.3-9.7%), endometrial (8.2-12.8%), lung (1.4-8.8%), and colorectal (4.9-6.2%) cancer (cbioportal.org) [23,24]. Among the 831 observed unique genetic ERBB3 alterations, the cBioPortal database lists only eight missense mutations as oncogenic and 41 variants affecting 11 additional amino acid residues as likely oncogenic (Supplementary Fig. 1).

Modified cell model for screening of activating variants of ERBB3 pseudokinase

Murine lymphoid Ba/F3 cells were selected as the cellular background for screening activating ERBB3 mutations as we have previously used the model to screen for activating EGFR [25] and ERBB4 [26] variants. The fact that the Ba/F3 cells are critically dependent on exogenous IL-3 for survival and that this dependency can be substituted by ectopic expression of an active RTK [34], allows the use of the cells for a functional readout of RTK-driven growth. However, overexpression of even the well-characterized oncogenic E928G variant of ERBB3 was not sufficient to promote IL-3-independent Ba/F3 cell growth (Fig. 1A), as expected for a pseudokinase in the context devoid of endogenously expressed heterodimeric partners. These observations indicated that Ba/F3 background could not be used to address differential transforming potential of ERBB3 variants in the absence of a kinase-competent partner to heterodimerize with ERBB3.

To modify the model with simultaneous co-expression of ERBB2, but to limit the role of ERBB2 to only serve as the receiver kinase in the ERBB3/ERBB2 dimer, a ERBB2 V956R mutant variant [29–31] was introduced to the Ba/F3 cells (Fig. 1B-C; Supplementary Fig. 2A-B). We had previously demonstrated that co-expression of wild-type ERBB2 together with wild-type ERBB3 in the same model was sufficient to readily promote IL-3-independent growth [27], precluding the use of wild-type ERBB2 as the heterodimer partner. The V956R variant with a valine-to-arginine mutation in the C-lobe of the ERBB2 kinase domain disrupts the capability of the receptor to serve as an activator but does not interfere with its function as a receiver in the transactivation process between two ERBB kinases (Fig. 1C). The “activator-incompetent” V956R was therefore also expected to be incapable of transactivating ERBB2 in a homodimeric ERBB2/ERBB2 complex. Consistent with reduced potential of the ERBB2 V956R mutant in transforming the Ba/F3 cells, it did not promote IL-3-independent growth even in the context of co-expression with wild-type ERBB3 when the complex was not activated by the NRG-1 ligand (Fig. 1B). However, indicating a window for a functional read-out, introducing the oncogenic mutant ERBB3 E928G together with the activator-incompetent ERBB2 V956R did result in IL-3-independent growth (Fig. 1B). The approach also seemed to differentiate between ERBB3 variants with variable transactivation potencies, as the other known oncogenic variants G284R and V104M were clearly less potent as compared to E928G in promoting growth and ERBB3 phosphorylation (Supplementary Fig. 2A-B).

An expression library of randomly mutated ERBB3 variants

To perform an unbiased screen to study thousands of ERBB3 mutations in a high-throughput assay, a randomly mutated ERBB3 cDNA library was created with error-prone PCR as previously described for the iSCREAM pipeline [25,26]. As a result, the average mutation frequency in the cDNA library (pDONR221-ERBB3 library) was estimated to be 1.3 mutations per a 4 029 bp ERBB3 cDNA insert by Sanger sequencing. The cDNA library was subsequently cloned into a pBABE-gateway retroviral mammalian expression vector. To characterize the distribution of ERBB3 variants in the library, the ERBB3 insert from the pBABE-puro-gateway-ERBB3 library was PCR amplified and deep sequenced using Illumina NovaSeq6000 platform. The analysis indicated that the library was comprised of 9 013 unique ERBB3 single nucleotide variants out of the 12 088 theoretically possible (74.6%). The resulting amino acid alterations indicated the presence of 8 055 out of the 8 276 theoretically possible amino acid changes (derived by altering a single nucleotide in a codon), resulting in the coverage of 97.3% of all possible ERBB3 missense or nonsense mutations. Specific missense or nonsense mutation distribution and different transition and transversion mutation distributions are shown in Supplementary Fig. 3.

Functional genetics screen with the modified iSCREAM pipeline

The ERBB3 mutation library was transduced into Ba/F3 cells expressing ERBB2 V956R. Cells expressing the mutant expression library or wild-type ERBB3 were cultured in the presence or absence of IL-3, in the presence of 20 ng/ml of the ERBB3 ligand NRG-1 without IL-3, or in the presence of 20 ng/ml of NRG-1 without IL-3 for 48 hours prior to complete depletion of both IL-3 and NRG-1 from the culture medium. The cells expressing the ERBB3 mutation library were able to survive in the complete absence of both IL-3 and NRG-1, while all cells expressing wild-type ERBB3 died (Supplementary Fig. 2C). These observations indicated that the ERBB3 cDNA library included variants that promoted Ba/F3 cell growth by transactivating ERBB2 V956R.

To identify ERBB3 variants enabling Ba/F3 cell transformation, genomic DNA was extracted from the surviving cell populations and the ERBB3 cDNA inserts were PCR amplified and deep sequenced (>1 300 000 X) on Illumina NovaSeq6000 platform. The read counts specific for each ERBB3 coding sequence variant (normalized to total number of reads at each ERBB3 locus) in the surviving cell populations were compared to the original transduced IL-3-dependent cell populations and the enrichment of each specific mutation (fold change) as well as the relative variant frequency of each mutation in the surviving cell pool was estimated (Supplementary Information).

ERBB3 variants promoting Ba/F3 cell transformation

The combined results from three independent samples demonstrated that cDNAs encoding 18 ERBB3 missense variants were enriched by at least 25 fold in the Ba/F3 cell populations surviving in the absence of all exogenous ligands (Fig. 2). Seven of the mutations (P212L, Y265C, L361P, L482P, A676T, D797V, and E928G) were enriched by over >100 fold (Fig. 2). The known oncogenic variant E928G was one of the main hits in the screen, validating the approach. In addition, the 18 enriched variants included the missense mutations K329R and E332K annotated as likely oncogenic by cBioPortal (Supplementary Fig. 1).

To validate the hits, 12 individual ERBB3 mutations, each reaching the enrichment level of at least 25 fold, were selected and transduced into Ba/F3 cells either alone or together with ERBB2 V956R. The 12 variants included all seven with enrichment of over 100 fold, as well as five others from the group of 11 reaching the enrichment level of 25-100 fold.

When expressed in Ba/F3 cells alone, none of the 12 mutants was able to promote IL-3-independent growth (Supplementary Fig. 4A-B), consistent with the lack of dimerization partner to serve as target for transactivation by ERBB3. However, when co-expressed with the activator-incompetent ERBB2 construct (Supplementary Fig. 4B), expression of E332K, A676T, or E928G induced emergence of IL-3-independent clones (Fig. 3A). When ERBB3 expression was first primed by a 48-hour incubation in the presence of 20 ng/ml of NRG-1, three additional ERBB3 mutants, K279N, N353T, and D797V, reproducibly also promoted IL-3-independence (Fig. 3A). Importantly, however, in none of the experimental repeats did an IL-3-independent clone emerge from vector- or wild-type ERBB3-transduced clones, regardless of the presence or absence of the heterodimerizing partner ERBB2 V956R.

The L361P mutant seemed to have a slightly aberrant molecular weight in western analyses (Supplementary Fig. 4B), and flow cytometry using an antibody against the extracellular domain of ERBB3 demonstrated lack of the epitope at the cell surface (Supplementary Fig. 4C), consistent with defective maturation and translocation of the variant to the cell surface.

Functional validation of the ERBB3 variants in MCF-10A and NIH 3T3 cells

The 12 ERBB3 mutations selected for further functional characterization were also tested for activity in human MCF-10A mammary epithelial cells and mouse NIH 3T3 fibroblasts. Again, when expressed alone, none of the variants were able to promote survival of serum-starved MCF-10A cells (Supplementary Fig. 5A-B). However, when analyzed in MCF-10A cells co-expressing ERBB2 V956R (Supplementary Fig. 5C), the presence of 50 ng/ml of NRG-1 significantly enhanced the survival and proliferation of cells expressing the ERBB3 variants K279N, E332K, N353T, L482P, A676T, and E928G (Fig. 3B). No significant ligand-independent growth advantage was observed for any of the ERBB3 variants in MCF-10A background (Fig. 3B).

To address the ability of the ERBB3 variants to promote focus formation in NIH 3T3 fibroblasts, the cells infected with retroviruses encoding the different ERBB proteins were seeded in 6-well plates and cultured in 3% serum for two weeks. When expressed alone, the ERBB3 variants K279N, E332K, and D797V were able to promote formation of foci that covered significantly more area than foci formed by cells expressing wild-type ERBB3 (Supplementary Fig. 6A-C). However, more potent focus formation was observed when ERBB3 variants were analyzed in cells co-expressing ERBB2 V956R (Fig. 3C-D; Supplementary Fig. 6D), and also a greater number of ERBB3 variants demonstrated activity. Indeed, 9 out of the 12 hits selected for validation showed transforming activity in the focus formation assay, including P212L, Y265C, K279N, K329R, E332K, N353T, L482P, A676T, and E928G (Fig. 3D).

ERBB3 mutations demonstrating activity across different cell backgrounds

Statistical analysis of the growth assays repeated for 3 to 8 times demonstrated that five of the 12 hit variants consistently promoted growth when co-expressed with ERBB2 V956R in all the three cell backgrounds studied (Fig. 4). These were K279N, E332K, N353T, A676T, and E928G. The ERBB3 variant L482P also showed enhanced growth compared to the wild-type cell line in MCF-10A and NIH 3T3 cells but promoted emergence of an IL-3-independent Ba/F3 clone in only one out of eight experiments, not reaching statistical significance. In addition, four variants, P212L, Y265C, K329R, and D797V promoted some activity in two cell backgrounds but reached significance only in one. Only the ERBB3 variants L361P and D1259Y were unable to stimulate growth across all the three cell line models. The mutations were relatively ineffective when expressed alone, with the exception of K279N, E332K, and D797V that induced statistically more colony formation in NIH 3T3 cells also without concomitant ERBB2 V956R overexpression (Supplementary Fig. 6A-B), suggesting that ERBB3 pseudokinase mutants can harbor gain-of-function properties in cells that, unlike the Ba/F3 cells, express other ERBBs endogenously.

Long-read sequencing identifies co-occurring mutations

As the mutation frequency in the mutation library was higher than 1 change per ERBB3 cDNA (i.e. 1.3), the clonal evolution of Ba/F3 cells could favor enrichment of cDNAs that harbor multiple ERBB3 mutations in cis. Relatively weak mutations could also function synergistically to achieve transforming activities. To address whether composite ERBB3 mutations occurred in cis during the clonal evolution in the Ba/F3 screen, the samples analyzed with Illumina NGS – producing reads with the length of approximately 100 bp – were reanalyzed with PacBio High-Fidelity Circular Consensus Sequencing [35] producing reads covering the whole ERBB3 inserts.

Indeed, when sequencing reads coding for any of the 18 variants enriched by ≥25 fold in the Ba/F3 screen were analyzed, only 28.0% of the full-length reads included only a single ERBB3 mutation (Fig. 5A; Supplementary Information). Instead, 40.0% of the samples included another missense ERBB3 mutation, and 32.0% three or more co-occurring mutations. This distribution of single vs. multiple mutations was significantly different from those ERBB3 mutations that were only enriched by 1 to 25 fold in the Illumina NGS analysis (P = 0.015; Pearson’s Chi-squared test) that more frequently occurred in contexts of single variants (Fig. 5A). Of the 12 variants that were functionally validated, only ERBB3 K279N, N353T, and E928G were predominantly found to be present alone, suggesting that the activity of these mutations did not benefit from additional mutations (Fig. 5B-C). Interestingly, these three mutations composed three out of the five mutations with most consistent functional activity in the validation analyses carried out by expression constructs verified to include only single ERBB3 mutations (Fig. 4). The rest of the mutations co-occurred along with one or several other ERBB3 mutations in cis in the same sequencing reads (Fig. 5B-C).

Functional significance of mutation co-occurrence

To test the hypothesis that some of the ERBB3 variants with weak individual transforming potential functioned synergistically with another alteration to gain activity, the following two co-occurring mutation pairs were selected for further analysis: K329R+E332K and L361P+L482P (Fig. 5B-C). To express the double mutations in cis in the same receptor, the mutation pairs were cloned into the same cDNA molecules using 1- or 2-step site-directed mutagenesis. The activity of the double mutations in promoting cell survival and growth was then analyzed in Ba/F3 and MCF-10A cells as above and compared to the respective single mutations. In both models, the double mutant ERBB3 K329R+E332K demonstrated significantly enhanced growth as compared to cells expressing either K329R or E332K alone (Fig. 6). Similar to the analyses of single mutants, the activity was restricted to the context of simultaneous overexpression of ERBB2 V956R (Supplementary Fig. 7 and 8). However, no enhanced growth was observed with cells expressing the double mutant L361P+L482P that, resembling the cells expressing the single mutant ERBB3 L361P, seemed to possess less transforming activity as compared to cells expressing L482P alone (Fig. 6). The L361P+L482P double mutant also phenocopied L361P alone in aberrant migration of the ERBB3 band in western analyses and low or no ligand-stimulated ERBB3 phosphorylation (Supplementary Fig. 7B; Supplementary Fig. 8B-C), consistent with lack of L361P cell surface expression (Supplementary Fig. 4C).

Structural analysis of the ERBB3 variants

The observed ERBB3 mutations were mainly clustered in the ECD region and the recent cryo-EM structures with PDB codes 7MN5 and 7MN6 [36] were used for structural analysis of the region (Fig. 7A). Here, we focus on the structural analysis of the ERBB3 variants that were consistently transforming in all three cellular backgrounds, with the exception of E928G for which previous analyses are available [37,38].

K279 of ERBB3 is located centrally within the globular extracellular region of the heterodimer, located within domain II just after the dimerization arm (β-hairpin loop), at the beginning of the disulphide containing module 5 of ERBB3 (Fig. 7A). This latter module is in direct contact with the dimerization arm of ERBB2 and, although K279 does not interact directly with residues of ERBB2, its aliphatic part of the side chain does interact with the disulfide bridge of module 5. K279 forms a salt bridge and a hydrogen bond with module 4 of domain II (Supplementary Fig. 9A). Modules 2-7 of domain II can be considered to function like a “spring”: stretching the spring could narrow the ligand-binding site and contracting the spring could make it more open (Fig. 7B). The “springs” are sensitive to alterations as demonstrated by the oncogenic mutation S310F of ERBB2 that stabilizes the dimerization arm of ERBB3 (PDB entry 7MN6 [36]). With K279N, close interactions with D252 would take place if modules 4 and 5 of domain II approach each other by an Ångstrom or more. On the other hand, K279N packs against P260 and stabilizing interactions with the main-chain carbonyl atoms of T278 (at the end of the dimerization arm) and V287 (located in module 5) may also take place. Thus, because domain II and structures surrounding K279 are so critical to heterodimer formation and receptor activation, it is certainly likely that the mutant K279N, by maintaining a strong interaction with D252, would affect the function of the ERBB3/ERBB2 complex and any stabilization of the complex that results would accentuate the normal enzymatic and signaling functions.

K329 is located at the beginning of extracellular globular domain III of ERBB3 but is not in direct contact with the closest domains (II and III) of ERBB2 (Fig. 7A, C; Supplementary Fig. 9B). The mutation K329R alone, would increase the hydrogen bonding possibilities due to the long length of the side chain and the guanidinium group but conserves the positive charge. The orientation in ERBB3 suggests that K329R could further stabilize the domain in ERBB3, the guanidinium group having the potential to form two hydrogen bonds to the main-chain oxygen atom of V352 (Fig. 7C).

E332, adjacent to the S-S bridge, is located near K329 but points in the opposite direction and towards the domain II of ERBB3 (Fig. 7A, C). In the native structure, E332 forms a salt bridge with R339. With the E332K mutant, several options are possible: e.g. (1) K332 could form strong hydrogen bonds with S338 (2.5 Å) and G333 (3.6 Å) or (2) a strong electrostatic inter-subunit link connecting K332 of ERBB3 and E280 of ERBB2 could be possible but would need changes in backbone conformations or a likely rigid body movement of the domain III half of the clamp (Fig. 7C). Moreover, in the presence of K329R (double mutant), added strength would further support the structural changes and explain why K329R exerts an effect in concert with the E332K mutation but not on its own (Fig. 7C).

The N353 is one of the ten N-glycosylation sites in ERBB3 [39]. There has been evidence that site-specific glycosylation can have distinct effects [40]. Western blot analysis of cells expressing the ERBB3 variant N353T demonstrates a small size shift of the protein (see e.g. Supplementary Fig. 5B; Supplementary Fig. 6D). The structures (PDB codes 7MN5 and 7MN6 [36]) of the ECD of the ERBB2/ERBB3 heterodimer shows that N353 is glycosylated and the identified mutation to threonine (N353T) would eliminate N-glycosylation (Supplementary Fig. 9C).

The JM region is susceptible to threonine phosphorylation [41]. T678 in EGFR is a conserved site in other ERBB receptors except ERBB3 (A676 for ERBB3) [42] and a mutation in T678 of EGFR has been studied to have substantial effects [43]. The ERBB3 A676T mutant very likely increases the dimer stability based on the analysis of the EGFR and ERBB2 structures and derived models: for example, 1) an intrasubunit hydrogen bond may form between the side-chain oxygen of A676T and the main-chain nitrogen of R679, respectively equivalent to that seen for T678 and R681 in the X-ray structure of EGFR (PDB code 3GOP [41]; Fig. 7D) or 2) an intersubunit hydrogen bond may form between the side chains of A676T and R669 equivalent to that seen between T686 (chain B) and R678 (chain A) in the NMR structure of ERBB2 (PDB code 2N2A [44]; Fig. 7E). Either of these interactions, located close to the plasma membrane in a helix of the juxtamembrane region, have the potential to support the kinase domain dimerization.

Sensitivity of mutant ERBB3 transactivation to ERBB inhibitors

The potential to pharmacologically target the identified ERBB3 variants and to use them as predictive biomarkers for ERBB targeting drugs was assessed in Ba/F3 cells. As ERBB3 variants alone did not promote ERBB3-dependent growth, and as there are no clinically relevant drugs available that would directly block ERBB3, the analyses were carried out in cells co-expressing ERBB2 V956R. The cells were cultured for 72 hours in the presence of different concentrations of the ERBB2 antibodies trastuzumab or pertuzumab, or the second-generation ERBB tyrosine kinase inhibitor neratinib. The two antibodies did not show an effect for any of the cell lines regardless of the ERBB3 variant expressed when the cells were stimulated with 20 ng/ml NRG-1 for the duration of the experiment (Fig. 8A-B; Supplementary Fig. 10A). This was demonstrated by IC50 values remaining at the >10 µg/ml level, similar to treatment of cells without any ERBB3 expression (vector control) that were cultured in the presence of IL-3, consistent with the well-documented role of NRG-1 in promoting resistance to ERBB drugs [19,45–47]. When the cells were cultured in the absence of NRG-1, significantly lower IC50’s were observed (in the 1–100 ng/ml range) but the sensitivity was similar between all ERBB3 variants.

The sensitivity to the TKI neratinib was similar between all ERBB3 variants (Fig. 8C; Supplementary Fig. 10). Interestingly, however, two of the single ERBB3 mutations, P212L and L482P, were completely resistant to neratinib, even when the cells were cultured in the absence of NRG-1 (Fig. 8C; Supplementary Fig. 10B). The P212L and L482P mutations are located respectively on domains II and III, on either side of the NRG-1–binding site, and hence may result in a receptor conformation mimicking the NRG-1 bound state and leading to resistance to neratinib even in the absence of NRG-1. Cells expressing the double mutation L361P+L482P were also an outlier as they responded similarly – and with relatively high IC50 – to neratinib both in the presence and absence of NRG-1, in accordance with the very low level of ERBB3 or ERBB2 phosphorylation (Supplementary Fig. 7B; Supplementary Fig. 8C). Taken together, these observations suggest that all the ERBB3 variants can be targeted in ERBB3/ERBB2 heterodimeric complexes with anti-ERBB2 antibodies, but that the sensitivity to ERBB TKI varies.

Annotation of predictive mutations is critical for evolution of precision oncology, including therapies targeting the ERBB signaling pathway [48–50]. Analyses of the role of ERBB family proteins as predictive biomarkers has focused on EGFR and ERBB2 while there is evidence that individual mutations in ERBB3 and ERBB4 could also have clinical significance [26, 51, 52]. Here, we addressed the functional significance of 97.3% of all theoretically possible missense and nonsense mutations of ERBB3. To functionally characterize activating mutations in a pseudokinase, a Ba/F3 cell-based model was developed that depended on the activator function of ERBB3’s kinase domain in stimulating transactivation of ERBB2 in an ERBB3/ERBB2 heterodimeric context. This model was used for a modified iSCREAM functional genetics screen to allow unbiased estimation of growth advantage conferred by thousands of random ERBB3 mutations in parallel.

The screen identified 18 ERBB3 variants that were enriched in the Ba/F3 cell population dependent on ERBB3 activator function for survival. Of these, 12 were studied individually in three different cell line models. These validation experiments suggested that 10 out of the 12 variants were more active than wild-type ERBB3 in at least one cell line model and five promoted significant growth advantages in all three cell backgrounds confirming that most of the hits identified in the screen were true positives. Of the 18 variants reaching the cut-off of 25 fold enrichment, three had previously been annotated as oncogenic (E928G) or likely oncogenic (K329R and E332K) by the cBioPortal database. In addition, the variant P212L has been reported as a somatic mutation in a clinical sample representing uterine clear cell carcinoma. These findings, together with the observation that all the ERBB3 variants responded to at least some ERBB inhibitor drugs in the context of ERBB3/ERBB2 complexes, suggest that the modified iSCREAM was able to identify clinically relevant new ERBB3 mutations.

Mechanistically, ERBB3 mutations can affect the function of the receptor differently, depending on the location of the mutation and/or the amino acid change [51]. The majority of the ERBB3 variants studied here was located in the ECD but there was also a mutation located in the intracellular juxtamembrane domain and two mutations located in the TKD. The mutations have the potential to affect hydrogen bonding and hydrophobic interactions, which could result in increased stability of ERBB3 heterodimer or extend the lifetime of the active receptor.

While the great majority of the identified ERBB3 variants were functionally validated in at least one cell model, only 6 out of 12 consistently promoted growth of Ba/F3 cells when reintroduced to cells from defined expression vectors. However, four additional variants promoted growth in at least once out of eight independent tests, which – albeit not statistically significant – suggested that varying expression levels or other stochastic experimental conditions could have played a role in allowing IL-3-independent growth. Importantly though, the wild-type ERBB3 was never sufficient to promote a growth advantage regardless of the simultaneous ERBB2 co-expression. However, the fact that two of the ERBB3 mutations did not recapitulate the Ba/F3 cell growth phenotype observed in the initial screen raised the question whether some of the hits from the screen were bystander passenger mutations that accidentally co-occurred in the same cDNA together with the actual driver mutation. In principle, this was allowed by the experimental set-up that was designed to produce an average of 1.3 mutations per each ERBB3 cDNA molecule in the library used for the screen.

Co-occurrence of mutations in the ERBB genes has been observed in clinical samples and can cause enhanced signaling and/or resistance to targeted therapy [21, 53]. Thus, the co-occurrence of ERBB3 mutations in cis in the enriched IL-3-independent Ba/F3 populations was addressed by long-read sequencing using PacBio High-Fidelity Circular Consensus Sequencing. The results indicated that, indeed, 72% of the observed hit mutations co-occurred with at least one other ERBB3 missense mutation, and only ERBB3 variants K279N, N353T, and E928G were identified to occur predominantly as single mutations alone. The co-occurrence of multiple mutations in the Ba/F3 cell clones gaining growth advantage was also found to be significantly more frequent as compared to clones displaying little or no enhanced growth. Moreover, functional characterization demonstrated growth-advantage for cells expressing the double mutant ERBB3 K329R + E332K over cells expressing the same two mutations individually. These results suggest, that ERBB3 mutations occurring in cis may be functionally synergistic in vitro, consistent with the observation that in clinical cancer samples, ERBB receptors are among the oncogenes that most frequently harbor multiple somatic mutations [53].

The two hits, L361P and D1259Y, identified in our screen that did not demonstrate any growth advantage in any of the validation models were also found to co-occur in cis with other ERBB3 variants in the long-read sequencing. However, validation experiments in which the co-occurring L482P was expressed in cis with L361P failed to indicate transforming activity that would have exceeded the activity gained by L482P alone. Flow cytometry analysis demonstrated that ERBB3 L361P was not able to translocate to the cell surface and the variant also exhibited defective ERBB3 phosphorylation and activation of downstream signaling as compared to wild-type ERBB3. We were not able to identify a plausible mechanism, except a role as a passenger of L482P, why this mutation was enriched in our screen. We did not formally study the co-occurring mutations that were identified together with ERBB3 D1259Y and cannot thus conclude whether any of these were of functional significance. However, these data suggest that the modified iSCREAM method may also produce false positive observations, such as passenger mutations, although at a relatively low frequency.

The screen also failed to identify all previously discovered activating ERBB3 variants, indicating the screen was not comprehensive for all the 8 055 protein variants included in the cDNA library. We did not e.g. identify the most frequently reported ERBB3 variant V104M or the third most frequent G284R (Supplementary Fig. 1), that are both also annotated oncogenic by cBioPortal. However, our control experiments in the Ba/F3 model used for the screen demonstrated that cells expressing V104M were unable to establish an IL-3-independent population and cells expressing G284R were relatively inefficient (Supplementary Fig. 2A). These findings are consistent with the conclusion that the design of our screen to identify mutations promoting transactivation activity by ERBB3 excludes identification of variants conferring oncogenicity by some, not yet determined, mechanisms of action. It is also possible in the experimental setup that cells harboring the most potently activating mutation outgrow cells expressing variants with less activity resulting in false negative observations for the latter.

In summary, the screen of activating mutations was able to identify several novel activating mutations with putative predictive value in the pseudokinase ERBB3.

Competing Interests

M. K. A. Koivu reports grants from Finnish Cultural Foundation during the conduct of the study. D. Chakroborty declares no potential competing interests. K. Elenius reports grants from Academy of Finland, Cancer Foundation Finland, and Turku University Central Hospital, and research agreements with Boehringer Ingelheim and Puma Biotechnology during the conduct of the study; and ownership in Abomics, Novo Nordisk, Orion, and Roche outside of the submitted work. K. J. Kurppa reports grants from Academy of Finland, Instrumentarium Science Foundation, Finnish Cultural Foundation, Sigrid Jusélius Foundation, and Jane and Aatos Erkko Foundation during the conduct of the study. T. T. Airenne and M. S. Johnson report grants from Academy of Finland, Sigrid Jusélius Foundation, and Tor, Joe, and Pentti Borg´s Memorial Fund during the conduct of the study.

Acknowledgement

We thank Jukka Lehtonen for the scientific IT support (Biocenter Finland Bioinformatics and Structural Biology networks), Mika Savisalo and Maria Tuominen for skillful technical assistance, the CSC IT Center for Science for the generous computational resources (project number 2002814), NordForsk Nordic POP (Patient Oriented Products), the Solutions for Health strategic area of Åbo Akademi University, and the InFlames flagship program of the Academy of Finland on inflammation, cancer and infection, University of Turku and Åbo Akademi University.

Author Contributions

MKAK was responsible for methodology, investigation, creating and analyzing data, interpreting results and writing the original draft and editing. DC was responsible for methodology, creating and analyzing data, writing and editing. TTA was responsible for creating and analyzing structural data, interpreting results and writing and editing. MSJ was responsible for creating and analyzing structural data, interpreting results, writing and editing and supervision. KJK was responsible for supervision, methodology and writing and editing. KE was responsible for supervision, funding acquisition, methodology and writing and editing.

Competing Interests

M. K. A. Koivu reports grants from Finnish Cultural Foundation during the conduct of the study. D. Chakroborty declares no potential competing interests. K. Elenius reports grants from Academy of Finland, Cancer Foundation Finland, and Turku University Central Hospital, and research agreements with Boehringer Ingelheim and Puma Biotechnology during the conduct of the study; and ownership in Abomics, Novo Nordisk, Orion, and Roche outside of the submitted work. K. J. Kurppa reports grants from Academy of Finland, Instrumentarium Science Foundation, Finnish Cultural Foundation, Sigrid Jusélius Foundation, and Jane and Aatos Erkko Foundation during the conduct of the study. T. T. Airenne and M. S. Johnson report grants from Academy of Finland, Sigrid, Jusélius Foundation, and Tor, Joe, and Pentti Borg´s Memorial Fund during the conduct of the study.

Data Availability Statement

All data generated or analyzed during this study are included in this published article and its supplementary information files.

Hunter T. Tyrosine phosphorylation: thirty years and counting. Curr Opin Cell Biol. 2009;21(2):140-146. doi:10.1016/j.ceb.2009.01.028
Sheetz JB, Lemmon MA. Looking lively: emerging principles of pseudokinase signaling. Trends Biochem Sci. 2022;47(10):875-891. doi:10.1016/j.tibs.2022.04.011
Manning G, Whyte DB, Martinez R, Hunter T, Sudarsanam S. The protein kinase complement of the human genome. Science (80- ). 2002;298(5600):1912-1934. doi:10.1126/science.1075762
Boudeau J, Miranda-Saavedra D, Barton GJ, Alessi DR. Emerging roles of pseudokinases. Trends Cell Biol. 2006;16(9):443-452. doi:10.1016/j.tcb.2006.07.003
Zeqiraj E, van Aalten DMF. Pseudokinases-remnants of evolution or key allosteric regulators? Curr Opin Struct Biol. 2010;20(6):772-781. doi:10.1016/j.sbi.2010.10.001
James C, Ugo V, Le Couédic JP, et al. A unique clonal JAK2 mutation leading to constitutive signalling causes polycythaemia vera. Nature. 2005;434(7037):1144-1148. doi:10.1038/nature03546
Keeshan K, He Y, Wouters BJ, et al. Tribbles homolog 2 inactivates C/EBPα and causes acute myelogenous leukemia. Cancer Cell. 2006;10(5):401-411. doi:10.1016/j.ccr.2006.09.012
Gentile A, Lazzari L, Benvenuti S, Trusolino L, Comoglio PM. Ror1 is a pseudokinase that is crucial for met-driven tumorigenesis. Cancer Res. 2011;71(8):3132-3141. doi:10.1158/0008-5472.CAN-10-2662
Grandinetti KB, Stevens TA, Ha S, et al. Overexpression of TRIB2 in human lung cancers contributes to tumorigenesis through downregulation of C/EBPα. Oncogene. 2011;30(30):3328-3335. doi:10.1038/onc.2011.57
Haikala HM, Jänne PA. Thirty Years of HER3: From Basic Biology to Therapeutic Interventions. Clin Cancer Res. 2021;27(13):3528-3539. doi:10.1158/1078-0432.CCR-20-4465
Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib. N Engl J Med. 2004;350(21):2129-2139. doi:10.1056/NEJMOA040938
Paez JG, Jänne PA, Lee JC, et al. EGFR mutations in lung, cancer: Correlation with clinical response to gefitinib therapy. Science (80- ). 2004;304(5676):1497-1500. doi:10.1126/science.1099314
Slamon DJ, Leyland-Jones B, Shak S, et al. Use of Chemotherapy plus a Monoclonal Antibody against HER2 for Metastatic Breast Cancer That Overexpresses HER2. N Engl J Med. 2001;344(11):783-792. doi:10.1056/nejm200103153441101
Junttila TT, Akita RW, Parsons K, et al. Ligand-Independent HER2/HER3/PI3K Complex Is Disrupted by Trastuzumab and Is Effectively Inhibited by the PI3K Inhibitor GDC-0941. Cancer Cell. 2009;15(5):429-440. doi:10.1016/j.ccr.2009.03.020
Holbro T, Beerli RR, Maurer F, Koziczak M, Barbas CF, Hynes NE. The ErbB2/ErbB3 heterodimer functions as an oncogenic unit: ErbB2 requires ErbB3 to drive breast tumor cell proliferation. Proc Natl Acad Sci. 2003;100(15):8933-8938. doi:10.1073/pnas.1537685100
Lee-Hoeflich ST, Crocker L, Yao E, et al. A central role for HER3 in HER2-amplified breast cancer: Implications for targeted therapy. Cancer Res. 2008;68(14):5878-5887. doi:10.1158/0008-5472.CAN-08-0380
Agus DB, Akita RW, Fox WD, et al. Targeting ligand-activated ErbB2 signaling inhibits breast and prostate tumor growth. Cancer Cell. 2002;2(2):127-137. doi:10.1016/S1535-6108(02)00097-1
Hanker AB, Brewer MR, Sheehan JH, et al. An acquired HER2 T798I gatekeeper mutation induces resistance to neratinib in a patient with HER2 mutant-driven breast cancer. Cancer Discov. 2017;7(6):575. doi:10.1158/2159-8290.CD-16-1431
Xia W, Petricoin EF, Zhao S, et al. An heregulin-EGFR-HER3 autocrine signaling axis can mediate acquired lapatinib resistance in HER2+ breast cancer models. Breast Cancer Res. 2013;15(5):1-15. doi:10.1186/bcr3480
Awada A, Dirix L, Manso Sanchez L, et al. Safety and efficacy of neratinib (HKI-272) plus vinorelbine in the treatment of patients with ErbB2- positive metastatic breast cancer pretreated with anti- HER2 therapy. Ann Oncol. 2013;24(1):109-116. doi:10.1093/annonc/mds284
Hanker AB, Brown BP, Meiler J, et al. Co-occurring gain-of-function mutations in HER2 and HER3 modulate HER2/HER3 activation, oncogenesis, and HER2 inhibitor sensitivity. Cancer Cell. 2021;39(8):1099-1114.e8. doi:10.1016/J.CCELL.2021.06.001
Murthy R, Borges VF, Conlin A, et al. Tucatinib with capecitabine and trastuzumab in advanced HER2-positive metastatic breast cancer with and without brain metastases: a non-randomised, open-label, phase 1b study. Lancet Oncol. 2018;19(7):880-888. doi:10.1016/S1470-2045(18)30256-0
Cerami E, Gao J, Dogrusoz U, et al. The cBio Cancer Genomics Portal: An open platform for exploring multidimensional cancer genomics data. Cancer Discov. 2012;2(5):401-404. doi:10.1158/2159-8290.CD-12-0095
Gao J, Aksoy BA, Dogrusoz U, et al. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal. 2013;6(269). doi:10.1126/scisignal.2004088
Chakroborty D, Kurppa KJ, Paatero I, et al. An unbiased in vitro screen for activating epidermal growth factor receptor mutations. J Biol Chem. 2019;294(24):9377-9389. doi:10.1074/jbc.RA118.006336
Chakroborty D, Ojala VK, Knittle AM, et al. An Unbiased Functional Genetics Screen Identifies Rare Activating ERBB4 MutationsFunctional Screen for Activating ERBB4 Mutations. Cancer Res Commun. 2022;2(1):10-27. doi:10.1158/2767-9764.CRC-21-0021
Koivu MKA, Chakroborty D, Tamirat MZ, Johnson MS, Kurppa KJ, Elenius K. Identification of Predictive ERBB Mutations by Leveraging Publicly Available Cell Line Databases. Mol Cancer Ther. 2021;20(3):564-576. doi:10.1158/1535-7163.MCT-20-0590
Merilahti JAM, Ojala VK, Knittle AM, Pulliainen AT, Elenius K. Genome-wide screen of gamma-secretase-mediated intramembrane cleavage of receptor tyrosine kinases. Mol Biol Cell. 2017;28(22):3123-3131. doi:10.1091/mbc.E17-04-0261
Zhang X, Gureasko J, Shen K, Cole PA, Kuriyan J. An Allosteric Mechanism for Activation of the Kinase Domain of Epidermal Growth Factor Receptor. Cell. 2006;125(6):1137-1149. doi:10.1016/J.CELL.2006.05.013
Ward MD, Leahy DJ. Kinase activator-receiver preference in erbb heterodimers is determined by intracellular regions and is not coupled to extracellular asymmetry. J Biol Chem. 2015;290(3):1570-1579. doi:10.1074/jbc.M114.612085
Pahuja KB, Nguyen TT, Jaiswal BS, et al. Actionable Activating Oncogenic ERBB2/HER2 Transmembrane and Juxtamembrane Domain Mutations. Cancer Cell. 2018;34(5):792-806.e5. doi:10.1016/j.ccell.2018.09.010
Ojala VK, Knittle AM, Kirjalainen P, et al. The guanine nucleotide exchange factor VAV3 participates in ERBB4-mediated cancer cell migration. J Biol Chem. 2020;295(33):11559-11571. doi:10.1074/jbc.ra119.010925
Guzmán C, Bagga M, Kaur A, Westermarck J, Abankwa D. ColonyArea: An ImageJ plugin to automatically quantify colony formation in clonogenic assays. PLoS One. 2014;9(3). doi:10.1371/journal.pone.0092444
Warmuth M, Kim S, Gu X ju, Xia G, Adrián F. Ba/F3 cells and their use in kinase drug discovery. Curr Opin Oncol. 2007;19(1):55-60. doi:10.1097/CCO.0b013e328011a25f
Wenger AM, Peluso P, Rowell WJ, et al. Accurate circular consensus long-read sequencing improves variant detection and assembly of a human genome. Nat Biotechnol 2019 3710. 2019;37(10):1155-1162. doi:10.1038/s41587-019-0217-9
Diwanji D, Trenker R, Thaker TM, et al. Structures of the HER2-HER3-NRG1β complex reveal a dynamic dimer interface. Nature. 2021;600(7888):339-343. doi:10.1038/S41586-021-04084-Z
Collier TS, Diraviyam K, Monsey J, Shen W, Sept D, Bose R. Carboxyl group footprinting mass spectrometry and molecular dynamics identify key interactions in the HER2-HER3 receptor tyrosine kinase interface. J Biol Chem. 2013;288(35):25254-25264. doi:10.1074/JBC.M113.474882
Littlefield P, Liu L, Mysore V, Shan Y, Shaw DE, Jura N. Structural analysis of the EGFR/HER3 heterodimer reveals the molecular basis for activating HER3 mutations. Sci Signal. 2014;7(354):ra114. doi:10.1126/scisignal.2005786
Black LE, Longo JF, Carroll SL. Mechanisms of Receptor Tyrosine-Protein Kinase ErbB-3 (ERBB3) Action in Human Neoplasia. Am J Pathol. 2019;189(10):1898-1912. doi:10.1016/j.ajpath.2019.06.008
Yokoe S, Takahashi M, Asahi M, et al. The Asn418-linked N-glycan of ErbB3 plays a crucial role in preventing spontaneous heterodimerization and tumor promotion. Cancer Res. 2007;67(5):1935-1942. doi:10.1158/0008-5472.CAN-06-3023
Red Brewer M, Choi SH, Alvarado D, et al. The Juxtamembrane Region of the EGF Receptor Functions as an Activation Domain. Mol Cell. 2009;34(6):641-651. doi:10.1016/j.molcel.2009.04.034
Guy PM, Platko J V., Cantley LC, Cerione RA, Carraway KL. Insect cell-expressed p180erbB3 possesses an impaired tyrosine kinase activity. Proc Natl Acad Sci. 1994;91(17):8132-8136. doi:10.1073/pnas.91.17.8132
Thiel KW, Carpenter G. Epidermal growth factor receptor juxtamembrane region regulates allosteric tyrosine kinase activation. Proc Natl Acad Sci U S A. 2007;104(49):19238-19243. doi:10.1073/PNAS.0703854104
Bragin PE, Mineev KS, Bocharova O V., Volynsky PE, Bocharov E V., Arseniev AS. HER2 Transmembrane Domain Dimerization Coupled with Self-Association of Membrane-Embedded Cytoplasmic Juxtamembrane Regions. J Mol Biol. 2016;428(1):52-61. doi:10.1016/j.jmb.2015.11.007
Wilson TR, Fridlyand J, Yan Y, et al. Widespread potential for growth-factor-driven resistance to anticancer kinase inhibitors. Nature. 2012;487(7408):505-509. doi:10.1038/nature11249
Yang L, Li Y, Shen E, et al. NRG1-dependent activation of HER3 induces primary resistance to trastuzumab in HER2-overexpressing breast cancer cells. Int J Oncol. 2017;51(5):1553-1562. doi:10.3892/ijo.2017.4130
Yonesaka K, Iwama E, Hayashi H, et al. Heregulin expression and its clinical implication for patients with EGFR-mutant non-small cell lung cancer treated with EGFR-tyrosine kinase inhibitors. Sci Rep. 2019;9(1). doi:10.1038/s41598-019-55939-5
Jiang P, Sinha S, Aldape K, Hannenhalli S, Sahinalp C, Ruppin E. Big data in basic and translational cancer research. Nat Rev Cancer. 2022;22(11):625-639. doi:10.1038/s41568-022-00502-0
Min HY, Lee HY. Molecular targeted therapy for anticancer treatment. Exp Mol Med. 2022;54(10):1670-1694. doi:10.1038/s12276-022-00864-3
Wahida A, Buschhorn L, Fröhling S, et al. The coming decade in precision oncology: six riddles. Nat Rev Cancer. Published online November 24, 2022:1-12. doi:10.1038/s41568-022-00529-3
Jaiswal BS, Kljavin NM, Stawiski EW, et al. Oncogenic ERBB3 Mutations in Human Cancers. Cancer Cell. 2013;23(5):603-617. doi:10.1016/j.ccr.2013.04.012
Kurppa KJ, Denessiouk K, Johnson MS, Elenius K. Activating ERBB4 mutations in non-small cell lung cancer. Oncogene. 2016;35(10):1283-1291. doi:10.1038/onc.2015.185
Saito Y, Koya J, Araki M, et al. Landscape and function of multiple mutations within individual oncogenes. Nature. 2020;582(7810):95-99. doi:10.1038/s41586-020-2175-2
Lehtonen J V., Still DJ, Rantanen V V., et al. BODIL: a molecular modeling environment for structure-function analysis and drug design. J Comput - Aided Mol Des. 2004;18(6):401-419. doi:10.1007/S10822-004-3752-4

Yes there is potential conflict of interest.

Download PDF

Journal Publication

published 28 May, 2024

Read the published version in Oncogene →

Editorial decision: revise
27 Mar, 2023
Review #2 received at journal
22 Mar, 2023
Review #1 received at journal
03 Mar, 2023
Review #3 received at journal
21 Feb, 2023
Reviewer #3 agreed at journal
21 Feb, 2023
Reviewer #2 agreed at journal
17 Feb, 2023
Reviewer #1 agreed at journal
16 Feb, 2023
Reviewers invited by journal
13 Feb, 2023
Submission checks completed at journal
13 Feb, 2023
First submitted to journal
10 Feb, 2023
Unknown event
10 Feb, 2023
Editor assigned by journal
09 Feb, 2023

You are reading this latest preprint version

Trans-activating mutations of the pseudokinase ERBB3

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Materials And Methods

Results

Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1