The molecular context and substrate profiles of human cell surface receptor tyrosine kinases

In a recent study published in the journal EMBO Reports, researchers discussed and identified the systemic functions and interactions of 52 active cell surface receptor tyrosine kinases (RTKs), providing insights on RTK-RTK interactions and substrates, RTK-protein complexes, and signaling pathways.

Study: Physical and functional interactome atlas of human receptor tyrosine kinases. Image Credit: ibreakstock/Shutterstock
Study: Physical and functional interactome atlas of human receptor tyrosine kinases. Image Credit: ibreakstock/Shutterstock


Cell-to-cell communication is a crucial process in determining cellular decisions that take place through chemical signaling using molecules or receptors. Most cell-to-cell communications are facilitated by protein kinases.

Of all the human protein kinases, 137 are tyrosine kinases associated with intercellular communications. RTKs regulate the properties of substrate proteins that help coordinate biological pathways and are also known to be associated with diseases such as cancer.

RTKs, which exist on the cellular membranes, function by phosphorylating their downstream cytoplasmic substrates during stimuli such as growth factors. Upon stimulation, the RTKs oligomerize and activate the signals. The intracellular domains then transphosphorylate the neighboring RTKs. The interactions and phosphorylation of RTKs with their substrates can either be stable or transient.

Mapping of protein-protein interactions (PPI) can help understand the role of RTKs in cellular signaling networks. However, a lack of knowledge of the proteins interacting with RTKs has hindered this mapping.

About the study

In the current study, researchers have used two PPI mapping methods, affinity purification coupled to mass spectrometry (AP-MS), to determine the stable interactions and proximity-dependent biotin identification (BioID) for detecting proximal and transient interactions.

A total of 52 RTK constructs obtained from different sources were cloned into expression vectors, AC-TAG-C, and pcDNATM-DEST40, using gateway cloning. To obtain UltraID compatible destination vectors, plasmids with tags; C-terminal: StrepIII/HA/UltraID) for BioID and N terminal: UltraID/HA/StrepIII for AP-MS were synthesized and inserted into MAC-tagged vectors, with MAC-tag removed.

Stable cell lines expressing MAC-tagged RTK baits were developed by co-transfecting the Flp-In 293T-REx cell lines with expression RTK vectors and pOG44 vectors. Transfected cell lines were selected using antibiotics, streptomycin, and hygromycin. Positive clones were pooled, amplified, and analyzed by AP-MS and BioID.

For ligand binding experiments, cells were treated with the RTK ligands, namely, Epidermal growth factor (EGF), Fibroblast growth factor (FGF), Insulin-like growth factor (IGF), Hepatocyte growth factor (HGF), Neurotrophin-3 (NT-3), Platelet-derived growth factor-BB subunits (PDGF-BB), Glial Cell-Derived Neurotrophic Factor (GDNF), for eight RTKs.

For the NTRK3 ultraID experiment, Flp-InTM293T-REx cell lines expressing ultraID-tagged NTRK3 were transfected with the expression vector and the pOG44 vector and further selected using the antibiotics. RTK interactors were affinity purified for AP-MS and BioID.

In vitro kinase assay was performed in the HEK293 cells. Briefly, the proteins from cells were determined by BCA assay followed by the inactivation of endogenous kinases. For kinase reaction, the cell lysates were treated with kinase and gamma [18O4]-ATP to determine the direct substrates of RTKs. Recombinant RTKs were used for the study. Further, phosphopeptides were enriched using immobilized metal ion affinity chromatography.

Liquid chromatography-mass spectrometry (LC-MS) was done. Data were filtered using Significance Analysis of INTeractome (SAINT) express version 3.6.0 and Contaminant Repository for Affinity purification tools. Known interactors were mapped using several databases. Expression of high-confidence interactors (HCIs) and protein baits were checked with the Human Protein Atlas database version 20.1.

Bioinformatic analyses were done to determine prey–prey cross-correlation, their association in the matrix, kinase domain sequence-based clustering, phosphotyrosine sites, AP-MS, BioID data clustering, and other parameters.

The signaling pathway was monitored by cignal 45-pathway reporter array. RTK-RTK interactions were determined by co-precipitation and Western blot.

Study findings

Cell lines were investigated for expression of RTKs, wherein it was found that HEK293 was one of the highly RTK expressing cell lines. Stringent statistical filtering led to the identification of 6,050 unique HCIs.

A total of 1,145 interactions were identified with AP-MS, 4,497 with BioID, and 408 with both methods. The interactors consisted of 1,521 unique proteins. Several RTK subfamilies showed similar interactors. It was observed that 15 RTKs had more than 150 identified interactors, and the other 37 had fewer interactors. Across the six databases, the known interactors were found mostly in BioGRID, IntACT, and PINA2 followed by String, bioplex and human cell map.

Shared interactions between RTKs and their subfamilies were observed in both the BioID and AP-MS methods. BioID showed higher interactions with all the subfamilies except the ROS. Moreover, the researchers detected 675 interactors shared within subfamilies and 728 shared with receptors in another subfamily.

Phosphorylation was observed in all the eight RTKs, indicating their active status. RTKs tagged with MAC were localized into the cellular membrane and other cellular compartments such as endoplasmic reticulum endosomes, confirmed by immunofluorescence confocal microscopy.

Further, the specificity of interactors in different cell lines or tissues was determined. The interactors were observed in all the tissues and cell lines in the atlas; fewer than 300 were found in many, whereas fewer than 100 in only one. The co-precipitation method detected 69 AP-MS interactions. Effect of pervanadate and ligand-induced interactions were compared wherein a total of 1,132 HCIs, including 595 unique proteins, were identified, of which 872 HCIs were observed in both the treatments. In ligand treated samples, 83 were detected with an average spectral count of over five, whereas in pervanadate treated samples, 25 HCIs were observed. Four interactions, EGFR-MET (known), EGFR-INSR, FGFR1-IGF1R, and FGFR1-MET, were not seen, whereas FGFR1-FGFR2 (known) and EGFR-Eph were seen with pervanadate treatment. These results thus indicated that pervanadate treatment might not produce RTK–RTK interactions.

In the NTRK3 experiment, the BioID and ultraID results agreed concerning PLCG1, PTPN1, and PTPN11 in ultraID samples. The results indicated that pervanadate treatment helps in enhancing the identification of typical RTK-dependent interactors, gaining more insights.

Furthermore, 77 RTK-RTK interactions were identified, of which 33 were between receptors of the same subfamily. Twenty-seven were detected by BioID, which indicated membrane areas and structures shared between RTKs. Sixteen were identified by both methods, whereas 28 by AP-MS, suggesting the stable RTK-RTK heterodimers. Experimentally, co-IP detected 27 RTK interactions, except four indicating indirect interaction.

Both the methods showed interactions between EphA2 and AXL, EphA3, EphA5, EphA6, EphA7 (previously known), and LTK RTKs. These interactions likely represented heterocomplexes, as HEK293 cells do not express Eph2 widely.

To determine the involvement of RTK interactors in complexes in a variety of cellular compartments, enrichment analysis of CORUM complexes of RTKs was done with gene ontology annotations. Out of a total of 208 unique enriched complexes assign, 59 were assigned localizations. Kinase maturation complexes and TNF-alpha/NF-kappa B signaling complexes were also enriched. LTC-PLC-gamma-1-p85-GRB2-SOS signaling complex was the most commonly enriched in 27 RTKs, followed by coat protein complex II (COPII) with 21 RTKs.

Moreover, 26 nuclear complexes were also detected. The researchers also identified 93 exclusive nuclear proteins with 40 different RTKs and 909 proteins with some nuclear activity. Amongst 40 RTKs, MER and FLT3 had the highest, i.e., 22 interactions. DDR1, a collagen receptor, had the lowest 12, whereas FGFR1 had 172 interactions.

Further, with interactions of 29 known RTKs baits with HSP90-related complexes, 15 were strong interactors, ten were weak, and four did not show interactions. CDC37, HSP90AA1, and HSP90AB showed interaction with 11 RTKs. Concerning the previous studies, all the results in this experiment were comparable; however, FGFR4 was found to be a new potential HSP90 interactor in this study.

The top two domains of HCIs were SH3 and SH2, playing a role in kinase signaling. Molecular functions associated with HCIs were mostly related to ATP binding, protein kinase binding, and heat shock protein binding. Further, it was found that the ERBB2 signaling pathway was significantly enriched in almost all RTKs, while the type I interferon was enriched in all except three RTKs, indicating that RTKs share several signaling pathways. COPII vesicle coating, budding, and cargo loading related proteins were found in this study dataset, which was lacking in the previous studies.

RTK interactors formed protein clusters having different functions. The researchers found a link between 2,020 unique protein pair associations. Out of these, 105 were already known, and 130 pairs were in the same reactome pathways. A total of 21 clusters were detected using AP-MS and BioID, which had three or more proteins associated. The largest cluster found in AP-MS played a role in small molecule transport, protein phosphorylation, and platelet signaling, whereas the proteins in the largest one found in BioID were critical in vesicle trafficking and endocytosis.

Proximal RTK associations were observed with functional protein networks related to RHO GTPase signaling, as well as MAPK and PI3K/AKT signaling in BioID clusters. The AP-MS clusters showed a direct association of RTKs in Notch and WNT signaling. Further, 46 HCLs shared an association with the EphA5 and EphA7 receptors. Other nine HCIs shared with ephrin receptors A5, A6, A7, and A8, while 16 shared between EphA4, A7, and A8.

The researchers found a total of 2,254 unique phosphorylated tyrosine sites, with 7,758 unique kinase-substrate interactions or 10,194 kinase-substrate phosphorylation sites. A total of 1,027 sites were detected with only one kinase; others had around 37 kinases. Clustering studies showed several kinase groups, especially the Ephrin receptor subfamily, clustered together based on phosphosites. Most were close to their position in the kinase domain sequence-based tree.

Further, the MAPK6/MAPK4 signaling (31 substrate proteins) and RHO GTPases Activate Formins (27 proteins) pathways were found with the highest number of identified proteins. The commonly enriched pathway was the VEGFA-VEGFR2 pathway. It was found that TGF-beta had the highest number of substrates.

Kinase deficient mutants (KD) were used to determine the dependence of interactions on RTK protein kinase. Wild type (WT) RTKs, such as AXL, EphA7, and MER, had more HCIs compared to KD, whereas in others, DDR2, the KD mutant had more HCIs. Some RTKs were found to lose some interactions; others benefitted by gaining the interactions. Proximity was observed to a wider variety of proteins, probably from the irregular cellular localization for the KD mutant. The researchers observed a different response between KD and WT with ATF6, MAP/JNK, MAPK/Erk, and NFKB pathways.

Further known roles of EGFR were identified by interactome analysis. The analysis provided an additional molecular context for EGFR actions and dynamics. The EphA7 and phosphorylome were characterized. The results indicated that KD mutation did not hinder the association with proteins in Ephrin signaling; however, it may affect specific receptor localization.


Overall, the researchers study the interactome and phosphorylome of human RTKs. The results provided information on protein complexes, molecular landscape, and signaling activity of RTK. This study thus provides additional knowledge on the previously limited data available. Despite the large study, the researchers highlighted limitations concerning the use of limited cell lines and studies on RTK behaviors. Further studies are therefore essential for a complete understanding of RTKs.  

Journal reference:
Prajakta Tambe

Written by

Prajakta Tambe

Prajakta Tambe, Ph.D. worked at Queen’s University Belfast on a project that focused on studying ‘Role of Tregs in Acute Respiratory Distress Syndrome'. Prajakta completed a Ph.D. in August 2020 at Agharkar Research Institute, University of Pune, India. Her work aimed to develop dendrimer-based nanoparticles for the targeted delivery of MCL-1 gene-specific siRNA to bring about apoptosis in breast and prostate cancer cells and in vivo breast cancer xenograft models.


Please use one of the following formats to cite this article in your essay, paper or report:

  • APA

    Tambe, Prajakta. (2022, April 13). The molecular context and substrate profiles of human cell surface receptor tyrosine kinases. News-Medical. Retrieved on June 22, 2024 from

  • MLA

    Tambe, Prajakta. "The molecular context and substrate profiles of human cell surface receptor tyrosine kinases". News-Medical. 22 June 2024. <>.

  • Chicago

    Tambe, Prajakta. "The molecular context and substrate profiles of human cell surface receptor tyrosine kinases". News-Medical. (accessed June 22, 2024).

  • Harvard

    Tambe, Prajakta. 2022. The molecular context and substrate profiles of human cell surface receptor tyrosine kinases. News-Medical, viewed 22 June 2024,


The opinions expressed here are the views of the writer and do not necessarily reflect the views and opinions of News Medical.
Post a new comment

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.

You might also like...
How COVID-19 disrupts social and temporal perception