Molecular Imaging of Nonsmall Cell Lung Carcinomas Expressing Active Mutant EGFR Kinase Using PET with [124I]-Morpholino-IPQA

Mutations in the kinase domain of epidermal growth factor receptor (EGFR) have high levels of basal receptor phosphorylation and are associated with clinical responsiveness to Iressa in patients with nonsmall cell lung cancer (NSCLC). This study aimed to assess the feasibility of morpholino-[124I]IPQA derivative as an in vivo PET imaging tool for the expression of different EGFR mutants in NSCLC. In vitro radiotracer accumulation and washout studies demonstrated a rapid accumulation and progressive retention after washout of morpholino-[131I]IPQA derivative in high EGFR-expressing H1299 NSCLC derivative cell lines (L858R and E746-A750 del cell lines), but not in EGFR-transfected H1299 cell line and vector-transfected H1299 cell line. Using the morpholino-[124I]IPQA derivative, we obtained noninvasive microPET images of EGFR activity in L858R and E746-A750 del subcutaneous tumor xenografts, but not in subcutaneous tumor xenografts grown form control cell line. Different EGFR mutant (activity) tumors have a different morpholino-[∗I]IPQA derivative uptake. However, it still needs to modify the structure of IPQA to increase its water solubility and reduce hepatobiliary clearance. Morpholino-[124I]IPQA derivative may be a potential probe for selection of the candidate patients suffering from NSCLC for the small molecule tyrosine kinase inhibitor therapy (e.g., Iressa) in the future.


Introduction
Nonsmall cell lung cancer (NSCLC) represents the majority of lung cancers [1]. Lung cancer treatment depends on several factors including tumor type, size, and patient's health. Surgery, radiation therapy, and chemotherapy using tumor specific targeted agents such as vascular endothelial growth factor (VEGF) inhibitor bevacizumab and epidermal growth factor receptor-tyrosine kinase inhibitors (EGFR TKIs) are the primary tools for treating lung cancer. These targeted agents are initially effective in certain small subpopulations of patients, but eventually nearly all patients turn out to be resistant to the further treatments [2]. The limitations in efficacy and safety associated with the existing treatments for NSCLC underscore the need for novel biomarkers and imaging approaches for identification of patients who may 2 BioMed Research International benefit from particular therapeutic agents and approaches with improved efficacy and safety profiles.
The importance of EGFR signaling pathway in the development and progression of NSCLC has been widely recognized [3,4]. EGFR overexpression is observed in tumors of more than 60% of patients with metastatic NSCLC and correlates with poor prognosis [5]. These observations have provided a rationale for the development of novel anticancer agents that target EGFR. Two classes of EGFR-targeted agents are currently being tested in clinical trials: antibodies that bind to the extracellular domain of EGFR (i.e., cetuximab, panitumumab, matuzumab, nimotuzumab, etc.) and small molecular inhibitors that bind to the ATP-binding site of the EGFR tyrosine kinase (i.e., gefitinib, erlotinib, lapatinib, canertinib, etc.) [6]. Clinical trials have revealed significant variability in response to EGFR TKI gefitinib (Iressa), with approximately 10% of Caucasian patients, 25-30% of Japanese patients, and 57% of Taiwanese patients [5,[7][8][9][10][11][12][13]. Sequencing test of the EGFR gene is associated with a majority of tumors responding to EGFR kinase inhibitors harbor mutations in the kinase domain of EGFR [9,14,15]. The response rate to gefitinib and erlotinib in patients with tumors exhibiting activating mutations of EGFR is approximately 75%, suggesting that these mutations, at least in part, may have caused malignant transformation and contribute in large to the tumor maintenance pathway [16,17]. Two subtypes of activating EGFR mutations have been described: tyrosine kinase domain mutation (45%-50%) in EGFR exons [18][19][20][21][22][23][24] and truncating mutations involving exon 2 to 7. The most frequently detected alterations are small deletions in exon 19 (35%-45%) that eliminate amino acids 747-750 (Leu-Arg-Glu-Ala), located around the active site of the tyrosine kinase, and point mutations in exon 21 that result in the amino acid substitution Leu858 → Arg, a residue located in the activation loop [9]. Mutations are most frequently detected in a subpopulation of NSCLC patients with characteristics associated with a better treatment outcome: women, nonsmokers, patients of southeast Asian and Japanese origins, and patients with adenocarcinoma histology and, in particular, bronchioloalveolar carcinoma [18,19]. The mutational status of EFGR kinase could be considered as a positive predictive biomarker of response to NSCLC [20,21].
Considerable efforts have been made over the past decade to develop radiolabeled agents for noninvasive imaging of EGFR expression and activity. These agents include radiolabeled antibodies to EGFR and radiolabeled small molecular compounds based on structures of known EGFR TKIs [22]. Most of the radiolabeled agents for molecular imaging with positron emission tomography (PET) are derived from 4-(anilino)quinazoline pharmacophore [23,24], which includes ML series [21,22,[25][26][27], as well as [4-(3-[ 124 I]iodoani-lino)-quinazolin-6-yl]amide-(3-morpholin-4-yl-propyl)-amide ([ 124 I]IPQA) [28], [29], [ 11 C]PD153035 [30], and [ 11 C]erlotinib [31]. Recently, clinical studies with [ 11 C]PD153035 have demonstrated some promise for imaging EGFR expression in NSCLC patients [32]; however, none of these imaging agents exhibits selectivity for detection of NSCLC expressing active mutant EGFR kinases or EGFR kinase mutations that confer resistance to inhibitors that are currently used in clinical practice. Therefore, we have been developing a PET radiotracer with preferential binding to active mutant EGFR kinases, not to EGFR kinase mutants conferring resistance to current small molecular inhibitors (i.e., gefitinib). PET imaging using such a selective radiolabeled agent should allow for visualization of primary and metastatic tumor lesions driven by activating mutations in EGFR kinase and selection of patients who may benefit from therapy with EGFR kinase inhibitors. Also, after the initial course of treatment with EGFR inhibitors, repetitive imaging with such a selective radiotracer could be used for monitoring the development of tumor lesions with acquired mutations (i.e., T790 M) conferring resistance to EGFR inhibitors.

Radiosynthesis of Morpholino-[ 124 I]IPQA Derivative.
Morpholino-[ 124 I]IPQA derivative was synthesized from but-2-enedioic acid(3-morpholin-4-propyl)-amide[4-(3-tributylstannyl-phenyl-amino)-quinazolin-6-yl]-amide (morpholino-IPQA derivative), as described previously [28]. The radiosynthesis of morpholino-[ 124 I]IPQA derivative was modified according to the earlier study [33]. The tin precursor (100 g; from MD Anderson Cancer Center, Houston, TX, USA) was dissolved in 0.02 mL of methanol, and then the [ 124 I]NaI (37∼55 MBq) was added to the solution of precursor and vortexed followed by the addition of a 0.018 mL mixture of 30% hydrogen peroxide/acetate acid (1 : 3). The reaction mixture was vortexed for 1 min and then allowed to stand for 5 min. Saturated sodium bicarbonate (0.2 mL) and 2N sodium thiosulphate (0.12 mL) were added to quench the reaction, and the reaction mixture was then loaded to a plus C-18 Sep-Pak cartridge system. The C-18 cartridge system (preconditioned with 10 mL of ethanol, then 10 mL of water) was eluted with water (30 mL), followed by 20% ethanol/water (25 mL), 40% ethanol/water (25 mL), and 60% ethanol/water (25 mL). The majority of the product was isolated during the elution with 60% ethanol/water. Each fraction was collected and assayed for radioactivity measurement, and the fractions that contain the final product were assayed by radio-thin-layer chromatography (radio-TLC). The radiochemical purity was determined by radio-TLC (silica gel 60 F254; eluent: chloroform/methanol = 6/1). The collected products with sufficient purity were combined and evaporated.

Radiosynthesis of Morpholino-[ 131 I]IPQA Derivative.
No carrier-added morpholino-[ 131 I]IPQA derivative was prepared using the same procedure as that of morpholino-[ 124 I]IPQA (50 g of the tin precursor was used). The radiolabeled product was isolated in 65% radiochemical yield (decay corrected) with a radiochemical purity of 92.3%.
Cells were grown in flasks with RPMI 1640 with 10% FBS and antibiotics at 37 ∘ C in humidified atmosphere with 5% CO 2 . Cells were kept in the log phase of proliferative activity. The L858R, E746-A750del, EGFR-transfected, and vector-transfected H1299 cells were kind gifts from Drs. Shih-Feng Tsai and Yi-Rong Chen (The National Health Research Institutes, Miaoli, Taiwan).

Irreversible Binding of Morpholino-[ 124 I]IPQA to Active
Mutant EGFR Kinase Domain. The irreversible and covalent bindings of morpholino-[ 131 I]IPQA to the EGFR kinase domain were evaluated in H1299 derivatives cells. The cells were grown in a 15 cm culture dish until 60-70% confluency and then incubated for one hour in fresh culture medium supplemented with 20% FBS and morpholino-[ 131 I]IPQA at 0.37 MBq/mL. Thereafter, the cells were harvested by scraping, pelleted by centrifugation at 1,000 rpm for 5 minutes, and lysed in 0.5 mL of buffer containing protein extraction reagent (Cytobuster, Novagen, USA) and aprotinin, leuppetin, pepstain (1 g/mL for each), Na 3 VO 4 , NaF, and PMSF (1 mM for each; Sigma-Aldrich, CA, USA). The cell lysate was cleared by 14,000 ×g centrifugation at 4 ∘ C for 15 minutes. The cell lysate supernatant was denatured by boiling with 4x Laemmli sample buffer and separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis using precast 8% Tris HCl gel cassettes (BioRad, CA, USA). After transferring proteins into a minitank electroblotter device, the membrane was exposed to AX film (Konica, Japan) for seven days at room temperature to produce an autoradiogram of 131 I-labeled protein bands. Thereafter, the nitrocellulose transfer membrane was immunostained with a rabbit polyclonal EGFR (1005)-sc-03 antibody (Santa Cruz, CA, USA) and visualized using the ECL kit (Amersham Biosciences, UK). The colocalization of 131 I-labeled proteins in the autoradiogram with protein bands stained with anti-EGFR antibody was assessed.

In Vitro Radiotracer Uptake and Washout
Assay. Radiotracer uptake and washout studies were performed in monolayer cultures of four NSCLC cell lines as described previously [28]. Briefly, tumor cells were grown in 15 cm culture dishes until 60∼70% confluent, at which point the cells were exposed to the culture medium without FCS to induce serum starvation (to inhibit EGFR kinase activity). The radiotracer morpholino-[ 131 I]IPQA at 0.18 MBq/mL was then added to fresh cell culture medium without FCS (serum-starved), and tumor cell monolayers were exposed to the radioactivitycontaining medium for 5, 10, 20, 30, and 60 minutes. In the first part of the experiment, the cells were harvested by gentle scraping at different time intervals, pelleted by centrifugation (3,500 rpm for 2 minutes). The cell pellet and 0.1 mL of radioactive supernatant were weighed and assessed for radioactivity using a Packard 5500 gamma counter (Perkin-Elmer, CA, USA); the radioactivity concentration was expressed as cpm/g cells and cpm/mL medium, respectively. In the second part of the experiment, tumor cells exposed to the morpholino-[ 131 I]IPQA containing medium for a given time interval (5, 10, 20, and 60 minutes) were washed with fresh (nonradioactive) medium for different time intervals before subsequent harvesting for measurement of retained radioactivity, as described above. Cells-to-medium radioactivity concentration ratios were calculated and plotted versus time to evaluate the radiotracer accumulation and washout kinetics.

Subcutaneous Tumor Xenografts and PET Image Acquisition and Data
Analysis. NOD/SCID mice weighing about 25∼30 g ( = 3) were injected subcutaneously into the left shoulder region with L858R, E746-A750 del, or wildtype EGFR-transfected H1299 cells (5 × 10 6 cells/mouse). In the right shoulder region of each mouse, 5 × 10 6 of vector-transfected H1299 cells were inoculated as a control tumor. The tumors grew to about 8∼10 mm in diameter after 4 weeks. At this point, the mice were anesthetized (2% isoflurane/98% oxygen mixture) and injected intravenously with morpholino-[ 124 I]IPQA (2.6 MBq/mouse). One-hour dynamic PET imaging was performed on MicroPET R4 system (Siemens, TN, USA) and followed with a 30-minute static imaging at 24 hours after i.v. administration of radiotracer. PET images were reconstructed using the ordered subsets expectation maximization iterative reconstruction algorithm. Regions of interest were drawn over tumors and other tissues of interest.

Statistical
Analysis. Group data were expressed as average ± SE and compared using analysis of variance, regression analysis, and group and paired Student's t tests; a value of < 0.05 was considered statistically significant. Volume of distribution at equilibrium expressed as cell-to-medium accumulation ratio (average ± standard error). 2 Coefficient (slope) of the exponential washout = A × exp (−BX) .

Radiochemistry.
The final product of morpholino-[ 124 I] IPQA derivative (Figure 1) was dissolved in saline (with 10% ethanol added). Radio-TLC showed the Rf value of morpholino-[ 131/124 I]IPQA derivative to be 0.5. The radiochemical yield was 50%, and the radiochemical purity was ≥90% (decay-corrected). Radio-high-performance liquid chromatography (radio-HPLC) using an Alltech Alltima C18LL column (250 × 4.6 mm; Fisher Scientific, USA) and a mobile phase consisting of 0.1 M acetate buffer (acetate/acetonitrile = 55/45) at a flow rate of 1.0 mL/min exhibited the product as a peak at about 10 minutes.  (Figure 2(a)) These bands were correspondent with the 172 kDa protein band stained with anti-EGFR antibody (Figure 2(b)). The above observations collectively suggest that in EGFR-transfected cells (cells with wild-type EGFR overexpression) the lower radioactive intensity of morpholino-[ 131 I]IPQA (Figure 2(a)) corresponds to a higher level of anti-EGFR immunoblotting (Figure 2(b)) compared to EGFR mutation cells.

Preferential Accumulation of Morpholino-[ 131 I]IPQA in NSCLC Cells with L858R and E746-A750
Del EGFR Mutations. All four types of cells showed a rapid accumulation of morpholino-[ 131 I]IPQA at the initial 10 minutes and thereafter reached a plateau at 60 minutes (Figures 3(a)-3(d)). In L858R and E746-A750 del cells, the cell-to-medium ratio (CMR) of morpholino-[ 131 I]IPQA was about 60-65 at 60 minutes, which was ∼1.6-fold higher than that of wild-type EGFR-transfected H1299 cells (Table 1).

In Vivo PET Imaging of Morpholino-[ 124 I]IPQA Exhibits
Differentials of Accumulation in NSCLC Tumors with Different EGFR Expressions. In vivo PET imaging was performed in 18 mice (6 per tumor pair) before and after treatment with Iressa. MicroPET images demonstrated that highest accumulation level of morpholino-[ 124 I]IPQA was observed in L858R tumor xenograft at 24 hours after radiotracer administration. The accumulation was 1.23-, 2.36-, and 3.08-folds higher than that of E746-A750 del, wild-type EGFR-transfected, and vector-transfected tumor xenograft in baseline group, respectively (Figures 4(a)-4(c), left panel, Figure 5). Pretreatment with Iressa (100 mg/kg 1 h before administration of morpholino-[ 124 I]IPQA) results in 53% and 38% decrease in the accumulation of radiotracer in L858R and E746-A750 del tumors, respectively (Figures 4(a)-4(b) right panels, Figure 5). There was a similar accumulation of radiotracer in wild-type EGFR-transfected and vector-transfected tumor xenografts before and after Iressa treatment (Figures 4(a)-4(c); Figure 5). The results of tumor-to-vector ratio at 24 hours after administration of morpholino-[ 124 I]IPQA was summarized in Table 5.

Biodistribution of Morpholino-[ 131 I]IPQA Derivative in NSCLC Tumor-Bearing Mice.
In the first hour, high % ID/g was observed in pancreas, kidney, stomach, lung, liver, and small and large intestines ( Table 2). Four tumors expressing different levels of EGFR activity/expression had the peak radioactivity concentration in the first hour and then showed a gradual decrease over time; however, blood, heart, liver, and spleen showed a redistribution 24 hours after administration of morpholino-[ 131 I]IPQA. The levels of radioactivity in blood or heart rapidly dropped to 0.1∼0.2% ID/g one hour after radiotracer injection and progressively decreased during 4 to 48 hours. We also observed a rapid excretion of radiotracer by liver. This pattern of hepatobiliary clearance was followed by a fast increased radioactivity in blood and liver and subsequent clearance by kidneys. Up to 24 hours after injection of radiotracer, the tumor-to-blood (TBR) and tumor-to-muscle concentration ratios (TMR) were not distinguishable among four tumor xenografts (Tables 3 and 4; all other three cell lines  were compared to L858R tumor cells at the same time point). The TBR and TMR in L858R tumor xenograft 24 hours after injection of radiotracer were 2.64 and 13.89, respectively (Tables 3 and 4). The E746-A750 del tumor xenograft had similar TBR and TMR compared to those of L858R group. The L858R tumor xenograft also had a better tumor-to-vector ratio 24 hours after administration of radiotracer when compared to those of other 2 xenografts (Table 5).

Discussion
Predicting the expression of EGFR mutation through noninvasive PET imaging with a specific EGFR kinase radiotracer would provide an assessment for the NSCLC patients who may be benefit from EGFR inhibitors therapeutic regiment. Previously, our collaborator reported that the PET imaging with morpholino-[ 124 I]IPQA, which could irreversibly and specifically bind to active form of EGFR kinase, allowed for identification of tumors with high EGFR kinase signaling activity (i.e., A431 highly expressing EGFR in NSCLC and U87 del EGFR cells expressing EGFRvIII mutants in brain gliomas) [28].
In the current study, by using morpholino-[ 124 I]IPQA with PET, we demonstrated that L858R and E746-A750 del EGFR mutated cells, which are the most frequent mutation in NSCLC [9], showed significant increased accumulation of radiotracer when compared to the wild-type EGFRtransfected and vector-transfected cells in vitro and in vivo. The locations of the L858R missense mutations are shown within the activating loop of the tyrosine kinase, whereas the in-frame deletion, E746-A750 del, is present within another loop which flanks the ATP cleft [14]. Those mutations are predicted to alter the position of these amino acids relative to that of phosphorylation status of the cells and the sensitivity of inhibitor (i.e., gefitinib).
The expression levels of phosphorylated EGFR in L858R and E746-A750 del EGFR mutated cells were similar, whereas wild-type EGFR-transfected cells showed the highest expression of total EGFR. Nevertheless, the mutation in active site of tyrosine kinase domain results in the enhanced EGFR signaling (phosphorylation status), which explains the increased accumulation of morpholino-[ 131 I]IPQA in L858R and E746-A750 del EGFR mutated cells. Similar interpretation can also   be applied to the preferential accumulation and retention of morpholino-[ 124 I]IPQA in these mutant EGFR cells in in vivo PET imaging.
In comparison with previous reports of EGFR imaging agents [22,25,26,34,35], the results of our in vivo PET imaging studies with morpholino-[ 124 I]IPQA in mice bearing four tumor xenografts expressing different levels of phosphorylated EGFR are more selective. Predominant accumulations of morpholino-[ 124 I]IPQA in EGFR-expressing L858R and E746-A750 del carcinoma tumor xenografts reflect the high level of phosphorylated EGFR expression and activity in those tumor cells, which are known to be responsive to therapy with small molecular inhibitors of EGFR (e.g., Gefitinib) [14]. The H1299 EGFR tumor xenograft expressing EGFR represented a model of therapy-resistant tumors. H1299 EGFR tumor had lower morpholino-[ 124 I]IPQA accumulation and produced similar images before and after treatment with EGFR inhibitor, Gefitinib (Iressa).
Our results also provided more information of tissue distributions at later time points (24 and 48 hours after radiotracer administration). Selective accumulations of radiotracer in the L858R and E746-A750 del EGFR mutants were observed when compared to the tumors with wildtype EGFR or vector-transfected cells. Similar to the results from dynamic PET imaging obtained from 0 to 70 minutes after administration of morpholino-[ 124 I]IPQA [28], the biodistribution data obtained from 1, 4, 24, and 48 hours after administration of morpholino-[ 131 I]IPQA showed a rapid clearance of radiotracer from blood and other major tissues such as heart, brain, liver, intestines, and kidney. The results demonstrated that the nonspecific binding of the radiotracer was washed out rapidly and caused increased tumor-tobackground ratio due to the irreversible and selective bindings to active state of EGFR kinase.
We also observed an unexpected high accumulation (3.2% ID/g at 24-hour time point) in pancreas at all time points. Mohammed et al. reported that gefitinib blocks EGFR signaling pathway in progression of pancreatic intraepithelial neoplasms (PanINs) to PDAC in conditional LSL-Kras G12D/+ transgenic mice model [36]. Kelley and Ko reported a significant survival benefit with the addition of the EGFR tyrosine kinase inhibitor erlotinib to gemcitabine chemotherapy for the first-line treatment of patients with advanced pancreatic cancer [37]. These results demonstrated that morpholino-[ 131 I]IPQA was entrapped in normal pancreatic cell (i.e., ductal cells and the islets of Langerhans) [38].
However, despite the lower lipophilicity of morpholino-[ 124 I]IPQA compared to previously reported EGFR imaging compounds such as the N-{4-[(4,5-dichloro-2-fluorophenyl) amino]quinazolin-6-yl}-acrylamide, 4-[(3,4-dichloro-6-fluorophenyl)amino]-6,7-dimethoxyquinazoline, and 4-(3bromoanilino)-6,7-dimethoxyquinazoline [22,25,26,34,35], the morpholino-[ 124 I]IPQA still exhibited a significant hepatobiliary clearance. Also, high accumulation of the radiotracer in abdomen area over time would decrease the possibility of using this radiotracer for imaging primary site of NSCLC in animal model or for colorectal metastases. The further improvement could focus on the optimization of its pharmacokinetic properties by additional chemical derivatization (i.e., decrease lipophilicity and extraction by the liver, increase0 the half-life in plasma and accumulation amount/retention0 time in tumor tissue, and increase water solubility0 and renal clearance for a better background-totarget ratio).0 The relatively low resolution of microPET system (spatial resolution ∼1.8 mm) may not allow for assessing the heterogeneity of morpholino-[ 124 I]IPQA radioactivity accumulation. Therefore, micro-PET/CT and autoradiography (with 10-30 M in plane resolution) are suitable to study the heterogeneity of morpholino-[ 124 I]IPQA-derived radioactivity accumulation and to compare it with the immunohistochemically stained (adjacent) tissue sections for the total EGFR or phosphorylated EGFR (Tyrosine 1068).
As a breakthrough of NSCLC treatment management in 2004, Lynch et al. [14] first reported that specific mutations in EGFR gene in NSCLC patients were correlated with clinical responsiveness to the tyrosine kinase inhibitor gefitinib. These findings were confirmed by Pao et al. in the tumor cells from the patients with L858R mutation but not for L747-S752 del EGFR mutant [14]. The study results from Sordella et al. also supported that EGF-independent autophosphorylation in these mutations caused high sensitivity to gefitinib and selectively activated downstream pathways [39]. Tracy et al. also proved that gefitinib-induced apoptosis in the L858R mutant lead dramatic response to gefitinib [40]. The authors also concluded that the affinity for the L858R mutant to gefitinib was 20-fold higher than that for wild type, which could be explained by its tighter binding to active conformation of the tyrosine kinase domain. Also, it could be expected to predict the efficacy in the treatment of the subgroup with gefitinib. Tracy et al. suggested that targeted agents developed to inhibit AKT pathway may be therapeutically more effective than those designed to inhibit the ERK 1/2 pathway in patients whose tumors contain EGFR L858R if the AKT pathway is consistently active [40]. Moreover, Brognard et al. reported that constitutive AKT activation has also been associated with resistance to chemotherapy and radiation in NSCLC cell lines [41]. Further study could be the mono/combinations of chemotherapy with gefitinib or inhibitors of the AKT pathway which should be tested in vitro in cell lines and in vivo in tumor xenografts with EGFR L858R mutations to determine whether these may be additive or synergistic.
In conclusion, an enhanced binding of morpholino-[ 124/131 I]IPQA derivatives to the ATP binding site of mutant kinase of L858R or E746-A750 del EGFR mutant warrants that PET imaging with morpholino-[ 124 I]IPQA has a potential for identification of tumors with high EGFR kinase activity in NSCLC and for the monitoring or selection of individual therapies with EGFR inhibitors. Further optimization of this class of radio compound would be necessary in order to lower lipophilicity and reduce hepatobiliary clearance.