We used 8-week-old female nude (nu/nu) mice and BALB/c mice housed in the animal facilities of The University of Texas M.D. Anderson Cancer Center or the Ludwig Institute for Cancer Research. All animal procedures were reviewed and approved by the Institutional Animal Care and Use Committees of these institutions.
The following antibodies were used: goat polyclonal anti-TfR IgG (R&D Systems), mouse monoclonal anti-TfR IgG (Fisher Scientific), rabbit polyclonal anti-bacteriophage (Sigma-Aldrich), rat monoclonal anti-CD31 (BD Pharmingen), and rabbit polyclonal anti-CD31 (AbCam). Secondary HRP-conjugated anti-rabbit IgG, Cy-3–conjugated anti-rabbit IgG, Cy-5–conjugated anti-hamster IgG, and FITC-conjugated anti-rat IgG were purchased from Jackson ImmunoResearch Laboratories Inc. Human TfR (>95% purity) and FragEL were purchased from Calbiochem, and human apo (>90% purity) and holo-Tf (>90% purity) were purchased from R&D Systems. All synthetic peptides were synthesized and cyclized according to our specifications by commercial vendors (AnaSpec or PolyPeptide Laboratories).
U87-MG human glioblastoma cells (ATCC), rat gliosarcoma cells (9L) (31
), and rat gliosarcoma cells expressing human TfR (9L 3.9) (31
) were maintained in DMEM supplemented with 10% FBS and antibiotics.
Phage display selection and phage homing in vivo.
In vivo phage screening was performed as described previously (26
). Animals received 1010
TU phage library i.v., and brains were collected after 24 hours of circulation. Bound phage were recovered by bacterial infection and were processed as described previously (26
). For homing of selected phage in vivo, animals (n
= 5, each group) were deeply anesthetized and received 1010
TU targeted phage or insertless control phage i.v. After 10 minutes, 30 minutes, 6 hours, or 24 hours, animals were perfused through the heart with 10 ml DMEM.
Isolation of brain blood vessels.
Fractionation of brain blood vessels was performed as described previously (29
). Mice aged 3–5 weeks were deeply anesthetized and perfused through the heart with 10 ml PBS. Brain tissue homogenate was diluted to a final volume of 5 ml, transferred to the top of a 30% (w/v) dextran solution, and centrifuged for 25 minutes at 3,000 g
(4°C). The myelin layer and pellet containing blood vessels were carefully collected. The final fractions were resuspended in an adequate volume and used for DNA or RNA extraction.
DNA and RNA isolation from tissue and quantitative real-time PCR.
Total DNA was isolated according to manufacturer’s instructions (DNeasy Blood and Tissue Kit, Qiagen). The amount of phage present in each tissue sample was determined by quantitative PCR (7500 Fast Real-Time PCR System instrument, Applied Biosystems), as described previously (49
). Each point of the curve, as well as each tissue sample of DNA, was determined in triplicate. For gene expression studies, 2 sets of total RNA were independently isolated from microvessel and parenchyma fractions with the RNeasy Mini Kit (Qiagen). DNA synthesis was performed with the SuperScript III First-Strand Synthesis System (Invitrogen). The following primer pairs were used: slc2a1 (glut-1), 5′ TTCTCTGTCGGCCTCTTTGT 3′ and 5′AGGCCAACAGGTTCATCATC 3′; VE-cadherin, 5′ TCATCAAACCCACGAAGTCC 3′ and 5′ TGTTTTTGCCTGAAGTGCTG 3′; Occludin: 5′ CCTACTCCTCCAATGGCAAA 3′ and 5′ GGCACCAGAGGTGTTGACTT 3′.
The construction of the human glioma TMA was previously described (50
). The TMA included primary human gliomas, with all histologic subtypes and grades of diffuse gliomas as codified in the WHO classification (51
), including GBM, AA, AO, O, AMOA, MOA, and GS. Low-grade astrocytomas were not analyzed here, because of their small sample size (n
= 4). Pathology scoring was performed by 2 independent observers (F.I. Staquicini and G.N. Fuller).
Immunohistochemistry on sections of fixed, paraffin-embedded mouse tissue was performed with a labeled streptavidin biotin (LSAB) plus peroxidase kit (DAKO). Slides were blocked for nonspecific protein binding, and a polyclonal rabbit anti-bacteriophage primary antibody was added (1:500 dilution), followed by 1-hour incubation with a peroxidase-conjugated anti-rabbit secondary antibody. Slides were developed with a specific substrate (chromogen DAB). Immunohistochemistry on glioma TMAs and CD31 staining were performed on an automated immunohistochemical autostainer (Lab Vision Corp.). Detection of apoptosis in paraffin-embedded specimens was performed with the FragEL DNA Fragmentation Detection Kit (Calbiochem). Hematoxylin was used for counterstaining.
Immunofluorescence was performed on 1% PFA-fixed cryostat sections (60- to 80-μm thickness). Tissues were washed 3 times with PBS and once with PBS containing 0.3% Triton X-100, followed by blocking for 1 hour in 5% appropriate normal serum diluted in PBS containing 0.3% Triton X-100. Tissue sections were incubated with specific antibodies diluted in PBS containing 1% appropriate normal serum and 0.3% Triton X-100 for 1 hour at room temperature. Sections were stained for 1 hour with Cy3-conjugated and FITC-conjugated secondary antibodies. Confocal images were acquired on a laser scanning confocal microscope (Zeiss LSM510) equipped with krypton-argon and helium-neon lasers. Image analysis was performed with the Zeiss LSM 3.2 software package.
In vitro phage binding assays.
TfR, apo-Tf, holo-Tf, and BSA were immobilized on microtiter wells of 96-well plates overnight (ON) at 4°C. Wells were blocked with PBS containing 3% BSA for 1 hour at room temperature and incubated with 109 TU phage. Bound phage were recovered by infection of host bacteria with 200 μl E. coli K91 Kan in log phase. To test phage binding inhibition by free iron, we used an initial solution of 10 mM iron (III) phosphate (FePO4) (Sigma-Aldrich) in double-distilled water. Final concentrations of iron ranged from 0.01 mM to 0.5 mM.
Phage binding assays on cells.
For cell-phage binding assays, 106
cells were incubated with 109
TU phage for 2 hours on ice in the presence of either apo- or holo-Tf (200 μg/ml) in 200 μl DMEM. Cells and phage were centrifuged through the organic phase, and the cell-bound phage were recovered by bacterial infection of host E. coli
as described previously (52
Steady-state fluorescence spectroscopy.
Steady-state tryptophan fluorescence spectra were obtained on a Spectrofluorimeter LS 50B (Perkin-Elmer). Increasing molar ratios of Fe+3, the targeted peptide CRTIGPSVC, or control peptides (control peptide 1, CGLPYSSVC; control peptide 2, CSGIGSGGC; control peptide 3, CRFESSGGC; and control peptide 4, CPQRGVTPC) were incubated ON at room temperature with 2 μM apo-Tf in buffer containing 100 mM HEPES, 10 mM NH4CO3, pH 7.4. Samples were excited at 295 nm, and emission scans were collected from 305 to 400 nm, with an excitation slit of 2.5 nm and an emission slit of 6 nm. All spectra were corrected for background fluorescence by subtraction of the appropriate blanks.
CD spectra were recorded on a Jasco J720 spectropolarimeter from 230 to 320 nm, with a bandwidth of 1 nm, and integrated for 1 second at 0.2-nm intervals. Samples were measured at room temperature using cuvettes with a 2-mm path length. Absorbance values of Fe+3 and peptide alone were subtracted from the results. Increasing concentrations of Fe+3, the targeted peptide CRTIGPSVC, or control peptides (control peptide 1, CGLPYSSVC; control peptide 2, CSGIGSGGC; control peptide 3, CRFESSGGC; and control peptide 4, CPQRGVTPC) were titrated in 100 μM apo-Tf in buffer containing 100 mM HEPES, 10 mM NH4CO3, pH 7.4. The data were expressed as mean residue ellipticity (θ) in deg cm2/dmol, which was calculated from (θ) = (d × s × M)/(c × l), where d denotes observed ellipticity (the displacement in cm from the baseline), s denotes sensitivity in mdeg/cm, M denotes the mean residue weight, c denotes protein concentration in mg/ml, and l denotes cell path length. The concentrations of protein were kept at 8 mg/ml.
Orthotopic human glioblastoma xenografts.
We used a guide-screw system to implant human glioma cells into the mouse brain (41
). Animals were kept warm until their recovery from anesthesia and were allowed to move around freely thereafter. In vivo homing experiments with targeted and control phage were performed approximately 12–15 days after tumor implantation.
Subcutaneous human glioblastoma xenografts.
A solution containing 3 × 106 U87-MG cells was subcutaneously injected into the right flank of nude mice (n = 5, each group). Tumor homing experiments were performed when palpable subcutaneous tumor xenografts reached 6–8 mm in diameter.
Targeted therapy and molecular-genetic imaging.
Orthotopic brain tumor-bearing animals received a single dose (1011 TU i.v. per mouse) of CRTIGPSVC AAVP–HSV-TK or control, approximately 7 days after tumor implantation. Treatment with GCV (80 mg/kg/d i.p.) was initiated 7 days after AAVP administration i.v. To image HSV-TK expression, PET and CT scans were performed 2 hours after i.v. administration of the radiolabeled nucleoside analog [18F]-FEAU. A microPET R4 (Concorde Microsystems), equipped with a computer-controlled positioning bed in a 10.8-cm transaxial and 8-cm axial field of view with no septa and operating in 3D list mode, was used. PET/CT imaging was performed with an Inveon micro-PET/CT scanner (Siemens Preclinical Solution). Glioma-bearing mice were anesthetized (with isoflurane 2% in 98% oxygen), and their temperature was kept at 38°C with a heat lamp. The microCT imaging parameters were as follows: x-ray voltage of 80 kVp, anode current of 500 μA, and exposure time of 300–350 milliseconds for each of the 360 rotational steps. Images were reconstructed by a 2D ordered subsets expectation maximization algorithm. PET and CT image fusion and image analysis were performed with vendor software ASIPro 22.214.171.124 (Siemens Preclinical Solution).
Radiolabeled substrate synthesis.
F]-FEAU was synthesized to radiochemical purity of more than 99% by using 5-ethyluracil-2,5-bis-trimethylsilyl ether as the pyrimidine base for condensation with 1-bromo-2-deoxy-2-[18F]fluoro-3,5-di-O
-benzoyl-α-D-arabinofuranose. For quantification of [18
F]-FEAU radioactivity, regions of interest were drawn on images, and the measured values were converted from nCi/mm3
into percentage of injected dose per gram of tissue (% ID/g) (22
All numerical data are expressed as mean ± SEM. We analyzed data sets for significance with Student’s t test and 2-way ANOVA. We considered P values of less than 0.05 as statistically significant.