SPECT Imaging of Herpes Simplex Virus Type1 Thymidine Kinase Gene Expression By [123I]FIAU1

Original Investigations

SPECT Imaging of Herpes Simplex Virus Type1 Thymidine Kinase Gene Expression By [123I]FIAU1 Seok Rye Choi, PhD, Zhi-Ping Zhuang, PhD, Ann-Marie Chacko, Paul D. Acton, PhD, Juri Tjuvajev-Gelovani, PhD Mikhai Doubrovin, MD, PhD, David C. K. Chu, PhD, Hank F. Kung, PhD

Rationale and Objectives. Introduction of suicide genes, such as herpes simplex virus type1 thymidine kinase (HSV1-tk), in tumor cells has provided a useful method for tumor gene therapy. Several L-nucleosides, such as Lamivudine (3TC) and Clevudine (L-FMAU), have been successfully tested as high-potency antiviral agents. To investigate the potential differences between D- and L-isomers of nucleosides, [125/123I]-2’-fluoro-2’-deoxy-1␤-D/L-arabino-furanosy-5-iodo-uracil (D/ L-FIAU) have been synthesized and evaluated as potential SPECT agents for imaging HSV1-tk gene expression. Materials and Methods. [125/123I]D- and L-FIAU were prepared by iododestannylation of the respective tin precursors with 125/123I-sodium iodide. In vitro cell uptake studies were performed by incubation of [125I]D- and L-FIAU in RG2 cells expressing HSV1-tk (RG2TK⫹). In vivo studies including biodistribution and SPECT were performed in RG2TK⫹ and RG2TK⫺ tumor-bearing nude mice using [123I]D- and L-FIAU. Results. Cell uptake and biodistribution studies indicated that [125/123I]L-FIAU did not show any high accumulation (sensitivity) or uptake ratios (selectivity) in HSV1-TK–positive (RG2TK⫹) tumors as compared to control tumors. In contrast, [125/123I]DFIAU displayed both sensitivity and selectivity to RG2TK⫹ tumors. The selective in vivo accumulation of [123I]D-FIAU increased with time and the tumor uptake ratios (RG2TK⫹/RG2TK⫺) for 2, 4, and 24 hours averaged 6.2, 22.7, and 58.8, respectively. High-resolution SPECT of four nude tumor-bearing mice demonstrated a very high uptake of [123I]D-FIAU in the RG2TK⫹ tumor, while no significant tracer accumulation was observed in the RG2TK⫺ tumor and other organs. Conclusion. The data suggest that only the D-isomer of [123I]FIAU is useful for imaging HSV1-tk gene expression in mice by high-resolution SPECT imaging. Key Words. D-FIAU; L-FIAU; HSV1-tk; SPECT; molecular imaging. ©

AUR, 2005

Salvage pathways are important cellular routes for recycling nucleosides as starting material for the intracellular synthesis of DNA or RNA. It is particularly important for Acad Radiol 2005; 12:798 – 805 1 From the Department of Radiology (S.R.C., Z.-P.Z., P.D.A., H.F.K.) University of Pennsylvania, 3700 Market Street, Room 305, Philadelphia, PA 19104 and the Department of Pharmacology (A.-M.C., J.F.K.), University of Pennsylvania, Philadelphia, PA 19104; Department of Experimental Diagnostic Imaging, The University of Texas M. D. Anderson Cancer Center, Houston, TX (J.T.-G.); Department of Radiation Oncology, Memorial SloanKettering Cancer Center, New York, NY (M.D.) The College of Pharmacy, University of Georgia, Athens, GA (D.C.K.C.). Received February 9, 2005; revision received April 7; revision accepted April 8. Address correspondence to H.F.K. e-mail: [email protected]

© AUR, 2005 doi:10.1016/j.acra.2005.04.010

798

viruses that may not have a full spectrum of enzymes for the de novo synthesis of nucleosides. Thymidine kinase (TK) is a key enzyme in this de novo pathway (TK denotes the enzyme, while tk refers to the gene). This ratelimiting enzyme catalyzes the monophosphorylation of thymidine to dTMP. The monophosphate, dTMP, is subsequently converted to diphosphate, dTDP, and triphosphate, dTTP. Ultimately, the triphosphate, dTTP, is a substrate for DNA polymerases. The dTTP is in a unique position in this metabolic pathway, because it is a substrate useful only for DNA synthesis and not for RNA synthesis. Human TK enzyme is one of the key enzymes controlling the rate of DNA synthesis. Many viruses also use this salvage pathway to replicate themselves, and

Academic Radiology, Vol 12, No 7, July 2005

TABLE 1 Nucleosides Used to Study HSV1-tk Gene Expression

more importantly, this pathway provides a rapid and direct link to DNA synthesis, which is critical for the survival of many viruses. Due to its pivotal role in virus survival, the viral TK enzymes are highly indiscriminating. Taking advantage of the differences in selectivity between human and viral TK enzymes, antiviral agents have been successfully developed. In this project, we propose to explore the differences in selectivity and develop labeled nucleosides for measuring HSV1-TK enzyme activity by in vivo imaging. Four stereoisomers, ␤-D, ␣-D, ␤-L and ␣-L, exist for nucleosides. Among them, ␤-D nucleosides are the natural building blocks for DNA and RNA synthesis. In the past 10 years, there has been significant progress in using L-nucleosides as anti-viral agents (1). Herpes simplex virus type 1 (HSV1) thymidine kinase (ATP, thymidine 5’-phosphotransferase; EC 2.7.1.21) is the first rate-limiting step in the salvage pathway for synthesis of triphosphate nucleoside precursors for viral DNA synthesis. HSV1-TK enzyme has been shown to phosphorylate ␤-L-thymidine as well as ␤-Dthymidine with equal efficiency. The Ki of ␤-L thymidine, 2 ␮M is almost as potent as the Km of the natural substrate, ␤-D-thymidine (Km ⫽ 2.8 ␮M). However, in vitro enzyme assays showed that ␤-L-thymidine is not recognized by human TK. Therefore, L-nucleosides appear to be more efficacious against HSV1 as compared to mammalian cells. Similarly, both ␤-D and ␤-L isoforms of carbocyclic analogs of 5-iodo- and 5-(2-bromovinyl)-2’-deoxyuridine showed potent inhibition of HSV replication and high binding affinity for HSV1-TK enzyme (2– 4). L-FMAU (Table 1) is a thymidine analog that has demonstrated a high potency in antiviral activity against hepatitis B virus (HBV) and Epstein Barr virus (EBV). L-FMAU, like the corresponding D-nucleoside, is phosphorylated by viral TK to the monophosphate, L-FMAU-MP. Subsequently, it is phosphorylated to the diphosphate and triphosphate by nucleoside monophosphate

IMAGING HSV1-TK GENE EXPRESSION BY [123I]FIAU

kinase and nucleoside diphosphate kinase, respectively. The antiviral activity resides with L-FMAU-TP, which is a potent inhibitor of the DNA polymerase of HBV and EBV (5,6). The selectivity on viral DNA polymerase confers the high potency toward the viruses while sparing normal tissues and organs from toxic side effects. Upon entry into the cell, monophosphorylation of the nucleoside by HSV1-TK is the only enzymatic reaction responsible for its intracellular trapping. The selective trapping of the nucleoside by HSV1-TK could potentially be translated into an imaging signal that reflects the presence of HSV1-TK (Table 1). The rationale for testing the L-isomer of FIAU as a potential imaging agent for HSV1-tk gene expression is based on the suitability of L-nucleosides for HSV1-TK due to the lack of enantioselectivity of the enzyme, in addition to allowing for the safer imaging of gene delivery (7). It is our objective to exploit the differences in membrane transportation, biochemical and metabolic processes between human and viral enzymes for the unnatural L-nucleosides. As such, it may be possible to develop labeled L-nucleosides that selectively accumulate in tumor cells expressing the HSV1-tk gene. In the past 20 years, 1-(2’-deoxy-2’-flouro-1’-␤-arabinofuranosyl)-5-iodouracil (FIAU) has been tested as an experimental antiviral agent (8,9). Unfortunately, it has shown a high toxicity in humans that limits its use as a therapeutic agent (10). More recently, radiolabeled D-FIAU (124I or 131I) has been used as an imaging agent for positron emission tomography (PET) or single photon emission computed tomography (SPECT) imaging of HSV1-tk gene expression. The imaging approach is very attractive in measuring gene expression in vivo and this topic has been extensively reviewed recently (11–18). The approach originated from cancer gene therapy based on HSV1-tk gene expression coupled with acyclovir (ACV) or ganciclovir (GCV) treatment. Several nucleosidebased–labeled pyrimidine and purine ribosyl or acyclic nucleosides, including FIAU, FHPG, FHBG, etc. (see Table 1), have been reported as potential imaging agents for detecting the expression of HSV1-tk gene. Recent reports have suggested that D-FIAU has several unique features making it an attractive tracer for imaging HSV1-tk gene expression in tumor tissues (19 –21). Indeed, [124I]DFIAU shows excellent uptake in xenograft HSV1-TK⫹ tumor tissue, and the tracer accumulation is severalfold higher than that of FHPG. The signal to noise ratio for FIAU is also superior to that of FHPG (20). However, the major disadvantage of using [124I]D-FIAU as a probe of gene expression is its radionuclide properties: I-124 has a

799

CHOI ET AL

relatively long half-life (T1/2 ⫽ 4 d) with un-wanted ␤-emission. The ␤-emission may be useful for radiation therapy, but for a diagnostic procedure it adds radiation burden. Earlier, [131I]D-FIAU was prepared as an imaging agent for studying the biodistribution of this tracer in mice or rats bearing RG2TK⫹ tumor (22). Planar and SPECT imaging of the HSV1-tk gene expression has been demonstrated. However, the physical characteristics of I-131 are not ideal for SPECT imaging; specifically, the relatively long half-life T1/2 ⫽ 8 d and a medium energy ␥-ray of 364 KeV. Since it is commonly known that I-123 (T1/2 ⫽ 13 hr and ␥-ray ⫽ 159 KeV) is ideally suited for SPECT imaging, we have prepared [123I]DFIAU as a potential high-resolution SPECT imaging agent for imaging HSV1-tk gene expression in mice.

MATERIALS AND METHODS Reagents used in the syntheses were purchased from Aldrich Co. or Fluka Co., and were used without further purification unless otherwise indicated. Preparative thinlayer chromatography (PTLC) was performed on silica gel plates with a fluorescent indicator that was visualized with light at 254 nm. 1H NMR spectra were obtained on a Bruker spectrometer (Bruker DPX 200; Bruker Biospin, Billerica, MA). The chemical shifts are reported as ␦ (ppm) values and referenced to CDCl3 (7.26 ppm for 1H). The coupling constants are reported in Hz. All animal experiments were carried out according to protocols approved by the Institutional Animal Care and Use Committee (IACUC) and the Radiation Safety Committee of the University of Pennsylvania. All procedures are in compliance with the National Institutes of Health Guide for Care and Use of Laboratory Animals. 2’-Fluoro-2’-deoxy-1-␤-D-arabinofuranosy-5-(trin-butyltin)-uracil: A mixture of 2’-fluoro-2’-deoxy-1␤-D-arabinofuranosy-5iodo-uracil, (D-FIAU) (10 mg, 0.37 mmol), bis-(tributyltin) (0.1 mL), Pd(Ph3P)4 (10 mg) and triethylamine (0.1 mL) in dioxane (2 mL) was stirred at 90°C for 3 hours. The solvent was removed and the residue was purified by preparative TLC (CH2Cl2: MeOH, 2:1) to give 8 mg of the desired product (58%). 1H NMR (200 MHz, CDCl ): ␦ 0.92 (t, J ⫽ 7.1 Hz, 3 9H), 1.20 –1.45 (m, 12H), 1.55–1.68 (m, 6H), 3.69 –3.74 (m, 2H), 3.92 (d,d, J ⫽ 9.6, 4.9 Hz, 1H), 4.31 (d,d,d, J ⫽ 18.5, 4.3, 2.5 Hz, 1H), 4.99 (d,d,d, J ⫽ 52, 3.9, 2.5 Hz,

800

Academic Radiology, Vol 12, No 7, July 2005

1H), 6.24 (d,d, J ⫽ 17.7, 3.9 Hz, 1H), 7.49 (d, J ⫽ 2.0 Hz, 1H). 2’-Fluoro-2’-deoxy-1-␤-L-arabinofuranosy-5-(trin-butyltin)-uracil: A similar procedure was used to prepare the L-derivative. A mixture of 2’-fluoro-2’-deoxy-1␤-L-arabinofuranosy-5-iodo-uracil, (L-FIAU) (10 mg, 0.37 mmol), bis-(tributyltin) (0.1 mL), Pd(Ph3P)4 (10 mg) and triethylamine (0.1 mL) in dioxane (2 mL) was stirred at 90°C for 3 hours. The solvent was removed and the residue was purified by preparative TLC (CH2Cl2: MeOH, 2:1) to give 10 mg of the desired product (68%). 1H NMR (200 MHz, CDCl ): ␦ 0.92 (t, J ⫽ 7.1 Hz, 3 9H), 1.20 –1.45 (m, 12H), 1.55–1.68 (m, 6H), 3.66 –3.74 (m, 2H), 3.92 (d,d, J ⫽ 9.5, 4.7 Hz, 1H), 4.31 (d,d,d, J ⫽ 14.7, 3.8, 2.5 Hz, 1H), 4.99 (d,d,d, J ⫽ 52, 3.8, 2.5 Hz, 1H), 6.24 (d,d, J⫽17.6, 3.8 Hz, 1H), 7.49 (d, J ⫽ 2.0 Hz, 1H). Radiolabeling Radioactive sodium iodides (I-125 and I-123) were purchased from Perkin Elmer Life and Analytical Sciences, Inc. (Boston, MA) and Nordion (Ottawa, Canada), respectively. Both of the radionuclides are obtained in no-carrier-added form. Radioactive-labeled D- and L-FIAU were prepared by the iododestannylation method reported previously for [124I]D-FIAU (23). Briefly, to a solution of 5-stannyl derivative of FIAU (100 ␮g in 50 ␮L methanol) was added Na125/123I followed by the addition of a 20 ␮L mixture of 30% hydrogen peroxide/acetic acid (1:3, v/v). To quench the reaction, saturated sodium bisulfite solution (0.1 mL) was added. The reaction mixture was loaded onto a C18 Sep-Pak cartridge (Waters, Milford, MA). The C18 cartridge was eluted with water, followed by methanol to isolate the radioiodine-labeled FIAU. The methanol was evaporated under a stream of N2 and the [125/123I]FIAU was formulated in saline solution. The radiochemical purity was determined by thinlayer chromatography (eluent: ethyl acetate/acetone/H2O, 14:8:1). The radiolabeled product was isolated in 60%80% radiochemical yield (radiochemical purities ⬎ 95%). The iodinated products are prepared under a no-carrieradded condition and it is likely that the specific activity is close to theoretical value: 2,200 and 240,000 Ci/mmol for I-125 and I-123, respectively. Cell Lines and Culture Rat glioma cells RG2, expressing the HSV1-tk gene (RG2TK⫹), were used for evaluation as reported (23).

Academic Radiology, Vol 12, No 7, July 2005

Non-transduced (wild-type) RG2 cells (RG2TK-) were used as controls. Cells were cultured in modified Eagle medium (GIBCO BRL, Grand Island, NY) plus 10% fetal bovine serum (FBS) (Hyclone, Logan,UT). Geneticin (G418, 250 mg/L) was added to the same basic medium for RG2TK⫹ cells. Cells were incubated at 37°C in a humidified atmosphere of 5% CO2/95% air. Cell Uptake Studies Twenty-four hours prior to the experiment, 1 ⫻ 106 cells were seeded in each well of a 24-well plate (Falcon; BD Biosciences, Bedford, MA). The cells were then washed with phosphate buffered saline (PBS) buffer and incubated for 5, 15, 30, 60, and 120 minutes at 37°C with [125I]D- or L-FIAU in MEM medium without FBS. After the incubation, cells were washed with PBS, released from the plate with trypsin/EDTA treatment, and resuspended in 1 mL of water. The cell-associated radioactivity was counted in a gamma counter (Cobra II Autogamma, Packard Instruments, Downers Grove, IL). The averaged radioactivity of 4 wells was used to calculate the % uptake. Protein concentration was determined using Lowry’s method with bovine serum albumin as a standard. Data were expressed as % uptake/100 ␮g protein.

IMAGING HSV1-TK GENE EXPRESSION BY [123I]FIAU

age of the injected dose per gram of tissue. Each value represents the mean ⫾ SEM of 3 mice. Small Animal MicroSPECT Imaging All scans were performed on an in-house microSPECT imaging system consisting of a Prism 3000XP tripleheaded gamma camera (Philips Medical Systems, Cleveland, OH), equipped with custom-made pinhole collimators (Nuclear Fields, Des Plaines, IL). The pinhole collimators had a focal length of 24 cm and a diameter of 2 mm. All the images were reconstructed using 10 iterations of a cone-beam iterative simultaneous algebraic reconstruction technique (SART) (24) with a correction for center-of-rotation error. Performance characteristics of this system have been described elsewhere (25). Four nude mice were injected with RG2TK⫹ cells in the left shoulder, and in the same animal wild-type RG2TK- cells were injected in the right shoulder as a negative control. When the tumor had grown to a size of approximately 1 cm, the animals were injected in the tail vein with 6 –13 mCi/animal [125I]FIAU, imaged for 2,6, and 24 hours post-injection, and then dissected.

RESULTS Subcutaneous Xenografts in Nude Mice as the Tumor Model Subcutaneous xenografts were produced in nu/nu CD-1 male nude mice (Charles River, Wilmington, MA) by a subcutaneous injection of 2 ⫻ 106 tumor cells in 100 ␮L of serum-free cell culture medium under anesthesia (a mixture of ketamine and xylazine, intraperitoneally). The transduced RG2TK⫹ cells were positioned in the left shoulder, while in the same animal wild-type RG2TK⫺ cells were injected in the right shoulder as a negative control. Animals were studied when the subcutaneous tumor xenografts reached a diameter of 1 cm (normally occurring on 14 –21 days after the subcutaneous implantation) (23). Biodistribution Studies Biodistribution studies were carried out in RG2TK⫺ and RG2TK⫹ tumor-bearing nude mice. Approximately 5 ␮Ci of [125I]L- or D-FIAU was injected into the tail vein. The animals were sacrificed at various time points postinjection by cardiac excision under anesthesia. Several tissue and organs, HSV1-tk-expressing and non-transduced parental tumors were removed and weighed. The radioactivity in each tissue was measured using a gamma counter (Packard Cobra). Results were expressed as the percent-

Radiolabeling of the D- and L-FIAU with I-125 and I-123 was successfully achieved by an iododestannylation reaction in good yield (60%– 80%) and high radiochemical purity (⬎ 95%) similar to that reported previously for [131I]FIAU (23). When D- or L-[125I]FIAU was injected (i.v.) into two groups of nude mice with xenografts of both RG2TK⫺ and RG2TK⫹ cells, the tumor uptake between RG2TK⫹ and the RG2TK⫺ showed dramatic differences (Figure 1A). The uptake values for the RG2TK⫹ tumor were 10 times higher than the RG2TK⫺ xenograft tumor. In addition, it was observed that the Lisomer showed a higher washout for all of the organs and tissues investigated (Figure 1). In an in vitro cell uptake study, either RG2TK⫺ or RG2TK⫹ cells were incubated with [125I]D- or L-FIAU. At 5, 15, 30, 60, and 120 minutes, the cellular uptake of the nucleoside clearly demonstrated that only the D-isomer displayed a significant cell uptake in the RG2TK⫹ cells (Figure 1B). No further study was done on the L-isomer. To further evaluate the biodistribution of the D-isomer of [125I]FIAU, the same xenografted nude mice were dissected at 2, 6, and 24 hours after intravenous injection

801

CHOI ET AL

Academic Radiology, Vol 12, No 7, July 2005

Figure 1. (a) Comparison of uptake of D- and L- forms of [125I]FIAU in tumor-bearing nude mice at 2 hours after an intravenous injection of the tracer. Subcutaneous xenografts were produced in nu/nu CD-1 male nude mice (Charles River) by a subcutaneous injection of 2 x 106 tumor cells in 100 ␮L of serum-free cell culture medium. Transduced RG2TK⫹ cells were positioned in the left shoulder and wild-type RG2 cells were injected in the right shoulder as a negative control. (b) Cell uptake studies of [125I]D- or L-FIAU in RG2TK⫺ or RG2TK⫹ cells were carried out at different time periods. The uptake was normalized to unit protein values. The results clearly showed that only the D-isomer displayed a significantly higher cell uptake in the RG2TK⫹ cells. It is likely that HSV1-TK enzyme is the dominating factor, after phosphorylation the monophosphate of [125I]D-FIAU was efficiently trapped inside the cells. Table 2 Biodistribution of [125I]D-FIAU in Tumor-bearing Nude Mice (%dose/g, n ⴝ 3)

Blood Heart Muscle Lung Kidney Spleen Liver Skin Brain Thyroid Tumor-RG2TK⫺ Tumor-RG2TK⫹ RG2TK⫹/Muscle RG2TK⫹/RG2TK⫺

2hr

6hr

24hr

0.57⫾0.63 0.37⫾0.03 0.24⫾1.82 0.45⫾0.04 0.76⫾0.16 0.38⫾0.01 0.46⫾0.35 0.48⫾1.09 0.06⫾0.02 14.8⫾0.10 0.55⫾0.10 3.0⫾0.3 21.4⫾16.0 6.2⫾2.3

0.09⫾0.04 0.04⫾0.00 0.02⫾0.05 0.07⫾0.00 0.09⫾0.01 0.14⫾0.01 0.07⫾0.01 0.10⫾0.09 0.01⫾0.00 34.4⫾0.14 0.12⫾0.03 2.1⫾0.2 91.1⫾36.6 22.7⫾12.7

0.002⫾0.001 0.002⫾0.000 0.001⫾0.004 0.016⫾0.001 0.009⫾0.000 0.033⫾0.002 0.016⫾0.002 0.027⫾0.06 0.001⫾0.00 17.0⫾0.08 0.016⫾0.004 0.89⫾0.04 705.4⫾181.5 58.8⫾20.8

Subcutaneous xenografts were produced in nu/nu CD-1 male nude mice (from Charles River) by a subcutaneous injection of 2 x 106 RG2TK⫺ or RG2TK⫹ tumor cells in 100 ␮L of serum-free cell culture medium.

(Table 2). The most noticeable change was that the ratio of RG2TK⫹ tumor versus RG2TK⫺ tumor showed a significant increase with time, while all the other organs or tissues showed a fast washout. The thyroid displayed consistently high levels of radioactivity, most likely due to in vivo deiodination (no pretreatment of the mice to block thyroid uptake) (Table 2).

802

Figure 2. (a) A representative microSPECT image of a nude mouse implanted with RG2TK⫹ tumor and RG2TK⫺ tumor at 2 and 24 hours after an intravenous injection of [123I]D-FIAU (active form). A total of four nude mice were imaged by a pinhole SPECT. (b) A necrotic core of a xenografted RG2TK⫹ tumor in a nude mouse was detected by microSPECT after injection of [123I]D-FIAU, while a RG2TK⫺ tumor implanted on the other side showed low uptake. Thyroid is not visible in the coronal and transaxial views selected for this figure.

We have tested the [123I]D-FIAU for high-resolution SPECT imaging. Our laboratory has recently developed an ultra-high–resolution small animal SPECT system, which is suitable for imaging small structures in the mouse (25–27). Using [123I]D-FIAU as a tracer, the feasibility of 3D imaging in xenograft tumor of nude mice was evaluated by this high-resolution SPECT system. Four nude mice were injected with a dose of the tracer (6-13 mCi/each) and imaged at 2, 6, and 24 hours postinjection. Similar to that observed in the biodistribution study, there was a consistent increase in signal with time in the RGTK⫹ tumor over expressing HSV1-TK enzyme. The RG2TK⫹/RG2TK⫺ tumor ratios were 10, 49, and 74 (by dissection) for 2, 6, and 24 hours, respectively. Interestingly, in one tumor-bearing mouse, the tumor displayed a ring of activity indicating that there is a different pattern of uptake in the middle of the tumor as compared to the periphery of the tumor. The disparity of tumor cell growth between active and apparently less active tumor tissues was clearly visualized by a high-resolution SPECT imaging (Figure 2). At the end of the 24-hour SPECT imaging study each mouse was dissected. The ratio of radioactivity in the xenograft RG2TK⫹ and RG2TK⫺ tumor tissue was counted. As expected, at 24 hours after intravenous injection of [123I]D-FIAU the ratio of uptake in RG2TK⫹/ RG2TK⫺ tumor displayed a very high value, 50 ⫾ 15; a value, which is comparable to that measured by SPECT, 74 ⫾ 31. The in vivo dissection produced data comparable to that of microSPECT imaging (Table 3).

IMAGING HSV1-TK GENE EXPRESSION BY [123I]FIAU

Academic Radiology, Vol 12, No 7, July 2005

Table 3 Ratios of Uptake Determined by Tissue Dissection and SPECT Imaging of Nude Mice at 24 hours After an Intravenous Injection of [123I]D-FIAU

Injected dose (mCi) Mouse#1 Mouse#2 Mouse#3 Mouse#4 AVG ⫾ SD

13 8 6 6.5

RG2TK⫹ tumor/muscle ratios

RG2TK⫹/RG2TK⫺ tumor ratios by ␥ counting (ex vivo)

RG2TK⫹/RG2TK⫺ tumor ratios by SPECT (in vivo)

723 2194 2306 1004 1556 ⫾ 810

49 30 54 66 50 ⫾ 15

39 57 99 101 74 ⫾ 31

DISCUSSION Results presented in this article suggest that the Lisomer of FIAU is inferior to the D-isomer in serving as a substrate for HSV1-TK. The L-isomer displays less favorable uptake and retention in the RG2TK⫹ tumor. In the in vitro cell uptake assay, the D-isomer is retained in the RGTK⫹ cells ⬎10-fold higher than that of the L-isomer (Figure 1). The minimal uptake of [125I]L-FIAU in RG2TK⫹ cells in vitro suggests that the tracer may have difficulty in transporting across the cell membrane, and/or the phosphorylation of L-FIAU by HSV1-TK may not be efficient enough to trap this tracer inside the transduced cells. As expected, the biodistribution data of [125I]D-FIAU in nude mice were comparable to that reported previously using [131I]D-FIAU (23). The uptake values for [125I]DFIAU in the RG2TK⫹ tumor implanted in nude mice were 10 times higher than those in the wild-type RG2TK⫺ xenograft tumor. In addition, it was observed that the L-isomer showed a higher washout for all of the organs and tissues investigated. It is likely that the Lisomer has faster kinetics in passing through the kidneys and hence, the majority of the dose is excreted rapidly. It is also likely that the L-isomer may not be efficiently “reabsorbed” in the renal tubular system. It is important to note that the HSV1-TK enzyme may be able to catalyze the phosphorylation of L-FIAU as well as D-FIAU. However, if the washout rate from the blood circulation of mice is too high, there may not be sufficient circulating L-FIAU for trapping in the TK enzyme–positive tumor tissue. It may also be possible that the L-FIAU simply cannot cross the cell membrane by the nucleoside transport mechanisms normally designed for the D-isomer. In view of this negative finding it is unlikely that L-FIAU would be useful as an imaging agent for detecting TK enzyme in the tumor cells. However, there is not enough

Tumor weight (g) 0.6759 0.7246 0.1542 0.1442

evidence to rule out other L-nucleosides as potential imaging agents for the same gene expression system. A fast washout rate from kidneys may not be equally applied to all of the L-nucleosides. In this study we used nude mice instead of nu/nu rats since the nu/nu mice are readily available from commercial sources. The resolution of the microSPECT system used in our laboratory (25, 26, 28) is sufficiently high to image a mouse. The 159 KeV ␥-ray emitted by I-123 is more suitable than I-131 (␥-ray ⫽ 364 KeV) for SPECT imaging; Therefore, it is advantageous to use [123I]DFIAU instead of the [131I]D-FIAU for high-resolution SPECT imaging of a mouse. It is also noted that the [123I]D-FIAU uptake ratios of RG2TK⫹ versus RG2TK⫺ tumor were very high (⬎ 10) at 2 hours after the tracer injection. The background counts were too low to accurately determine the ratios by SPECT imaging. However, the ratios measured by tissue dissection (see Tables 2 and 3) confirmed the high-contrast observation by SPECT imaging (Figure 2A). Upon imaging one of the tumor-bearing mice, a ring on the periphery of the tumor tissue was revealed, suggesting that the interior portion of the tumor may have reached a low metabolic profile and was undergoing necrosis. The disparity between tumor cells and their uptake was clearly visualized by highresolution SPECT imaging (Figure 2B). Imaging would thus have a clear advantage over tissue counting to measure the uptake and retention of the tracer. A simple tissue counting technique would not detect the intercellular differences in the tumor that account for the donut-shaped uptake pattern observed by SPECT imaging. In the past 10 years, significant progress has been made in using PET and SPECT imaging methods to detect gene expression in living animals (29 –31). Using the HSV1-tk gene expression system and pyrimidine and purine ribosyl or acyclic nucleosides as a reporter gene/reporter probe combination is the most successful molecular

803

CHOI ET AL

imaging paradigm in demonstrating the feasibility of this approach (18,32–34). However to optimize the reporter gene/probe pair, several major factors must be considered which may control the processes of uptake and retention of nucleosides in xenograft-tumor cells expressing HSV1TK. Pharmacokinetics of nucleosides (such as the D- and L-isomers) and their analogs may be different, which may lead to differences in delivery between normal and tumor tissue. The differences may be largely due to a change of regional blood flow to the tumor tissue, a factor not directly related to gene expression. In addition, permeation of nucleosides and analogs through the cell membrane of normal and tumor tissues may not share the same mechanism(s). The transport mechanisms, CNT (concentrative nucleoside transport) or ENT (equilibrium nucleoside transport) (35,36) in tumor tissue have not been fully elucidated. At least 3 subtypes of these 2 groups of nucleoside transporters have been cloned (37). Depending on the status of the tumor cells, different membrane transport systems may be employed for various nucleoside analogs. It is likely that the cellular transport mechanism(s) in tumor tissue will be different from that of the normal cells. Enzyme kinetics and substrate preference between HSV1-TK and mammalian TK enzyme will be different. It is known that these two enzymes have different substrate preferences. This disparity is the foundation of selective suicide gene therapy. Such disparities can be enhanced with mutant HSV1-TK enzymes, such as HSV1sr39TK (38,39), which lead to high intracellular trapping of [18F]FHBG due to improved enzyme-substrate specificity. The differences in substrate specificity between Dand L- isomers have not been investigated. It is entirely possible that D- and L-nucleosides may have different Vmax and Km values for HSV1-TK and mutated HSV1-TK enzyme. An improved reporter gene/probe combination may significantly enhance signal to noise ratio determined by PET or SPECT imaging. There are other factors potentially involved in the uptake and retention of the imaging signal. For example, phosphate derivatives of FIAU or other nucleosides may be metabolized subsequent to phosphorylation by HSV1-TK. The metabolic fate of the nucleosides may significantly alter the retention and uptake of the tracers in the tumor. It is also important to consider the carrier level and endogenous ligand competition (ie, thymidine); differences in Ki values between the nucleosides and thymidine may affect the kinetics of the enzymatic reaction. The in vitro and in vivo studies presented in this article clearly suggested that only the

804

Academic Radiology, Vol 12, No 7, July 2005

D-isomer can be readily phosphorylated and trapped in the cells expressing HSV1-TK enzyme, but it may be possible that the kinetics of nucleosides in the blood circulation and the nucleoside transporters on the tumor cell surface may also play a role, in addition to the phosphorylation reaction catalyzed by TK enzymes. Recently, expression of HSV1-tk gene by transiently infected adenovirus CMV-promoter– driven mutant HSV1-tk gene expression (Ad CMVHSV1-sr39tk) showed a dramatically higher uptake and retention using a tracer, [18F]FHBG, as compared to [3H]D-FIAU (40). It was suggested that [18F]FHBG/Ad CMVHSV1-sr39tk as a probe/ gene expression pair is more appealing than other combinations. The results indicated that either total HSV1-TK enzyme is less efficient or the levels in stably expressed cells such as RG2TK⫹ cells were lower as compared to those of Ad CMVHSV1-sr39tk gene transiently expressed cells. The differences of Km/Vmax in these two systems, ie, HSV1-TK versus HSV1-sr39TK, may be important for the higher efficient trapping of [18F]FHBG in the Ad CMVHSV1-sr39tk transiently expressed cells. It is likely that there may be additional changes in cell membrane transport for these two probes in stably expressed cells as compared to Ad CMVHSV1-sr39tk gene transiently expressed cells. The validation of HSV1-tk as a reporter gene for in vivo imaging assays has been of clear benefit in gene therapy trials. In a clinical gene therapy trial of liposome-gene complex (LIPO-HSV1-tk), followed by GCV administration in 5 recurrent glioblastoma patients, preliminary findings showed that PET imaging using [18F]FIAU was able to identify vector-mediated gene expression and monitor the therapeutic effect of GCV treatment (30,41). Gambhir et al constructed triple fusion reporter genes compatible with bioluminescence, fluorescence, and PET imaging. A triple fusion reporter vector consisted of a bioluminescence synthetic Renilla luciferase (hrl) reporter gene, a reporter gene encoding the monomeric red fluorescence protein (mrfp1), and a mutant herpes simplex virus type 1 sr39 thymidine kinase. They validated the activities of all three proteins encoded by the fusion gene in cell culture. They imaged living mice bearing 293T cells transiently expressing the hrlmrfp-ttk vector by microPET and using a highly sensitive cooled charge-coupled device camera compatible with both bioluminescence and fluorescence imaging (40,42). In conclusion, a comparison study of D- and L-isomers of [125I]FIAU in cell uptake and retention using RG2TK⫹ tumor cells showed that only the D-isomer displayed a significantly higher uptake and retention based on the ex-

Academic Radiology, Vol 12, No 7, July 2005

pression of HSV1-TK enzyme. In vivo imaging study using xenograft RG2TK⫹ tumor in nude mice also confirmed that only the D-isomer of [123I]FIAU is highly concentrated in the TK⫹ tumor. Undoubtedly, high-resolution microSPECT imaging can provide a useful tool in monitoring HSV1-tk gene expression using [123I]D-FIAU as the reporter probe. This study suggests a useful approach in molecular imaging by microSPECT. REFERENCES 1. Gumina G, Chong Y, Choo H, Song GY, Chu CK. L-nucleosides: antiviral activity and molecular mechanism. Curr Top Med Chem 2002; 2: 1065–1086. 2. Eriksson S, Munch-Petersen B, Johansson K, Eklund H. Structure and function of cellular deoxyribonucleoside kinases. Cel Mol Life Sci 2002; 59:1327–1346. 3. Wang J, Choudhury D, Chattopadhyaya J, Eriksson S. Stereoisomeric selectivity of human deoxyribonucleoside kinases. Biochemistry 1999; 38:16993–16999. 4. Johansson NG, Eriksson S. Structure-activity relationships for phosphorylation of nucleoside analogs to monophosphates by nucleoside kinases. Acta Biochimica Polonica 1996; 43:143–160. 5. Chu CK, Boudinot FD, Peek SF, et al. Preclinical investigation of L-FMAU as an anti-hepatitis B virus agent. Antivir Ther 1998; 3(Suppl 3):113–121. 6. Chong Y, Chu CK. Understanding the unique mechanism of -FMAU (Clevudine) against hepatitis B virus: molecular dynamics studies. Bioorg Med Chem Lett 2002; 12:3459 –3462. 7. Magrassi L, Finocchiaro G, Milanesi G, Benvenuto F, Spadari S, Focher F. Lack of enantioselectivity of herpes virus thymidine kinase allows safer imaging of gene delivery. Gene Ther 2003; 10:2052–2058. 8. Mercer JR, Xu LH, Knaus EE, Wiebe LI. Synthesis and tumor uptake of 5-82Br- and 5-131I-labeled 5-halo-1-(2-fluoro-2-deoxy-beta-D-ribofuranosyl)uracils. J Med Chem 1989; 32:1289 –1294. 9. Watanabe KA, Reichman U, Hirota K, Lopez C, Fox JJ. Nucleosides. 110. Synthesis and antiherpes virus activity of some 2’-fluoro-2’-deoxyarabinofuranosylpyrimidine nucleosides. J Med Chem 1979; 22:21–24. 10. Gordon M. Severe toxicity of fialuridine (FIAU). N Engl J Med 1996; 334:1136 –1138. 11. Gambhir SS. Molecular imaging of cancer with positron emission tomography. Nat Rev Cancer 2002; 2:683– 693. 12. Liang Q, Gotts J, Satyamurthy N, et al. Noninvasive, repetitive, quantitative measurement of gene expression from a bicistronic message by positron emission tomography, following gene transfer with adenovirus. Mol Ther 2002; 6:73– 82. 13. Nichol C, Kim EE. Molecular imaging and gene therapy. J Nucl Med 2001; 42:1368 –1374. 14. Ray P, Bauer E, Iyer M, et al. Monitoring gene therapy with reporter gene imaging. J Nucl Med 2001; 31:312–320. 15. Sun X, Annala AJ, Yaghoubi SS, et al. Quantitative imaging of gene induction in living animals. Gene Ther 2001; 8:1572–1579. 16. Haberkorn U, Altmann A. Imaging methods in gene therapy of cancer. Curr Gene Ther 2001; 1:163–182. 17. Gambhir SS, Herschman HR, Cherry SR, et al. Imaging transgene expression with radionuclide imaging technologies. Neoplasia 2000; 2(12): 118 –38. 18. Blasberg RG, Tjuvajev JG. Herpes simplex virus thymidine kinase as a marker/reporter gene for PET imaging of gene therapy. Q J Nucl Med 1999; 43: 163–169. 19. Brust P, Haubner R, Friedrich A, et al. Comparison of [18F]FHPG and [124/125I]FIAU for imaging herpes simplex virus type 1 thymidine kinase gene expression. Eur J Nucl Med 2001; 28:721–729. 20. Gelovani Tjuvajev J, Doubrovin M, Akhurst T, et al. Comparison of radiolabeled nucleoside probes (FIAU, FHBG, and FHPG) for PET imaging of HSV1-tk gene expression. J Nucl Med 2002; 43:1072–1083.

IMAGING HSV1-TK GENE EXPRESSION BY [123I]FIAU

21. Haubner R, Avril N, Hantzopoulos PA, Gansbacher B, Schwaiger M. In vivo imaging of herpes simplex virus type 1 thymidine kinase gene expression: early kinetics of radiolabelled FIAU. Eur J Nucl Med 2000; 27:283–291. 22. Tjuvajev JG, Finn R, Watanabe K, et al. Noninvasive imaging of herpes virus thymidine kinase gene transfer and expression: a potential method for monitoring clinical gene therapy. Cancer Res 1996; 56:4087– 4095. 23. Tjuvajev JG, Doubrovin M, Akhurst T, et al. Comparison of radiolabeled nucleoside probes (FIAU, FHBG, and FHPG) for PET imaging of HSV1-tk gene expression. J Nucl Med 2002; 43: 1072–1083. 24. Mueller K, Yagel R, Wheller JJ. Fast implemntation of algebraic methods for three-dimensional reconstruction from cone-beam data. IEEE Trans Med Imag 1999; 18:538 –548. 25. Acton PD, Choi SR, Plossl K, Kung HF. Quantification of dopamine transporters in the mouse brain using ultra-high resolution single-photon emission tomography. Eur J Nucl Med 2002; 29:691– 698. 26. Acton PD, Hou C, Kung MP, Plossl K, Keeney CL, Kung HF. Occupancy of dopamine D2 receptors in the mouse brain measured using ultra-high-resolution single-photon emission tomography and [123]IBF. Euro J Nucl Med Mol Imag 2002; 29:1507–1515. 27. Acton PD, Kung HF. Small animal imaging with high resolution single photon emission tomography. Nucl Med Biol 2003; 30:889 – 895. 28. Auricchio A, Acton PD, Hildinger M, et al. In vivo quantitative noninvasive imaging of gene transfer by single-photon emission computerized tomography. Hum Gene Ther 2003; 14:255–261. 29. Sharma V, Luker GD, Piwnica-Worms D. Molecular imaging of gene expression and protein function in vivo with PET and SPECT. J Magn Reson Imaging 2002; 16:336 –351. 30. Shah K, Jacobs A, Breakefield XO, Weissleder R. Molecular imaging of gene therapy for cancer. Gene Ther 2004; 13:13. 31. Blasberg RG. Molecular imaging and cancer. Mol Cancer Ther 2003; 2: 335–343. 32. Tjuvajev JG, Stockhammer G, Desai R, et al. Imaging the expression of transfected genes in vivo. Cancer Res 1995; 55:6126 – 6132. 33. Gambhir SS, Barrio JR, Herschman HR, Phelps ME. Assays for noninvasive imaging of reporter gene expression. Nucl Med Biol 1999; 26: 481– 490. 34. Deng W-P, Yang WK, Lai W-F, et al. Non-invasive in vivo imaging with radiolabelled FIAU for monitoring cancer gene therapy using herpes simplex virus type 1 thymidine kinase and ganciclovir. Euro J Nucl Med Mol Imag 2004; 31:99 –109. 35. Hyde RJ, Cass CE, Young JD, Baldwin SA. The ENT family of eukaryote nucleoside and nucleobase transporters: recent advances in the investigation of structure/function relationships and the identification of novel isoforms. Mol Membr Biol 2001; 18: 53– 63. 36. Cabrita MA, Baldwin SA, Young JD, Cass CE. Molecular biology and regulation of nucleoside and nucleobase transporter proteins in eukaryotes and prokaryotes. Biochem Cell Biol 2002; 80: 623– 638. 37. Yao SY, Ng AM, Vickers MF, et al. Functional and molecular characterization of nucleobase transport by recombinant human and rat equilibrative nucleoside transporters 1 and 2. Chimeric constructs reveal a role for the ENT2 helix 5-6 region in nucleobase translocation. J Biol Chem 2002; 277: 24938 –24948. 38. Liang Q, Nguyen K, Satyamurthy N, et al. Monitoring adenoviral DNA delivery, using a mutant herpes simplex virus type 1 thymidine kinase gene as a PET reporter gene. Gene Ther 2002; 9: 1659 –1666. 39. Gambhir SS, Bauer E, Black ME, et al. A mutant herpes simplex virus type 1 thymidine kinase reporter gene shows improved sensitivity for imaging reporter gene expression with positron emission tomography. Proc Natl Acad Sci U S A 2000; 97: 2785–2790. 40. Min JJ, Iyer M, Gambhir SS. Comparison of [18F]FHBG and [14C]FIAU for imaging of HSV1-tk reporter gene expression: adenoviral infection vs stable transfection. Euro J Nucl Med Mol Imag 2003; 30: 1547– 60. 41. Jacobs A, Voges J, Reszka R, et al. Positron-emission tomography of vector-mediated gene expression in gene therapy for gliomas. Lancet 2001; 358(9283): 727–729. 42. Ray P, De A, Min JJ, Tsien RY, Gambhir SS. Imaging tri-fusion multimodality reporter gene expression in living subjects. Cancer Res 2004; 64: 1323–30.

805