To develop a molecular probe for MRI detection of human tumor telomerase reverse transcriptase (hTERT) mRNA expression. Uniformly phosphorothioate-modified hTERT antisense oligonucleotide (ASON) homing hTERT mRNA was labeled with gadolinium (Gd) through the bifunctional chelator 1,4,7, 10-tetraazacyclododecane-N, N', N'', N'''-tetraacetic acid (DOTA) stirred within 45 minutes at 60°C. The Gd labeled probes were characterized in vitro. The cellular uptake rate and biodistribution of 99mTc-DOTA-ASON was measured instead of that of Gd-DOTA-ASON. A549 lung adenocarcinoma model was established in BALB/c nude mice and Gd-DOTA-ASON was injected intraperitoneally and MR images were acquired using 7.0T Micro-MRI (Bruker Biospec, Ettlingen, Germany) at different time points. Immunohistochemical analysis of telomerase activity of each xenograft was operated two days after in vivo imaging. The binding efficiency of Gd-DOTA-ASON reached as high as 71.7 ± 4.5% (n = 6). Gd-DOTA-ASON displayed perfect stability in fresh human serum at 37°C for 24 h. Compared with normal lung cells, A549 cells showed an obviously higher uptake of 99mTc-DOTA-ASON than that of lung cells (10.5 ± 2.7% vs. 4.8 ± 2.6%, P < 0.05). The signal intensity of A549 xenografts can be enhanced by Gd-DOTA-ASON and the signal to noise ratio (SNR) of tumor to muscle reached 2.37 and maintained a relatively high level within 6 h after injection. The activity of hTERT in A549 tumors can be suppressed by Gd-DOTA-ASON in pathological slices. The results of this study show that Gd-DOTA-ASON can be a promising intracellular MR contrast probe for targeting telomerase-positive carcinomas. (Cancer Sci, doi: 10.1111/j.1349-7006.2012.02316.x, 2012)
Telomeres, the ends of human chromosomes, are nucleoprotein structures, in which non-coding DNA sequences are arranged in tandemly repeated units of TTAGGG hexanucleotides, which may contain up to 25 000 bp.[1, 2] These terminal structures protect chromosomes against degradation, rearrangement and fusion to other chromosomal ends. In the process of DNA replication, conventional DNA polymerases are unable to completely replicate to the end of linear DNA molecular, thus leading to the loss of 50–200 bp of telomeric DNA per round of replication. When the length of telomeres reaches the Hayflick limit, cells cease to divide and undergo cellular senescence or mortality stage 1, and this stage is controlled by the p53 and pRB tumor suppressor proteins, which constitute a potent anticancer mechanism.[5, 6] Dysfunction of p53 or pRB (or both) can reverse replicative senescence and such post-senescent cells continue to proliferate until they reach crisis or mortality stage 2, characterized by widespread apoptosis and chromosomal instability, including end-to-end chromosomal fusions and nonreciprocal translocations.[7, 8] Cells that survive crisis maintain stable telomere lengths and become immortal, which is an essential step for malignant progression. Telomere maintenance is evident in virtually all types of malignant tumors. Telomeres in cancer cells are maintained either by the activation of telomerase or a telomerase-independent mechanism of telomere maintenance called alternative lengthening of telomeres (ALT), which is a recombination-based interchromosomal exchange of sequence information that maintains telomere length.[10, 11]
Telomerase is a ribonucleoprotein complex consisting of an RNA template (TERC), which contains a template region complementary to the telomeric repeat sequence and a reverse transcriptase catalytic subunit (TERT). The telomerase holoenzyme binds to the telomere through alignment of the RNA template region with the telomere repeat sequence. The protein component acts as a reverse transcriptase and catalyses the addition of telomeric repeats to the ends of chromosomes using the RNA subunit as a template.
Telomerase is activated in 85–90% of human carcinomas, but not in normal somatic cells; therefore, its detection holds promise as a diagnostic marker or prognostic index for cancer.[13, 14] Because human tumor telomerase reverse transcriptase (hTERT) expression correlates with telomerase activity and its presence is essential for enzymatic activity, the detection of hTERT mRNA is considered to be a reliable marker for the presence of cancer cells in clinical samples.[15, 16]
Conventional imaging paradigms rely on the detection of anatomical changes in disease that are preceded by molecular genetic changes that go otherwise undetected. With the advent of molecular imaging, several studies were undertaken on hTERT expression in vivo;[17-19] however, both single photon emission computed tomography and positron emission radionuclide-based imaging are associated with poor anatomical resolution and radioactive risk. MRI has the advantages of having very high spatial resolution (10–100 μm) and the ability to measure more than one physiological parameter using different radiofrequency pulse sequences. These features make MRI very attractive for imaging gene expression with high anatomical resolution.[20-22]
In the current study, we synthesized gadolinium (Gd)-DOTA-hTERT antisense oligonucleotide (ASON) and demonstrated the applicability and efficacy of these magnetic probes for tumor telomerase expression in vivo in a small animal model.
Materials and Methods
Uniform phosphorothioate ASON and sense oligonucleotide (SON) were obtained from Genscript Biotechnologies (Nanjing, China) for the present study, both with a primary amino on the 3′-end via a 7-member methylene carbon spacer designed for covalent conjugation with the chelator, and both purified with PAGE according to a previous report. The sequence of ASON is 5′-TAGAGACGTGGCTCTTGA-3′ and that of SON is 5′-TCAAG-AGCCACGTCTCTA-3′, with each chain having a molecular weight of approximately 5.5 kDa. The amine ASON/SON was processed under sterile conditions. The DOTA chelator was provided by Tokyo Kasei Kogyo Co. Ltd (Tokyo, Japan). The N-Hydroxysulfosuccinimide sodium (S-NHS) used as an intermediate and Gd–DTPA were purchased from GE Healthcare (Shanghai, China).
The 1-(3-Dimethylaminoptoply)-3-ethylcarbodimide hydrochloride (EDC.HCl) used as a condensing agent was purchased from Sigma Aldrich. The Gd (III) chloride hexahydate (GdCl3.6H2O) was purchased from Aladdin. The Spectra/Por Float-A-Lyzer G2 (MWCO: 3.5–5.0 kDa) was purchased from Spectrum Laboratories (Houston, TX, USA) and was used as a purification instrument. All other chemicals were reagent grade and were used without purification. The 99mTc-pertechnetate was eluted from a 99Mo–99mTc generator (China Institute of Atomic Energy, Beijing, China).
DOTA-ASON/SON conjunction and verification
ASON/SON was conjugated with DOTA-NHS via the derivatized amine on the 3′-end. DOTA-NHS-ester was prepared following a previously reported procedure.[23, 24] In brief, DOTA (4.0 mg, 9.8 μmol), EDC.HCl (2 mg, 10.4 μmol) and S-NHS (2 mg, 9.2 μmol) were added to a centrifuge tube containing 200 μL pH 5.5 PBS. The tube was vortexed immediately and incubated for 30 min at room temperature (RT). The semi-stable amino-reactive intermediate DOTA-NHS-ester was then dropped into 1.6 mL pH 8.5 PBS containing 66 μg ASON/SON (22.6 nmol) with a DOTA-NHS-ester to ASON/SON molar ratio of 226:1. The achromatous mixture was stirred at RT for 4 h. Then, the reaction solution was transferred into a dialysis tube, and dialyzed for 9 h to eliminate the unreacted materials using pH 7.0 PBS as dialyzate; the dialysis solution was changed every 3 h. Finally, the purified DOTA-hTERT ASON/SON was stored at −20°C for future use.
The verification of DOTA-ASON/SON conjunction was performed following the procedure described by Derek W. Bartlett et al.[24, 25] In brief, equimolar amounts of Gd chloride (GdCl3) and either DOTA-ASON/SON or DOTA were added into a microcentrifuge tube and mixed with 0.2 M PBS (pH 7.5). The microcentrifuge tubes were incubated for 45 min at 60°C for the labeling reaction. Half of the labeling reaction was combined with a double volume of 0.5 mM arsenazo III and a quadruple volume of 0.2 mM hydrogen chloride (HCl), which were added into 24-well plates (wells without GdCl3 were set as blank controls), and the absorbance at 630 nm was measured to determine the concentration of free Gd3+ using a UV spectrophotomete (BioTek MQX200; Gene Company Limited, Hong Kong, China) compared with an established standard curve derived from different concentration of GdCl3. The concentration of free Gd as well as the binding rate of Gd to DOTA-ASON/SON can be calculated using the standard curve. The Gd labeled DOTA-ASON/SON was purified using Sephadex G25 gel filtration column before in vitro and in vivo use.
In vitro stability of gadolinium-DOTA-antisense oligonucleotide
To assess the stability of Gd-DOTA-ASON at RT, as well as in normal saline (NS) and in fresh human serum at 37°C (incubated in fresh human serum or NS at 37°C and 5% CO2 at 10 μg/mL), a UV spectrophotometer (BioTek MQX200) was used. Wavelength of 630 nm was selected to determine the amount of free Gd3+ for evaluating the stability of Gd-DOTA-ASON and 260 nm for the intact ASON under different conditions.
Cell culture and uptake tests and in vitro MRI study
The pulmonary adenocarcinoma A549 and normal human lung cell lines were generous gifts from the Laboratory of Nuclear Medicine, West China Hospital of Sichuan University. The cells were cultured in high glucose DMEM (GIBCO, Beijing, China) containing l-glutamine (2 mM), sodium bicarbonate (1.5 mg/L), nonessential amino acids (0.1 mM) and sodium pyruvate (1.2 mM), supplemented with 10% FBS, 100 U/mL of penicillin and 100 μg/mL of streptomycin. The cell growth was monitored by inverted microscopy with a cell count plate. Logarithmic growth phase cells were seeded in 96-well plates (105 cells/hole), and incubated for 2 days before adding 20 μL 99mTc-DOTA-ASON/SON (37 MBq/mL) for uptake or 20 μL Gd-DOTA-ASON/SON (10 μg/mL) in 200 μL media for in vitro MRI (and 10 μg/mL Gd-DTPA for control). The plate was reincubated at 37°C and 5% CO2 for 4 h, and then the media was removed and washed three times with PBS. All of the removed radioactive media and washing liquid were collected and the counts were measured under a γ-well counter (Xi'an nuclear instrument factory, China). For in vitro imaging, the cells were suspended in PBS and T1-weighted images were acquired under 7.0T MRI.
Animals and tumor formation
Adult BALB/c nude mice (18–20 g, 6–8-weeks old) were purchased from the Department of Laboratory Animal Center, Sichuan University. All animal manipulations were carried out using sterile techniques and were approved by the Animal Protection Committee of Sichuan Province. A549 cells were trypsinized, suspended in PBS and injected subcutaneously into the right buttocks of BALB/c mice at 1–2 × 106 cells per mouse. Tumors were allowed to grow to 3–5 mm in diameter (1 week after inoculation) before micro MRI, and for biodistritution the diameter was up to 1 cm.
Biodistribution of 99mTc-DOTA-antisense oligonucleotide in A549 tumor-bearing nude mice
The biodistribution of 99mTc labeled DOTA-ASON in A549-bearing nude mice was observed instead of Gd-DOTA-ASON. The 99mTc-DOTA-ASON was prepared as follows: first, 1 mL (3700 kBq/100 μL) of 99mTc-pertechnetate was eluated from a 99Mo–99mTc generator and 100 μL SnCl2 (10 μg) was added into the eluate. Then, 100 μL stannous eluate was dropped into DOTA-ASON solution (approximately 10 μg); the final pH value was adjusted to approximately 7.0 with PBS (pH 8.0) and stirred for 30 min at RT. After reaction, the 99mTc labeled DOTA-ASON mixture was transferred into a 2 × 20 cm Sephadex G25 gel filtration column with 0.2 mM PBS (pH 7.2) for purification, and the radiochemical purity of 99mTc-DOTA-ASON was measured following the report by Liu et al. A total of 30 adult A549 tumor-bearing BALB/c nude mice (21 ± 2.4 g) were included and randomized into five groups. The diameter of tumors ranged from 0.8 to 1.2 cm (mean 0.9 cm). Each mouse was injected with 200 μL of 99mTc-DOTA-ASON (740 kBq) intraperitoneally. All mice were killed at 0.5, 1, 2, 4 and 8 h post-injection, respectively, and the tissues of interest (heart, liver, spleen, lung, kidney, stomach, small intestine, bladder, skeletal muscle, blood and tumor) were cut off and weighed. The radioactivity of each specimen was measured using a γ-well counter and recorded as percentage injected activity per gram (%ID/g) of tissue with a correction for background and decay.
In vivo MRI study
To investigate the magnetic resonance (MR) antisense probe's targeting potential and the MRI sensitivity of Gd-DOTA-ASON in vivo, A549 tumor-bearing mice MRI was performed with a 7.0T Micro-MRI (Bruker Biospec). Three groups of A549 (five mice per group) tumor-bearing nude mice were prepared as follows: the MR antisense probe was used for one group and Gd-DTPA and Gd-DOTA-SON were injected into the mice in the other two groups; mice were anesthetized with 1–2% inhaled isoflurane anesthesia (Nine Sent Pharmaceutical Co. Ltd, Hebei, China) in 1:2 of O2:N2; and Gd-DTPA, Gd-DOTA-ASON and Gd-DOTA-SON (1 μmol of Gd3+/kg) were then injected intraperitoneally. The MRI frame consisted of a nonmagnetic stereotactic wrist coil with a cylindric surface coil (5 cm internal diameter) positioned directly over the mouse pelvis. T1-weighted multiple slice multiple echo plus fat repression imaging were performed before injection, using the following parameters: repetition time (TR), 561 ms; echo time (TE), 14 ms; field of view, 4.0 cm; section thickness, 1 mm, 15 slices; matrix, 256 × 256. MR scanning using the same parameters were given at 0.5, 1, 2, 4, 6 and 24 h post-injection. Signal intensities were measured in a defined region of interest, which were in similar locations within the tumor center. ParaVision 5.0 was used for all pictures and data.
The tumor-bearing mice were killed 2 days after MRI scans. Tumor tissues were collected and placed into formaldehyde solution for 48 h, and the specimens were then embedded in paraffin block and cut into 10-μm-thick slices. Afterwards, immunohistostaining was performed to detect the telomerase activity of each sample.
All quantitative data were analyzed with SPSS11.0 software (SPSS Co. Ltd, Chicago, IL, USA) and the results were depicted as mean ± SD. Means were compared using 1-way anova and Student t-test. P-values < 0.05 were considered statistically significant.
Conjunction and verification of DOTA-antisense oligonucleotide
DOTA was activated with EDC and S-NHS, and the DOTA-sulfosuccinimide ester was conjugated to ASON to yield DOTA-ASON. The binding efficiency of Gd to DOTA-ASON/SON was 71.7 ± 4.5% (n = 6), at 45 min at 60°C.
In vitro stability of gadolinium-DOTA-ASON
The optical density of undecomposed ASON and concentration of free Gd iron under different conditions is displayed in Figure 1. The intact ASON remained at 90 ± 2.5% (n = 6) (Fig. 1A) and Gd-DOTA-ASON displayed perfect stability at RT and 37°C, both in NS and serum: the conjunction efficiency of Gd to DOTA-ASON was maintained at 69.5 ± 1.7, 68.4 ± 2.1 and 67.4 ± 2.8% over 24 h (Fig. 1B).
Cell culture and uptake tests and in vitro MRI study
The labeling rate of 99mTc-DOTA-ASON/SON was approximately 85%; after purification, the radiochemical purity reached 94 ± 3.5% (n = 6). The cellular uptake of 99mTc-DOTA-ASON in A549 cells was obviously higher than that of lung cells (15.5 ± 4.7 vs 6.8 ± 2.6%, P < 0.05, n = 12), and the uptake of 99mTc-DOTA-SON in A549 cells and lung cells was 4.21 ± 1.65 and 2.27 ± 1.32% respectively, which was significantly lower than that of 99mTc-DOTA-ASON (P < 0.001, n = 12). The in vitro MR images showed that the signal intensity of cells incubated with Gd-DOTA-ASON was significantly higher than that of Gd-DTPA or Gd-DOTA-SON (2.43 ± 0.40 × 106vs 1.48 ± 0.11 × 106 and 1.65 ± 0.23×106, P < 0.001, n = 8) (Fig. 2). All this evidence reveals that these antisense probes have potential for in vivo imaging.
Biodistribution of 99mTc-DOTA-antisense oligonucleotide in A549 tumor-bearing nude mice
The biodistribution of 99mTc-DOTA-ASON is shown (Fig. 3). The radiolabeled probe displayed a relatively higher uptake in kidney and liver (Fig. 3A), followed by tumor and small intestine (Fig. 3B) at different time points after injection. The high accumulation of 99mTc-DOTA-ASON in kidney indicated that the probe can be cleared out via the urinary tract. Bone marrow displayed a relatively low uptake, with 0.65 ± 0.12%ID/g at 0.5 h, down to 0.08 ± 0.02%ID/g at 24 h. The biodistribution result showed a quick blood clearance, with 1.86 ± 0.25%ID/g at 0.5 h after injection, which was retained at 0.63 ± 0.08%ID/g in the first 4 h. By contrast, the tumor uptake of 99mTc-DOTA-ASON was at a high level within 0.5 h, and maintained a comparably high and stable trend over 8 h, before gradually diminishing. A similar tendency was observed in the 99mTc-DOTA-SON group, but with a significantly lower uptake rate than that of the 99mTc-DOTA-ASON group (P < 0.05).
In vivo MRI study
A549 tumors were clearly enhanced by Gd-DOTA-ASON or Gd-DTPA within 0.5 h after injection. The Gd-DOTA-ASON group showed a constant enhancement within 6 h, whereas this was not seen 2 h post-injection in the Gd-DTPA group (Fig. 4). The highest signal to noise ratio (SNR) of tumor to neighboring muscle in each group appeared at 0.5 h post-injection. The SNR of tumor to muscle in the Gd-DOTA-ASON group was significantly higher than that of the Gd-DOTA-SON group (P < 0.05) at 1, 2, 4, 6 h post-injection, but gradually fell to a similar level after 24 h (Fig. 5). This result indicates that Gd-DOTA-ASON can be translated into cells in vivo and can be used as an intracellular MR contrast reagent.
Immunohistochemistry staining was performed on A549 tumors with anti-hTERT (Fig. 6), indicating A549 cells an overexpression status. The staining intensity of the positive area in the Gd-DOTA-ASON-treated tumor is lower than that with Gd-DTPA; DOTA-ASON can decrease the activity of telomerase in A549 tumor cells in vivo.
Telomeres and telomerase are overexpressed in the majority of tumor cells (85–90%),[10, 11] and represent an extremely attractive target for malignant tumor therapy and imaging. Using 99mTc-MAG3-ASON targeting hTERT mRNA in carcinomas, Liu et al. paved the way for in vivo imaging of hTERT expression. Although radionuclide-based imaging has poor spacial and anatomical resolution, the nuclear imaging does have high sensitivity with the sub nM to pM concentration of tracer, and these methods are suitable for tumor screening (particular whole body scans) and for patients who cannot undergo or refuse to accept MRI. Given the small number of reports on Gd-based imaging probes homing[27-30] and the absence of in vivo imaging of hTERT expression with MR antisense technology in recent years, in the present study, we synthesized DOTA-ASON and labeled it with Gd3+ for targeted imaging of hTERT expression in malignant tumors. The stability, cell uptake rate, biodistribution and tumor accumulation of Gd-DOTA-ASON were similar to those of 99mTc-MAG3-ASON, due to some shared affinity to plasma albumin and physiochemical in serum between MAG3 and DOTA. The conjunction rate of Gd-DOTA-ASON was much improved compared with a report by Bartlett et al. Although detailed reasons are yet to be clarified, these results are partially attributed to higher pH and longer reaction time, which facilitate the chelating process.[24, 25, 31]
MRI is an advanced and preferred imaging modality in the diagnosis of a wide range of malignancies, particularly when supplemented with Gd-DTPA dynamic contrast enhancement (DCE).[32, 33] Nevertheless, use of MRI DCE as an extracellular contrast agent depends largely on the tumor blood supply, and the acquisition process is limited by time. In this study, both in vitro and in vivo imaging result of the labeled ASON provided evidence that the molecular probe can be used as an intracellular contrast reagent, and with which, MRI can be performed without perfusion dependence and time-course limits. The mechanisms of Gd-DOTA-ASON uptake are far less understood, and endocytosis or pinocytosis are hypothesized for there are no direct oligonucleotides receptors in the cell membrane. If this is the case, the uptake of Gd-DOTA-ASON is non-specific and known to be concentration and energy dependent.
The highest tumor to muscle SNR at 0.5 h after injection might reflect the blood perfusion of tumors because all three groups showed similar accumulation of contrast agents with different tumor clearances. The coincidence between the SNR-Time result of Gd-ASON and biodistribution of 99mTc-ASON in tumors attested that the two labeled probes shared the same dynamics in tumors. The SNR time result also indicated that Gd-DOTA-ASON can be applied in tumor homing imaging within a convenient scanning timeframe.
Compared with superparamagnetic iron oxide (SPIO)-labeled or radionuclide-labeled probes, a relatively higher concentration is required for Gd-based imaging. In addition, unlike SPIO-labeled probes, which can be directly detected through iron stain in pathological sections, the staining of Gd-based probes is still technically difficult. Further studies on new labeling materials or probes might provide some insight into detection of telomerase expression. In the previous study, a 13-mer-N3′–P5′thio-phosphoramidate oligonucleotide, GRN163,[35, 36] was used as a telomerase template antagonist and it showed higher cellular uptake, and may be a potential probe for targeting telomerase.
Apart from telomerase, other mechanisms for maintaining telomere length have been found in human tumors. Telomerase activity is absent in approximately 15% of tumors and the telomere lengthening is obtained by recombination events between telomeres, known as alternative lengthening of telomere. Therefore, other imaging methods may be explored to visualize these telomerase-negative carcinomas.
The present study reveals Gd-DOTA-ASON to be a promising intracellular MR probe targeting hTERT mRNA expression.
We are grateful to Professors Lin Li and ChenZhong Fan (Department of Nuclear Medicine, West China Hospital of Sichuan University) for presenting cell lines and to Professor FaBao Gao (Department of Molecular Imaging, West China Hospital of Sichuan University,) for giving good ideas in the process of animal imaging.
The authors have no conflict of interest to declare.