The soluble form of pan-RTK inhibitor and tumor suppressor LRIG1 mediates downregulation of AXL through direct protein-protein interaction in Glioblastoma
Abstract
Background: Targeted approaches for inhibiting epidermal growth factor receptor (EGFR) and other receptor tyrosine kinases (RTKs) in glioblastoma (GBM) have led to therapeutic resistance and little clinical benefit, raising the need for the development of alternative strategies. Endogenous LRIG1 (Leucine-rich Repeats and ImmunoGlobulin-like domains protein 1) is an RTK inhibitory protein required for stem cell maintenance, and we previously demonstrated the soluble ectodomain of LRIG1 (sLRIG1) to potently inhibit GBM growth in vitro and in vivo. Methods: Here, we generated a recombinant protein of the ectodomain of LRIG1 (sLRIG1) and determined its activity in various cellular GBM models including patient-derived stem-like cells and patient organoids. We used proliferation, adhesion and invasion assays, and performed gene and protein expression studies. Proximity ligation assay and NanoBiT® complementation technology were applied to assess protein-protein interactions. Results: We show that recombinant sLRIG1 downregulates EGFRvIII but not EGFR, and reduces proliferation in GBM cells, irrespective of their EGFR expression status. We find that sLRIG1 targets and downregulates a wide range of RTKs, including AXL, and alters GBM cell adhesion. Mechanistically, we demonstrate that LRIG1 interferes with AXL but not with EGFR dimerization.
Conclusion: These results identify AXL as a novel sLRIG1 target and show that LRIG1-mediated RTK downregulation depends on direct protein interaction. The pan-RTK inhibitory activity of sLRIG1 warrants further investigation for new GBM treatment approaches.
IMPORTANCE OF THE STUDY
Endogenous RTK inhibitors are poorly studied and have so far not been exploited in a therapeutic perspective for cancer patients. We describe for the first time the anti-tumor activity of a recombinant RTK inhibitory protein (LRIG1) that strongly alters cell proliferation and adhesion, which provides the basis for a novel protein-based therapeutic concept targeting multiple RTKs. We confirm that soluble LRIG1 downregulates multiple RTKs, and identifiy AXL as a novel LRIG1 target, known as a driver of GBM progression and resistance to anti-EGFR therapy. The finding that downregulation of AXL is dependent on direct protein-protein interaction provides novel insight into the mechanism of LRIG1- induced RTK regulation. Our data indicate that recombinant LRIG1 is endowed with pan-RTK activity and has strong potential for overcoming therapeutic resistance against small molecule inhibitors in GBM.
1.Introduction
Aberrant growth factor signaling through receptor tyrosine kinases (RTKs) is a hallmark of cancer, and triggers abnormal cell proliferation, enhanced motility and therapeutic resistance of many solid tumors including glioblastoma (GBM) 1. The Cancer Genome Atlas (TCGA) project confirmed alterations in RTK genes or downstream pathways as essential drivers of GBM, detectable in >80% of patients 2. In particular, the epidermal growth factor receptor (EGFR) is predominantly amplified/mutated. The most common structural variant in GBM, EGFR variant III (EGFRvIII), is characterized by a truncated extracellular domain lacking the ligand-binding site, is constitutively active and highly oncogenic 3. Altered signaling via EGFR and/or EGFRvIII is involved in GBM proliferation and invasion 4. At the molecular level, EGFR and EGFRvIII interplay 5 or interact with other RTKs 6,7 to fine-tune these oncogenic processes.Anti-EGFR therapies consequently have been in the focus of attention: antibodies, tyrosine kinase inhibitors (TKIs), or vaccines were tested in GBM patients, but largely remained unsuccessful 8-10, due to limited drug delivery to the brain, tumor heterogeneity and acquired resistance. Resistance to EGFR targeting was associated with activation of other RTKs such as EGFRvIII 11, InsR/IGF1R 12 or PDGFRβ13. In addition, RTKs converge on common signaling pathways, which further hinders their specific inhibition by targeted drugs.LRIG1 (Leucine-rich Repeats and ImmunoGlobulin-like domains 1) acts as a negative regulator of numerous RTKs, and modulates proliferation, invasion, and angiogenesis 14. LRIG1 was identified as a bonafide tumor suppressor 15,16 and stem cell marker 17,18, further associated with good prognosis in several cancers 19.
Although the mechanism of action remains a matter of debate, this membrane protein is thought to downregulate RTK signaling through receptor ubiquitination, internalization and degradation 14. The extracellular part of LRIG1 can be shed from the membrane and released in the extracellular space as soluble LRIG1 (sLRIG1), and is sufficient to inhibit EGFR signaling 20,21. We previously reported that the treatment of GBM patient-derived orthotopic xenografts (PDOX) with sLRIG1 reduced tumor growth and improved mouse survival, in GBM with or without EGFR amplification. sLRIG1 also impaired proliferation of U87-derived cells with different EGFR expressionstatus. These data suggested that sLRIG1 reduced GBM progression at least partially through EGFR- independent mechanisms, potentially targeting other mediators of tumor growth 22.Here, we have engineered, produced and purified a recombinant human sLRIG1 protein (rh-sLRIG1) that reduces proliferation of GBM cells and patient-derived organoids in vitro, and impacts cell adhesion. sLRIG1 downregulates EGFRvIII, but not wild-type EGFR. We show that sLRIG1 affects multiple RTKs at once, and we identify AXL as a novel target. Mechanistically, sLRIG1 hinders AXL dimerization, but does not interfere with EGFR.
2.Materials and methods
Cell lines and primary patient-derived GBM organoids. Glioblastoma stem-like cells (GSCs) and U87-derived cell lines were cultured as described previously 22,23. Collection and use of patient tumor tissue samples was performed after patients provided informed consent, and approved by the appropriate local ethics committee (National Ethics Committee for Research (CNER), Luxembourg, REC-LRNO-20110708). Establishment of orthotopic patient-derived xenografts (P3, T16 and T188) and animal care were described previously 22,23. Surgical procedures were performed in accordance with the regulations of the European Directive on animal experimentation (2010/63/EU). At tumor endpoint, brain tissue was harvested and processed with Neural Tissue Dissociation kit (Miltenyi) followed by Mouse Cell Depletion kit (Miltenyi). 1000 single cells were seeded per well (384-well plate) for tumor organoid formation.Production of recombinant human sLRIG1 (rh-sLRIG1) and control IgG, and cell treatment. The LRIG1 ectodomain (A35-S779) was C-terminally tagged with a Histidine tag, shuttled in a baculovirus vector (Gateway) and expressed in Sf9 cells. The recombinant protein was purified using an Imac column on an Äkta purifier (GE Healthcare), eluted using an imidazole gradient and dialyzed against phosphate-buffered saline (PBS). For cell-based assays, rh-sLRIG1 protein or IgG control protein were added to the culture medium at a concentration of 15 µg/mL (= 0.18 µM), and incubated for 6 days.Circular dichroism (CD) and fluorescence spectroscopy.
Far-UV CD spectra (195-260 nm) were recorded with a Jasco J-810 spectropolarimeter, with a protein concentration of 0.1 mg/ml (1.15 µM). Data are presented as the molar residue ellipticity ([Ɵ]MRW). Intrinsic fluorescence spectrum was recorded with a Cary Eclipse spectrofluorometer (Varian), with a protein concentration of 0.01 mg/ml. The excitation wavelength was 280 nm and emission was recorded in the 300-440 nm range, at a rate of 600 nm·min-1 (See Supplementary Methods).Cell proliferation assay. Proliferation rates were obtained by plating 1×105 cells per well in 6-well plates. After six days, cells were counted using an automated cell counter (Countess, Life Technologies), and the cell number fold change was calculated.Viability and cytotoxicity assays. PDOX-derived cells (P3, T16 and T188) were freshly isolated and seeded in 384-well plates. After spheroid formation, we applied rh-sLRIG1 or R428 (S2841, SelleckChem) and performed CellTiter-Glo®2.0 and CellToxTM-Green assays (Promega), according to the manufacturer’s instructions. IgG and DMSO were used as negative controls. Analysis and IC50 determination were performed via GraphPad Prism 7 software. (See Supplementary Methods).Invasion assay. Boyden chamber assays were performed using inserts coated with collagen type I (Sigma) and ECM proteins (Sigma).
After 16 hours, invaded cells were fixed with 4% paraformaldehyde and stained with Crystal Violet. Cells were counted in five representative fields/insert, countings were corrected for proliferation to obtain the percentage of invasion. (See Supplementary Methods).Gene expression analysis. Total RNA of U87-EGFRvIII and U87-EGFRvIII-sLRIG1 cells was extracted using the RNeasy Mini Kit (Qiagen). Total RNA was hybridized on Human Gene 2.0 ST Arrays (Affymetrix) in accordance with the manufacturer’s instructions.. A list of differentially expressed genes (DEG) was created by analysis of variance (ANOVA) with FDR<0.01 and an absolute fold change FC>2. Raw data are accessible on GEO (E-MTAB-7474). (See Supplementary Methods).Reverse transcription and qRT-PCR. cDNA was synthesized using iScript Reverse Transcriptase (BioRad) and applied for real-time PCR reaction in a Via7 instrument using Fast SYBR Green (Applied Biosystems) and specific primers (Eurogentec) (Table S1). (See Supplementary methods).Western blot and antibody arrays. Protein extracts were resolved in NuPage 4-12 % BisTris gels (ThermoFisher), and blotted onto a PVDF membrane according to standard protocols. Blots were probed with primary antibodies at 4°C overnight. Secondary antibodies (Jackson ImmunoResearch) were applied, and blots were developed with a chemiluminescent substrate (ThermoFisher).
Human phospho-RTK antibody array (ARY001B, R&D Systems) was performed following the manufacturer’s instructions. (See Supplementary Methods).Immunofluorescence. GBM cells were fixed with 4% paraformaldehyde and blocked for 1 hour with PBS supplemented with 0.1 % Triton X-100 (PBS-T) and 10% FBS. Primary antibodies were diluted in PBS-T and incubated for 2 hours at room temperature. After PBS washing, cells were incubated with Alexa Fluor 488, 555 or 647-conjugated antibodies (ThermoFisher). Image acquisition and analysis were performed using a LSM880 Confocal microscope and ZEN2 software (Zeiss) (See Supplementary Methods).In situ proximity ligation assay. Proximity ligation assay (PLA) was performed to detect interaction between sLRIG1 and AXL. Cells were incubated overnight with primary antibodies, and DuoLink® detection probes (Sigma) were used according to the manufacturer’s instructions. Image acquisition was performed using Ni-E microscope (Nikon) and analysis was done with Image J software (See Supplementary Methods).
NanoBiT® complementation assay. RTK dimerization and interaction with LRIG1 were monitored by NanoLuc complementation assay (NanoBiT, Promega) 24-26. U87 cells were transfected with pNBe vectors containing human EGFR, AXL or LRIG1, C-terminally fused to LgBiT or SmBiT. For competition experiments, pIRES plasmids containing untagged EGFR, AXL, LRIG1 or sLRIG1 were co-transfected. 48 hours post-transfection, cells were harvested and distributed into white 96-well plates, incubated with ligands of interest and then with Nano-Glo Live Cell substrate. RTK dimerization or interaction with LRIG1 were evaluated with a ClarioStar luminometer (BMG LabTech). The signal is reported as a ratio to “untreated” control condition (without ligand), being set to 1. (See Supplementary Methods).Statistical analysis. Data were analyzed using the GraphPad Prism 7 software. Results are reported as mean ± standard error of the mean, with the n described as the number of biological replicates. Data were submitted to Student t tests or ANOVA (two-tailed), and statistical significance was set at p < 0.05. 3.Results Recombinant soluble LRIG1 downregulates EGFRvIII, and reduces proliferation of GBM cells and patient-derived organoids, independent of their EGFR status. We have previously shown that the soluble fragment of LRIG1 (sLRIG1, Fig 1. A) inhibits GBM proliferation in vivo and in vitro in different GBM cells and patient-derived orthotopic xenografts (PDOX), irrespective of their EGFR status 22. We established a GBM cell line expressing sLRIG1 (Fig S1. A) that presented a significant proliferation defect (Fig 1. B), which was associated with a dramatic reduction in EGFRvIII protein levels (Fig 1. C). To further validate the effect of sLRIG1 in other GBM cell types, we generated a purified recombinant human sLRIG1 protein (rh-sLRIG1), allowing cell treatment in a reproducible and standardized fashion (human IgG was used as a negative control protein). Folding of rh-sLRIG1 was confirmed (Fig S1. B, C), consistent with the published X-ray structure of the LRIG1 ectodomain 27. We applied rh-sLRIG1 to U87 cell lines with different EGFR status, and verified that rh- sLRIG1 was captured by cells under treatment (Fig S1, D-E). After 6 days of treatment, we observed a significant reduction in cell number compared to control, in U87-EGFRvIII, U87-EGFR, and U87 (Fig 1. D-F). In agreement with our previous results, EGFRvIII levels were significantly reduced upon treatment with rh-sLRIG1 (Fig 1. G). In contrast, the level of wild-type EGFR was unchanged in U87-EGFR (Fig 1. H) and slightly increased in U87 (Fig 1. I). Similar to what was observed for full-length LRIG1 28, rh- sLRIG1 seems to impact more strongly on EGFRvIII expression compared to wild-type. EGFRvIII downregulation was concentration-dependent and most prominent 6 days after treatment, with 15 µg/mL giving the strongest effect (Fig S1. F-G). We further validated the anti-proliferative activity of rh-sLRIG1 in patient-derived GBM stem-like cells (GSCs), growing as spheres in serum-free medium (Fig 1. J). After 6 days of rh-sLRIG1 treatment, proliferation was significantly reduced in NCH465 (Fig 1. J, K) and NCH601 (Fig 1. J, L). EGFR expression remained unchanged in NCH465 (Fig 1. M), and undetectable in NCH601 (Fig 1. N), further supporting an EGFR-independent effect of sLRIG1. Similar to U87- EGFRvIII cells, the strongest effect was seen with 15 µg/mL rh-sLRIG1 (Fig S1. H-I). We validated the growth-inhibitory activity of rh-sLRIG1 in different patient-derived organoids, derived from patients P3 (no EGFR amplification), T16 (EGFR-amplified) and T188 (EGFR-amplified), treated for 6 days in vitro before measuring Viability/Cytotoxicity. We observed that rh-sLRIG1 reduced the luminescence signal emitted from viable cells (referred to as “viability”) in P3, but also in T16 and T188, in a concentration-dependent manner (Fig 1. O-Q). We attributed this effect to a reduced cell proliferation, since rh-sLRIG1 hardly induced cell death (referred to as “cytotoxicity”) below 30 ug/mL. Altogether, these results validate rh-sLRIG1 as an efficient compound endowed with anti-proliferative activity against GBM, and suitable for soluble application. The EGFR-independent effect was in line with our in vivo study 22, and prompted us to ask whether other RTKs might be involved in the functional impact of sLRIG1. We performed a human phospho-RTK antibody array, and observed that sLRIG1 expression or treatment with rh-sLRIG1 both reduced phosphorylation of known LRIG1 targets and numerous additional RTKs (Figs 2. A, S1. J-K). Western-blot analysis showed that total protein level of ErbB2, Met, but also novel LRIG1 targets, e.g. PDGFRβ and AXL receptors were significantly reduced in NCH465 and NCH601 upon treatment with rh-sLRIG1 (Fig 2. B-C). Noteworthy, AXL belongs to the TAM family of receptors (Tyro3-AXL-Mer) 29, is involved in epithelial-to-mesenchymal transition (EMT) in cancer 30, and is described as a regulator of cell migration/invasion, especially in GBM 31,32. Protein analysis showed that AXL levels were significantly downregulated by rh-sLRIG1 addition in U87-derived cells, independent of their EGFR status (Fig 2. D), and also upon sLRIG1 expression (Fig 2. E). These results identify multiple novel LRIG1 targets. Among them, AXL receptor is downregulated at the protein level, independent of EGFR expression in the cells. To investigate the genome-wide impact of sLRIG1, we performed transcriptomic analysis of sLRIG1 overexpressing cells. From 750 differentially expressed genes (DEGs) (FDR ≤ 0.01, fold change ≥ |2|), the top DEGs were validated by qPCR (Fig S2. A-K). RTKs known to be targeted by LRIG1 were not downregulated at the mRNA level, neither after sLRIG1 overexpression (Fig S2. L) nor after rh-sLRIG1 treatment (Fig S2. M-O). Gene ontology (GO) analysis of the protein-coding DEGs by WebGeSTALT indicated (1) cell adhesion, (2) extracellular matrix organization, and (3) migration as the main biological processes influenced by sLRIG1 (Fig 3. A, S3. A). Heatmaps displaying DEGs in these enriched biological processes showed clear distinct clustering (Fig S3. B-D). The analysis of TCGA data via the GlioVis® platform 33 showed that in GBM patients, AXL expression is highly correlated with most of the genes that are included in these GO categories, whereas EGFR shows a contrasting correlation profile (Fig 3. B). These results corroborated the morphological changes observed in sLRIG1-expressing cells: they displayed a more condensed shape and formed clusters. We observed a reorganization of vimentin filaments, and a collapse of the actin cytoskeleton (Fig 3. C). Based on the involvement of AXL in cell adhesion and invasion, we analysed AXL contribution to GBM cell morphology and confirmed its colocalization with vinculin at the tip of actin filaments (Fig 3. D). AXL staining was associated with actin at cell protrusions in GBM cells, which were lost in the presence of sLRIG1 (Fig S4, A-B). We also studied the effect of rh-sLRIG1 on the EGFR-negative GSC line NCH601 plated on ECM by quantifying the number of low-adherent (rounded) versus adherent cells. We found that rh-sLRIG1 treatment induced a dramatic change in cell adhesion compared to IgG (Fig 3. E, F), recapitulating the previous results. A similar albeit lower effect was observed upon treatment with the AXL inhibitor R428 compared to DMSO (Fig 3. E, G).. These data prompted us to ask whether sLRIG1 could also affect GBM cell invasion. We found that sLRIG1 overexpression dramatically impaired cell invasion through an ECM- coated membrane (Fig 3. H), and this effect was reproduced after treatment with rh-sLRIG1 (Fig 3. I). Together with recent reports 34,35, these data endorse AXL as a regulator of cell motility, proliferation and survival, and suggest that AXL might be involved in the cell remodelling observed upon sLRIG1 treatment. We further confirmed that AXL inhibition by R428 potently reduced cell viability in a concentration-dependent manner, in NCH601 cells and in P3 and T16 patient-derived organoids (Fig 3. J-L). At the molecular level, sLRIG1 interferes with AXL but not with EGFR dimerization. Immunofluorescence staining indicated colocalization of AXL with LRIG1 in GBM cells (Fig 4. A). By in situ proximity ligation assay we observed a relative increase in the number of interaction foci in sLRIG1 expressing cells, confirming a close interaction between both proteins, (Fig 4. B). Based on the observation that AXL was downregulated by LRIG1, whereas EGFR was not, we aimed to determine the differential mechanism of action. We assessed the real-time interaction between LRIG1 and RTKs, using a nanoluciferase-based complementation assay (NanoBiT, Promega) 24 in GBM cells transfected with plasmids encoding for AXL or LRIG1, both C-terminally fused to a part of the luciferase enzyme (LgBiT or SmBiT) (Fig 4. C). As reported previously 36, AXL-AXL interaction (dimerization) generated a strong signal, even in absence of ligand. Of note, the AXL-LRIG1 coupling also generated a strong signal, indicating putative protein-protein interaction. Such signal was not observed in the LRIG1-AXL configuration, which might be explained by the different sizes of the C-terminal tails of the two proteins fused to the fragments of the split nanoluciferase (Figs 4. D, S5. A). In the presence of Gas6, AXL dimerization increased in a concentration-dependent manner (Fig 4. E-F), but no ligand-induced increase in AXL-LRIG1 interaction was recorded for the two configurations tested (Fig 4. F). To circumvent the possible bias linked to the size of the C-terminal tails of the proteins for efficient complementation, we monitored AXL dimerization in the presence of non-tagged competitors that were introduced in increasing concentration (Fig 4. G). In the presence of untagged AXL monomer in excess, the Gas6-induced signal increase could no longer be observed, indicating an interference with dimerization of the tagged receptors. Similar results were obtained with an excess of LRIG1, as well as in the presence of sLRIG1 (the secreted form of LRIG1) (Figs 4. H, S5. B). These results indicate that both sLRIG1 and full-length LRIG1 directly interact with AXL in a ligand-independent manner, resulting in impaired AXL dimerization. We applied the same approach to determine LRIG1 interaction with EGFR (Fig 4. I). Without EGF, we observed ligand-independent dimerization of EGFR 37. However, the low signal for EGFR-LRIG1 interaction indicated a very low rate of molecular encountering (Fig 4. J), despite both proteins being expressed after transfection (Fig S5. C). In the presence of EGF, we observed a concentration-dependent increase in EGFR dimerization (Fig 4. K-L), which was not observed for EGFR-LRIG1 and other couplings (Fig 4. L). The signal corresponding to EGFR dimerization was abrogated upon excess of untagged EGFR competitor, but not affected by increasing amounts of LRIG1 or sLRIG1 (Fig 4. M-N), indicating that they do not interfere with EGFR dimerization. In summary, we demonstrate a direct protein-protein interaction of LRIG1 (and sLRIG1) with AXL, but not with EGFR, suggesting that LRIG1-induced downregulation of RTKs requires molecular interactions. 4.Discussion As a negative regulator of RTKs, in particular EGFR and other members of the ErbB family, LRIG1 represents a promising anticancer agent for various malignancies, including GBM. Our previous work showed that interstitial delivery of the soluble part of LRIG1 (sLRIG1) potently inhibited GBM growth in vivo independent of EGFR status 22, suggesting that additional RTKs may be involved. In this study, we introduce recombinant human sLRIG1 protein (rh-sLRIG1) as an active protein, suitable for soluble application. We show that rh-sLRIG1 efficiently affects growth and adhesion of GBM cells and patient- derived organoids, and has pan-RTK inhibition activity. We highlight AXL as a novel sLRIG1 target and show a direct LRIG1-AXL interaction that appears to be required for receptor regulation. Our findings enlarge the panel of LRIG1 targets to 10 RTK members including EGFR, ErbB2, ErbB3, ErbB4, Met, Ret, Ron, PDGFRα, IGF1R 14,38 and AXL, non-TK receptors 39 and tumor-specific RTKs such as EGFRvIII. A previous report showed that LRIG1 expression induced a stronger downregulation of EGFRvIII protein levels compared to EGFR wild-type 28, which we confirm here with rh-sLRIG1 treatment. The sensitivity of EGFRvIII to sLRIG1 contrasts with its resistance to small molecule inhibition 40, and likely attributable to its truncated ectodomain, possibly responsible for a better interaction with sLRIG1, or to specific signaling related to its constitutive activity. Along the same line, a recent study showed that the tumor-suppressive effect of LRIG1 was stronger in lung cancer cells harboring EGFR mutations, compared to EGFR wild-type cells 38. Our results show that sLRIG1 does not interact with EGFR, but its interaction with EGFRvIII could not be firmly tested in our assay (this receptor does not respond to ligand). In the case of AXL, our results strongly correlate the protein-protein interaction (and interference with receptor dimerization) with the RTK downregulation at the protein level. We speculate that (soluble) LRIG1 exerts a receptor-specific anticancer activity that might be depending on receptor sequence, 3D-structure, or third-party interactors. The signaling events that occur downstream of LRIG1-RTK interactions can therefore be diversified accordingly and need to be comprehensively untangled. We show that rh-sLRIG1 not only affects cell proliferation but impairs adhesion and invasion. The cytoskeleton plays a pivotal role in cell adhesion, intracellular organization, differentiation, and division, and cytoskeleton defects directly impact cell survival 42. RTKs are recognized modulators of actin dynamics and other cytoskeleton proteins. Among them, AXL modulates actin polymerisation, promotes cell-matrix adhesion, connects with focal adhesions in lung cancer, Schwannoma or GBM cells 34,35,43, and is generally associated with a migratory phenotype. Here, we confirm that AXL inhibition affects GBM cell adhesion, which may underlie the sLRIG1 effect, although at this stage we cannot rule out additional AXL-independent effects. Our data also shows that sLRIG1 affects GBM cell invasion, but the extent of RTK involvement remains to be elucidated. EGFR, EGFRvIII and AXL contribute to GBM invasion (either individually or in combination) 4,44, while concurrently impacting other cell phenotypes. The redundancy of RTK signaling and the development of resistance to targeted RTK inhibitors are well- known challenges in cancer. Several reports indicated AXL as a major driver of resistance to anti-EGFR therapies, e.g. in lung cancer 46 and in GBM 47. At the molecular level, AXL-EGFR heterodimerization or transactivation diversify downstream signaling into additional pathways, beyond those triggered by individual receptors, which limits the efficacy of EGFR targeting strategies 44,48 and stresses the potential of AXL inhibition in the treatment of EGFR-driven GBM. With regard to the dismal outcome of RTK- targeting in GBM, a broader perspective on RTK inhibition could be considered. Pan-RTK targeting based on drug combination approaches 49, Pan-ErbB inhibitors 50 or antibodies 51 was shown to overcome resistance and to suppress tumorigenesis more efficiently than single receptor targeting. We have demonstrated that sLRIG1 is endowed with pan-RTK inhibitory activity, and may be a valid candidate in this context. Ultimately, a better insight in the molecular determinants of sLRIG1 activity will help to design a more effective and clinically suitable ONO-7475 sLRIG1-based therapeutic against RTK- dependent cancers.