Immunoassays for the quantification of ALK and
`phosphorylated ALK support the evaluation of on-target
`ALK inhibitors in neuroblastoma
`Elizabeth R. Tucker1†, Jennifer R. Tall1,2†, Laura S. Danielson1, Sharon Gowan3, Yann Jamin4,
`Simon P. Robinson4, Udai Banerji2 and Louis Chesler1
`
`1 Paediatric Solid Tumour Biology and Therapeutics Team, Division of Clinical Studies and Cancer Therapeutics Division, The Institute of
`Cancer Research, Sutton, Surrey, UK
`2 Clinical Pharmacodynamic Biomarker Team, Cancer Therapeutics Division, The Institute of Cancer Research, Sutton, Surrey, UK
`3 Tumour Biology and Metastasis, Cancer Therapeutics Division, The Institute of Cancer Research, Sutton, Surrey, UK
`4 Division of Radiotherapy and Imaging, The Institute of Cancer Research, Sutton, Surrey, UK
`
`Targeted inhibition of anaplastic lymphoma kinase (ALK) is a successful
`approach for the treatment of many ALK-aberrant malignancies; however,
`the presence of resistant mutations necessitates both the development of
`more potent compounds and pharmacodynamic methods with which to
`determine their efficacy. We describe immunoassays designed to quantitate
`phosphorylation of ALK, and their use in preclinical models of neuroblas-
`toma, a pediatric malignancy in which gain-of-function ALK mutations
`predict a poor overall outcome to conventional treatment. Validation of
`the immunoassays is presented using a panel of neuroblastoma cell lines
`and evidence of on-target ALK inhibition provided by treatment of a
`genetically engineered murine model of neuroblastoma with two clinical
`ALK inhibitors, crizotinib and ceritinib, highlighting the superior efficacy
`of ceritinib.
`
`Keywords
`ALK; neuroblastoma
`
`Correspondence
`L. Chesler, Paediatric Solid Tumour Biology
`and Therapeutics Team, Division of Clinical
`Studies and Cancer Therapeutics Division,
`The Institute of Cancer Research, 15
`Cotswold Road, Sutton, Surrey SM2 5NG,
`UK
`Tel: +44 0208 722 4176
`E-mail: louis.chesler@icr.ac.uk
`
`†Joint first authors
`
`(Received 16 December 2016, revised 24
`March 2017, accepted 11 April 2017,
`available online 31 May 2017)
`
`doi:10.1002/1878-0261.12069
`
`1. Introduction
`
`The rapid development of targeted therapeutics against
`anaplastic
`lymphoma kinase
`(ALK) has already
`resulted in significant changes to the up-front treat-
`ment of patients with ALK-rearranged non-small-
`cell lung cancer (NSCLC) (Camidge et al., 2012), and
`a current aim is to bring these personalized therapies
`to the benefit of childhood cancer patients, including
`those with neuroblastomas that harbor point muta-
`tions of ALK (Chen et al., 2008; George et al., 2008;
`Janoueix-Lerosey et al., 2008; Mosse et al., 2008). The
`
`broad range of ALK mutations found either as a sec-
`ondary resistance mechanism to ALK inhibition in
`NSCLC or as a primary resistance mechanism in neu-
`roblastoma necessitates thorough interrogation of the
`ability of individual compounds to inhibit ALK phos-
`phorylation and subsequent survival signaling down-
`stream of ALK, enabling the rapid translation of the
`most promising compounds from the laboratory to
`clinical trials (Bresler et al., 2011, 2014). To achieve
`this, highly relevant preclinical models should be cou-
`pled with accurate methods to report pharmacody-
`namic responses.
`
`Abbreviations
`ALK, anaplastic lymphoma kinase; MSD
`
`â
`, Meso Scale Discovery.
`
`996
`
`Molecular Oncology 11 (2017) 996–1006 ª 2017 The Authors. Published by FEBS Press and John Wiley & Sons Ltd.
`This is an open access article under the terms of the Creative Commons Attribution License, which permits use,
`
`distribution and reproduction in any medium, provided the original work is properly cited.
`
`Lassen - Exhibit 1061, p. 1
`
`

`

`E. R. Tucker et al.
`
`Immunoassays to quantitate ALK
`
`We have therefore developed sensitive immunoassays
`to detect and quantitate the expression of total ALK
`protein and phosphorylated forms of ALK using the
`Ò
`) platform. This format
`mesoscale technology (MSD
`allows for the development of assays that are appropri-
`Ò
`ate for both in vitro and in vivo studies, and MSD
`assays are already routinely incorporated into clinical
`studies in order to measure pharmacodynamic end
`points (Basu et al., 2015). We sought to validate our
`ALK immunoassays using both neuroblastoma cell
`lines and tumor tissue from the Th-ALKF1174L/MYCN
`transgenic model, which has previously demonstrated
`the inadequacy of the first-generation ALK inhibitor,
`crizotinib,
`to elicit
`therapeutic responses
`in ALK
`F1174L-driven neuroblastomas (Berry et al., 2012).
`Our results demonstrate the successful application of
`Ò
`immunoassays to measure ALK and phospho-
`MSD
`rylated ALK as pharmacodynamic biomarkers follow-
`ing treatment with small-molecule ALK inhibitors in
`both in vitro and ex vivo tissues. We show that in
`addition to autophosphorylation of ALK at Y1278
`and Y1604, following the fate of phosphorylated ALK
`at Y1586 also provides a marker of active ALK levels.
`
`2. Materials and methods
`
`2.1. Cell lines
`
`Neuroblastoma cell lines and HeLa cells were obtained
`from the American Type Culture Collection, CLB-GA
`was a gift from V. Combaret (Lyon), and these were
`shown to be mycoplasma-free using a PCR-based assay
`(Minerva Biolabs, Berlin, Germany). The Ba/F3 ALK
`F1174L cells were a gift from R. George (Boston, USA)
`and were transduced as described previously (George
`et al., 2008). Cells were cultured in RPMI 1640 media
`supplemented with 2 mM glutamine (Gibco, Thermo
`Fisher Scientific, Waltham, MA, USA), 1 9 MEM
`nonessential amino acids, and 10% FBS (Gibco) and
`grown at 37 °C with 5% CO2 in a humidified incubator.
`
`2.2. Drug treatments
`Cell lines were treated at 50–70% confluency at the
`indicated drug concentrations or with DMSO at a con-
`centration that matched the greatest DMSO for the drug-
`treated cells. Crizotinib and ceritinib were purchased
`from Shanghai Haoyuan Chemexpress Company.
`
`2.3. GI50 determination
`
`In order to calculate the half maximal growth inhibi-
`tory concentration (GI50) of individual compounds,
`
`neuroblastoma tumor cells were seeded into 96-well
`plates in a total volume of 100 lL and allowed to
`attach overnight. Compound (dissolved in DMSO)
`was added to wells in six replicates of 100 lL, across a
`concentration gradient including a DMSO-only con-
`trol, the next day. The cells were exposed to drug for
`72 h. Thereafter, the cell number in treated versus con-
`trol wells was estimated after cell fixation with 10%
`trichloroacetic acid and staining with sulforhodamine
`B in 1% acetic acid. The GI50 was calculated as the
`drug concentration that inhibits cell growth by 50%
`compared with control growth, according to nonlinear
`regression analysis, using GRAPHPAD PRISM (La Jolla,
`CA, USA).
`
`2.4. Preparation of protein lysates
`
`Cell lines were harvesting by scraping, spun at 500 g
`for 5 min, and washed once in phosphate-buffered sal-
`ine, and the cell pellets were resuspended in CHAPS
`lysis buffer [50 mM Tris/HCl pH 8.0, 1 mM EDTA,
`150 mM NaCl, 1% CHAPS, 0.2 mM PMSF, 1 : 50
`Phosphatase Inhibitor Cocktail 2 and 3 (Sigma-Aldrich,
`St. Louis, MO, USA), 1 : 100 Protease Inhibitor Cock-
`tail
`(Sigma-Aldrich)]. Frozen tissue
`samples were
`homogenized in CHAPS lysis buffer prepared as for cell
`lysates. After incubation for 30 min on ice, lysates were
`spun at 16 000 g for 15 min and the supernatant was
`collected. Protein concentrations were determined using
`BCA protein assay (Thermo Fisher Scientific) by com-
`parison with bovine serine albumin standard.
`
`2.5. ALK Meso Scale Discovery
`
`â
`
`immunoassays
`
`Multiarray 96-well plates (Meso Scale Discovery) were
`coated overnight at 4 °C with 0.5 lg mL
`1 mouse
`total ALK antibody (Clone 31F12; Cell Signaling
`Technology Inc., Danvers, MA, USA) diluted in
`50 mM carbonate buffer. Plates were washed 5 9 in
`wash buffer (0.1% Tween 20 in Tris-buffered saline)
`and incubated for 1 h with blocking buffer (5% BSA
`in wash buffer). After washing, samples were added
`lysates being diluted to 20 lg per well
`with cell
`(in vitro) or 30 lg per well (in vivo) in 1 9 Tris Lysis
`Buffer (Meso Scale Discovery) or recombinant ALK
`protein (Thermo Fisher Scientific) diluted in phos-
`phate-buffered saline and incubated overnight at 4 °C.
`After washing, the plates were then incubated for 1 h
`with the appropriate antibody (0.4 lg mL
`1
`total
`ALK D5F3, 0.2 lg mL
`1 pY1278 ALK D59G10,
`1 pY1586 ALK 3B4, 0.2 lg mL
`0.4 lg mL
`1 pY1604
`ALK D96H9, or 0.2 lg mL
`1 phospho-tyrosine P-
`Tyr-1000,
`all
`purchased
`from Cell
`Signaling
`
`Molecular Oncology 11 (2017) 996–1006 ª 2017 The Authors. Published by FEBS Press and John Wiley & Sons Ltd.
`
`997
`
`Lassen - Exhibit 1061, p. 2
`
`

`

`Immunoassays to quantitate ALK
`
`E. R. Tucker et al.
`
`Technology Inc.), washed, and incubated for a further
`1 h with 0.5 lg mL
`1 anti-rabbit SULFO-tag anti-
`body. The plates were washed, 2 9 Read Buffer T
`(Meso Scale Discovery) was added to wells, and elec-
`trochemiluminescence counts were made using a MSD
`SECTOR Imager 6000.
`
`2.6. Immunoblotting
`Western blotting of 20 lg of denatured lysates was
`carried out using precast 4–12% Bis/Tris gels in 1 9
`MOPS running buffer [Thermo Fisher Scientific (Invi-
`trogen)]. A prestained molecular weight marker was
`loaded alongside the experimental samples (Invitro-
`gen). The gels were transferred to PVDF membranes
`at 30 V for 3 h, or at 0.05 mA overnight. The mem-
`branes were blocked for 1 h in 5% (w/v) nonfat dry
`milk in Tris-buffered saline with 0.1% Tween 20
`(TBST), and then incubated with primary antibody
`overnight at 4 °C in 2.5% milk TBST (ALK, pY1278
`ALK, pY1604 ALK, pY1586 ALK (all equivalent to
`Ò
`immunoassay), ERK1/2,
`the
`respective MSD
`pERK1/2, Akt, pAkt, GAPDH, all from Cell Signal-
`ing Technology Inc.). Membranes were incubated with
`the appropriate horseradish peroxidase-linked sec-
`ondary antibody and proteins were visualized using
`electrochemical luminescence (ECL plus) [GE Health-
`care Life Sciences (Amersham Biosciences), Buckin-
`ghamshire, UK] detection system on a LAS-3000
`Imaging System [GE Healthcare Life Sciences (Fuji-
`film), Buckinghamshire, UK]. Quantification of blots
`was performed by densitometry using IMAGEJ software
`analysis (Schneider et al., 2012).
`
`2.7. Immunoprecipitation
`
`Cell lysates were subjected to preclearing before direct
`incubation with the antibody of interest (ALK; IgG)
`for an hour at 4 °C. Protein G beads were added to
`the lysate and incubated overnight at 4 °C. The next
`day, the samples were spun down for 30 s at 4 °C and
`the supernatant was discarded. The bead pellet was
`washed five times with lysis buffer, before denaturing
`by heating to 95 °C for 5 min in SDS sample buffer.
`
`2.8. Mouse models
`
`All experiments, including the breeding of transgenic
`animals, were performed in accordance with the local
`ethical review panel, the UK Home Office Animals
`(Scientific procedures) Act 1986, the ARRIVE (Animal
`Research: Reporting of In Vivo Experiments) guideli-
`(Kilkenny et al., 2010) and the UK NCRI
`nes
`
`(Workman et al., 2010). Th-ALKF1174L/
`guideline
`MYCN tumor-bearing animals were enrolled into ther-
`apeutic trials when their abdominal tumors reached
`5 mm in diameter according to palpation. Volumetric
`MRI was performed as previously described (Jamin
`et al., 2013), with each animal undergoing imaging on
`day 0 and day 7. The tumor volume at each time point
`was then calculated. For in vivo oral dosing on days
`1–7, crizotinib was dissolved in sterile water with 10%
`Tween 20. Ceritinib was dissolved in 0.5% methylcel-
`lulose, 0.5% Tween 80 with sterile water. Two hours
`following the final dose of either compound, tumor tis-
`sue was excised and snap-frozen prior to analysis.
`
`3. Results
`
`3.1. Detection of recombinant ALK (rALK) protein
`with immunoassays
`Ò
`
`platform, we optimized immunoas-
`Using the MSD
`says to detect phosphorylated or total ALK protein.
`Confirmation of the ability of the assays to detect
`ALK or phosphorylated ALK species was sought
`through the use of a kinase active recombinant ALK
`(rALK) protein (Fig. 1). Using a titrating amount of
`rALK, we found that pan-pY ALK, pY1278 ALK,
`pY1586 ALK, pY1604 ALK, and total ALK were
`detected in a linear fashion in each assay (Fig. 1A).
`The reproducibility of the assays to detect rALK was
`assessed both within a single experiment (Fig. 1B) and
`across
`at
`least
`three
`independent
`experiments
`(Fig. 1C). In all cases, the intra-assay percentage coef-
`ficient of variation for pY1278, pY1604, and total
`ALK was less than 2.4%, 6.2%, and 6.2%, respec-
`tively. In four of five samples for the pY assay and the
`pY1586 assay, the coefficient of variation was less than
`2.5% and 5.4%, respectively. The interassay variability
`was less than 30% in all assays, indicating high repro-
`ducibility of the assay signals.
`
`3.2. Immunoassays quantitate ALK and
`phosphorylated ALK forms in neuroblastoma cell
`lines
`
`Gain-of-function mutations of ALK, which lead to
`constitutive ALK phosphorylation, are known to con-
`tribute to the aggressive nature of pediatric neuroblas-
`toma tumors, and therefore, the detection of ALK
`activity in available neuroblastoma cell
`lines is of
`increasing importance. We employed a panel of these
`cell lines harboring either one of the two most frequent
`ALK mutations (F1174L or R1275Q), or wild-type
`ALK, and assessed the total and phosphorylated
`
`998
`
`Molecular Oncology 11 (2017) 996–1006 ª 2017 The Authors. Published by FEBS Press and John Wiley & Sons Ltd.
`
`Lassen - Exhibit 1061, p. 3
`
`

`

`E. R. Tucker et al.
`
`Immunoassays to quantitate ALK
`
`pALK
`
`pY1278 ALK
`pY1586 ALK
`pY1604 ALK
`
`R 2 = 0.9187
`R 2 = 0.9296
`
`R 2 = 0.9581
`
`2.5
`
`2.0
`
`1.5
`
`1.0
`
`0.5
`
`0.0
`
`MSD ECL counts (×105)
`
`pY ALK
`
`R 2 = 0.9634
`
`0
`
`5
`
`20
`15
`10
`rALK (pg·µL–1)
`
`25
`
`0
`
`5
`
`20
`15
`10
`rALK (pg·µL–1)
`
`25
`
`7.0
`
`6.0
`
`5.0
`
`4.0
`
`3.0
`
`2.0
`
`1.0
`
`0.0
`
`MSD ECL counts (×105)
`
`Total ALK
`
`R 2 = 0.9562
`
`0
`
`50
`
`150
`100
`rALK (pg·µL–1)
`
`200
`
`Interassay variation
`
`pY
`
`pY1278
`
`pY1586
`
`pY1604
`
`Total ALK
`
`30
`
`25
`
`20
`
`15
`
`10
`
`C
`
`% CV
`
`05
`
`[rALK]
`
`Intra-assay variation
`
`pY
`
`pY1278
`
`pY1586
`
`pY1604
`
`Total ALK
`
`10.0
`9.0
`8.0
`7.0
`6.0
`5.0
`4.0
`3.0
`2.0
`1.0
`0.0
`
`A
`
`MSD ECL counts (×105)
`
`30
`
`25
`
`20
`
`15
`
`10
`
`B
`
`% CV
`
`05
`
`[rALK]
`
`Fig. 1. Immunoassays to quantitate phosphorylated and total ALK protein. (A) Titration of recombinant ALK protein (rALK) in MSD
`assays
`to quantify pan-phospho-tyrosine ALK, phospho-Y1278 ALK, phospho-Y1586 ALK, phospho-Y1604 ALK, and total ALK protein levels.
`Mean  SD from ≥ 3 independent
`repeats. R-squared values for
`linear
`regression from each of
`the assays are indicated. Assay
`â
`assay with increasing
`intra-assay variability calculated across triplicate wells of each ALK MSD
`reproducibility was assessed by (B)
`concentrations of recombinant ALK protein (same as A) and (C) interassay variability calculated across from ≥ 3 independent repeats using
`increasing concentrations of recombinant ALK protein (same as A) and is presented as percentage coefficient of variation (CV).
`
`â
`
`abundance of ALK by immunoblotting, in addition to
`pathways downstream of ALK (Fig. 2A). When com-
`pared to the Ba/F3-transduced cell
`lysates,
`it was
`apparent that detection and therefore quantification of
`phosphorylated ALK is challenging in the neuroblas-
`toma cell panel by western blotting as bands were only
`clearly visible for pY1278 and pY1586 ALK in CLB-
`GA and LAN-5 lysates. Phosphorylated Akt and
`phosphorylated ERK 1/2 signals were detected more
`strongly in CLB-GA and LAN-5 again, as well as in
`three of the four ALK F1174L cell
`lines, excluding
`Kelly cells. However, analyzing the same cell lysates in
`the ALK immunoassays we developed showed detect-
`able levels of pY1278, pY1586, pY1604, and total
`ALK in all the neuroblastoma cell lines tested, whereas
`no signals were obtained for HeLa cell lysate in any of
`the ALK assays consistent with the lack of ALK
`expression (Fig. 2B). When the levels of phosphory-
`lated ALK in the neuroblastoma cell lines were nor-
`malized to total ALK signals from the immunoassays
`and compared with densitometry of the immunoblots,
`there was a positive correlation that was statistically
`significant for the total ALK and pY1278 and pY1586
`
`ALK assays, and although not significant, a positive
`trend between densitometry and immunoassay for the
`detection of pY1604 ALK was seen (Fig. 2C). Further-
`more, upon immunoprecipitation of ALK and detec-
`tion of pan-pY from the same lysates (Fig. 2D), there
`was a stronger pY signal obtained for the two cell
`lines harboring the R1275Q ALK mutation (CLB-GA
`and LAN-5). This finding is consistent with CLB-GA
`and LAN-5 cell lysates giving the highest signals out
`of the cell
`line panel
`in the pY1278, pY1586, and
`pY1604 ALK assays (Fig. 2B) and the pY ALK
`Ò
`assay (Fig. 2E). Together, these data indicate
`MSD
`that the ALK immunoassays can be used to character-
`ize basal levels of ALK activity in neuroblastoma cells
`in a more quantitative manner and with greater sensi-
`tivity than immunoblotting.
`
`3.3. ALK immunoassays quantify the dose- and
`time-dependent response of neuroblastoma cell
`lines to therapeutic ALK inhibition
`
`the phospho-ALK
`To further assess the ability of
`MSD assays to measure changes in ALK activity, we
`
`Molecular Oncology 11 (2017) 996–1006 ª 2017 The Authors. Published by FEBS Press and John Wiley & Sons Ltd.
`
`999
`
`Lassen - Exhibit 1061, p. 4
`
`

`

`Immunoassays to quantitate ALK
`
`E. R. Tucker et al.
`
`made use of two clinical small-molecule ALK inhibi-
`tors, crizotinib and ceritinib. Crizotinib was less suc-
`cessful against ALK F1174L neuroblastomas in clinical
`trial, but showed efficacy in the treatment for ALK
`
`R1275Q neuroblastomas (Mosse et al., 2013). Ceritinib
`is a second-generation ALK inhibitor, currently being
`evaluated in pediatric clinical studies. We used two
`neuroblastoma cell
`lines to compare crizotinib and
`
`6
`
`4
`
`2
`
`r = 0.9936
`P < 0.0001***
`
`0
`0.00
`
`012345
`
`0.15
`0.10
`0.05
`Immunoblot quantitation
`(pY1278 ALK/tALK)/GAPDH
`
`0.20
`
`r = 0.7254
`P = ns
`
`0.00
`
`0.06
`0.04
`0.02
`Immunoblot quantitation
`(pY1604 ALK/tALK)/GAPDH
`
`0.08
`
`Hela
`
`IMR-32
`
`LAN-5
`
`MSD ECL counts (×103)
`
`MSD ECL counts (×103)
`
`r = 0.9202
`P = 0.0033**
`
`0.0
`
`1.0
`0.5
`Immunoblot quantitation
`tALK/GAPDH
`
`1.5
`
`r = 0.9465
`P = 0.0012**
`
`0.0
`
`0.4
`0.3
`0.2
`0.1
`Immunoblot quantitation
`(pY1586 ALK/tALK)/GAPDH
`
`0.5
`
`8
`
`6
`
`4
`
`2
`
`0
`
`10
`
`02468
`
`D
`
`C
`
`MSD ECL counts (×105)
`
`MSD ECL counts (×103)
`
`Hela
`
`IMR-32
`
`LAN-5
`
`CLB-GA
`
`Kelly
`
`SK-N-SH
`
`LAN-1
`
`SH-SY5Y
`
`Ba/F3 ALKF1174L
`
`A
`
` pY1278 ALK
`
`pY1586 ALK
`
`pY1604 ALK
`
`Total ALK
`
`pS473 Akt
`
`Total Akt
`
`pERK 1/2
`
`ERK 1/2
`
`GAPDH
`
`B
`
`12.0
`
`10.0
`
`pY1278 ALK
`pY1586 ALK
`pY1604 ALK
`
`CLB-GA
`
`Kelly
`
`SK-N-SH
`
`LAN-1
`
`SH-SY5Y
`
`Ba/F3 ALKF1174L
`
`pY ALK
`
`HeLa
`
`IMR-32
`
`LAN-5
`
`CLB-GA
`
`Kelly
`
`SK-N-SH
`
`LAN-1
`
`SH-SY5Y
`
`Total ALK
`
`IP ALK
`
`IP pY
`
`E
`
`4.0
`
`3.0
`
`2.0
`
`1.0
`
`0.0
`
`MSD ECL counts (x104)
`
`Hela
`
`IMR-32
`
`LAN-5
`
`CLB-GA
`
`Kelly
`
`SK-N-SH
`
`LAN-1
`
`SH-SY5Y
`
`Total ALK
`
`HeLa
`
`IMR-32
`
`LAN-5
`
`CLB-GA
`
`Kelly
`
`SK-N-SH
`
`LAN-1
`
`SH-SY5Y
`
`8.0
`
`6.0
`
`4.0
`
`2.0
`
`0.0
`
`MSD ECL counts (x103)
`
`8.0
`7.0
`6.0
`5.0
`4.0
`3.0
`2.0
`1.0
`0.0
`
`MSD ECL counts (x105)
`
`ALK status:
`
`F1174L
`
`R1275Q
`
`WT
`
`-
`
`ALK status:
`
`F1174L
`
`R1275Q
`
`WT
`
`-
`
`1000
`
`Molecular Oncology 11 (2017) 996–1006 ª 2017 The Authors. Published by FEBS Press and John Wiley & Sons Ltd.
`
`Lassen - Exhibit 1061, p. 5
`
`

`

`E. R. Tucker et al.
`
`Immunoassays to quantitate ALK
`
`in a dose-dependent and time-
`ceritinib treatment
`dependent manner using the total ALK, pY ALK,
`pY1278, pY1604, and pY1586 ALK immunoassays
`(Fig. 3). The drug concentrations were decided using
`the GI50 values for each compound (Table 1). Kelly
`cells
`(ALK F1174L) demonstrated a similar dose
`response to both drugs (Fig. 3A) with inhibition of
`ALK phosphorylation at sub-GI50 concentrations, and
`this similarity was reflected in the accompanying immu-
`noblot
`(pY1604 and pY1278 immunoblots did not
`detect bands and are therefore not included). CLB-GA
`cells (ALK R1275Q) (Fig. 3B) were more sensitive to
`treatment with ceritinib than with crizotinib, with over
`50% inhibition of phosphorylation seen with all phos-
`pho-ALK assays at 10 nM ceritinib. This result is also
`seen in the accompanying immunoblot (pY1604 immu-
`noblot did not detect bands and is therefore not
`included) (Fig. 3B). In both cell lines, loss in pY1586
`ALK signal follows the pattern of decreasing pY ALK,
`pY1604 ALK, and pY1278 ALK (CLB-GA only) with
`ALK inhibition and a reduction in band intensity in
`the accompanying immunoblots for pY1278 ALK.
`Although ALK phosphorylation was inhibited at sub-
`GI50 concentrations of crizotinib or ceritinib, signaling
`pathways downstream of ALK, such as AKT and
`ERK1/2, were not inhibited until higher concentrations
`at 3-hour treatment (immunoblots in Fig. 3A,B). The
`results of the dose–response experiments led to the
`choice of drug concentrations to be used in the subse-
`quent time course experiments, as the dose that led to a
`percentage of DMSO treated at near 20% phosphory-
`lated ALK.
`There was a rapid decrease in the phosphorylated/to-
`tal ALK ratio (P/T ratio)
`in Kelly cells (Fig. 3C)
`(graphed as percentage of zero-hour control sample) fol-
`lowing the treatment of the cells with either crizotinib or
`ceritinib. This effect was sustained for 24 h and is
`reflected in the accompanying immunoblots of the same
`lysates, where sustained inhibition of phosphorylated
`ERK1/2, a downstream effector of ALK, is also appar-
`ent
`following the treatment with either compound.
`Subtle inhibition of phosphorylation of Akt at S473 was
`also evident following the treatment of Kelly cells with
`
`crizotinib or ceritinib. Treatment of CLB-GA cells
`(Fig. 3D) with either crizotinib or ceritinib also resulted
`in a rapid reduction in P/T ALK ratio from 10-min
`treatment, an effect that was also sustained for the entire
`24-h time course. At 50 nM, there was some evidence
`that the downstream effectors of ALK were differen-
`tially affected following the treatment with either com-
`pound. Dephosphorylation of Akt at S473 occurred
`more rapidly after treatment of the cells with crizotinib;
`however, although treatment with either compound
`gave rise to a loss of ERK1/2 phosphorylation quickly,
`this effect was transient and not sustained for the dura-
`tion of the experiment. Together, our results indicate
`that a sustained dephosphorylation of ALK at multiple
`phosphorylation sites by both crizotinib and ceritinib
`can be detected quantitatively by immunoassay in neu-
`roblastoma cell lines with ALK mutations at F1174L or
`R1275Q.
`
`3.4. Treatment of Th-ALKF1174L/MYCN tumor-
`bearing animals demonstrates increased efficacy
`of ceritinib over crizotinib and a greater
`pharmacodynamic response
`
`Finally, we tested frozen tissue specimens from Th-
`ALKF1174L/MYCN animals
`(in which spontaneous
`tumors arise, driven by coexpression of the human
`ALKF1174L and human MYCN transgenes), which had
`been treated with either crizotinib or ceritinib for
`7 days. Treatment of animals with ceritinib (100 mg kg
`1
`per day) results in delayed growth or tumor regression
`[established using noninvasive MRI volumetric measure-
`ments (Fig. 4B depicts representative images)], compared
`to delayed growth only following the treatment with
`crizotinib (100 mg kg
`1 per day) (Fig. 4A). The quanti-
`tation of pY1586 ALK and total ALK by immunoassay
`is
`shown next
`to the corresponding immunoblots
`(Fig. 4C,D). The ratio of pY1586 to total ALK revealed
`a mean decrease to 0.34 following crizotinib treatment
`versus 0.17 following ceritinib treatment (each normal-
`ized to the mean of their respective vehicle controls)
`(Fig. 4E), indicating a larger pharmacodynamic effect
`with ceritinib, consistent with the tumor volume data
`
`Fig. 2. Basal ALK activity in neuroblastoma cell lines. (A) Immunoblots of lysates from neuroblastoma cell line panel, including lysate from
`Ba/F3 ALK F1174L as a positive control for ALK expression and lysate from Hela cells as negative control for ALK expression. Arrows
`indicate ALK band of interest at 220 kDa. (B) Quantitation of phospho-Y1278 ALK, phospho-Y1586 ALK, phospho-Y1604 ALK, and total ALK
`protein levels in a cell line panel using immunoassays. Mean  SD from ≥ 3 biological repeats. (C) Correlation of MSD
`â
`ECL counts with
`quantitation of total ALK (tALK) (P = 0.0033**) and phospho-ALK (pY1278 ALK P = < 0.0001***; pY1586 ALK P = 0.0012**; pY1604 ALK
`P = ns (not significant)) immunoblots for neuroblastoma cell lysates. r values for Pearson’s rank correlation of each graph are indicated. (D)
`Immunoprecipitation of total ALK from the same neuroblastoma cell line panel lysates (as in (A)), to allow the assessment of pan-phospho-
`line panel by immunoassay. Mean  SD from ≥ 3 biological
`tyrosine ALK status. (E) Quantitation of pan-phospho-tyrosine ALK, in a cell
`repeats.
`
`Molecular Oncology 11 (2017) 996–1006 ª 2017 The Authors. Published by FEBS Press and John Wiley & Sons Ltd.
`
`1001
`
`Lassen - Exhibit 1061, p. 6
`
`

`

`E. R. Tucker et al.
`
`Total ALK
`pY ALK
`pY1278 ALK
`pY1586 ALK
`pY1604 ALK
`
`200
`180
`160
`140
`120
`100
`80
`60
`40
`20
`
`Ceritinib (nM)
`
`160
`140
`120
`100
`80
`60
`40
`20
`0
`–20
`
`Percentage of DMSO treated
`
`200
`180
`160
`140
`120
`100
`80
`60
`40
`20
`
`Crizotinib (nM)
`
`CLB-GA
`
`Crizotinib
`
`Ceritinib
`
`nM
`
`0 5 20 50 200 0 5 20 50 200
`
`Immunoassays to quantitate ALK
`
`B
`
`160
`140
`120
`100
`80
`60
`40
`20
`0
`–20
`
`Percentage of DMSO treated
`
`Total ALK
`pY ALK
`pY1586 ALK
`pY1604 ALK
`
`600
`550
`500
`450
`400
`350
`300
`250
`200
`150
`100
`50
`
`Ceritinib (nM)
`
`160
`140
`120
`100
`80
`60
`40
`20
`0
`–20
`
`Percentage of DMSO treated
`
`Kelly
`
`600
`550
`500
`450
`400
`350
`300
`250
`200
`150
`100
`50
`
`Crizotinib (nM)
`
`A
`
`160
`140
`120
`100
`80
`60
`40
`20
`0
`–20
`
`Percentage of DMSO treated
`
`Crizotinib
`
`Ceritinib
`
`nM
`
`0 20 50 200 500 0 20 50 200 500
`
`pY1278 ALK
`
`pY1586 ALK
`
`Total ALK
`
`pS473 Akt
`
`Total Akt
`
`pERK 1/2
`
`ERK 1/2
`
`GAPDH
`
`pY1278 ALK
`pY1586 ALK
`pY1604 ALK
`pY ALK
`
`24 h
`6 h
`3 h
`1 h
`10 min
`0 min
`24 h
`6 h
`3 h
`1 h
`10 min
`0 min
`24 h
`6 h
`3 h
`1 h
`10 min
`0 min
`24 h
`6 h
`3 h
`1 h
`10 min
`0 min
`24 h
`6 h
`3 h
`1 h
`10 min
`0 min
`24 h
`6 h
`3 h
`1 h
`10 min
`0 min
`24 h
`6 h
`3 h
`1 h
`10 min
`0 min
`24 h
`6 h
`3 h
`1 h
`10 min
`0 min
`
`D
`
`100
`
`80
`
`60
`
`40
`
`20
`
`0
`
`P/T ratio (% 0 h control)
`
`Crizotinib
`
`CLB-GA
`
`Ceritinib
`
`50 nM crizotinib
`0 10 min 1 h 3 h 6 h 24 h
`
`50 nM ceritinib
`
`0 10 min 1 h 3 h 6 h 24 h
`
`Time
`
`pY1278 ALK
`
`pY1586 ALK
`
`Total ALK
`
`pS473 Akt
`
`Total Akt
`
`pERK 1/2
`
`ERK 1/2
`
`GAPDH
`
`pY1586 ALK
`
`Total ALK
`
`pS473 Akt
`
`Total Akt
`
`pERK 1/2
`
`ERK 1/2
`
`GAPDH
`
`pY1586 ALK
`pY1604 ALK
`pY ALK
`
`24 h
`6 h
`3 h
`1 h
`10 min
`0 min
`24 h
`6 h
`3 h
`1 h
`10 min
`0 min
`24 h
`6 h
`3 h
`1 h
`10 min
`0 min
`24 h
`6 h
`3 h
`1 h
`10 min
`0 min
`24 h
`6 h
`3 h
`1 h
`10 min
`0 min
`24 h
`6 h
`3 h
`1 h
`10 min
`0 min
`
`Crizotinib
`
`Ceritinib
`
`Kelly
`
`200 nM crizotinib
`
`200 nM ceritinib
`
`Time
`
`0 10 min 1 h 3 h 6 h 24 h
`
`0 10 min 1 h 3 h 6 h 24 h
`
`C
`
`100
`
`80
`
`60
`
`40
`
`20
`
`0
`
`P/T ratio (% 0 h control)
`
`pY1586 ALK
`
`Total ALK
`
`pS473 Akt
`
`Total Akt
`
`pERK 1/2
`
`ERK 1/2
`
`GAPDH
`
`1002
`
`Molecular Oncology 11 (2017) 996–1006 ª 2017 The Authors. Published by FEBS Press and John Wiley & Sons Ltd.
`
`Lassen - Exhibit 1061, p. 7
`
`

`

`E. R. Tucker et al.
`
`Immunoassays to quantitate ALK
`
`Fig. 3. Dose–response and time course comparison of ALK inhibition by crizotinib or ceritinib. Dose response in Kelly (A) and CLB-GA (B)
`â
`neuroblastoma cells with 3-hour treatment of crizotinib or ceritinib. Phosphorylated and total ALK levels were quantified by MSD
`immunoassays, and mean  SD from three independent repeats are shown. Corresponding immunoblots of Kelly and CLB-GA cell lysates
`following the treatment with either crizotinib or ceritinib are shown below as indicated. Vertical dashed lines indicate GI50. Time course
`treatment of Kelly cells (C) with crizotinib or ceritinib (both 200 nM) and CLB-GA cells (D) with crizotinib or ceritinib (both 50 nM).
`â
`immunoassays (three independent repeats), and the phosphorylated ALK/
`Phosphorylated and total ALK levels were quantified by MSD
`total ALK ratio as a % of untreated is plotted. Corresponding immunoblots of Kelly and CLB-GA cell lysates following the treatment with
`either crizotinib or ceritinib are shown below as indicated.
`
`(Fig. 4A). Our results suggest that these immunoassays
`are suitable for the quantification of ALK as a biomarker
`following the targeted therapy in vivo.
`
`4. Discussion
`
`In this report, we have validated immunoassays to
`detect ALK and phosphorylated ALK in vitro, success-
`fully applied them to an in vivo model of neuroblas-
`toma, and demonstrated a quantitative difference in
`on-target pharmacodynamic changes between a first-
`and second-generation ALK inhibitor.
`To date, the majority of studies investigating the
`mechanism of activation and phosphorylation of ALK
`utilize oncogenic ALK fusion proteins, such as NPM-
`ALK. Activation of full-length ALK has been shown to
`result in the phosphorylation of Y1278 alongside Y1282
`and Y1283, which lie in the Y’Ras’YY autophosphory-
`lation motif within the activation loop (Tartari et al.,
`2008). Y1604 is found at the C-terminal tail of the recep-
`tor and has been reported to be important for transfor-
`mation activity and docking of phospholipase Cy in
`studies on NPM-ALK. Subsequently, Y1604 and Y1586
`were together identified as tyrosine sites phosphorylated
`following ALK activation in a phosphoproteomics
`study of full-length ALK (Sattu et al., 2013), and the
`detection of dephosphorylation at Y1586 by our
`immunoassay, following the treatment with ALK inhi-
`bitors, confirms the role of this phosphorylation site in
`the constitutive activity of ALK.
`Our panel of neuroblastoma cell lines harbor either
`the ALK F1174L or the ALK R1275Q mutation or
`
`Table 1. 72-Hour GI50 of crizotinib or ceritinib in neuroblastoma cell
`lines (nM).
`
`Cell lines
`
`Crizotinib
`
`SH-SY5Y
`LAN-1
`SK-N-SH
`Kelly
`CLB-GA
`LAN-5
`IMR-32
`
`480.6  93.0
`236.6  9.8
`3059.5  272.2
`432.4  28.6
`167.6  48.3
`132.0  22.8
`511.7  29.1
`Mean  SD from three independent repeats.
`
`Ceritinib
`
`82.0  25.4
`46.8  36.3
`1311.5  252.4
`542.4  62.2
`43.4  7.3
`47.3  6.3
`1203.8  386.7
`
`express wild-type ALK only and therefore represent
`two of the three hotspots of ALK mutation sites in
`et al., 2014). Our
`neuroblastoma tumors
`(Bresler
`cross-comparison of
`immunoblots with the ALK
`immunoassays
`reveals
`strong
`correlation of
`the
`detection of total ALK and two of three ALK phos-
`phorylation sites (Fig. 2, Section 3.2.). The immuno-
`precipitation of ALK and detection of pY furthermore
`reflect the results from the pY ALK immunoassay.
`Measurement of phosphorylated ALK by immunoas-
`say has several advantages over western blotting such
`as greater quantitative power and higher-throughput
`Ò
`platform we used also
`analysis of samples. The MSD
`provides
`superior
`sensitivity and a much greater
`dynamic range in which to detect changes in ALK
`activity.
`Interestingly, by analyzing a neuroblastoma cell line
`panel, we found that cells with the ALK R1275Q
`mutation had the strongest ALK phosphorylation sig-
`nals. R1275 accounts for 43% of ALK mutations in
`neuroblastoma, yet mutations at the F1174 locus (ac-
`counting for 30% of ALK mutations) are reported to
`have the strongest effect on nonphosphorylated ALK
`tyrosine kinase domain in vitro (Bresler et al., 2014). It
`is possible that ALK activity is different in a whole-
`cell environment and acclimatization to cell culture
`conditions alters the dominant cell signaling pathways
`and accounts for this discrepancy between purified
`mutated ALK tyrosine kinase domain experiments and
`our cell-based approaches. Furthermore, we show that
`dephosphorylation at the pY1586 site closely followed
`dephosphorylation at
`the pY1278
`and pY1604
`autophosphorylation sites upon ALK inhibition, and
`therefore, phosphorylation of ALK at any of these
`three
`sites
`could be used as pharmacodynamic
`biomarkers for ALK activity.
`Crizotinib and ceritinib are both under clinical evalu-
`ation for the treatment of ALK-positive neuroblastomas
`with genetic ALK alterations, and our immunoassays
`allow for a direct pharmacodynamic preclinical compar-
`ison of the two compounds (Mosse et al., 2013). The
`observation that ceritinib has a lower GI50 in CLB-GA
`than crizotinib is supported by the results of
`the
`immunoassay in these cells, where we see inhibition of
`
`Molecular Oncology 11 (2017) 996–1006 ª 2017 The Authors. Published by FEBS Press and John Wiley & Sons Ltd.
`
`1003
`
`Lassen - Exhibit 1061, p. 8
`
`

`

`Immunoassays to quantitate ALK
`
`E. R. Tucker et al.
`
`*P = 0.0254
`
`***P = 0.0008
`
`B
`CrizoƟnib
`vehicle
`
`CrizoƟnib
`
`CeriƟnib
`vehicle
`
`Ceritinib
`
`Day 0
`
`Day 7
`
`Vehicle
`
`Crizotinib
`
`Vehicle
`
`Ceritinib
`
`20
`
`15
`
`10
`
`5
`
`0
`
`D
`
`MSD ECL counts (x104)
`
`Total ALK
`
`GAPDH
`
`Vehicle
`
`Crizotinib
`
`Vehicle
`
`Ceritinib
`
`Vehicle
`
`Crizotinib
`
`Vehicle
`
`Ceritinib
`
`ns
`
`*P = 0.0256
`
`300
`
`200
`
`100
`
`0
`
`–100
`
`2500
`
`2000
`
`1500
`
`1000
`
`500
`
`0
`
`MSD ECL counts
`
`A
`
`% Volume change at day 7
`
`C
`
`pY1586 ALK
`
`GAPDH
`
`2.0
`
`1.5
`
`1.0
`
`0.5
`
`0.0
`
`P/T ratio
`
`Normalized
`
`E
`
`Vehicle Crizotinib Vehicle
`
`Ceritinib
`
`â
`
`
`immunoassay quantitation of Th-ALKF1174L/MYCN tumors