CT Imaging of Cell-Line and Patient-Derived EGFR

Jun 26, 2019 - Prior to microSPECT, microCT was performed with routine parameters ... Mice were then sacrificed under anesthesia and blood, and select...
0 downloads 0 Views 3MB Size
Download PDF
Article Cite This: Mol. Pharmaceutics XXXX, XXX, XXX−XXX

pubs.acs.org/molecularpharmaceutics

MicroSPECT/CT Imaging of Cell-Line and Patient-Derived EGFRPositive Tumor Xenografts in Mice with Panitumumab Fab Modified with Hexahistidine Peptides To Enable Labeling with 99mTc(I) Tricarbonyl Complex Anthony Ku,† Conrad Chan,† Sadaf Aghevlian,† Zhongli Cai,† David Cescon,§ Scott V. Bratman,¶,⊥ Laurie Ailles,⊥ David W. Hedley,§ and Raymond M. Reilly*,†,#,‡ Downloaded via BUFFALO STATE on July 21, 2019 at 03:58:56 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

†

Department of Pharmaceutical Sciences, University of Toronto, 144 College Street, Toronto, ON M5S 3M2, Canada Departments of §Medical Oncology, ¶Radiation Oncology, and ⊥Medical Biophysics, Princess Margaret Cancer Centre, 610 University Avenue, Toronto, ON M5G 2M9, Canada # Department of Medical Imaging, University of Toronto, 263 McCaul Street, Toronto, ON M5T 1W7, Canada ‡ Toronto General Research Institute and Joint Department of Medical Imaging, University Health Network, 200 Elizabeth Street, Toronto, ON M5G 2C4, Canada S Supporting Information *

ABSTRACT: We aimed to investigate the feasibility of conjugating synthetic hexahistidine peptides (His6) peptides to panitumumab Fab (PmFab) to enable labeling with [99mTc(H2O)3(CO)3]+ complex and study these radioimmunoconjugates for imaging EGFR-overexpressing tumor xenografts in mice by microSPECT/CT. Fab were reacted with a 10-fold excess of sulfo-SMCC to introduce maleimide functional groups for reaction with the terminal thiol on peptides [CGYGGHHHHHH] that harbored the His6 motif. Modification of Fab with His6 peptides was assessed by SDSPAGE/Western blot, and the number of His6 peptides introduced was quantified by a radiometric assay incorporating 123 I-labeled peptides into the conjugation reaction. Radiolabeling was achieved by incubation of PmFab-His6 in PBS, pH 7.0, with [99mTc(H2O)3(CO)3]+ in a 1.4 MBq/μg ratio. The complex was prepared by adding [99mTcO4]− to an Isolink kit (Paul Scherrer Institute). Immunoreactivity was assessed in a direct (saturation) binding assay using MDA-MB-468 human triplenegative breast cancer (TNBC) cells. Tumor and normal tissue uptake and imaging properties of 99mTc-PmFab-His6 (70 μg; 35−40 MBq) injected i.v. (tail vein) were compared to irrelevant 99mTc-Fab 3913 in NOD/SCID mice engrafted subcutaneously (s.c.) with EGFR-overexpressing MDA-MB-468 or PANC-1 human pancreatic ductal carcinoma (PDCa) cellline derived xenografts (CLX) at 4 and 24 h post injection (p.i.). In addition, tumor imaging studies were performed with 99m Tc-PmFab-His6 in mice with patient-derived tumor xenografts (PDX) of TNBC, PDCa, and head and neck squamous cell carcinoma (HNSCC). Biodistribution studies in nontumor bearing Balb/c mice were performed to project the radiation absorbed doses for imaging studies in humans with 99mTc-PmFab-His6. PmFab was derivatized with 0.80 ± 0.03 His6 peptides. Western blot and SDS-PAGE confirmed the presence of His6 peptides. 99mTc-PmFab-His6 was labeled to high radiochemical purity (≥95%), and the Kd for binding to EGFR on MDA-MB-468 cells was 5.5 ± 0.4 × 10−8 mol/L. Tumor uptake of 99mTcPmFab-His6 at 24 h p.i. was significantly (P < 0.05) higher than irrelevant 99mTc-Fab 3913 in mice with MDA-MB-468 tumors (14.9 ± 3.1%ID/g vs 3.0 ± 0.9%ID/g) and in mice with PANC-1 tumors (5.6 ± 0.6 vs 0.5 ± 0.1%ID/g). In mice implanted orthotopically in the pancreas with the same PDCa PDX, tumor uptake at 24 h p.i. was 4.2 ± 0.2%ID/g. Locoregional metastases of these PDCa tumors in the peritoneum exhibited slightly and significantly lower uptake than the primary tumors (3.1 ± 0.3 vs 4.2 ± 0.3%ID/g; P = 0.02). In mice implanted with different TNBC or HNSCC PDX, tumor uptake at 24 h p.i. was variable and ranged from 3.7 to 11.4%ID/g and 3.8−14.5%ID/g, respectively. MicroSPECT/CT visualized all CLX and PDX tumor xenografts at 4 and 24 h p.i. Dosimetry estimates revealed that in humans, the whole body dose from administration continued...

Received: Revised: Accepted: Published: © XXXX American Chemical Society

A

April 19, 2019 June 25, 2019 June 26, 2019 June 26, 2019 DOI: 10.1021/acs.molpharmaceut.9b00422 Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Article

Molecular Pharmaceutics

of 740−1110 MBq of 99mTc-PmFab-His6 would be 2−3 mSv, which is less than for a 99mTc-medronate bone scan (4 mSv). KEYWORDS: panitumumab, 99mTc, single photon emission computed tomography (SPECT), epidermal growth factor receptors (EGFR), triple-negative breast cancer, pancreatic cancer, head and neck squamous cell carcinoma

■

INTRODUCTION

al. in 199925 and offers several advantages including avoiding disruption of the structural integrity of the mAb (a risk for disulfide reduction), simple and rapid radiolabeling that is stable, which takes advantage of His6 affinity tags that are present in many recombinant proteins. The method has been applied to labeling bivalent anti-EGFR (scFv)2 fragments and nanobodies with 99mTc.26,27 In addition, a kit for facile and efficient synthesis of the [99mTc(H2O)3(CO)3]+ complex was designed and manufactured by Mallinckrodt Pharmaceuticals (Isolink) further enhancing the attractiveness of this approach to 99mTclabeling of mAb fragments.28 However, a practical limitation is the need for a His6 tag incorporated into the fragments to complex [99mTc(H2O)3(CO)3]+. There has been a recent rapid expansion in the development of mAbs as anticancer agents,29 and many pharmaceutical quality mAbs are available as templates to design novel MI probes for cancer. However, these pharmaceutical mAbs do not often incorporate a His6 tag for labeling with [99mTc(H2O)3(CO)3]+. To overcome this limitation, we report here for the first time the conjugation of synthetic His6 peptides to panitumumab Fab fragments that enabled labeling with [99mTc(H2O)3(CO)3]+. To illustrate the utility of this method, we studied these 99mTc-labeled panitumumab Fab for imaging cell-line derived tumor xenografts (CLX) and primary patient-derived tumor xenografts (PDX) of TNBC, PDCa, and HNSCC in mice by microSPECT/CT. These studies also provide the first direct comparison of the imaging properties of 99mTc-labeled panitumumab Fab in CLX and more clinically relevant PDX mouse tumor models for three cancer types that overexpress EGFR.

Epidermal growth factor receptor (EGFR) overexpression is found in several epithelial-derived solid tumors, including triplenegative breast cancer (TNBC),1 pancreatic ductal carcinoma (PDCa),2 and head and neck squamous cell carcinoma (HNSCC).3 Radiological imaging (CT, mammography, MRI, and ultrasound) is used to diagnose and stage these cancers, but molecular imaging (MI), which includes single photon emission computed tomography (SPECT) and positron emission tomography (PET), may have an important role in characterizing lesions as EGFR-positive. MI could identify patients who would benefit from EGFR-targeted therapies. Anti-EGFR monoclonal antibodies (mAbs), panitumumab (Vectibix; Amgen), and cetuximab (Erbitux; Bristol-Myers Squibb) combined with chemotherapy have shown promise for treatment of TNBC,4,5 and panitumumab combined with chemotherapy and radiation improved the outcome of patients with HNSCC.6−8 Anti-EGFR mAbs have not proven effective for treatment of PDCa due to downstream KRAS mutation,9,10 but we recently reported that radioimmunotherapy (RIT) with 177 Lu-labeled panitumumab inhibited the growth of KRAS mutant PANC-1 human PDCa xenografts in NODRag1nullIL2rgnull (NRG) mice, suggesting that RIT can overcome KRAS mutation and may be effective for treatment of EGFRpositive PDCa in humans.11 Imaging of EGFR expression on tumors has mostly been performed with mAbs or their fragments labeled with 111In or 99m Tc for SPECT12−17 or with 89Zr or 64Cu for PET.18−22 The higher sensitivity and better spatial resolution of PET and the ability to more accurately quantify tumor uptake are advantages compared to SPECT. Nonetheless, SPECT remains the most prevalent imaging modality in nuclear medicine. In Canada, there are 10 times more SPECT systems than PET tomographs in clinical use.23 Moreover, 99mTc is the most widely used radionuclide in nuclear medicine for several reasons: (i) on-site production at low cost using a 99Mo/99mTc generator, (ii) a γphoton energy with high abundance [Eγ = 140 keV (98.9%)] that is optimal for imaging with the γ-camera, (iii) a short physical half-life (t1/2p = 6 h), which minimizes the radiation doses from imaging procedures, and (iv) highly efficient (>90%) chelation radiochemistry that enables kit formulation. Since 99m Tc is short-lived, mAb fragments (e.g., F(ab′)2, Fab or scFv) are used for MI since these quickly localize in tumors and are rapidly eliminated from the blood and most normal organs (except kidneys) to generate high tumor/blood (T/B) and tumor/normal tissue (T/NT) ratios for imaging. Several approaches have been studied for labeling mAbs and their fragments with 99mTc.24 These include (i) reduction of some disulfide bonds on the intact IgG to generate free thiols that complex 99mTc, (ii) modification with hydrazinenicotinamide (HYNIC) to complex 99mTc-labeled coligands (e.g., 99m Tc-tricine or 99mTc-glucoheptonate), (iii) conjugation of preformed 99mTc-mercaptoacetylglycylglycylglycine (99mTcMAG3), and (iv) binding of a 99mTc(I)-carbonyl complex [99mTc(H2O)3(CO)3]+ to a hexahistidine (His6) “tag” that is often incorporated into recombinant proteins for metal affinity purification. This latter method was first described by Waibel et

■

MATERIALS AND METHODS Cell Culture. MDA-MB-468 human triple-negative breast cancer (TNBC) cells and PANC-1 human pancreatic adenocarcinoma (PDCa) cells were purchased from the American Type Culture Collection (Manassas, VA). MDAMB-468 cells were cultured in RPMI 1640 medium (SigmaAldrich) with 10% fetal bovine serum (FBS, Invitrogen), penicillin (100 U/mL), and streptomycin (100 μg/mL) under a 5% CO2 atmosphere at 37 °C. PANC-1 cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% FBS, penicillin (100 U/mL), and streptomycin (100 μg/mL) under a 5% CO2 atmosphere at 37 °C. Tumor Xenograft Mouse Models. Cancer cell line tumor xenografts (CLX) were established in mice using MDA-MB-468 TNBC cells and PANC-1 PDCa cells. MDA-MB-468 cells express 1.3 × 106 EGFR/cell,12 while PANC-1 cells express 3.7 × 104 EGFR/cell.21 Briefly, NOD/SCID mice were inoculated subcutaneously (s.c.) with 3 × 106 MDA-MB-468 or PANC-1 cells suspended in 100 μL of serum-free medium on the right hind leg. Tumors were allowed to reach 0.8−1.2 cm in diameter. Patient-derived PDCa, BC, and head and neck squamous cell carcinoma (HNSCC) xenografts (PDX) were established by coauthors Drs. David Hedley, David Cescon, or Laurie Ailles. Briefly, OCIP23 PDX were prepared by orthotopic implantation into the pancreas of NOD/SCID mice through a small incision in the upper quadrant of the left abdomen, as previously reported.30 OCIP23 cells express 4.3 × 105 EGFR/cell.21 B

DOI: 10.1021/acs.molpharmaceut.9b00422 Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Article

Molecular Pharmaceutics

was incubated with horseradish peroxidase (HRP)-conjugated rabbit anti-histidine antibodies (ThermoFisher) diluted 1000× in Tris buffer for 1 h followed by the addition of 3,3′diaminobenzidine tetrahydro chloride (DAB; Sigma-Aldrich) and 30% H2O2 in Tris buffer. Synthesis of [99mTc(H2O)3(CO)3]+ and 99mTc Labeling of PmFab-His6. [99mTc(H2O)3(CO)3]+ complex was synthesized by adding 740 MBq (1.0 mL) of 99mTc sodium pertechnetate ([99mTcO4]−) into an IsoLink kit (Paul Scherrer Institute). The IsoLink kit consists of a lyophilized formulation of 8.5 mg sodium tartrate (Na2C4H4O6), 2.85 mg sodium tetraborate (Na2B4O7), 7.15 mg sodium carbonate (Na2CO3), and 4.5 mg sodium boranocarbonate (Na2H3BCO2) in a 10 mL N2-flushed glass vial. After adding 99mTc, the kit vial was heated in a boiling water bath for 20 min, then cooled to room temperature (RT). The kit solution was then adjusted to pH 7.0 by addition of 140 μL of 1 M HCl into the vial. [99mTc(H2O)3(CO)3]+ complex was produced in high radiochemical yield (>93%) assessed by thin layer chromatography (TLC) on plastic backed silica gel 60 F254 plates (Millipore) developed in 1% HCl in methanol.28 In this system, three 99mTc species are separated (Rf = 0.0 for 99m Tc-colloids; Rf = 0.2−08 for [99mTc(H2O)3(CO)3]+; and Rf = 0.9 for [99mTcO4]−). Radioactivity on the TLC plate was visualized using a Cyclone Plus Phosphor Imager (PerkinElmer), and the image was analyzed by the PerkinElmer OptiQuantTm Image Analysis Software. 99m Tc-labeling was achieved by incubation of PmFab-His6 in PBS, pH 7.0, in a 1.5 mL Eppendorf tube with [99mTc(H2O)3(CO)3]+ in a 1.4 MBq/μg ratio for 1 h at 37 °C. The final radiochemical purity (RCP) of 99mTc-PmFab-His6 was measured by instant thin layer-silica gel chromatography (ITLC-SG; Agilent Technologies), developed in 100 mM sodium citrate, pH 5.5. This system separates 99mTc-PmFab-His6 (Rf = 0.0) from unbound [99mTc(H2O)3(CO)3]+ and free [99mTcO4]− (both Rf = 1.0).28 Radioactivity distribution on the ITLC-SG strips (Rf = 0−0.5 and Rf = 0.5−1.0) was determined by cutting the strips in two equal sections and measuring each section in a radioisotope dose calibrator (Capintec Model CRC15R) or γcounter. An irrelevant control Fab (Fab 3913) donated by Dr. Sachdev Sidhu at the Toronto Recombinant Antibody Centre (TRAC), University of Toronto recognizing an epitope of the Ebola virus and containing an endogenous His6-tag at the Cterminus of the heavy chain was similarly labeled with [99mTc(H2O)3(CO)3]+. EGFR Binding Assay. The affinity of binding of 99mTcPmFab-His6 to EGFR was determined in a direct (saturation) binding assay. Briefly, increasing concentrations of 99mTcPmFab-His6 (0.25−150 nmol/L) were incubated with 1 × 106 MDA-MB-468 cells (1.3 × 106 EGFR/cell)31 in Eppendorf tubes for 3 h at 4 °C. Cells were rinsed with PBS (pH 7.4) then centrifuged at 1000 × g for 5 min. The supernatant was removed. The total radioactivity bound to the cells (TB) and unbound radioactivity in the supernatant were assayed in a γ-counter. This was repeated in the presence of a 50-fold molar excess of unlabeled PmFab-His6 (0.25−150 nmol/L) to estimate nonspecific binding (NSB). Specific binding (SB) was calculated by subtracting NSB from TB. The amount of 99mTc-PmFab-His6 bound (pmol) was plotted vs the concentration (nmol/L) of unbound 99mTc-PmFab-His6. The binding curve was fitted to a one-site receptor-binding model using Prism Ver. 4.0 software (GraphPad). MicroSPECT/CT Imaging and Biodistribution Studies. MicroSPECT/CT imaging was performed at 4 and 24 h post i.v.

HNSCC tumors derived from four different patient specimens (#459, #391, #128, #191) were used to establish PDX. These PDX were generated from patient primary tumor specimens that were dissected into small fragments (∼1 mm3) and engrafted s.c. on the flank of NRG mice. These xenografts were serially propagated by dissociating the excised tumors into single cell suspensions and s.c. inoculating 2 × 105 cells in the right and or left flank of NRG mice. Breast cancer PDX were generated from surgical resection specimens of primary tumors from women with ER, PR, and HER2-negative or TNBC by orthotopic implantation of small tumor fragments (2−3 mm) in female NSG mice. Following initial engraftment, tumors were serially propagated by s.c. implantation of tumor fragments in female C.B.-17 SCID mice. Panitumumab Fab (PmFab). Panitumumab Fab (PmFab, MW = 55 kDa) were prepared by proteolytic digestion of 5 mg of panitumumab IgG (Vectibix, Amgen, MW = 147 kDa) using 10 mL of immobilized-papain slurry (Pierce Biotechnology). Briefly, immobilized-papain was equilibrated in digestion buffer (20 mM NaH2PO4, 10 mM EDTA, and 80 mM of cysteine HCl, pH 4.5) then combined with 5 mg of panitumumab IgG and incubated for 20 h at 37 °C in a Excella E24 Incubator Shaker (New Brunswick Scientific) at 300 rpm. The cysteine HCl was incorporated into the digestion buffer immediately prior to adding panitumumab IgG. Conditions were optimized to achieve maximum Fab yield with no residual undigested IgG. Following IgG digestion, the resin suspension was centrifuged at 1000 × g and rinsed 3 times with phosphate buffered saline (PBS) pH 7.4. The pooled supernatant was passed through a Millex-GS 0.22 μm syringe filter unit (Millipore). Purified PmFab were then concentrated to 8.0 mg/mL and buffer exchanged by ultrafiltration into PBS using an AMICON UltraCentrifugal Unit (Millipore; MWCO = 30 kDa). The purity and homogeneity of PmFab were assessed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) on a 7.5% Mini-Protean Tris/Glycine mini-gel (Biorad) under reducing and nonreducing conditions stained with Biosafe Coomassie Blue G-250 stain (Biorad). Conjugation of PmFab to Hexahistidine (His6) Peptides. PmFab were reacted with a 10-fold molar excess of sulfosuccinimidyl 4-(N-maleimidomethyl) cyclohexane-1-carboxylate (sulfo-SMCC) (ThermoFisher Scientific) at room temperature (RT) for 1 h to introduce maleimide groups. The reaction mixture was purified from excess sulfo-SMCC by ultrafiltration (Amicon; MWCO = 1 kDa), then reacted at 4 °C overnight with a 100-fold molar excess of hexahistidine (His6)containing peptides (CGYGGHHHHHH, MW = 1.32 kDa; BioBasic) that presented a terminal thiol for reaction with the maleimide group on PmFab. In addition, the peptides incorporated a tyrosine (Y) for radioiodination to enable quantification of the number of His6 conjugated to PmFab (see Supporting Information). The resulting panitumumab Fab-His6 (PmFab-His6) were purified by ultrafiltration (Amicon; MWCO = 30 kDa) and analyzed by SDS-PAGE and Western blot probed with horseradish peroxidase (HRP)-conjugated rabbit anti-His6 antibodies. The conjugation of His6 to Pm-Fab was confirmed by Western blot. Briefly, 10 μg of PmFab or PmFab-His6 were subjected to SDS-PAGE analysis on a 7.5% Tris/Glycine minigel. The protein bands were electrophoretically transferred to an ImmunBlot polyvinylidene fluorine (PVDF) membrane (BioRad). Nonspecific binding sites were blocked by overnight incubation at 4 °C with 50 mM NaCl/20 mM Tris buffer, pH 7.4 with 3% BSA. To reveal His6-containg proteins, the membrane C

DOI: 10.1021/acs.molpharmaceut.9b00422 Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Article

Molecular Pharmaceutics

Statistical analysis. Results were expressed as mean ± SEM and were tested for statistical significance using an unpaired ttest (P < 0.05).

(tail vein) injection of 35−40 MBq (70 μg) 99mTc-PmFab-His6 in the following groups: (i) NOD/SCID mice bearing MDAMB-468 (n = 3) or PANC-1 (n = 4) CLX, (ii) two C.B.-17 SCID mice bearing different TNBC PDX, (iii) four NRG mice bearing different HNSCC PDX, and (iv) NOD/SCID mice bearing OCIP23 PDX (n = 4). Mice were anaesthetized using 2% isoflurane in O2 and were imaged in a supine position. MicroSPECT images were acquired on a Bioscan nanoSPECT/CT (Mediso) equipped with four NaI detectors fitted with 1.4 mm nine-pinhole collimators. Imaging was acquired for 45 min (60 s/projection) at 4 h p.i. and for 90 min at 24 h p.i. (120 s/projection) for a total of 24 projections in a 256 × 256 matrix. Prior to microSPECT, microCT was performed with routine parameters on the Bioscan nanoSPECT/CT (45 kVp, 177 μA, 180 projections) or a GE eXplore Locus Ultra MicroCT (GE Healthcare) (80 kVp, 50 mA, 1000 projections). Reconstruction of microSPECT images was performed with an ordered subset expectation maximization (OSEM) algorithm and isotropic voxel size of 300 μm using InVivoScope 1.43 (Bioscan) software. CT images were reconstructed using the Spatiotemporal Targeting and Amplification of Radiation Response (STTARR) program in-house GPU reconstruction program with an isotropic voxel size of 154 μm in a DICOM format. Co-registration of microSPECT and CT images were performed using VivoQuant 2.5 (inviCRO) software. Mice were then sacrificed under anesthesia and blood, and selected tissue samples were excised, collected, weighed, and transferred to γcounting tubes. Tissue radioactivity was measured in a γcounter, decay-corrected to the time of injection and expressed as percent injected dose/g (%ID/g). PDX were analyzed similarly but excised tumors were also subjected to immunohistochemical (IHC) staining with antihuman EGFR antibodies (Invitrogen) to confirm EGFR positivity. Normal Organ Dosimetry. To project normal organ absorbed radiation doses in humans from 99mTc-PmFab-His6, these doses were estimated based on the uptake and elimination of radioactivity from normal organs in mice. Briefly, 4 groups of 5 nontumor bearing Balb/c mice were injected i.v. with 35 MBq (70 μg) of 99mTc-PmFab-His6 and were sacrificed under anesthesia at 3, 6, 16, and 24 h p.i, respectively. Selected normal source organs were harvested, transferred to γ-counting tubes, and their radioactivity measured in a γ-counter. The radioactivity in the normal organs of human adults was proportionately extrapolated from that in mice using the %kg/g method, i.e., (%ID/organ) human = [(%ID/organ) mouse × (mouse body weight/mouse organ weight) × (human organ weight/ human body weight)].32 The time-integrated radioactivity in the normal source organs from 0 to 24 h (A0−24h) was estimated using Prism Ver. 4.0 software (GraphPad) from the AUC0−24h (Bq × sec) derived from the radioactivity in the normal source organs. Assuming radioactive decay was the only source of elimination of radioactivity from 24 h to infinity, the timeintegrated activity (A24h‑∞) was calculated by dividing the radioactivity at 24 h p.i. by the decay constant for 99mTc (0.115 h−1). The combined A (0−∞h) for each organ was used to predict the radiation-absorbed doses in human adults (mGy/ Bq) using OLINDA/EXM 1.0 software.33 All animal studies were conducted under a protocol (#2843.8) approved by the Animal Care Committee at the University Health Network following Canadian Council on Animal Care (CCAC) guidelines.

■

RESULTS Preparation of PmFab and Conjugation to His6Peptides. SDS-PAGE demonstrated one major band for PmFab under nonreducing conditions with the expected size for Fab (MW ≈ 51 kDa) and one major band under reducing conditions representing the dissociated heavy and light chains (MW ≈ 25 kDa) (Figure 1A). Conjugation of His6 peptides to

Figure 1. (A) SDS-PAGE analysis of PmFab (lane 1) and PmFab-His6 (lane 2) under nonreducing conditions or reducing conditions (lanes 3, 4) on a 7.5% Tris/Glycine mini-gel stained with Coomassie blue. (B) Western blot analysis probed with HRP-conjugated rabbit antihistidine antibodies of PmFab (lane 1) and PmFab-His6 (lane 3) under nonreducing conditions or reducing conditions (lanes 2, 4). A protein ladder of standard molecular weights (MW) is also shown.

PmFab caused an apparent increase in the size of PmFab-His6 immunoconjugates (MW ≈ 53 kDa) under nonreducing conditions and under reducing conditions (MW ≈ 28 kDa) (Figure 1A). This band shift was consistent with the results of a radiometric assay, which incorporated 123I-labeled His6 peptides into the conjugation reaction (see Supporting Information) and revealed that there were 0.80 ± 0.03 His6 peptides (MW = 1.32 kDa each) conjugated per PmFab molecule. Western blot analysis probed with HRP-conjugated rabbit antihistidine antibodies showed a prominent immunopositive band for PmFab-His6 and no immunopositive band for unconjugated PmFab under reducing or nonreducing conditions, respectively (Figure 1B). 99m Tc-Labeling of PmFab-His6 and EGFR Binding Assay. The IsoLink kit converted [99mTcO4]− to [99mTc(H2O)3(CO)3]+ with high RCP (>93%) determined by TLC with autoradiographic measurement of radioactivity on the TLC plate (see Supporting Information, Figure S1). 99mTc-PmFabHis6 was labeled to high RCP (≥95%), and the specific activity was 0.45−0.55 MBq/μg. No postlabeling purification was required. Control, irrelevant Fab 3913 incorporating a His6 affinity tag was labeled to ≥90% RCP and a specific activity of 0.36−0.43 MBq/μg. A saturation binding assay was performed to estimate the dissociation constant (Kd) and the maximum number of binding sites (Bmax) for binding of 99mTc-PmFab-His6 to EGFR on MDA-MB-468 cells. Fitting of the specific binding (SB) curve to a one-site receptor binding model for two independent assays with duplicate samples at each concenD

DOI: 10.1021/acs.molpharmaceut.9b00422 Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Article

Molecular Pharmaceutics tration for each assay revealed a mean Kd = 5.5 ± 0.4 × 10−8 M and Bmax = 5.9 ± 0.2 × 106 EGFR/cell (Figure 2).

Table 1. Tumor and Normal Tissue Uptake of 99mTc-PmFabHis6 and Nonspecific 99mTc-Fab 3913 at 24 h Postinjection in NOD/SCID Mice with s.c. MDA-MB-468 Human Breast Cancer or PANC-1 Human Pancreatic Cancer Xenografts percent injected dose/g (mean ± SEM)a MDA-MB-468 99m

tissue heart lungs stomach pancreas small intestine spleen liver kidneys bone skin muscle blood tumor tumor/ blood ratio

Figure 2. Direct (saturation) binding of 99mTc-PmFab-His6 to EGFR on MDA-MB-468 cells in the absence of (total binding; TB) or presence (nonspecific binding; NSB) of a 50-fold molar excess (2500 nmol/L) of unlabeled PmFab-His6. Specific binding (SB) was calculated by subtracting NSB from TB. Curves were fitted to a onesite receptor binding model using Prism Ver. 4.0 software (GraphPad). Error bars represent the mean ± SD of duplicate samples. In the representative assay shown, Kd = 5.5 ± 0.4 × 10−8 mol/L and Bmax = 5.9 ± 0.2 × 106 EGFR/cell.

Tc-PmFabHis6

2.2 ± 0.05 2.0 ± 0.4 1.4 ± 0.2 0.5 ± 0.05b 0.7 ± 0.3 3.7 ± 0.9 8.4 ± 2.4b 22.2 ± 4.1b 0.9 ± 0.2 2.4 ± 2.2 0.8 ± 0.2 1.3 ± 0.3 14.9 ± 3.1b 11.7 ± 1.4b

PANC-1

99m

99m

99m

Tc-Fab 3913

TcPmFab-His6

Tc-Fab 3913

2.0 ± 0.2 3.0 ± 0.4 1.2 ± 0.4 1.3 ± 0.1 1.9 ± 0.3

1.9 ± 0.2 1.6 ± 0.3 1.4 ± 0.2b 1.3 ± 0.2b 1.1 ± 0.2

1.4 ± 0.4 1.5 ± 1.6 0.4 ± 0.1 0.4 ± 0.1 0.7 ± 0.4

10.0 ± 5.1 14.1 ± 0.9 114.1 ± 12.1 0.70 ± 0.03 1.6 ± 0.4 0.9 ± 0.3 2.6 ± 0.5 3.0 ± 0.9 1.1 ± 0.2

3.4 ± 0.7 9.4 ± 2.7b 36.6 ± 5.7 0.6 ± 0.4 2.2 ± 0.5b 0.9 ± 0.3 1.7 ± 0.2 5.6 ± 0.6b 3.3 ± 0.2b

7.9 ± 3.6 5.9 ± 2.5 88.3 ± 38.5 0.5 ± 0.1 0.6 ± 0.2 0.4 ± 0.1 0.2 ± 0.03 0.5 ± 0.1 0.3 ± 0.07

a

n = 3 for MDA-MB-468 CLX and n = 4 for PANC-1 CLX. Significantly different compared to irrelevant 99mTc-Fab 3913 (P < 0.05).

b

1.4 ± 0.2 vs 0.4 ± 0.1, P < 0.001; 2.2 ± 0.5 vs 0.6 ± 0.2; P < 0.001). For comparison with CLX, the tumor and normal tissue uptakes of 99mTc-PmFab-His6 were also studied in mice implanted with EGFR-positive PDX, established by engrafting tumor specimens from patients with TNBC, PDCa, or HNSCC. In C.B.-17 SCID mice with s.c. TNBC PDX tumor uptake of 99m Tc-PmFab-His6 at 24 h p.i. in two mice engrafted with different specimens were 3.7%ID/g and 11.4%ID/g. One of these mice exhibited similar tumor uptake as in mice engrafted with MDA-MB-468 CLX (11.7 ± 1.4% ID/g; Table 1), while the other mouse exhibited 3-fold lower tumor uptake, demonstrating that uptake varies in PDX tumor models engrafted with different patient specimens. Blood radioactivity in mice with TNBC PDX was 2.1%ID/g and 0.7%ID/g, respectively. These values were similar to those in mice with MDA-MB-468 tumors (1.3 ± 0.3%ID/g; Table 1). T/B ratios in these two mice (1.8 and 17.3, respectively) were widely different due to the differences in tumor uptake. Uptake in the kidneys in mice with TNBC PDX were 21.5%ID/g and 24.0%ID/g, respectively, while liver uptake was 7.0%ID/g and 7.9%ID/g, respectively. These values were similar to those in mice engrafted with MDA-MB-468 tumors (22.2 ± 4.1%ID/g and 8.4 ± 2.4%ID/g, respectively; Table 1). In a group of NOD/SCID mice (n = 4) engrafted orthotopically with the same patient-derived PDCa tumor (OCIP23), uptake of 99mTc-PmFab-His6 in the primary tumor implanted into the pancreas was 4.2 ± 0.3%ID/g, while blood radioactivity was 1.2 ± 0.4%ID/g. Tumor uptake was 1.8-fold significantly lower than in mice engrafted s.c. with PANC-1 human PDCa tumors (5.6 ± 0.6%ID/g; P = 0.006; Table 1), but the T/B ratio for mice with PDX was not significantly different than in mice with CLX (3.9 ± 1.2 vs 3.3 ± 0.2; P = 0.3582) since the blood radioactivity in mice with CLX was slightly higher (1.7

Biodistribution Studies. In NOD/SCID mice with s.c. EGFR-positive MDA-MB-468 human TNBC tumors (n = 3), tumor uptake of 99mTc-PmFab-His6 at 24 h p.i. was 14.9 ± 3.1% ID/g, which was 5-fold significantly higher than control irrelevant 99mTc-Fab 3913 (3.0 ± 0.9; P < 0.05; Table 1). The concentration of radioactivity in the blood in mice injected with 99m Tc-PmFab-His6 was 1.3 ± 0.3%ID/g, which was 2-fold significantly lower than 99mTc-Fab 3913 (2.6 ± 0.5% ID/g; P = 0.017). The tumor/blood (T/B) ratio for 99mTc-PmFab-His6 (11.7 ± 1.4) was 10.4-fold significantly higher than 99mTc-Fab 3913 (1.1 ± 0.2, P = 0.002). Kidney and liver uptake were 22.2 ± 4.1%ID/g and 8.4 ± 2.4%ID/g, respectively, for 99mTc-PmFabHis6. 99mTc-Fab 3913 exhibited 5.1-fold and 1.7-fold higher uptake in the kidneys and liver than 99mTc-PmFab-His6 (114.1 ± 12.1%ID/g, P < 0.001; 14.1 ± 0.9%ID/g, P = 0.019, respectively). Pancreas uptake for 99mTc-PmFab-His6 was 2.6fold significantly lower than 99mTc-Fab 3913 (0.5 ± 0.05 vs 1.3 ± 0.1%ID/g; P < 0.001). In NOD/SCID mice with s.c. EGFR-positive PANC-1 human PDCa (n = 4), tumor uptake of 99mTc-PmFab-His6 at 24 h p.i. was 5.6 ± 0.6%ID/g, which was 11-fold significantly higher than irrelevant 99mTc-Fab 3913 (0.5 ± 0.1%ID/g, P < 0.001; Table 1). Blood radioactivity was 1.7 ± 0.2%ID/g for 99mTc-PmFabHis6 vs 0.15 ± 0.035%ID/g for 99mTc-Fab 3913 (P < 0.0001). T/ B ratios were 3.3 ± 0.2 and 0.3 ± 0.07 for 99mTc-PmFab and 99m Tc-Fab 3913 (P < 0.0001), respectively). Kidney uptake of 99m Tc-PmFab-His6 was 36.6 ± 5.7%ID/g, which was 2.4-fold significantly lower than 99mTc-Fab 3913 (88.3 ± 38.5%ID/g, P = 0.038). Liver uptake of 99mTc-PmFab-His6 (9.4 ± 2.7%ID/g) was 1.6-fold significantly higher than 99mTc-Fab 3913 (5.9 ± 2.5%ID/g, P < 0.0001). Uptake of 99mTc-PmFab-His6 in the pancreas, stomach, and skin were slightly but significantly greater than 99mTc-Fab 3913 (1.3 ± 0.2 vs 0.4 ± 0.1, P < 0.001; E

DOI: 10.1021/acs.molpharmaceut.9b00422 Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Article

Molecular Pharmaceutics ± 0.2%ID/g; Table 1). Locoregional metastases in the peritoneum from OCIP23 tumors implanted in the pancreas exhibited a tumor uptake of 3.1 ± 0.3% ID/g, which was slightly and significantly lower than the primary tumor (P = 0.02). Uptake in the kidneys and liver in mice with OCIP23 tumors were 58.8 ± 13.1%ID/g and 5.5 ± 1.1%ID/g, respectively, which were 1.6-fold and 1.7-fold significantly higher than in mice with CLX (36.6 ± 5.7%ID/g and 9.4 ± 2.7%ID/g, respectively; P = 0.02 and P = 0.04, respectively; Table 1). In a group of four NRG mice bearing HNSCC PDX, each of which was engrafted s.c. with a different patient specimen, tumor uptake ranged from 2.6 to 14.5%ID/g and blood radioactivity ranged from 0.5 to 2.0%ID/g (Table 2), resulting in T/B ratios

24 h p.i. of 99mTc-PmFab-His6 (Figure 3C and G, respectively). Imaging with irrelevant 99mTc-Fab-3913 did not visualize MDAMB-468 or PANC-1 tumors at 4 h (Figure 3B and F, respectively) or at 24 h p.i. (Figure 3D and H, respectively). Blood pool in the mediastinum and large vessels was noted at 4 h p.i. in mice with MDA-MB-468 or PANC-1 xenografts due to circulating 99mTc-PmFab- His6. This blood pool radioactivity decreased substantially at 24 h p.i. Kidneys, liver, and bladder were visualized at 4 h p.i. of 99mTc-PmFab-His6 or 99mTc-Fab3913, but the intensity of these organs decreased slightly at 24 h p.i. Background radioactivity was much lower at 24 h p.i. The kidneys were the most avid normal organ, likely due to renal elimination of the 99mTc-labeled Fab. Primary human TNBC (Figure 4), PDCa (OCIP23; Figure 5), and HNSCC PDX (Figure 6) were imaged at 4 or 24 h p.i. of

Table 2. Tumor and Normal Tissue Uptake of 99mTc-PmFabHis6 at 24 h Postinjection in NRG Mice with s.c. HNSCCa Patient-Derived Xenografts 99m

Tc-PmFab-His6 (%ID/g)

specimen

#459

#391

#128

#191

heart lungs stomach pancreas small intestine spleen liver kidneys bone skin muscle blood tumor tumor/blood ratio

1.8 1.8 1.1 0.8 0.8 3.5 7.2 15.4 0.7 2.0 0.6 1.1 3.1 2.8

0.9 0.6 0.5 0.4 0.4 1.7 4.6 13.1 0.3 0.8 0.3 0.5 2.6 5.2

4.8 3.6 1.4 1.3 1.1 3.2 10.6 26.1 0.9 2.8 2.1 0.9 6.9 7.7

4.3 1.5 2.0 1.8 2.2 3.8 16.0 34.0 1.4 2.8 1.5 2.0 14.5 7.3

Figure 4. Posterior whole-body microSPECT/CT images at 4 h p.i. of 99m Tc-PmFab-His6 in mice bearing two different human TNBC PDX (A) #941 and (B) #196 and the corresponding images at 24 h p.i. (C,D). The location of the tumors is indicated by the white arrow. Also visualized are the kidneys (green arrowheads) and liver (white arrowhead). The heart is visualized at 4 h p.i. (blue arrowhead) but is not visible at 24 h p.i. Some interstitially injected radioactivity is seen in A and C (white circle). Immunohistochemical (IHC) staining of both tumors ex vivo demonstrated strong EGFR positivity (E).

a

n = 1 for each HNSCC specimen. 99m

Tc-PmFab-His6 by microSPECT/CT. All PDX tumors were strongly positive for EGFR expression by IHC (Figures 4E, 5E, and 6E). OCIP23 tumors metastasize locally in the peritoneal cavity in mice, thus uptake of 99mTc-PmFab-His6 was visualized as diffuse radioactivity in the subphrenic region, due to most tumor nodules having a size less than the spatial resolution of the imaging system (