Abstract
As apoptosis occurs over an interval of time after administration of apoptosis-inducing therapy in tumors, the changes in technetium 99m (99mTc)-tricarbonyl (CO)3 His-annexin A5 (His-ann A5) accumulation over time were examined. Colo205-bearing mice were divided into six treatment groups: (1) control, (2) 5-fluorouracil (5-FU; 250 mg/kg), (3) irinotecan (100 mg/kg), (4) oxaliplatin (30 mg/kg), (5) bevacizumab (5 mg/kg), and (6) panitumumab (6 mg/kg). 99mTc-(CO)3 His-ann A5 was injected 4, 8, 12, 24, and 48 hours posttreatment, and micro–single-photon emission computed tomography was performed. Immunostaining of caspase-3 (apoptosis), survivin (antiapoptosis), and LC3-II (autophagy marker) was also performed. Different dynamics of 99mTc-(CO)3 His-ann A5 uptake were observed in this colorectal cancer xenograft model, in response to a single dose of three different chemotherapeutics (5-FU, irinotecan, and oxaliplatin). Bevacizumab-treated mice showed no increased uptake of the radiotracer, and a peak of 99mTc-(CO)3 His-ann A5 uptake in panitumumab-treated mice was observed 24 hours posttreatment, as confirmed by caspase-3 immunostaining. For irinotecan-, oxaliplatin-, and bevacizumab-treated tumors, a significant correlation was established between the radiotracer uptake and caspase-3 immunostaining (r = .8, p < .05; r = .9, p < .001; r = .9, p < .001, respectively). For 5-FU- and panitumumabtreated mice, the correlation coefficients were r = .7 (p = .18) and r = .7 (p = .19), respectively. Optimal timing of annexin A5 imaging after the start of different treatments in the Colo205 model was determined.
APOPTOSIS, or programmed cell death, plays a crucial role in the development of cancer as well as in successful treatment of cancer patients.1,2 After caspase-3 activation, there is a rapid redistribution and exposure of the anionic phospholipid phosphatidylserine (PS) on the apoptotic cell membrane, which is normally restricted to the inner leaflet of the cell membrane. This protein offers an ideal target for apoptosis detection because exposure of PS is a universal event in apoptosis, is independent of the apoptosis-inducing trigger, occurs within hours after apoptotic stimuli, and represents an abundant target on the extracellular side of the cell membrane. 3
Annexin A5, a human protein, has a selective and high affinity (Kd = 7 nM) for membrane-bound PS and has already been radiolabeled for in vivo apoptosis detection. 4 However, during the necrotic process, PS binding sites become accessible for annexin A5 and other types of cell death are also accompanied by externalization of PS, including autophagy, anoikis, and mitotic catastrophe.5,6 Consequently, annexin A5–based probes do not exclusively detect apoptotic events. Radiolabeled annexin A5 has been proven to provide early indication of tumor response in cancer patients.7,8 Importantly, as apoptosis is characterized by peaks of apoptotic cells (with concurrent peaks in PS expression), which then become rapidly engulfed by macrophages and neighboring cells, knowledge of the optimal imaging time is a requisite.
The current study was performed to determine the dynamics of apoptosis in a colorectal tumor model in response to 5-fluorouracil (5-FU); irinotecan; oxaliplatin; a monoclonal antibody (mAB) against vascular endothelial growth factor (VEGF), bevacizumab; and an mAB against epithelial growth factor receptor (EGFR), panitumumab. Technetium 99m (99mTc)-tricarbonyl (CO)3 His-annexin A5 (His-ann A5) was administered 4, 8, 12, 24, and 48 hours after the start of different treatments. The value of this radiotracer was previously evaluated, by our research group, in vitro and in vivo. His-tagged annexin A5 was successfully labeled with 99mTc-(CO)3, with yields of approximately 70 to 85% and high radiochemical purities (> 95%). Specific radioactivities of approximately 5,180 to 5,920 MBq/µg protein were obtained for 99mTc-(CO)3 His-ann A5. Stability tests confirmed good thermodynamic stability of the radiotracer, and in vitro binding characteristics proved that 99mTc-(CO)3 His-ann A5 retained its PS binding capacity, which was confirmed by annexin A5–fluorescein isothiocyanate flow cytometry. 9
Changes in 99mTc-(CO)3 His-ann A5 accumulation in the tumor over time were compared to changes in the caspase-3 levels in the tumor. Expression levels of the inhibitor of apoptosis protein (IAP) survivin and the auto-phagocytic marker of microtubule-associated protein 1 light chain 3-II (LC3-II) were also investigated.
Materials and Methods
Colo205 Cell Culture
Colo205 cells, a semiadherent human colorectal adenocarcinoma cell line (product number ATCC-CCL-222, American Type Culture Collection [ATCC], Manassas, VA), were grown in RPMI 1640 medium enriched with 10% fetal bovine serum (Invitrogen, Merelbeke, Belgium) and supplemented with antibiotic-antimycotic (100×, Invitrogen) at 37°C and 5% CO2, according to ATCC recommendations. Cells were kept in exponential phase by routine passage every 2 to 3 days (split ratio: 1/2–1/6).
Human Colorectal Cancer Xenograft Model
Female CD-1 nude mice (strain code 086), obtained from Charles River Inc (Wilmington, MA), were purchased at the age of up to 42 days and were treated in accordance with Belgian regulations. Animal experiments were approved by the local ethics committee. Animals were housed in individually ventilated cage conditions under a 12-hour dark/light cycle at constant humidity and temperature. For the Colo205-bearing mouse model, Colo205 cells in the exponential phase were harvested by trypsinization, washed three times with sterile phosphate-buffered saline (PBS), and resuspended in sterile PBS at a concentration of 1 × 107cells/mL. Cells were counted using the trypan blue method. Nude mice were injected subcutaneously in the right hind leg with 1 × 106 Colo205 cells in a volume of 100 µL. Owing to our study design, the use of different animals at the different time points (no serial imaging in the same animal), longitudinal tumor growth curves cannot be obtained. However, tumor volumes were measured using a Vernier caliper, and tumor volume was calculated as V = ab2/2, where a and b are the long and short axis of the tumor, respectively. Tumor volumes were measured at baseline (at the start of the experiment) and at the different time points after the start of treatment (4, 8, 12, 24, or 48 hours after drug administration).
After 12 to 14 days of inoculation, when the tumors reached a volume of approximately 500 to 1,000 mm3, animals were randomized into six different groups, receiving a single-dose treatment (intraperitoneally): group 1, control, 0.9% NaCl; group 2, 5-FU, 250 mg/kg; group 3, irinotecan, 100 mg/kg; group 4, oxaliplatin, 30 mg/kg; group 5, bevacizumab, 5 mg/kg; and group 6, panitumumab, 6 mg/kg. (Chemo)therapeutics were diluted in saline to the appropriate volume of 100 µL.
99mTc-(CO)3 His-Ann A5 Preparation
Briefly, 1 mL of freshly eluted 99mTcO4− (740–3,700 MBq) was added to the IsoLink vial (Mallinckrodt Medical B.V., Petten, the Netherlands) containing the following lyophilized products: 8.5 mg sodium tartrate, 2.85 mg Na2B4O7.10H2O, 7.15 mg sodium carbonate, and 4.5 mg sodium boranocarbonate. The vial was placed in a boiling water bath for 25 minutes. Four hundred microliters of the precursor was brought at pH 7 by adding stepwise 1 M HCl, and pH was checked on pH paper (Vel, Leuven, Belgium). His-ann A5 (50–60 µg, kindly provided by Prof. Dr. Reutelingsperger, Cardiovascular Research Institute, Maastricht, the Netherlands) was added to the precursor and incubated at 37°C for 1 to 1.5 hours while blown dry with nitrogen gas to reduce total volume. The vial was washed with 0.9% NaCl, and the reaction mixture was purified on an activated Sephadex PD-10 column (GE Healthcare Bio-Sciences AB, Uppsala, Sweden) with PBS/0.5% bovine serum albumin (Sigma-Aldrich-Fluka, Bornem, Belgium) and eluted with fractions of 500 µL PBS. The radioactivity of each fraction was measured in a dose calibrator (VIK-202, Veenstra, AC Joure, the Netherlands). The radiochemical purity of the 99mTc-(CO)3 His-ann A5 was determined by ascending instant thin-layer chromatography with silica gel–coated fiberglass sheets (Life Sciences, Pall Corporation, Leuven, Belgium) using physiologic saline as the mobile phase. Five microliters of the samples was spotted on instant thin-layer chromatography strips, and radioactivity was counted using an automated gamma counter (Cobra II series, Canberra Packard, Meriden, CT).
MicroSPECT Imaging of Cell Death in Colo205-Bearing Mice
For in vivo imaging of apoptosis, Colo205-bearing mice were injected via a tail vein with 18.5 MBq 99mTc-(CO)3 His-ann A5 (in approximately 100 µL) 4, 8, 12, 24, and 48 hours after the start of treatment. Single-photon emission computed tomographic (SPECT) imaging at different time intervals was performed in different animals (n = 3 at each time point). Three and a half hours after injection of the radiotracer, mice were positioned in the scanner and were maintained under isoflurane anesthesia for the duration of the experimental procedure while the body temperature was kept constant by a heating bed. Our previous in vivo biodistribution study of 99mTc-(CO)3 His-ann A5 revealed that 3.5 hours after injection of 18.5 MBq 99mTc-(CO)3 His-ann A5 would be an optimal timing for this radiotracer to allow sufficient blood pool clearance and optimal tumor imaging. 9 Static high-resolution tomographic images were acquired in four frames of 5 minutes with a focus on the tumor using the Milabs U-SPECT-II (Milabs, Utrecht, the Netherlands). This microSPECT scanner is equipped with collimators consisting of a tungsten cylinder with five rings of 15 pinhole apertures of 1.0 mm diameter delivering reconstructed images with an iteratively reconstructed resolution better as 1 mm. All pinholes focused on a single volume in the center of the tube. For this study, the animal bed was translated in three dimensions using an XYZ stage into two different bed positions focused on the tumor. A 20% main photopeak was centered at 140 keV to reconstruct the 99mTc images on 0.75 mm3 voxels by six iterations of 16 OSEM (ordered subsets expectation maximization) subsets. A syringe with a known amount of radioactivity was scanned along with the mice to allow semiquantification of the results. Region of interest (ROI) analysis and quantification of 99mTc-(CO)3 His-ann A5 accumulation were performed in PMod software (PMOD Technologies, Adliswil, Switzerland). All tumors showed visible radiotracer uptake, and ROI were drawn manually around the tumor on all consecutive slices encompassing the tumor. When defining the tumor ROI, particular care was taken to avoid inclusion of any erroneous signal originating from the bladder. For the purpose of quantification, tumor lesion counts in tomographic slices were summed, thus incorporating the entire dimension of the tumor. Percentage uptake was obtained using the following equation: (total lesion radioactivity counts at time of SPECT × radioactivity of standard at time of SPECT scan/standard counts at time of SPECT × total radioactivity injected) × 100% and normalized to tumor weight (g) and body weight of the mice (kg). Directly after the scan, the animals were killed by cervical dislocation, tumors were removed, formalin was fixed, and paraffin was embedded.
Immunohistochemistry of Caspase-3 (Apoptosis), Survivin (Antiapoptosis), and LC3-II (Autophagy)
Sections 4 µm thick were mounted on SuperFrost microscope slides (Menzel-Glaser, Braunschweig, Germany), which were deparaffinized in xylene and rehydrated in a downgraded series of ethanol. Heat-induced antigen retrieval was performed for 10 minutes in citrate buffer (pH 6), and endogenous peroxidase activity was blocked for 5 minutes with 0.3% hydrogen peroxide (Dako, Glostrup, Denmark) on each tissue slide. Each section was blocked with blocking solution (Tris-Buffered saline 0.1% Tween [TBST]/5% normal goat serum) for 1 hour at room temperature. Primary antibody cleaved caspase-3 (Cell Signaling Technology, Leiden, the Netherlands, #9661, 1/400 dilution in PBST [phosphate-buffered saline Tween 0.1%]/5% normal goat serum [NGS]), survivin (Cell Signaling Technology, #2808, 1/400 dilution in PBST/5% NGS), or LC3-II (Novus Biologicals, Littleton, CO, NB100-2331, 1/400 dilution in PBST/5% NGS) was then incubated overnight at 4°C (caspase-3 and survivin) or for 1 hour at room temperature (LC3-II). The tissue sections were incubated for 30 minutes at room temperature with a labeled polymer–horseradish peroxidase antirabbit secondary antibody (Dako). The color reaction was developed using the chromogen 3, 3-diaminobenzidine+ (DAB) (Dako) for 10 minutes. The tissue sections were counterstained with Mayer hematoxylin. TBST/5% NGS or isotype control instead of the primary antibody was used as the negative control. The number of caspase-3-positive (apoptotic) cells present in each tumor specimen was expressed as a fraction of the total number of cells. The apoptotic index was determined by an Optronicscolor digital camera (Olympus Corporation, Tokyo, Japan) and specialized software (Cell D Olympus Imaging Solutions, Münster, Germany). Ten ROI were chosen at random at a magnification of 200×. High necrotic regions were excluded for analysis. For survivin, nuclear and cytoplasmatic localization was analyzed. For LC3-II staining, only membranous staining was scored. The expression levels of survivin and LC3-II were assessed semiquantitatively by the product of scores estimated (intensity × percentage positively stained). The intensity of staining was graded from 0 to 3 (0 = no staining, 1 = weak, 2 = moderate, 3 = intense) and percentage of positively stained cells ranging from 1 to 5 (1 = 0–20%, 2 = 21–40%, 3 = 41–60%, 4 = 61–80%, 5 = 81–100%). The mean values of these levels were determined as immunohistologic survivin and LC3-II expression levels. Scores were obtained by two independent researchers in a blind manner.
Statistical Analysis
Statistical analysis was performed using SPSS 15.0 software (SPSS Inc, Chicago, IL). All data are expressed as mean ± SD. Power analysis was based on the following assumptions: a normal distribution of baseline (known from previous studies) and posttherapy annexin A5 uptake in tumors, a comparable standard deviation for annexin A5 uptake for all treatment groups, one-way analysis of variance (ANOVA) test (I factor crossed), five levels (five groups), a difference between minimum and maximum uptake of annexin A5 of 60% (based on available data in the literature 10 ) following therapy in animal models (which is the lowest level reported) of differences (going up to 200%), and a standard deviation of 10%, respectively, yielding a delta or effect size of 6.0 and a significance level of .05. Power analysis using these assumptions will show that using a sample size of three/group will yield a power of 0.998, which is far above the 0.8 standard used. As we could not confirm normal distribution of our data, the less powerful Kruskal-Wallis test for statistical analysis was used. Nonparametric Mann-Whitney testing was performed to assess the significance of the data obtained for both therapy-treated and control data, followed by the Bonferroni test. The Spearman rank correlation r was used to calculate the correlations between uptake of 99mTc-(CO)3 His-ann A5 and caspase-3-positive cells. Two-sided p values < .05 level were considered statistically significant.
Results
Tumor Volumes and Tumor Uptake of 99mTc-(CO)3 His-Ann A5
Tumor volumes were measured using a caliper at the start of the therapy (baseline) and at the different time points (4, 8, 12, 24, or 48 hours after the start of therapy). No significant difference in tumor volumes, in the different time groups, before or after therapy, was observed (Figure 1).
Tumor volumes (mm3), measured at baseline (before therapy) and after therapy at the different time points (4, 8, 12, 24, and 48 hours). No significant differences in tumor volumes between the different time groups exist (Kruskal-Wallis test, n = 18).
The accumulation of 99mTc-(CO)3 His-ann A5 in the tumors 4, 8, 12, 24, and 48 hours after a single dose of therapy over controls is shown in Figure 2. For 5-FU (250 mg/kg), a first peak of 153 ± 14% ID/g/kg over controls was observed 8 hours posttreatment. The increase remained at 12 and 24 hours posttreatment (129 ± 13 and 144 ± 44% ID/g/kg over controls, respectively). At 48 hours posttreatment, 99mTc-(CO)3 His-ann A5 uptake dropped to baseline levels (see Figure 2A).
Time course of 99mTc-(CO)3 His-ann A5 uptake (%ID/g/kg) and caspase-3-positive cells (%), in tumors of Colo205-bearing mice, normalized to controls. Colo205-bearing mice received a single dose (intraperitoneally) of 5-FU (A), irinotecan (B), oxaliplatin (C), bevacizumab (D), or panitumumab (E) (n = 3 at each time point). *p < .05, **p < .01. Error bars represent SD.
For the irinotecan- (100 mg/kg) treated mice (see Figure 2B), a large peak (163 ± 13% ID/g/kg) 24 hours posttreatment was observed. At 48 hours posttreatment, 99mTc-(CO)3 His-ann A5 uptake in the tumor was still slightly elevated (126 ± 34% ID/g/kg). Oxaliplatin-treated mice (30 mg/kg) showed a gradual increase in 99mTc-(CO)3 His-ann A5 uptake in the tumors (see Figure 2C), with a large peak (179 ± 25% ID/g/kg) 24 hours posttreatment and a sustained elevated increase 48 hours posttreatment (132 ± 16% ID/g/kg).
Animals receiving a single dose of the VEGF mAB bevacizumab (5 mg/kg) did not show any increase in 99mTc-(CO)3 His-ann A5 uptake in the tumors at any time point (see Figure 2D).
Colo205-bearing mice injected with the EGFR mAB panitumumab (5 mg/kg) showed an increase in accumulation of the radiotracer in the tumor (see Figure 2E) 4, 12, and 24 hours posttreatment (112 ± 30%, 123 ± 52%. and 162 ± 19% ID/g/kg, respectively). Representative SPECT images of control and treated tumors are shown in Figure 3.
MicroSPECT images of Colo205 tumors 3.5 hours after injection of 18.5 MBq 99mTc-(CO)3 His-ann A5 in the tail vein. Mice received (A) 0.9% NaCl (control) 4 hours before administration of the radiotracer, (B) 5-FU 4 hours before administration of the radiotracer, (C) oxaliplatin 12 hours before administration of the radiotracer, and (D) irinotecan 24 hours before administration of the radiotracer. Representative sagittal slices demonstrate accumulation of 99mTc-(CO)3 His-ann A5 in the tumors of untreated (control) and treated mice.
Immunostaining of Caspase-3 in Tumors
Caspase-3-positive cells, representing apoptotic cells, were observed in the tumor specimens obtained from both the control mice and the treated mice (Figure 4). The time course of caspase-3-positive cells in treated tumors over controls is shown in Figure 2. For the 5-FU-treated tumors, the time course of caspase-3-positive cells is similar to that of 99mTc-(CO)3 His-ann A5 uptake in the tumors; however, 48 hours posttreatment, levels of caspase-3 were still elevated, whereas uptake of the radiotracer was dropped to baseline levels (see Figure 2A). Irinotecan-treated tumors showed a significant increase 8 hours posttreatment (120 ± 49%) and a similar increase (326 ± 102%), compared to the radiotracer uptake, 24 hours posttreatment, which remained elevated 48 hours posttreatment (303 ± 98%), whereas uptake of the 99mTc-(CO)3 His-ann A5 at this time point was decreased to control level (see Figure 2B). Oxaliplatin-treated mice showed a similar pattern compared to the radiotracer uptake, with a gradual increase in caspase-3-positive cells in the tumors (see Figure 2C) with a large increase 24 hours posttreatment (391 ± 18%, respectively) and a sustained elevated increase 48 hours posttreatment (179 ± 30%). Bevacizumab-treated mice did not show any increase in 99mTc-(CO)3 His-ann A5 uptake in the tumors at any time point; however, a small increase in caspase-3 cells was observed 4 hours posttreatment (113 ± 14%) and a large increase 24 hours posttreatment, although neither was significant (201 ± 28%; see Figure 2D). For panitumumab-treated mice, a significant increase in caspase-3-positive cells 4 hours posttreatment was observed (167 ± 77%) and a later, not significant one 24 hours posttreatment (140 ± 28%). However, with 99mTc-(CO)3 His-ann A5, two peaks were observed: a small peak 12 hours posttreatment and a larger peak 24 hours posttreatment (see Figure 2E).
Cleaved caspase-3 immunostaining (× 400 original magnification) of tumor specimens in (A) control group (12 hours), (B) 5-FU-treated (8 hours), (C) irinotecantreated (24 hours), (D) oxaliplatin-treated (24 hours), (E) bevacizumab-treated (24 hours), and (F) panitumumab-treated (12 hours) mice. Cells stained brown were positive for caspase-3 immunostaining.
Correlation of Caspase-3 with 99mTc-(CO)3 His-Ann A5 Tumor Uptake
The tumor uptake of 99mTc-(CO)3 His-ann A5 was compared to the percentage of apoptotic cells from the caspase-3 immunostaining to evaluate the correlation between the two modalities (Figure 5). For irinotecan-, oxaliplatin-, and bevacizumab-treated tumors, a significant correlation was established (r = .8, p < .05; r = .9, p < .001; r = .9, p < .001, respectively). For 5-FU-treated and panitumumab-treated mice, the correlation coefficients observed between 99mTc-(CO)3 His-ann A5 tumor uptake and percentage of caspase-3 cells were r = .7 (p = .18) and r = .7 (p = .19), respectively.
Correlation of uptake of 99mTc-(CO)3 His-ann A5 and number of positive caspase-3 cells in the tumor. Regression analysis demonstrated a significant correlation between 99mTc-(CO)3 His-ann A5 and caspase-3-positive cells in the (B) irinotecan-, (C) oxaliplatin-, and (D) bevacizumab-treated groups (r = .8, p < .05; r = .9, p < .001; r = .9, p < .001, respectively). For the (A) 5-FU- and (E) panitumumabtreated tumors, correlation coefficients were r = .70 (p = .18) and r = .70 (p = .19), respectively.
Expression Pattern of the Antiapoptotic Protein Survivin in Tumors
Immunohistochemical analysis revealed that all control Colo205 tumors expressed high levels of survivin (mean = 12.9 ± 0.5). In both control and treated tumors, survivin was predominantly localized in the nucleus (Figure 6). All survivin scores of treated tumors were lower than that of control tumors, suggesting decreased expression of survivin in response to the different treatments (Figure 7, Table 1). For 5-FU-treated tumors, no significant decrease in survivin expression was observed. For the irinotecantreated tumors, a significant decrease in survivin was observed at 4 and 8 hours (p = .0007, p = .0469, respectively). For oxaliplatin-treated tumors, a significant decrease was observed 4, 12, 24, and 48 hours posttreatment (p = .0007, p = .0076, p = .0157, and p = .0307, respectively). For the bevacizumab-treated tumors, a significant decrease in survivin was observed at 4 and 12 hours (p = .05 and p = .03, respectively), whereas for the panitumumab-treated tumors, significant decrease was observed at 4 and 24 hours (p = .05 and p = .0156, respectively). IAP members, including survivin, block apoptosis by inhibiting caspase-3 and −7. Therefore, the correlation between survivin and caspase-3 was analyzed, but no significant correlation was confirmed for any of the treatment groups.
Immunohistochemical scores for survivin staining in the control and 5-FU-, irinotecan-, oxaliplatin-, bevacizumab-, and panitumumabtreated tumors.
Immunostaining of survivin (×400 original magnification) of tumor specimens in (A) the control group (24 hours), (B) 5-FU-treated (8 hours), (C) irinotecan-treated (24 hours), (D) oxaliplatin-treated (4 hours), (E) bevacizumab-treated (12 hours), and (F) panitumumabtreated (4 hours) mice. Cells stained brown were positive for survivin immunostaining.
Immunohistochemical Scores for Survivin and LC3-II Staining
5-FU 5 5-fluorouracil; I = intensity of staining; LC3-II = light chain 3-II; P = percentage of positively stained cells.
Data are expressed as mean ± SD (n = 3).
p < .05.
Expression of the LC3-II Protein (Autophagocytic Marker) in Tumors
The membrane-bound LC3-II expression level is regarded as a marker for the formation of autophagosomes. The control and bevacizumab- and panitumuma-btreated tumors were analyzed through immunohistochemical analysis with the LC3-II antibody to investigate if autophagy is activated in colorectal cancer (CRC) and whether (chemo)therapy alters the activation of autophagy (Figure 8). Overall, the Colo205 tumors (untreated-treated) showed high expression of LC3-II.
Immunostaining of LC3-II (×400 original magnification) of tumor specimens in the (A) control group (24 hours), (B) bevacizumab-treated (4 hours), and (C) panitumumabtreated (48 hours) mice. Cells stained brown were positive for LC3-II immunostaining.
An increase, although not significant (p = .06), in LC3-II expression was observed 24 hours following bevacizumab treatment. In panitumumabtreated mice, an increase (p = .05) in LC3-II expression was demonstrated 4 hours posttreatment (Figure 9, see Table 1).
Immunohistochemical scores for LC3-II staining in the control and bevacizumab- and panitumumabtreated tumors. An increase (p = .06) in LC3-II expression was observed 24 hours after bevacizumab treatment. In panitumumabtreated mice, an increase (p = .05) in LC-3 expression was demonstrated 4 hours posttreatment.
Discussion
Our data outline the importance of posttherapy timing of 99mTc-(CO)3 His-ann A5 imaging. Different dynamics of 99mTc-(CO)3 His-ann A5 uptake were observed in this CRC xenograft model, in response to a single dose of three different chemotherapeutics, the mAB against VEGF bevacizumab, and the mAB against EGFR, panitumumab. These therapeutics are already clinically used for the treatment of CRC. It is not surprising that the kinetics of tumor cell death vary widely in response to the (chemo)therapeutics because all exert their cytotoxic effects through different mechanisms. Different therapies induce apoptosis at different rates and extent, which may provide relevant information on disease activity and therapeutic efficacy.
Activation of the apoptotic pathway implies activation of caspases (
In addition to classic chemotherapy for the treatment of CRC, novel targeted therapies have been developed that inhibit crucial biologic pathways and key molecules involved in tumor growth and progression. The goal of this study was to investigate the early proapoptotic functions of all agents (until 48 hours after administration), including the VEGF antibody (bevacizumab) because VEGF inhibitors not only have antiangiogenic functions but also possess direct proapoptotic functions (via the PI3K-Akt pathway). 12 It has been demonstrated that a single dose of bevacizumab can already exert its effects (tumor growth inhibition, decrease in microvessel density, vessel maturation) within 48 hours.13,14
Panitumumab, a human mAB against EGFR, has been approved by the Food and Drug Administration for the treatment of EGFR-expressing metastatic CRC patients after failure of standard chemotherapy. Additionally, wild-type K-ras expression in CRC patients is obligatory for the use of panitumumab. 15 In the Colo205 cells, EGFR expression levels were evaluated and a low expression level was detected. However, it should be mentioned that on testing of different cell lines with variable expression levels of EGFR, the levels of EGFR expression did not significantly correlate with tumor response to cetuximab treatment (p = .2458). 16 In addition, wild-type K-ras was found in the Colo205 cell line. 17 Panitumumab, a human mAB against EGFR, induces apoptosis by the PI3K signaling pathway and the downstream Akt. 18 For the bevacizumab-treated tumors, a small and a large peak of caspase-3-positive cells were observed, whereas no increase in 99mTc-(CO)3 His-ann A5 in the tumors was found. For the panitumumabtreated tumors, only a small increase in accumulation of the radiotracer in the tumor 4 and 12 hours posttreatment was demonstrated, whereas a larger one was detected 24 hours posttreatment. However, a large increase in caspase-3-positive cells 4 hours posttreatment was observed and a smaller one 24 hours posttreatment. These observations confirm that the time course of biochemical events (PS exposure/caspase-3 activation) often differs depending on a wide variety of factors, including apoptosis-inducing agents, drug concentration, and exposure time. In particular, bevacizumab-induced apoptosis occurs possibly without concurrent PS exposure. Another explanation for the significant discrepancy between annexin A5 accumulation and caspase-3 staining in the bevacizumab- and panitumumabtreated tumors is that the peaks of PS exposure were missed as it has been described that the first peak of PS exposure can occur within the first 2 hours after the start of treatment (4-hour caspase-3 peak). The second peak in the bevacizumab-treated tumors may have taken place between 12 and 24 hours, before the caspase-3 activation.
The discrepancy in the annexin A5 uptake (SPECT) and the caspase-3 immunostaining results is not entirely clear, nor is the biologic significance of the demonstrated correlation. These findings warrant further investigation. The discrepancy in the results of the annexin A5 uptake and the caspase-3 immunostaining may also be attributed to the involvement of other types of cell death, induced by bevacizumab and panitumumab. Apoptosis and autophagic cell death are programmed cell deaths that have a fundamental role in tumorigenesis. In general, autophagy promotes survival to stress; however, its role in cancer development and response to therapy are complex. 19 Detection of autophagy in tissue can be performed by immunostaining LC3-II, which is localized in preautophagosomes and autophagosomes, making this protein an autophagosomal marker. 20 LC3-II has been shown to be highly expressed in gastrointestinal cancers. 21 Overall, the Colo205 tumors (untreated-treated) showed high expression of LC3-II. An increase, although not significant (p = .06) in LC3-II expression, was observed 24 hours after bevacizumab treatment, and an increase (p = .05) was demonstrated 4 hours after panitumumab treatment. At these time points, an increase in caspase-3-positive cells in the tumors was present, whereas tumor uptake of 99mTc-(CO)3 His-ann A5 was not significantly increased. A recent study confirmed the activation of autophagy in different CRC cell lines after treatment with panitumumab. 22
The expression of the IAP family member survivin was investigated in all Colo205 tumors by immunohistochemistry. Survivin exerts its antiapoptotic functions by binding and inhibiting active caspase-3, −7, and −9; by binding the proapoptotic Smac/Diablo proteins, inhibiting their pro-apoptotic functions; and by binding X-linked inhibitor of apoptosis protein (XIAP), which stabilizes their IAP function. 23 Moreover, survivin plays a key role in the regulation of apoptosis and cell division and is overexpressed in CRC patients. This overexpression in CRC is correlated with reduced tumor cell apoptosis and increased resistance to cancer therapy and predicts poor prognosis and shorter survival. 24 Recent studies indicate that the nuclear pool of survivin facilitates cell division and the cytosolic form interacts with Smac/Diablo and other IAP members, enabling the IAP function of survivin. 25 Untreated Colo205 tumors showed a high survivin expression, which has also been demonstrated in CRC patients. 24 Immunohistochemical staining for survivin demonstrated a predominantly nuclear staining pattern. For all therapies, levels of survivin are lower than in the nontreated, control tumors, suggesting an inhibition of the antiapoptotic protein survivin, a known defense system of CRC cells. In addition, expression of survivin is low at 8 hours in the 5-FU-treated tumor and at 4 and 24 hours in the panitumumabtreated tumors, where, at the same time, a peak in apoptosis (increased caspase-3 and/or 99mTc-(CO)3 His-ann A5 uptake) is observed. Hence, apoptosis induced by 5-FU and panitumumab is accompanied by parallel activation of the caspase cascade and a reduction in the antiapoptosis protein survivin. For the other therapies, no significant relationship of survivin expression and caspase-3 was found. This is in line with a study performed in thyroid carcinoma. 26 Possibly, other members of the IAP family are involved in the antiapoptotic pathway of Colo205 cells. In contrast, expression of survivin was strongly associated with reduced apoptotic index (assessed by TUNEL staining) in CRC. 24
Overall, the pharmacokinetics of this novel annexin A5 radiotracer, 99mTc-(CO)3 His-ann A5, obtained in our previous work, are somewhat similar to those of 99mTc-Hynic-ann A5 in mice, rats, rabbits, and swine.9,27–29 However, for 99mTc-Hynic-ann A5, other organs that concentrated radiolabeled annexin A5 include the spleen, bone marrow, stomach, and lungs.27,28 For 99mTc-(CO)3 His-ann A5, these organs did not show visible uptake. Limited available studies suggest that there are at least two peaks of PS expression: an early peak occurring 1 hour following chemotherapy and a second one appearing between 24 and 72 hours after completion of treatment.30–32 In the experimental model of Takei and colleagues, KDH-8- (hepatoma cells) bearing rats were treated with a single dose of cyclophosphamide. 31 The detection of apoptotic tumor response with 99mTc-Hynic-ann A5 required 20 hours after the start of treatment. Biodistribution studies in lymphoma-bearing mice treated with a single dose of doxorubicin revealed a biphasic increase in 99mTc-ann A5 (the mutant annexin V-117 with a built-in site for technetium labeling at the N-terminus), with an early peak between 1 and 5 hours after treatment and a second longer sustained rise from 9 to 24 hours after therapy. Both time points were associated with a 50 to 300% increase in average annexin A5 uptake over controls. 32 These studies, in concert with our study, demonstrate not only the significance of timing after initiation of the therapy but also that the increased apoptosis in tumors is a dynamic course and not simply cumulative.
To date, PS is the most promising biomarker for measuring the extent of cell death, and annexin A5 is the ligand with the highest affinity for PS. Experiences with annexin A5 have demonstrated that aspects such as biodistribution and target to background ratio require further improvement. The favorable biodistribution, low effective dose, and ability to image apoptosis in vivo make 99mTc-(CO)3 His-ann A5 a suitable candidate. Other apoptosis-targeting probes, including radiolabeled caspase inhibitors and substrates, have great potential for molecular imaging of apoptosis. However, no clinical data are yet available and thorough toxicologic studies are required before they can be applied in clinical studies.
In our preclinical model, Colo205 cells were injected subcutaneously so that tumor growth could be easily followed up. However, microenvironment and blood supply are different in these xenograft models compared to the clinical situation; therefore, the time course of apoptosis might differ substantially from patient data. The final goal of radiolabeled annexin A5 in the clinical setting is to evaluate the efficacy of anticancer therapy shortly after the start of the therapeutic regimen. Contradicting results previously reported may be due to an inadequate interval chosen between the administration of the therapy and imaging, so the peak of activity is missed.33–35 Many factors affect the timeline of apoptotic events, including the cell line used for the xenograft model, or tissue examined apoptosis-inducing agent, drug concentration, and exposure time. Consequently, in the clinical setting, it is crucial to initially characterize the time course of radiolabeled annexin A5 uptake so that the peak of apoptosis is detected and the early tumor response to therapy can be evaluated.
Conclusion
Quantitative 99mTc-(CO)3 His-ann A5 tumor uptake correlated well with the number of apoptotic cells as determined by caspase-3 immunostaining at various time points following treatment instigation. Our data outline the importance of posttherapy timing of 99mTc-(CO)3 His-ann A5 imaging.
Footnotes
Acknowledgments
We would like to acknowledge Covidien for providing us with the IsoLink kits. Also, we would like to express our gratitude to Philippe Joye for his assistance in animal handling and Steven Deleye for his technical assistance in handling the U-SPECT-II.
Financial disclosure of authors: Part of this work was financially supported by the European Union through a Euregional PACT II grant from the Interreg IV program of Grensregio Vlaanderen-Nederland (IVA-VLANED-1.20).
Financial disclosure of reviewers: None reported.