Tumour cell responses to new fibroblast growth factor receptor tyrosine kinase inhibitors and identification of a gatekeeper mutation in FGFR3 as a mechanism of acquired resistance

Abstract

Fibroblast growth factor receptors (FGFRs) can act as driving oncoproteins in certain cancers, making them attractive drug targets. Here we have characterized tumour cell responses to two new inhibitors of FGFR1–3, AZ12908010 and the clinical candidate AZD4547, making comparisons with the well-characterized FGFR inhibitor PD173074. In a panel of 16 human tumour cell lines, the anti-proliferative activity of AZ12908010 or AZD4547 was strongly linked to the presence of deregulated FGFR signalling, indicating that addiction to deregulated FGFRs provides a therapeutic opportunity for selective intervention. Acquired resistance to targeted tyrosine kinase inhibitors is a growing problem in the clinic but has not yet been explored for FGFR inhibitors. To assess how FGFR-dependent tumour cells adapt to long-term FGFR inhibition, we generated a derivative of the KMS-11 myeloma cell line (FGFRY373C) with acquired resistance to AZ12908010 (KMS-11R cells). Basal phosphorylated FGFR and FGFR-dependent downstream signalling were constitutively elevated and refractory to drug in KMS-11R cells. Sequencing of FGFR3 in KMS-11R cells revealed the presence of a heterozygous mutation at the gatekeeper residue, encoding FGFR3V555M; consistent with this, KMS-11R cells were cross-resistant to AZD4547 and PD173074. These results define the selectivity and efficacy of two new FGFR inhibitors and identify a secondary gatekeeper mutation as a mechanism of acquired resistance to FGFR inhibitors that should be anticipated as clinical evaluation proceeds.

Introduction

The fibroblast growth factors (FGFs) act via four receptor tyrosine kinases (FGF receptors (FGFRs)) to elicit diverse physiological responses.1, 2 FGFR signalling is important during embryonic development, has been implicated in proliferation, differentiation, angiogenesis and wound repair, and deregulated FGFR signalling is associated with chondrodysplasia and craniosynostosis syndromes.3 There is growing interest in inhibiting FGFR signalling in cancer. FGFR signalling promotes tumour angiogenesis, and deregulated expression of FGFs can drive tumour cell proliferation, survival and chemoresistance;4 furthermore, increasing evidence indicates that FGFRs can act as driving oncogenes in certain cancers.5, 6 FGFRs are deregulated in a variety of cancers, including multiple myeloma (MM),7 gastric,8 breast9 and urothelial10 cancers due to gene amplification, receptor overexpression and/or activating point mutations. Translocations can also result in FGFR1 fusion proteins with constitutive kinase activity.11 Finally, alternative splicing of receptors can sensitize them to FGFs that they would not normally be responsive to.12 Inappropriate activation of FGFRs and their downstream effector signalling pathways (for example, RAS-RAF-MEK1/2-ERK1/2, phosphoinositide-3 kinase, phospholipase-C gamma (PLCγ) and signal transducer and activator of transcription 3 (STAT3)) drives cell proliferation, survival and invasion.1, 5, 6 Consequently, there is a growing interest in developing FGFR-selective tyrosine kinase inhibitors (TKIs).

The success of FGFR inhibitors (FGFRi) will require knowledge of their cellular selectivity, determinants of intrinsic tumour cell sensitivity/resistance and the mechanisms by which tumour cells adapt (acquired resistance). Here we have investigated these issues with two new FGFR TKIs, the clinical lead compound AZD454713 and the closely related molecule AZ12908010. In a panel of 16 human cell lines from four cancer types, their biological activity is confined to those cells with deregulated FGFR2 or FGFR3; absence of deregulated FGFR or the presence of other oncogenic drivers (human epidermal growth factor receptor 2 (HER2) or RAS) is associated with intrinsic resistance. Using KMS-11 MM cells, we show for the first time that acquired resistance to FGFRi can arise due to the emergence of a second site FGFR3V555M gatekeeper mutation, suggesting that such FGFR mutations should be anticipated as clinical evaluation proceeds.

Results

Cellular characterization of AZ12908010

Initially we focused on three new ATP-competitive FGFR TKIs: AZD4547,13 which has recently entered clinical trials, and AZ12576089 and AZ12908010, which were discovered during the development of AZD4547. AZ12576089 (hereafter AZ6089) exhibits in vitro inhibitory concentration (IC)50 values of 57, 5, 78 and 435 nM against FGFR1–4, respectively, and causes >70% inhibition of 14 other protein kinases, including cyclin-dependent kinase-2 (CDK2)–cyclin A, YES, SRC and LCK at 1 μM.14 AZ12908010 (hereafter AZ8010) is more potent than AZ6089 (in vitro IC50 values of 8, 1 and 17 nM against FGFR1, FGFR2 and FGFR3) and has a more favourable selectivity profile, inducing 70% inhibition of FLT1, IGF1R, YES, TBK1 and CAMKK among 75 other protein kinases.

Initial cellular characterization of AZ6089 and AZ8010 was in FGF-responsive CCl39 fibroblasts15 (Supplementary Figure 1A). FGF2-driven cell proliferation was inhibited by 100 nM AZ6089 and abolished by 100 nM AZ8010; the epidermal growth factor receptor inhibitor, AG1478, served as a negative control (Supplementary Figure 1A). AZ8010 was more potent (IC50 of 3 nM compared with 30 nM for AZ6089; Supplementary Figure 1B), consistent with their relative in vitro potencies. FGF2-driven extracellular signal regulated kinase 1/2 (ERK1/2) phosphorylation and cyclin D1 expression was abolished by AZ6089 and AZ8010 (Supplementary Figures 1C and D), whereas responses to thrombin or epidermal growth factor were unaffected (Supplementary Figures 2A–C). Some FGFRi, including PD173074, exhibit activity against platelet-derived growth factor receptor (Supplementary Figure 2A), whereas AZ6089 and AZ8010 had only a marginal effect (Supplementary Figures 2A and D).

Assessing the anti-proliferative effects of AZ8010 and AZD4547 in FGFR-dependent tumour cell lines

It is anticipated that tumour cells with deregulated FGFR signalling evolve to be ‘addicted’ to their particular FGFRs,16 making them more sensitive to FGFRi, thereby providing a therapeutic window. Accordingly, we examined a panel of 16 human tumour cells lines, including multiple myeloma, urothelial, breast and gastric cancer in which FGFRs are deregulated. For each, we first validated them by characterizing FGFR expression/activity and their sensitivity to the established FGFRi PD173074. We then examined their sensitivity to AZ8010 or AZD4547, a highly potent inhibitor of FGFR1, 2 and 3 in vitro (IC50 values of 0.2, 2.5 and 1.8 nM, respectively) with weaker activity against FGFR4 kinase (IC50 165 nM).13 AZD4547 is selective for FGFR versus vascular endothelial growth factor receptor-2 (kinase insert domain receptor) in vitro (IC50=24 nM) and in cells; exhibits excellent selectivity across a range of tyrosine and serine/threonine kinases, including insulin growth factor receptor (>2900-fold), CDK2 (>50 000-fold) and p38 (>50 000-fold); and does not inhibit ALK, CHK1, epidermal growth factor receptor, mitogen-activated protein kinase-1, MEK1, p70S6K, platelet-derived growth factor receptor, protein kinase B (PKB), Src, Tie2 or PI3k at 100 nM. For these studies, we set aside AZ6089 due to lower potency (Supplementary Figure 1) and poorer selectivity. Results with AZ8010 or AZD4547 are described below for each tumour type.

Multiple myeloma

High-level FGFR3 expression is seen in MM patients with the t(4;14)(p16;q32) reciprocal translocation,7 whereas a subset also carry point mutations within FGFR3.17 KMS-11, KMS-18, OPM2 and H929 cells all contain the t(4;14)(p16.3;q32) translocation and overexpressed FGFR3; MM.1S cells are t(4;14)(p16.3;q32)-negative18 (Figure 1a). The relative expression levels of FGFR3 were consistent with published data.19, 20 KMS-11, KMS-18 and OPM2 cells overexpress FGFR3Y373C, FGFR3G384D and FGFR3K650E, respectively, whereas H929 cells express wild-type (WT) FGFR3.18

Figure 1
figure1

Biochemical and biological activity of AZ8010 in KMS-11 myeloma cell line. (a) KMS-11, KMS-18, OPM2, H929 and MM.1S cells maintained in complete medium (10% heat inactivated fetal bovine serum (FBS)) were treated with either dimethyl sulphoxide (DMSO) vehicle control or 100 nM PD173074 (PD) for 1 h. For the last 10 min, 10 ng/ml FGF2 was added to the medium. Cell lysates were then fractionated by SDS–polyacrylamide gel electrophoresis (PAGE) and subjected to western blot analysis with the indicated antibodies. (b) KMS-11 cells were treated for 1 h with increasing concentrations of AZ8010 or 100 nM PD before a 10-min treatment with 10 ng/ml FGF2 (F). Cell lysates were then fractionated by SDS–PAGE and subjected to western blot analysis with the indicated antibodies. (c) KMS-11 cells were treated for 24 h with the indicated concentrations of AZ8010 in medium containing 10% FBS. Thymidine (5 μM) containing 0.5 μCi [3H]thymidine was added for the final 6 h of treatment, and [3H]thymidine incorporation was quantified by liquid scintillation counting. (d, e) KMS-11 cells were treated for (d) 24 h or (e) 96 h with DMSO vehicle control, the indicated concentration of AZ8010 or 100 nM PD in medium containing 10% FBS. Cells were then stained with propidium iodide and their cell cycle distributions assessed by flow cytometry. In ce, data points represent mean±s.d. of triplicate determinations. In all cases, results are from a single representative experiment that was carried out n=3 times.

FGFR signalling is only partially ligand-independent in KMS-11 and OPM2 cells, and is completely dependent on ligand in KMS-18 cells.20 Indeed, FGF2 increased phosphorylation of the FGFR substrate, FRS2α, in both KMS-11 and OPM2 cell lines, and this was abolished by PD173074 (Figure 1a), whereas P-FRS2α was not detected in KMS-18 cells, H929 cells (low levels of WT FGFR3) or MM.1S cells. PD173074 inhibited the proliferation and survival of KMS-11 and OPM2 cells, had a marginal effect on KMS-18 and H929, and had no effect on MM.1S cells (Supplementary Figures 3A and B).

To test the efficacy of AZ8010, we focused on KMS-11 cells. FGFR3 signalling (phosphorylated FGFR (P-FGFR) and phospho-ERK 1/2 (P-ERK1/2)) was abolished by 10 nM AZ8010 (Figure 1b), whilst cell proliferation was also inhibited (IC5010 nM) (Figure 1c). Cells arrested in G1 of the cell cycle within 24 h (Figure 1d), whereas dead cells (sub-G1 DNA) were observed after 96 h (Figure 1e). Similar results were observed in OPM2 cells (data not shown) and with PD173074 (Supplementary Figure 3C). In addition, AZD4547 also inhibited signalling downstream of FGFR3 and blocked the proliferation of KMS-11 cells (see below, Figure 5c). In all assays, sensitivity to PD173074, AZ8010 or AZD4547 was confined to those cells with deregulated FGFR, but KMS-18 cells were notably less sensitive; the absence of deregulated FGFR (MM.1S) or the presence of NRASG13N (H929) was associated with resistance to FGFRi.

Urothelial cancer

Mutant FGFR3 is found in 60% of low-grade and low-stage urothelial cancer (UC),10, 21, 22 and in a subtype of high-grade UC,23 whereas FGFR3 overexpression occurs in both forms.22 In agreement with published mRNA expression data,24 FGFR3 protein levels were high in 97–7 cells, moderate in RT4 and RT112 cells, and low in 97–29 and T24 cells (Figure 2a); of these, 97–7 and 97–29 cells express FGFR3S249C. P-FRS2α was observed in cell lines overexpressing FGFR3 (97–7, RT4 and RT112) but was absent from 97–29 or T24 cells (Figure 2a). PD173074 inhibited the proliferation of 97–7, RT4 and RT112, causing a G1 cell cycle arrest, but had little effect on 97–29 cells and no effect on T24 cells (Supplementary Figures 4A–C).

Figure 2
figure2

Biochemical and biological activity of AZ8010 and AZD4547 in urothelial cancer cell lines. (a) 97–7, RT-4, RT112, 97–29 and T24 cells maintained in complete medium (10% fetal bovine serum (FBS)) were treated with either dimethyl sulphoxide (DMSO) vehicle control or 100 nM PD173074 (PD) for 1 h. For the last 10 min, 10 ng/ml FGF2 was added to the medium. Cell lysates were then fractionated by SDS–polyacrylamide gel electrophoresis and subjected to western blot analysis with the indicated antibodies. (b, c) 97–7, RT-4 and T24 cells were treated for 24 h with the indicated concentrations of (b) AZ8010 or (c) AZD4547 in medium containing 10% FBS. Thymidine (5 μM) containing 0.5 μCi [3H]thymidine was added for the final 6 h of treatment, and [3H]thymidine incorporation was quantified by liquid scintillation counting. (d) 97–7, RT4 and T24 cells were treated for 24 h with DMSO vehicle control or the indicated concentration of AZD4547 in medium containing 10% FBS. Cells were then stained with propidium iodide and their cell cycle distributions assessed by flow cytometry. In bd, data points represent mean±s.d. of triplicate determinations. In all cases, results are from a single representative experiment that was carried out n=3 times.

Both AZ8010 and AZD4547 inhibited the proliferation of 97–7 and RT4 cells (Figures 2b and c). The 97–7 and RT4 cells responded to AZD4547 with a dose-dependent G1 cell cycle arrest with no evidence of cell death (Figure 2d). In contrast, T24 cells (WT FGFR3, HRASG12V) were 100-fold less sensitive to AZ8010 (Figure 2b) and completely refractory to AZD4547 (Figures 2c and d). Thus, sensitivity to PD173074, AZ8010 and AZD4547 was confined to those UC cells that exhibited high expression of either mutant (97–7) or WT (RT4, RT112) FGFR3. Low-level expression of FGFR3S249C (97–29 cells) did not confer strong FGFR dependency, and the absence of deregulated FGFR and the presence of mutant HRASG12V (T24 cells) was associated with resistance to FGFRi.

Breast cancer

The 8p11-p12 amplicon encompassing FGFR1 is observed in 10–15% breast cancer patients, and a subset of these overexpress FGFR1 protein.9 Amplification and overexpression of FGFR2 is observed in 4–12% breast tumours,9 including triple-negative breast cancer,25 and is associated with poor prognosis.26 FGFR1 protein was not expressed in Sum52-PE breast cancer cells, whereas FGFR2 was expressed at very high levels27 (Figure 3a), whereas SKBR3 and T47D cells expressed FGFR3, but not FGFR1, FGFR2 or FGFR4 (Figure 3a). P-FGFR and P-FRS2α were only detected in Sum52-PE cells (correlating with FGFR2). PD173074 treatment inhibited ERK1/2 phosphorylation (Figure 3b) and proliferation of Sum52-PE (Supplementary Figure 5A) causing a G1 cell cycle arrest (Supplementary Figure 5B); SKBR3 and T47D cells were unaffected. Thus, Sum52-PE cells were addicted to FGFR activity to maintain P-FRS2α and ERK1/2 activity. Using these cells as a model system, we found that AZ8010 caused a dose-dependent inhibition of P-ERK1/2 and P-PKB. Inhibition of both pathways was observed at 10–30 nM (Figure 3c) within 1 h of treatment (Figure 3d) and was followed by expression of p27KIP1 and the de-phosphorylation and loss of RB (Figure 3d). Consistent with this, both AZ8010 and AZD4547 caused a dose-dependent inhibition of cell proliferation with IC50s of 5 nM (Figure 3e).

Figure 3
figure3

Biochemical and biological activity of AZ8010 and AZD4547 in the Sum52-PE breast cancer (BC) cell line. (a) Sum52-PE, SKBR3 and T47D BC cell lines were treated for 1 h with dimethyl sulphoxide (DMSO) vehicle control (d) or 100 nM PD173074 (PD) in complete medium. For anti-FGFR1 and FGFR4 antibodies, a positive control comprised extracts from HEK293 cells transiently transfected with relevant cDNAs; KMS-11 myeloma cell lysates were used to control for FGFR3 expression. The differences in the banding pattern of FGFR3 in BC cell lines compared with the KMS-11 cells is likely to reflect distinct FGFR3 splice forms or glycosylation patterns. (b) The same cell lines were treated for 1 h with DMSO vehicle control (−), 1 μM Selumetinib/AZD6244 (Sel), or 100 nM PD in serum-containing medium. (c, d) Sum52-PE cells were treated with (c) the indicated dose of AZ8010 or 100 nM PD for 24 h or (d) with 100 nM AZ8010 for the times indicated. (ad) Cell lysates were fractionated by SDS–polyacrylamide gel electrophoresis and subjected to western blot analysis with the indicated antibodies. (e) Sum52-PE cells were treated for 24 h with the indicated concentrations of AZ8010 or AZD4547 in medium containing 10% FBS. Thymidine (5 μM) containing 0.5 μCi [3H]thymidine was added for the final 6 h of treatment and [3H]thymidine incorporation was quantified by liquid scintillation counting. In e, data points represent mean±s.d. of triplicate determinations. In all cases, results are from a single representative experiment that was carried out n=3 times.

Gastric cancer

FGFR2 amplification, found in the more aggressive diffuse gastric cancer,28, 29, 30 is observed in 10% of primary GCs.31 KatoIII and Snu16 cells, which exhibit amplification and overexpression of FGFR2,32 displayed strong constitutive P-FGFR and P-FRS2α, which were strongly inhibited by AZD4547 (Figure 4a); in contrast, NCI-N87 had normal FGFR2 expression but exhibit HER2 amplification. The strong P-FGFR, P-FRS2α, P-STAT3, P-PLCγ, P-PKB and P-ERK1/2 signals in KatoIII and Snu16 cells were abolished by AZD4547 (IC50s of 10 nM) (Figure 4b; data not shown). Proliferation of KatoIII and Snu16 cells was abolished by AZD4547 (IC50 5 nM), whereas NCI-N87 cells were insensitive (Figure 4c). In KatoIII cells, AZD4547 treatment inhibited the expression of cyclin D1, CDK4 and CDK2, and induced p27KIP1 expression (Figure 4b) thereby causing a G1 cell cycle arrest; in contrast, NCI-N87 cells were unaffected (Figure 4d).

Figure 4
figure4

Biochemical and biological activity of AZD4547 in gastric cancer cell lines. (a) KatoIII, NCI-N87 and Snu16 cells were treated with dimethyl sulphoxide (DMSO) vehicle control (−) or 100 nM AZD4547 (+) in complete medium (10% fetal bovine serum (FBS)) for 24 h. (b) KatoIII cells were treated with the indicated doses of AZD4547 in complete medium (10% FBS) for 24 h. (a, b) Cell lysates were fractionated by SDS–polyacrylamide gel electrophoresis and subjected to western blot analysis with the indicated antibodies. The asterisk in the P-PKB blot indicates the position of P-PKB as distinct from the lower cross-reacting band. (c) KatoIII, Snu16 and NCI-N87 cells were treated for 24 h with the indicated concentrations of AZD4547 in medium containing 10% FBS. Thymidine (5 μM) containing 0.5 μCi [3H]thymidine was added for the final 6 h of treatment, and [3H]thymidine incorporation was quantified by liquid scintillation counting. (d) KatoIII and NCI-N87 cells were treated for 24 h with DMSO vehicle control or the indicated concentration of AZD4547 in medium containing 10% FBS. Cells were then stained with propidium iodide and their cell cycle distribution assessed by flow cytometry. In c, d, data points represent mean±s.d. of triplicate determinations. In all cases, results are from a single representative experiment that was carried out at least n=3 times.

In summary, characterization of 16 tumour cell lines revealed that: (i) sensitivity to AZ8010 and AZD4547 was confined to cells with deregulated FGFR activity, whether due to mutation, translocation or amplification (Table 1); (ii) within these, KMS-18 MM and RT112 UC cells were somewhat less sensitive, and (iii) the absence of deregulated FGFR or the presence of alternate oncogenic drivers (NRASG13N in H929 cells, HRASG12V in T24 cells or HER2 in NCI-N87 cells) was associated with intrinsic resistance.

Table 1 Summary of tumour cell sensitivity to FGFRi

Modelling acquired resistance to FGFRi

The success of TKIs, such as gefitinib, erlotinib and imatinib, has been marred by acquired resistance.33, 34 As new FGFRi enter clinical trials, it is timely to anticipate possible mechanisms of acquired resistance. As clinically relevant resistance mechanisms can be discovered by modelling in cell culture,34, 35 we generated AZ8010-resistant derivatives of the KMS-11 MM cell line (KMS-11R cells). KMS-11R cells grew at the same rate as parental KMS-11 cells (Supplementary Figure 6A), were 100-fold resistant to AZ8010 in proliferation assays (Figure 5a) and were also resistant in cell death assays (Figure 5b). KMS-11R cells were not cross-resistant to the multi-drug resistance transporter substrates cisplatin, etoposide and paclitaxel (Supplementary Figure 6B), arguing against a role for drug efflux pumps.

Figure 5
figure5

KMS-11R cells with acquired resistance to AZ8010 are cross-resistant to AZD4547. KMS-11 or KMS-11R cells were treated for 24 h with the indicated concentrations of (a, b) AZ8010 or (c, d) AZD4547 in medium containing 10% fetal bovine serum (FBS). (a, c) Thymidine (5 μM) containing 0.5 μCi [3H-methyl] thymidine was added for the final 6 h of treatment, and [3H] thymidine incorporation was quantified by liquid scintillation counting. (b, d) Cells were harvested, fixed, stained with propidium iodide and the fraction of dead cells (sub-G1 DNA) determined by flow cytometry. In all cases, data points represent mean±s.d. of triplicate determinations and are from a single representative experiment that was carried out n=3 times.

KMS-11R cells were cross-resistant to AZD4547 (Figures 5c and d) and PD173074 (Supplementary Figure 6C), suggesting an ‘on-target’ resistance mechanism. KMS-11R exhibited elevated basal P-FGFR, P-FRS2α and P-ERK1/2, and these markers were unaffected by AZ8010, except at 1 μM, where P-FRS2α was partially inhibited (Figure 6a). AZ8010-dependent changes in the expression of cyclin D1, D2 and p27KIP1 in KMS-11 cells were not observed in KMS11-R cells. Similar results were seen with AZD4547 (Figure 6b); basal P-FRS2α, P-ERK1/2, P-PKB, P-PLCγ and P-STAT3 were constitutively upregulated and non-responsive to drug in the KMS-11R cells compared with parental KMS-11 cells. The pro-apoptotic protein BIMEL, which is phosphorylated and inactivated by ERK1/2,36, 37 was hypo-phosphorylated and increased in abundance when KMS-11 cells were treated with AZD4547, whereas it was constitutively phosphorylated and drug refractory in KMS-11R cells. The increase in P-FGFR and FGFR signalling in KMS-11R cells was not due to increased expression of FGFR3 (Figure 6a) or other FGFRs (Figure 6c).

Figure 6
figure6

KMS-11R cells exhibit increased basal FGFR-dependent signalling that is refractory to both AZ8010 and AZD4547. (a, b) KMS-11 or KMS-11R cells were treated for 24 h with the indicated concentrations of AZ8010 (a) or AZD4547 (b). Cell lysates were then fractionated by SDS–polyacrylamide gel electrophoresis (PAGE) and subjected to western blot analysis with the indicated antibodies. Results are from a representative experiment that was carried out n=3 for each drug. (c) KMS-11 and KMS-11R cell lysates were fractionated by SDS–PAGE and subjected to western blot analysis with FGFR1, FGFR2 and FGFR4 antibodies. For FGFR2, SUM52-PE cell lysates were used as a positive control. For FGFR1 and FGFR4,+ indicates lysates from HEK293 cells transfected with the respective receptors, and indicates non-transfected HEK293 cell lysates.

Gatekeeper mutations in several tyrosine kinases confer increased activity as well as being drug refractory.38 Sequencing confirmed that both KMS-11 and KMS-11R contained the FGFRY373C mutation (Figure 7a), but we also found a heterozygous mutation encoding FGFRV555M in KMS-11R cells that was not detected in parental KMS-11 cells (Figure 7b). Val555 of FGFR3 has the same role as Val561 of FGFR1 and Val564 of FGFR2 as the gatekeeper residue in the ATP-binding site (shown in yellow in Figure 7c, upper panels); mutations at this site frequently impart resistance to TKIs.38, 39, 40 As residues comprising the ATP-binding pocket are completely conserved among FGFR1–3, with the exception of Ala559 in FGFR3 (see Methods), we used the structure of PD173074 in complex with FGFR1 to model this mutation.41 Our model suggests that Met, with a 25% bulkier side-chain, will restrict access to the cavity adjacent to the adenine ring-binding region (shown in yellow in Figure 7c, lower panel). In particular, the equivalent residue of FGFR1, Val561, makes van der Waals contact with the PD173074,41 whereas the bulkier side chain of Met561 would disrupt binding of PD173074 (Figure 7c, lower panel). Taken together, our results suggest that a secondary gatekeeper mutation accounts for acquired resistance to AZ8010, AZD4547 and PD173074 in KMS-11R cells.

Figure 7
figure7

KMS-11R cells harbour a Val555Met gatekeeper mutation in FGFR3. (a, b) Genomic DNA from KMS-11 or KMS-11R cells was extracted, and PCR was performed using M13-tagged forward and reverse primers to amplify (a) exon 9 including codon 373 or (b) exon 13 including codon 555 from FGFR3. PCR products were sequenced in both directions by dye-terminator sequencing using the ABI3730 capillary sequencer. Both KMS-11 and KMS-11R cells harbour the driving FGFR3Y373C mutation, whereas only KMS-11R harbour the FGFR3V555M gatekeeper mutation. (c) Cartoon and surface-rendered representation of the ATP-binding pocket of human FGFR1 in complex with PD173074 (sticks) (2fgi.pdb). Gatekeeper residue Val561, equivalent to FGFR3 Val555, is shown in yellow in upper panels. A model of the Val561-to-Met mutation, highlighted in yellow, is shown in the lower panels. Pocket surface contributed by Val561 or Met561 is shown in panels on right.

Genome-wide mRNA expression analysis identifies genes involved in myeloid differentiation and myeloma survival in KMS-11R cells

To identify changes in gene expression between KMS-11 and KMS-11R cells, we conducted genome-wide mRNA expression analysis using Human Illumina bead arrays. Candidate genes were selected on the basis of a false discovery rate of <0.05 and an absolute change of at least two-fold (Table 2). From this list, five (RUNX1T1, PAPSS2, SPINT2, LDCO1L and CEBPA) were validated by quantitative reverse-transcription PCR (qRT–PCR). Interleukin-6 receptor-α (IL6Rα) expression was also elevated in KMS-11R cells and although not considered significant by the above criteria, it was also analysed by qRT–PCR. In all cases, the pattern of expression in the qRT–PCR (Figure 8) was consistent with that in the arrays: PAPSS2, SPINT2, LDOC1L and CEBPA were downregulated in KMS-11R cells relative to KMS-11, whereas RUNX1T1 (runt-related transcription factor 1 translocated to 1) and IL6R were both upregulated. Notably, RUNX1T1, IL6R and CEBPA are implicated in leukaemias and myeloma.

Table 2 Genes that are differentially expressed in KMS-11R cells versus parental KMS-11 cells
Figure 8
figure8

Deregulated expression of CEBPA, RUNX1T1 and IL6R in KMS-11R cells. RNA from KMS-11 or KMS-11R cells was converted to cDNA and used as template in a qPCR reaction with SYBER green, and primers to amplify the gene products indicated. The mean expression values of glyceraldehyde-3-phosphate dehydrogenase and ribosomal protein L13A were used to normalize target gene transcript levels. Gene expression in KMS-11 cells was normalized to 1, and data points represent relative expression levels in KMS-11R cells. Error bars represent a 95% confidence interval. *All data points have a P-value of <0.008 determined by one sample t-test with the Bonferroni correction. Results include n=4 biological replicates, each with n=3 technical replicates.

Discussion

The incidence of FGFRs as oncogenic drivers in certain forms of cancer has prompted the search for new, selective FGFRi. AZD4547 has recently entered clinical trials, and its biochemical and biological activity in three FGFR-dependent tumour cell lines, including tumour xenograft studies, was recently reported.13 To broaden this analysis, we used a more diverse panel of 16 human tumour cell lines, representing four tumour types wherein FGFR deregulation is common, and also expanded our analysis to include the related compound AZ8010, providing a richer FGFR-based pharmacology. We also examined the issue of acquired resistance to FGFRi.

Deregulated FGFR2 or FGFR3 confers sensitivity to FGFRi: oncogene addiction and therapeutic opportunity

Genetic ablation of FGFR3S249C in 97–7 cells24 and anti-FGFR3 antibody treatment of OPM2 and KMS-11 cells42 supports the suggestion that tumour cells evolve to become addicted to their deregulated FGFRs. Our results using AZD4547, AZ8010 and PD173074 provide further strong support for the notion that such ‘oncogene addiction’16 provides an opportunity for tumour-selective intervention by FGFRi. Irrespective of tumour type, the 16 cell lines we studied could be divided into: (i) those that were very sensitive to FGFRi; (ii) those that were moderately sensitive to FGFRi, and (iii) those that were completely resistant to FGFRi. All those that were very sensitive exhibited high expression of WT or mutant FGFR2 or FGFR3 and have presumably evolved to be addicted to signalling by these receptors for proliferation and survival. The sensitivity of KG1a cells (FGFR1 gene fusion) to AZD4547 indicates that this also applies to FGFR1.13 These results underline the exquisite dependency of these cells upon their oncogenic FGFRs (FGFR addiction) and the extent to which this provides a therapeutic opportunity for selective intervention by an FGFRi such as AZD4547. Similar results have been reported for other novel FGFRi.43

The reduced sensitivity of KMS-18 cells agrees with previous reports wherein FGFRG384D failed to transform NIH3T3 cells20 and KMS-18 were less sensitive to the TKI CHIR-258/dovitinib.44 Clearly, FGFR3 mutations are graded in terms of their activation capability and this determines the degree of addiction and sensitivity to FGFRi. The moderate sensitivity of RT112 cells contrasts with a recent report;45 this may reflect differences in assays for proliferation/viability. For example, we employed a 24-h assay, whereas Lamont et al.45 used a 5-day assay and even the moderately sensitive KMS-18 cells responded to the FGFRi after 7 days of drug treatment.18

All cells that were insensitive lacked deregulated FGFR and presumably have not evolved any dependence upon FGFR signalling. Tumour cell responses to FGFRi were also correlated with the presence of intrinsic resistance factors, such as alternative oncogenic drivers (HER2 in NCI-N87, NRASG13N in H929 and HRASG12V in T24 cells). FGFR3 or RAS mutations are mutually exclusive in UC,46 and as mutant RAS activates many of the same effector pathways, it may circumvent the requirement for FGFR signalling upstream. KRAS mutations predict a poor response to anti-epidermal growth factor receptor antibodies in colon cancer;47 it will be interesting to see if relationships between RAS status and responses to FGFRi emerge as clinical trials progress.

Acquired resistance to FGFRi arising from a gatekeeper mutation in FGFR3

Acquired resistance to TKIs can arise through multiple mechanisms.33, 34, 35 KMS-11R cells with acquired resistance to AZ8010 were cross-resistant to AZD4547 and PD173074, and sequencing revealed a point mutation at the gatekeeper residue (FGFR3V555M). Gatekeeper mutations typically inhibit drug binding in the ATP-binding site and also increase the activity of the kinase domain by facilitating assembly of an enzymatically active kinase conformation.38 Indeed, FGFR1V561M exhibits a higher level of auto-phosphorylation.48 Our results indicate that FGFR3 can also be ‘activated’ as well as being rendered drug-resistant by a gatekeeper mutation. Although gatekeeper mutations in FGFRs have previously been engineered,48 we believe this is the first instance of their emergence in any FGFR by drug selection and suggests that this mechanism of resistance to FGFRi should be anticipated as clinical evaluation proceeds, perhaps through the development of second-generation inhibitors or covalent FGFRi.39, 48

A number of TKIs that inhibit FGFR are currently in development, including mixed kinase inhibitors, such as CHIR-258/dovitinib,44 brivanib alaninate49 and BIBF-1120,50 wherein FGFR may not be the dominant pharmacology. As these compounds progress, it will become increasingly important to understand their primary targets to interpret clinical trial results. Our analysis across a panel of cell lines suggests that AZD4547 is an FGFR-selective small molecule inhibitor, and in vitro pharmacology and lack of effect on vascular endothelial growth factor receptor-2-dependent markers in vivo13 are consistent with this. The demonstration that KMS-11R cells with the FGFR3V555M gatekeeper mutation are 100-fold resistant to AZD4547, AZ8010 and PD173074 provides further strong genetic evidence that the primary target of these compounds in this system is indeed FGFR3.

Deregulated expression of IL6R, RUNX1T1 and CEBPA in KMS-11R cells

Among the changes in gene expression associated with FGFR3Y373C;V555M in KMS-11R cells (Table 2 and Figure 8), three are of particular note. CEBPA encodes CCAAT/enhancer binding protein-α, a tumour suppressor that promotes myeloid cell differentiation. The CCAAT/enhancer binding protein-α is inactivated or downregulated in leukaemias and myeloma, and its re-expression restores their differentiation programme.51 RUNX1T1 (also called ETO) encodes a transcriptional repressor, which is fused to RUNX1 (AML1) in 8q22;21q22 translocation-positive acute myeloid leukaemia. The resulting RUNX1/RUNX1T1 (AML1–ETO) fusion gene is an oncogene that promotes leukaemic progression. Interestingly, RUNX1–RUNX1T1 downregulates CEBPA expression,52 raising the possibility that the decrease in CEBPA in KMS-11R cells might be linked to the increase in RUNX1T1. Finally, IL-6 promotes myeloma proliferation and survival53 so that the increase in IL6R expression in KMS-11R cells might facilitate IL6 signalling in these cells. It is notable that three genes implicated in myeloid malignancies are deregulated in KMS-11R cells and future experiments should aim to address how they are regulated by FGFR3Y373C;V555M signalling and their role, if any, in FGFRi resistance in KMS-11R cells. For example, RUNX1T1 is also overexpressed in imatinib-resistant chronic myeloid leukaemia cells.54

In summary, the biological activity of two new FGFRi, AZ8010 and AZD4547, is confined to tumour cells with deregulated FGFR2, FGFR3 (herein) and FGFR1.13 The absence of deregulated FGFR or the presence of other driving oncogenes (for example, HER2, NRAS or HRAS) is associated with intrinsic resistance. These results support the hypothesis that FGFR addiction provides a window for selective therapeutic intervention and highlights the efficacy and selectivity of these new compounds. We also show for the first time that acquired resistance to FGFRi can arise through the emergence of gatekeeper mutations in FGFR3. This in turn provides further genetic evidence of the FGFR-selective effects of AZ8010 and AZD4547, and suggests that gatekeeper mutations should be anticipated and planned for as clinical trials progress.

Materials and methods

Cells and routine cell culture

Culture of CCl39 fibroblasts was described previously.55 MM cell lines were provided by Jonathan Keats, Mayo Clinic, Scottsdale, AZ, USA, and maintained in RPMI-1640 medium. Urothelial carcinoma cell lines, RT4 and RT112, were provided by Joyce Nutt (Northern Institute for Cancer Research, Newcastle upon Tyne, UK); 97–7 and 97–29 were provided by Margaret Knowles (Cancer Research UK Clinical Centre, Leeds, UK) and T24 cells were provided by Andrew Garner, AstraZeneca (Alderley Edge, UK). Urothelial cell lines were maintained in RPMI-1640 or Dulbecco’s modified Eagle’s medium/Ham’s F12 (1/1) medium. Breast cancer cell lines, Sum52-PE cells, were purchased from Asterand (Royston, UK) and together with SKBR3 and T47D cells were maintained in Ham’s F12, RPMI-1640 or Dulbecco’s modified Eagle’s medium. Gastric cancer cell lines, Snu16 cells (ATCC), KatoIII (ECCC) and NCI-N87 (Matthias Ebert, Klinikum recht der Isar, Munich, Germany) were maintained in RPMI-1640 medium.

Inhibitors and growth factors

AZ6089, AZ8010, AZD454713 and Selumetinib/AZD6244 (Sel)56 were provided by AstraZeneca, Alderley Park, UK. PD173074, AG1478, FGF2, PDGF-AA, thrombin and epidermal growth factor were purchased from Calbiochem, Feltham, UK.

Preparation of cell extracts and western blotting

Cells were lysed in TG lysis buffer, harvested and cell extracts fractionated by SDS–polyacrylamide gel electrophoresis as described previously.36, 55 Antibodies used for western blotting were from Santa Cruz Biotechnology (Insight Biotechnology, Wembley, UK): FGFR1 SC-121, FGFR2 SC-6938, FGFR3 SC-123 FGFR4 SC-124, RB SC-50, CDK2 SC-163, CDK4 SC-601, Cyclin D2 SC56305; Cell Signalling Technology (Hitchin, UK): Phospho-FGFR 3476, P-ERK1/2 9106, P-PKB (S473) 9271, PKB 9272, P-STAT 3 9138, P-PLCγ 2821, PLCγ 2822, P-FRS2α 3864; Calbiochem: Cyclin D1 CC12, p27 NA35; Becton Dickinson (Oxford, UK): p21 556431, ERK1 610031; and Millipore (Feltham, UK): Bim AB17003. Immune complexes were detected by enhanced chemiluminescence (Amersham, GE Healthcare, Little Chalfont, UK).

Assays of cell proliferation, cell cycle and cell death

[3H]thymidine incorporation was assayed as described previously,55 except for non-adherent cells, which were harvested on to Unifilter plates (Perkin Elmer, Cambridge, UK) using a TomTec harvester (Hamden, CT, USA) before scintillation counting. Distribution of cells in G1, S or G2/M and dead cells (sub-G1) was determined by propidium iodide staining and flow cytometry.55

Generation of drug-resistant KMS11 cells

Cells in exponential growth phase were exposed to 30 nM AZ8010. After 4 days, they were passaged (1:3) into fresh medium containing 30 nM AZ8010 and 3 days later, the drug concentration was increased to 100 nM. AZ8010 dose was increased incrementally up to 3 μM, which took 11 weeks. Resistant cells were then maintained in 3 μM AZ8010 and named ‘KMS-11R’ cells.

Modelling the impact of the FGFR3V555M mutation on PD173074 binding to the ATP-binding pocket, based on the FGFR1:PD173074 structure

Sequence alignment of FGFR1–3 employed Jalview.57 Residues in the ATP-binding pocket that contact the inhibitor PD173074 and second tier residues were identified from the crystal structure of the complex (2fgi.pdb)41 by measuring proximity of surrounding residues to PD173074 using the programme PyMOL (www.pymol.org). These 65 residues were almost completely conserved among the three FGFR proteins. Notably, pocket residues equivalent to Ala559 in FGFR1 and FGFR2 (Ser565 and Ser568, respectively) contribute to the pocket surface only via their main-chain h side-chain. Modelling mutation of FGFR1 residue Val561 to Meth, equivalent to the FGFR3 Val555Met, used the FGFR1:PD173074 crystal structure and electron density map (2fgi.pdb) and the programme COOT.58 The rotamer of the methione that did not clash with surrounding residues was chosen for the model. Surface and cartoon renderings of the PD173074-binding pocket were made using the programme PyMOL (www.pymol.org).

Analysis of FGFR3 for mutations

Genomic DNA from KMS-11 and KMS-11R cells was extracted and PCR was performed to amplify the required regions from FGFR3 using M13-tagged forward and reverse primers. PCR products were sequenced in both directions by dye-terminator sequencing using the ABI3730 capillary sequencer (Applied Biosystems, Warrington, UK). Sequence traces were analysed for variation manually after assembly and quality calling with polyphred/phrap/consed packages. Primers used to amplify FGFR3 regions encompassing Tyr373 and Val555 were:

Exon 9, forward: 5′-IndexTermACTGTAAAACGACGGCCAGTGAGACCCTCCAGACAAGGC-3′

Exon 9, reverse: 5′-IndexTermACCAGGAAACAGCTATGACCGCAGAGAGGGCTCACACAG-3′

Exon 13, forward: 5′-IndexTermACTGTAAAACGACGGCCAGTGGGAAACACAAAAACATCATCAA-3′

Exon 13, reverse: 5′-IndexTermACCAGGAAACAGCTATGACCCTGAAGCCTCTCCACCTCTC-3′

Bold sequence indicates M13 tags.

mRNA expression arrays

Triplicate dishes of KMS-11 and KMS-11R cells were lysed in TRIZOL (Life Technologies, Glasgow, UK). Samples were sent to Cambridge Genomic Services where gene expression was measured with Illumina Human-6 v2 BeadChips (Illumina UK, Saffron Walden, UK) following standard protocols. Raw data was normalized using Genespring standard differential gene analysis and a list of candidate differentially expressed genes (Table 2) were selected on the basis of a false discovery rate of <0.0559 and an absolute change of at least two-fold. Microarray data are available in the ArrayExpress database (http://www.ebi.ac.uk/arrayexpress) under accession number E-TABM-1222.

Quantitative reverse-transcription PCR

RNA was isolated from KMS-11 and KMS-11R cells and converted to cDNA by using the Applied Biosystems Reverse Transcription kit (Applied Biosystems). The cDNA was then used as a template in a qPCR reaction with SYBER green (Applied Biosystems) and primers to amplify the specific gene products. Mean expression values of glyceraldehyde-3-phosphate dehydrogenase and ribosomal protein L13A were used to normalize target gene transcript levels.60 Primers used were:

Glyceraldehyde-3-phosphate dehydrogenase, forward: 5′-IndexTermTGCACCACCAACTGCTTAGC, reverse: 5′-IndexTermGGCATGGACTGTGGTCATGAG-3′; ribosomal protein L13A, forward: 5′-IndexTermCCTGGAGGAGAAGAGGAAAGAGA-3′, reverse: 5′-IndexTermTTGAGGACCTCTGTGTATTTGTCAA-3′; PAPSS2, forward: 5′-IndexTermGGAGAAGGAGTACTTACAGGTTATGC-3′, reverse: 5′-IndexTermCTTCCAGCCGTGTCTTATCCT-3′; LDOC1L, forward: 5′-IndexTermGCTGACCTTGTTACGCACCA-3′, reverse: 5′-IndexTermGAGAGGTTGGAGGTGTGGTCA-3′; RUNX1T1, forward: 5′-IndexTermTGACTCCTCCAACAATGCCA-3′, reverse: 5′-IndexTermTCAAGGCTGTAGGAGAATGGCT-3′; IL6R, forward: 5′-IndexTermAGGCTGTGCTCTTGGTGAGG-3′, reverse: 5′-IndexTermGAATACTGGCACGGCTCCTG-3′; SPINT2, forward: 5′-IndexTermAACAGCAATAATTACCTGACC-3′, reverse: 5′-IndexTermAAGGATGCACGGCAAGGC-3′; CEBPA, forward: 5′-IndexTermTCGGTGGACAAGAACAG-3′, reverse: 5′-IndexTermGCAGGCGGTCATTG-3′. Gene expression in KMS-11 cells was normalized to 1 and data points were expressed as relative expression levels in KMS-11R cells.

Accession codes

Accessions

ArrayExpress

References

  1. 1

    Klint P, Claesson-Welsh L . Signal transduction by fibroblast growth factor receptors. Front Biosci 1999; 4: D165–D177.

  2. 2

    Eswarakumar VP, Lax I, Schlessinger J . Cellular signaling by fibroblast growth factor receptors. Cytokine Growth Factor Rev 2005; 16: 139–149.

  3. 3

    Ornitz DM, Marie PJ . FGF signaling pathways in endochondral and intramembranous bone development and human genetic disease. Genes Dev 2002; 16: 1446–1465.

  4. 4

    Pardo OE, Arcaro A, Salerno G, Raguz S, Downward J, Seckl MJ et al. Fibroblast growth factor-2 induces translational regulation of Bcl-XL and Bcl-2 via a MEK-dependent pathway. J Biol Chem 2002; 277: 12040–12046.

  5. 5

    Jeffers M, LaRochelle WJ, Lichenstein HS . Fibroblast growth factors in cancer: therapeutic possibilities. Expert Opin Ther Targets 2002; 6: 469–482.

  6. 6

    Knights V, Cook SJ . De-regulated FGF receptors as therapeutic targets in cancer. Pharmacol Ther 2010; 125: 105–117.

  7. 7

    Keats JJ, Reiman T, Belch AR, Pilarski LM . Ten years and counting: so what do we know about t(4;14)(p16;q32) multiple myeloma. Leuk Lymphoma 2006; 47: 2289–2300.

  8. 8

    Jang JH, Shin KH, Park JG . Mutations in fibroblast growth factor receptor 2 and fibroblast growth factor receptor 3 genes associated with human gastric and colorectal cancers. Cancer Res 2001; 61: 3541–3543.

  9. 9

    Ray ME, Yang ZQ, Albertson D, Kleer CG, Washburn JG, Macoska JA et al. Genomic and expression analysis of the 8p11-12 amplicon in human breast cancer cell lines. Cancer Res 2004; 64: 40–47.

  10. 10

    Billerey C, Chopin D, Aubriot-Lorton MH, Ricol D, Gil Diez de Medina S, Van Rhijn B et al. Frequent FGFR3 mutations in papillary non-invasive bladder (pTa) tumors. Am J Pathol 2001; 158: 1955–1959.

  11. 11

    Reiter A, Sohal J, Kulkarni S, Chase A, Macdonald DH, Aguiar RC et al. Consistent fusion of ZNF198 to the fibroblast growth factor receptor-1 in the t(8;13)(p11;q12) myeloproliferative syndrome. Blood 1998; 92: 1735–1742.

  12. 12

    Moffa AB, Tannheimer SL, Ethier SP . Transforming potential of alternatively spliced variants of fibroblast growth factor receptor 2 in human mammary epithelial cells. Mol Cancer Res 2004; 2: 643–652.

  13. 13

    Gavine PR, Mooney L, Kilgour E, Thomas AP, Al-Kadhimi K, Beck S et al. AZD4547: an orally bioavailable, potent and selective inhibitor of the fibroblast growth factor receptor tyrosine kinase family. Cancer Res 2012; 72: 2045–2056.

  14. 14

    Klinowska T, Gavine PR, Mooney L, Parveen N, Al-Kadhimi K, Thomas A et al. The characterization of novel, potent and selective small molecule inhibitors of the FGFR tyrosine kinase in vitro and in vivo. Gordon Research Conference 2–7 March 2008; Lucca, Italy.

  15. 15

    Kahan C, Seuwen K, Meloche S, Pouysségur J . Coordinate, biphasic activation of p44 mitogen-activated protein kinase and S6 kinase by growth factors in hamster fibroblasts. Evidence for thrombin-induced signals different from phosphoinositide turnover and adenylylcyclase inhibition. J Biol Chem 1992; 267: 13369–13375.

  16. 16

    Weinstein IB, Joe A . Oncogene addiction. Cancer Res 2008; 68: 3077–3080.

  17. 17

    Chesi M, Brents LA, Ely SA, Bais C, Robbiani DF, Mesri EA et al. Activated fibroblast growth factor receptor 3 is an oncogene that contributes to tumor progression in multiple myeloma. Blood 2001; 97: 729–736.

  18. 18

    Trudel S, Ely S, Farooqi Y, Affer M, Robbiani DF, Chesi M et al. Inhibition of fibroblast growth factor receptor 3 induces differentiation and apoptosis in t(4;14) myeloma. Blood 2004; 103: 3521–3528.

  19. 19

    Lombardi L, Poretti G, Mattioli M, Fabris S, Agnelli L, Bicciato S et al. Molecular characterization of human multiple myeloma cell lines by integrative genomics: insights into the biology of the disease. Genes Chromosomes Cancer 2007; 46: 226–238.

  20. 20

    Ronchetti D, Greco A, Compasso S, Colombo G, Dell'Era P, Otsuki T et al. Deregulated FGFR3 mutants in multiple myeloma cell lines with t(4;14): comparative analysis of Y373C, K650E and the novel G384D mutations. Oncogene 2001; 20: 3553–3562.

  21. 21

    Kimura T, Suzuki H, Ohashi T, Asano K, Kiyota H, Eto Y . The incidence of thanatophoric dysplasia mutations in FGFR3 gene is higher in low-grade or superficial bladder carcinomas. Cancer 2001; 92: 2555–2561.

  22. 22

    Tomlinson DC, Baldo O, Harnden P, Knowles MA . FGFR3 protein expression and its relationship to mutation status and prognostic variables in bladder cancer. J Pathol 2007; 213: 91–98.

  23. 23

    Al-Ahmadie HA, Iyer G, Janakiraman M, Lin O, Heguy A, Tickoo SK et al. Somatic mutation of fibroblast growth factor receptor-3 (FGFR3) defines a distinct morphological subtype of high-grade urothelial carcinoma. J Pathol 2011; 224: 270–279.

  24. 24

    Tomlinson DC, Hurst CD, Knowles MA . Knockdown by shRNA identifies S249C mutant FGFR3 as a potential therapeutic target in bladder cancer. Oncogene 2007; 26: 5889–5899.

  25. 25

    Turner N, Lambros MB, Horlings HM, Pearson A, Sharpe R, Natrajan R et al. Integrative molecular profiling of triple negative breast cancers identifies amplicon drivers and potential therapeutic targets. Oncogene 2010; 29: 2013–2023.

  26. 26

    Sun S, Jiang Y, Zhang G, Song H, Zhang X, Zhang Y et al. Increased expression of fibroblastic growth factor receptor 2 is correlated with poor prognosis in patients with breast cancer. J Surg Oncol 2011; 105: 773–779.

  27. 27

    Tannheimer SL, Rehemtulla A, Ethier SP . Characterization of fibroblast growth factor receptor 2 overexpression in the human breast cancer cell line SUM-52PE. Breast Cancer Res 2000; 2: 311–320.

  28. 28

    Hattori Y, Odagiri H, Nakatani H, Miyagawa K, Naito K, Sakamoto H et al. K-sam, an amplified gene in stomach cancer, is a member of the heparin-binding growth factor receptor genes. Proc Natl Acad Sci USA 1990; 87: 5983–5987.

  29. 29

    Kuniyasu H, Yasui W, Kitadai Y, Yokozaki H, Ito H, Tahara E . Frequent amplification of the c-met gene in scirrhous type stomach cancer. Biochem Biophys Res Commun 1992; 189: 227–232.

  30. 30

    Tsujimoto H, Sugihara H, Hagiwara A, Hattori T . Amplification of growth factor receptor genes and DNA ploidy pattern in the progression of gastric cancer. Virchows Arch 1997; 431: 383–389.

  31. 31

    Mor O, Ranzani GN, Ravia Y, Rotman G, Gutman M, Manor A et al. DNA amplification in human gastric carcinomas. Cancer Genet Cytogenet 1993; 65: 111–114.

  32. 32

    Kunii K, Davis L, Gorenstein J, Hatch H, Yashiro M, Di Bacco A et al. FGFR2-amplified gastric cancer cell lines require FGFR2 and Erbb3 signaling for growth and survival. Cancer Res 2008; 68: 2340–2348.

  33. 33

    Wheeler DL, Dunn EF, Harari PM . Understanding resistance to EGFR inhibitors-impact on future treatment strategies. Nat Rev Clin Oncol 2010; 7: 493–507.

  34. 34

    Milojkovic D, Apperley J . Mechanisms of resistance to imatinib and second- generation tyrosine inhibitors in chronic myeloid leukemia. Clin Cancer Res 2009; 15: 7519–7527.

  35. 35

    Hammerman PS, Janne PA, Johnson BE . Resistance to epidermal growth factor receptor tyrosine kinase inhibitors in non-small cell lung cancer. Clin Cancer Res 2009; 15: 7502–7509.

  36. 36

    Ley R, Balmanno K, Hadfield K, Weston C, Cook SJ . Activation of the ERK1/2 signaling pathway promotes phosphorylation and proteasome-dependent degradation of the BH3-only protein, Bim. J Biol Chem 2003; 278: 18811–18816.

  37. 37

    Ewings KE, Hadfield-Moorhouse K, Wiggins CM, Wickenden JA, Balmanno K, Gilley R et al. ERK1/2-dependent phosphorylation of BimEL promotes its rapid dissociation from Mcl-1 and Bcl-xL. EMBO J 2007; 26: 2856–2867.

  38. 38

    Azam M, Seeliger MA, Gray NS, Kuriyan J, Daley GQ . Activation of tyrosine kinases by mutation of the gatekeeper threonine. Nat Struct Mol Biol 2008; 15: 1109–1118.

  39. 39

    Daub H, Specht K, Ullrich A . Strategies to overcome resistance to targeted protein kinase inhibitors. Nat Rev Drug Discov 2004; 3: 1001–1010.

  40. 40

    Azam M, Daley GQ . Anticipating clinical resistance to target-directed agents: the BCR-ABL paradigm. Mol Diagn Ther 2006; 10: 67–76.

  41. 41

    Mohammadi M, Froum S, Hamby JM, Schroeder MC, Panek RL, Lu GH et al. Crystal structure of an angiogenesis inhibitor bound to the FGF receptor tyrosine kinase domain. EMBO J 1998; 17: 5896–5904.

  42. 42

    Qing J, Du X, Chen Y, Chan P, Li H, Wu P et al. Antibody-based targeting of FGFR3 in bladder carcinoma and t(4;14)-positive multiple myeloma in mice. J Clin Invest 2009; 119: 1216–1229.

  43. 43

    Squires M, Ward G, Saxty G, Berdini V, Cleasby A, King P et al. Potent, selective inhibitors of fibroblast growth factor receptor define fibroblast growth factor dependence in preclinical cancer models. Mol Cancer Ther 2011; 10: 1542–1552.

  44. 44

    Trudel S, Li ZH, Wei E, Wiesmann M, Chang H, Chen C et al. CHIR-258, a novel, multitargeted tyrosine kinase inhibitor for the potential treatment of t(4;14) multiple myeloma. Blood 2005; 105: 2941–2948.

  45. 45

    Lamont FR, Tomlinson DC, Cooper PA, Shnyder SD, Chester JD, Knowles MA . Small molecule FGF receptor inhibitors block FGFR-dependent urothelial carcinoma growth in vitro and in vivo. Br J Cancer 2011; 104: 75–82.

  46. 46

    Jebar AH, Hurst CD, Tomlinson DC, Johnston C, Taylor CF, MA Knowles . FGFR3 and Ras gene mutations are mutually exclusive genetic events in urothelial cell carcinoma. Oncogene 2005; 24: 5218–5225.

  47. 47

    Normanno N, Tejpar S, Morgillo F, De Luca A, Van Cutsem E, Ciardiello F . Implications for KRAS status and EGFR-targeted therapies in metastatic CRC. Nat Rev Clin Oncol 2009; 6: 519–527.

  48. 48

    Zhou W, Hur W, McDermott U, Dutt A, Xian W, Ficarro SB et al. A structure-guided approach to creating covalent FGFR inhibitors. Chem Biol 2010; 17: 285–295.

  49. 49

    Huynh H, Ngo VC, Fargnoli J, Ayers M, Soo KC, Koong HN et al. Brivanib alaninate, a dual inhibitor of vascular endothelial growth factor receptor and fibroblast growth factor receptor tyrosine kinases, induces growth inhibition in mouse models of human hepatocellular carcinoma. Clin Cancer Res 2008; 14: 6146–6153.

  50. 50

    Hilberg F, Roth GJ, Krssak M, Kautschitsch S, Sommergruber W, Tontsch-Grunt U et al. BIBF 1120: triple angiokinase inhibitor with sustained receptor blockade and good antitumor efficacy. Cancer Res 2008; 68: 4774–4782.

  51. 51

    Radomska HS, Bassères DS, Zheng R, Zhang P, Dayaram T, Yamamoto Y et al. Block of C/EBP alpha function by phosphorylation in acute myeloid leukemia with FLT3 activating mutations. J Exp Med 2006; 203: 371–381.

  52. 52

    Pabst T, Mueller BU, Harakawa N, Schoch C, Haferlach T, Behre G et al. AML1-ETO downregulates the granulocytic differentiation factor C/EBPalpha in t(8;21) myeloid leukemia. Nat Med 2001; 7: 444–451.

  53. 53

    Kawano M, Hirano T, Matsuda T, Taga T, Horii Y, Iwato K et al. Autocrine generation and requirement of BSF-2/IL-6 for human multiple myelomas. Nature 1988; 332: 83–85.

  54. 54

    Binato R, Mencalha A, Pizzatti L, Scholl V, Zalcberg I, Abdelhay E . RUNX1T1 is overexpressed in imatinib mesylate-resistant cells. Mol Med Rep 2009; 2: 657–661.

  55. 55

    Todd DE, Densham RM, Molton SA, Balmanno K, Newson C, Weston CR et al. ERK1/2 and p38 cooperate to induce a p21CIP1-dependent G1 cell cycle arrest. Oncogene 2004; 23: 3284–3295.

  56. 56

    Davies BR, Logie A, McKay JS, Martin P, Steele S, Jenkins R et al. AZD6244 (ARRY-142886), a potent inhibitor of mitogen-activated protein kinase/extracellular signal-regulated kinase kinase 1/2 kinases: mechanism of action in vivo, pharmacokinetic/pharmacodynamic relationship, and potential for combination in preclinical models. Mol Cancer Ther 2007; 6: 2209–2219.

  57. 57

    Waterhouse AM, Procter JB, Martin DMA, Clamp M, Barton GJ . Jalview Version 2 a multiple sequence alignment editor and analysis workbench. Bioinformatics 2009; 25: 1189–1191.

  58. 58

    Emsley P, Lohkamp B, Scott WG, Cowtan K . Features and development of Coot. Acta Crystallogr D Biol Crystallogr 2010; 66: 486–501.

  59. 59

    Benjamini Y, Hochberg Y . Controlling the false discovery rate: a practical and powerful approach to multiple testing. J Roy Statist Soc Ser B (Methodological) 1995; 57: 289–300.

  60. 60

    Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, De Paepe A et al. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol 2002; 3: : RESEARCH0034.

Download references

Acknowledgements

We would like to thank Anne Segonds-Pichon (Babraham Bioinformatics Group) for statistical analysis of qRT–PCR data and Jonathan Keats, Janet Nutt, Margaret Knowles, Matthias Ebert and Andrew Garner for provision of some of the cell lines used in this study. We would especially like to thank Andrew Garner for initiating this project, and for discussions and encouragement in its early stages. We are grateful to Teresa Klinowska, Nigel Brooks, Elaine Kilgour and Paul Smith for many useful discussions and suggestions throughout. This work was supported by a CASE PhD studentship (VC, née Victoria Knights) funded by the Biotechnology and Biological Sciences Research Council (BBSRC) and AstraZeneca, and a sponsored research agreement between AstraZeneca and the Babraham Institute.

Author information

Affiliations

Authors

Corresponding author

Correspondence to S J Cook.

Ethics declarations

Competing interests

Laura Blockley, Mark Hampson and Paul Gavine are paid employees of AstraZeneca. All other authors declare no conflict of interest.

Additional information

Supplementary Information accompanies the paper on the Oncogene website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Chell, V., Balmanno, K., Little, A. et al. Tumour cell responses to new fibroblast growth factor receptor tyrosine kinase inhibitors and identification of a gatekeeper mutation in FGFR3 as a mechanism of acquired resistance. Oncogene 32, 3059–3070 (2013). https://doi.org/10.1038/onc.2012.319

Download citation

Keywords

  • acquired resistance
  • ERK1/2
  • FGFR
  • PKB
  • tyrosine kinase inhibitors

Further reading