Reporter PET Images Bortezomib Treatment-Mediated Suppression of Cancer Cell Proteasome Activity

Proteasomal protein degradation is a promising target for cancer therapy. Here, we developed a positron emission tomography (PET) technique based on the sodium-iodide symporter (NIS) gene fused with the carboxyl-terminal of ornithine decarboxylase (cODC) that noninvasively images cancer cells with inhibited proteasome activity. A retroviral vector was constructed in which the murine cODC degron was fused to the human NIS gene (NIS-cODC). Transiently transduced CT26 and HT29 colon cancer cells and stably expressing CT26/NIS-cODC cells were prepared. In cancer cells transiently transduced with NIS-cODC, NIS expression and transport activity was low at baseline, but NIS protein and 125I uptake was significantly increased by inhibition of proteasome activity with bortezomib. Stable CT26/NIS-cODC cells also showed increased cytosolic and membrane NIS by bortezomib, and four different stable clones displayed bortezomib dose-dependent stimulation of 125I and 99mTc-04− uptake. Importantly, bortezomib dose-dependently suppressed survival of CT26/NIS-cODC clones in a manner that closely correlated to the magnitudes of 125I and 99mTc-04− uptake. CT26/NIS-cODC tumors of bortezomib-treated mice demonstrated greater 124I uptake on PET images and increased NIS expression on tissue staining compared to vehicle-injected animals. NIS-cODC PET imaging may allow noninvasive quantitative monitoring of proteasome activity in cancer cells treated with bortezomib.

In this study, we constructed a novel reporter system consisting of the human NIS gene fused to the cODC degron. Cancer cells transiently or stably transfected with the construct were assessed for NIS expression and substrate transport activity in response to proteasome inhibition. We further investigated the capacity of the NIS-cODC reporter to image tumors in mice treated with bortezomib with radioiodine PET.

Proteasome inhibition of transduced cells increases NIS accumulation and substrate transport.
Figure 1 illustrates our pQCXIN retroviral expression vector in which the carboxyl terminus 37 amino acids of the murine cODC degron was fused to the NIS gene (NIS-cODC). The vector was first tested by transient transfection in CT26 and HT29 colon cancer cells. Proteasome activity of these cells was completely abrogated by treatment with bortezomib ( Fig. 2A). Transfected CT26 cells showed very low NIS expression at baseline, supporting the rapid degradation of NIS-cODC. However, 16 h inhibition of proteasome activity with 4 μM bortezomib induced a marked increase of NIS accumulation that was 17.5 ± 1.1 fold higher than non-transfected cells ( Fig. 2A). Furthermore, whereas NIS-cODC transfected cells displayed no increase or only mild increases in radioiodine uptake at baseline, bortezomib stimulated 125 I uptake to 594.8 ± 73.5% of control level for HT29 cells, and 125 I and 124 I uptake to 272.4 ± 29.4% and 236.3 ± 22.8% for CT26 cells (Fig. 2B). Lower baseline uptake level for HT29 compared to CT26 cells suggesting lower leakiness of expression might be explained by the 66.5 ± 1.4% greater proteasome activity for HT29 compared to CT26 cells ( Fig. 2A). Using CT26/NIS-cODC cells (clone #9), we compared the in-vitro activity of the PET reporter system to proteasome activity during treatment with graded doses of PS341. The results showed that proteasome activity was substantially reduced to 24.8% of untreated level by 16 h treatment with 12.5 nM of PS341, and almost completely inhibited by 25 nM or 50 nM (3.1% and 1.2% of untreated level, respectively). Under these conditions, NIS protein level gradually increased to 303.4 ± 68.4%, 398.9 ± 42.4%, and 951.7 ± 36.4% of untreated control level. Similarly, 125 I uptake level gradually increased to 124.1 ± 2.4%, 150.6 ± 13.2%, and 357.9 ± 14.1% of untreated control level (Fig. 4A). Correlation analysis showed that NIS expression and 125 I uptake increased exponentially as log proteasome activity decreased. Furthermore, 125 I uptake level showed a close linear relation to NIS expression level (r = 0.98; P < 0.0001; Fig. 4B).
Importantly, we compared bortezomib's effect on NIS substrate transport and its effect on cell survival. In four different stable CT26/NIS-cODC clones, there was a close correlation between 125 I uptake in cells exposed − uptake in cells exposed to 25 nM bortezomib for 48 h faithfully recapitulated this close correlation with cell loss by treatment with 25 nM (r = 0.959, P < 0.05) or 50 nM bortezomib (r = 0.980, P < 0.05; Fig. 5B).

Discussion
Construction of reporters that contain the cODC sequence provides a unique opportunity to monitor proteasome activity. Although cODC-fused fluorescent reporters are useful for identifying cancer cells with poor proteasome activity in vitro, poor signal penetrance limits their value for imaging living bodies. Here, we introduce a noninvasive PET reporter system based on cODC-fused NIS that allows quantitative imaging of cancer cells with suppressed proteasome activity in vivo.
The cODC consists of a 37-residue terminal-sequence that is recognized and rapidly degraded by the 26S proteasome. Because this occurs in a ubiquitin-independent manner, no posttranslational modifications are required for substrate preparation. Hence, cODC-tagged constructs readily act as substrates to probe proteasome-mediated degradation in vitro as well as in vivo. This degradation tag functions autonomously and can be appended to other proteins for proteasome targeting 9 . The cODC sequence is highly conserved, and tagging with the murine cODC leads to efficient degradation of proteins from various species by mammalian proteasomes 17 .
Although CT26 and HT29 cells are derived from colon tumors, they do not show significant endogenous NIS expression 13,18,19 . In our study, we transduced cells with a retroviral vector driven by the CMV promoter, which results in constitutive expression cODC-NIS, uninfluenced by the environment. In our study, CT26 and HT29 cells showed complete abrogation of proteasome activity in response to bortezomib treatment. In accordance with the rapid degradation of cODC-fused proteins under steady-state conditions, cells transiently transduced with NIS-cODC showed very low NIS expression at baseline. However, the cells displayed a substantial increase of NIS protein when proteasome activity was inhibited by bortezomib. The minimal amount of NIS-cODC protein at baseline allowed reporting of low proteasome activity with increases of NIS-mediated uptake of radioiodine and radio-technetium substrates.
Treatment of stably transfected CT26/NIS-cODC clones to bortezomib caused dose-dependent increases of 99m Tc-O 4 − and 125 I uptakes. In clones 1 and 20, 125 I uptake was higher by treatment with 25 nM of PS341 compared to 12.5 nM, but 99m Tc-O 4 − uptake was not. Although the cause for this finding is not clear, it is noteworthy that these 2 clones had apparently lower NIScODC transfection efficiency compared to clones 9 and 8. 99m Tc-O 4 − has been reported to rapidly efflux out of cancer cells transfected with the NIS gene 20 . It might therefore be possible − efflux in these cells is more rapid than that of radioiodine, causing less apparent differences in 99m Tc-O 4 − accumulation according to NIS expression in cancer cells with low transfection efficiency. The crucial role proteasomes play in degrading regulatory proteins that suppress cancer cell proliferation has led to clinical trials of proteasome inhibitors for cancer treatment. The ability to monitor proteasome activity of tumors in vivo can contribute to the development of new drugs that are more effective and apply to a broader range of malignancies 2,6,7 . We investigated this issue in CT26 cells engineered to stably express NIS-cODC. Our results showed that graded concentrations of bortezomib resulted in substantial and dose-dependent reductions of cell number by bortezomib for both CT26 and CT26/NIS-cODC cells. In contrast to a previous study where stable ZsGreen-cODC expression was found to be stressful for cancer cells 11 , we found that NIS-cODC expression by itself did not exert cellular stress. Importantly, the amount of increase of NIS substrate transport in bortezomib-treated CT26/NIS-cODC cells closely correlated to the magnitude of suppression in cell survival. This finding supports the possibility that the level of stimulated radioiodine uptake might help monitor cancer cell response to proteasome inhibitor treatment.
Finally, we investigated the ability of our reporter system to image inhibited proteasome activity in tumors of living mice. For this purpose, CT26 cells conveniently produced a syngeneic tumor model in wild type BALB/c mice. PET/CT imaging of CT26/NIS-cODC tumors revealed significantly increased radioiodine uptake after only 2 doses of bortezomib. This drug is considered a potent and specific proteasome inhibitor compared to the less specific and weaker predecessors, making it suitable for in vivo application 21 . Bortezomib is the first-in-class therapeutic proteasome inhibitor approved for human use, and several clinical studies have established its efficacy in multiple types of refractory cancers 22 . In our experiments, CT26/NIS-cODC tumors of bortezomib-treated animals displayed clearly increased radioiodine uptake on PET and showed strong NIS expression on tissue staining, which demonstrates the capacity of our reporter to noninvasively image tumor response to proteasome inhibition.
In clinical trials, bortezomib is administered at doses of up to 2 mg/m 2 for a cumulative dose of 40 mg/ m 2 23 . Assuming a body-mass index of 25 kg/m 2 , these doses would amount to approximately 0.08 and 1.6 mg/ kg, respectively. Extrapolation of human doses to mice cannot be done based on body weight alone, because species-dependent biochemical and functional systems alter pharmacokinetics. Small animals require larger drug dose compared with humans on a weight basis, and for mice, a multiplication correction factor of 12.3 is recommended 24 . In our case, this yields a single dose of approximately 1 mg/kg, which is the dose that we used. In a previous study, Momose et al. imaged mice harboring HEK293 tumors that stably expressed ZsGreen-cODC after injecting a single dose of bortezomib 12 . In their study, tumor fluorescence was greatest with 4 mg/kg bortezomib and weak with 1 mg/kg. Therefore, it is likely that higher doses of bortezomib would have resulted in higher levels of tumor radioiodine uptake. Bars on the right are mean ± SD of % uptake of triplicate samples. ‡ P < 0.001, **P < 0.01, **P < 0.05, compared to controls. (B) Correlations between NIS expression (left) or 125 I uptake (middle) and log proteasome activity, and between NIS expression and 125 I uptake (right). Proteasome activity is measured as rate of increase in fluorescent intensity (arbitrary units). Correlation coefficients (r) and P values are from regression analysis.
In our study, radioiodine uptake assessed in vivo by PET was highly accurate in comparison to uptake confirmed from extracted tissues. Whereas optical imaging of fluorescent signals is limited by depth of light penetration and scatter, gamma rays generated from positron annihilation have high tissue penetrance. Hence,  25 . Therefore, our NIS reporter PET system may provide highly accurate quantitative measurements and is expected to work just as well for deep seated orthotopic tumor models as in subcutaneous tumor models. Together, our results demonstrate the potential usefulness of NIS-cODC reporter PET for noninvasive monitoring of tumor response to proteasome inhibitors.
A limitation of this study is that we did not compare radioiodine uptake level stimulated by bortezomib to actual antitumor effects. Although this experiment was beyond the scope of the present study, which was to investigate the capacity of our reporter PET system to image proteasome inhibition, the issue deserves exploration in future studies.
In conclusion, we have developed a novel reporter PET technology based on the NIS gene fused with the cODC degron, which allows robust monitoring of suppressed cancer cell proteasome activity in response to treatment. Accumulation of NIS-cODC protein by low proteasome activity led to increased tumor radioiodine uptake that was noninvasively imaged by PET in a quantitative manner. This technique may therefore be useful for noninvasive monitoring of tumor response to proteasome inhibitors, which could facilitate new drug development. . We removed the ZsGreen gene sequence from this vector by digestion with Bam HI and Eco RI restriction enzymes (Takara, CA). The human NIS gene sequence cut out with identical restriction enzymes was then cloned into the vector to produce pQCXIN-NIS-cODC, which expresses the NIS-cODC fusion protein. The wild type human NIS that we used efficiently transports radioiodine in various in vitro and in vivo systems 13,14,16 . This gene sequence is comprised of 2490 nucleotides that encode a protein of 643 amino acids as previously described 26 . pQCXIN-NIS-cODC was transfected into PT67 packaging cells (Clontech, CA) in high-glucose DMEM media (Lonza, Basel, Switzerland) to produce retroviral vectors.

Cell culture and transient transfection. CT26 mouse colon cancer cells and HT29 human colon can-
cer cells from the American Type Culture Collection were maintained in RPMI (Lonza) supplemented with 10% fetal bovine serum (Serana, Germany) and 1% penicillin/streptomycin (Gibco Laboratories, Gaithersburg, MD) at 37 °C and 5% CO 2 in a humidified atmosphere. Cells were sub-cultured twice a week. To prepare transient transfectants, 5 × 10 4 cells were seeded into 24-well plates and transduced with pQCXIN-NIS-cODC or empty pQCXIN plasmids (control) by adding 2 μg DNA per well mixed with 2 μl lipofectamine LTX (Invitrogen, Carlsbad, CA) in opti-MEM (Gibco). Proteasome inhibition was performed by adding bortezomib (PS341; Calbiochem, MA) to the culture media.
Stably expressing cell lines. Stably-expressing CT26 cells were prepared because, unlike HT29 tumors that require immune-deficient mice, CT26 tumors are conveniently generated in wild-type mice.  1 mM ethylenediaminetetraacetic acid, 10 μg/mL aprotinin, and 1 mM phenylmethylsulfonylfluoride. Following removal of cell debris, the supernatant was incubated with 0.1 M Na 2 CO 3 at 4 °C for 1 h. The samples were filled with buffer containing sucrose (0.0856 g/mL), 10 mM HEPES, and 1 mM MgCl 2 and then centrifuged at 42,000 rpm for 1 h at 4 °C. After centrifugation, the membrane fraction pellet was suspended in phosphate buffered saline (PBS).
Cell lysate (30 μg) or membrane fraction protein (10 μg) were separated by electrophoresis on a 10% sodium dodecyl sulfate polyacrylamide gel, followed by transfer to a polyvinylidene difluoride membrane (Bio-RAD, CA). The membrane was blocked with 5% nonfat milk in Tris-buffered saline and polysorbate-20 for 1 h at room temperature and incubated overnight at 4 °C with primary antibodies against NIS or β-actin (Santa Cruz Biotechnology, TX). Membranes were then incubated at room temperature for 1 h with secondary anti-rabbit IgG antibody for NIS and anti-mouse IgG antibody for β-actin (Cell Signaling, MA). Immune reactive proteins were detected by chemiluminescence, and band intensities were quantified on a GS-800 densitometer using Quantity One software (Bio-Rad Laboratories). Cellular radioiodine and radio-technetium uptake measurements. Cells were incubated for 1 h with 74 kBq 125 I (Perkin Elmer, MA), 110 kBq 124 I, or 740 kBq 99m Tc-O 4 − added to the culture media in 5% CO 2 at 37 °C. Cells were rapidly washed twice with cold PBS, lysed with 0.1 N NaOH, and measured for cell-bound radioactivity on a high energy γ-counter (Perkin-Elmer) for 124 I or 99m Tc-O 4 − and γ-counter (Wallac) for 125 I. Uptake levels were normalized to protein content for each sample.
Proteasome activity assay. To measure proteasome activity, 1.2 × 10 6 cells in 6-well plates were washed and collected by scraping with 0.5% NP-40 in distilled water. After incubation for 30 min at −20 °C and centrifugation at 13,000 rpm for 10 min at 4 °C, supernatants were measured for proteasome activity with a fluorometric assay kit (Bio Vision, CA). Briefly, cell extracts were mixed with 100 μl assay buffer and 1 μl Suc-LLVY-7-amino-4-methylcoumarin (Suc-LLVY-AMC) substrate. Fluorescence was measured at 350 nm excitation and 440 nm emission wavelengths in a microplate reader at 37 °C for 30 min. Proteasome activity was measured as rate increase of fluorescent signals from Suc-LLVY-AMC degradation.
Sulforhodamine B (SRB) survival assay. Surviving cell content was measured by SRB assays. Briefly, cells seeded overnight on a 96-well plate were treated for the indicated durations of time, and cell monolayers were fixed with 10% (wt/vol) trichloroacetic acid at 4 °C. After cells were stained with SRB dye (Sigma-Aldrich) for 30 min, excess dye was removed by repeated washing with 1% (v/v) acetic acid. Protein-bound dye was finally dissolved in 10 mM Tris base solution and subject to spectrophotometric measurement of absorbance at 510 nm using a microplate reader.
In vivo 124 I PET imaging and ex vivo tumor uptake. Animal experiments were performed in accordance with the National Institutes of Health Guide for Care and Use of Laboratory Animals, and approved by the Institutional Animal Care and Use Committee of Samsung Medical Center. Three groups of tumor-bearing animals were prepared in 8-week-old wild type male BALB/c mice by subcutaneous injection of 1 × 10 8 cancer cells into the right shoulder region. The control group had tumors formed by non-transduced CT26 cells (n = 5). The second and third groups had CT26/NIS-cODC tumors and were intraperitoneally injected with DMSO vehicle (n = 6) or 1 mg/kg bortezomib (n = 8), respectively, at 12 and 24 h prior to PET imaging. When tumor size reached approximately 1.5 cm in diameter, animals were injected with 4 MBq of 124 I through the tail vein. Imaging was performed 1 h later under isoflurane anesthesia, using a small-animal PET/CT scanner (Inveon; Siemens, Germany) without respiratory gating. PET imaging was followed by non-enhanced CT without attenuation correction. Care was taken to maintain an environment of 23-25 °C.
A square ROI was placed on the tumor region with highest activity on coronal PET, and a mirrored ROI was placed on the contralateral shoulder as background. Tumor-to-background ratios of mean and maximum standardized uptake values (SUVmean and SUVmax) were used as indices of uptake. Immediately after PET imaging, animals were sacrificed by cervical dislocation. Tumor and background tissues were extracted, weighed, and measured for 124 I radioactivity on a high-energy γ-counter. Ex vivo measurements were expressed as weight-corrected tumor-to-muscle count ratios.

Immunohistochemistry staining of tissue sections for NIS expression. After micro-sectioning
formalin-fixed and paraffin-embedded tissues at 4 μm thickness, heat-induced antigen retrieval was performed for 3 min in pH 9.0 EDTA buffer (Dako). Sectioned slides were then incubated overnight with an anti-human NIS rabbit polyclonal antibody (1:100 dilution; Santa Cruz Biotechnology) at 4 °C. This was followed by incubation with HRP-labeled polymer-conjugated secondary antibodies against rabbit IgG (Dako, CA) for 30 min at room temperature. A color reaction was induced using the ready-to-use 3,3′-diaminobenzidine substrate chromogen solution (Dako) for 5 min, followed by washing with distilled water. Sections were lightly counterstained with hematoxylin for 30 seconds before dehydration and mounting.
Statistical analysis. Data are presented as mean ± standard deviation (SD). Significance of difference between groups was analyzed by two-tailed unpaired Student's t-tests for two groups and ANOVA with Tukey's post-hoc tests for three or more groups. Correlation was evaluated by linear regression or Spearman's rank-order analysis. P values < 0.05 were considered statistically significant. Data Availability. All data generated or analyzed during this study are included in this published article.