AC Research
Accelerated Articles Anal. Chem. 1997, 69, 275-280
Supercritical Fluid Extraction of 11C-Labeled Metabolites in Tissue Using Supercritical Ammonia Gunilla B. Jacobson,†,‡ Robert Moulder,‡ Li Lu,‡ Mats Bergstro 1 m,‡ Karin E. Markides,§ and Bengt La˚ngstro 1 m*,†,‡
Uppsala University PET Centre, Department of Analytical Chemistry, Department of Organic Chemistry, Institute of Chemistry, and The Subfemtomole Biorecognition Project,| Uppsala University, 751 21 Uppsala, Sweden
Supercritical fluid extraction (SFE) of 11C-labeled tracer compounds and their metabolites from biological tissue was performed using supercritical ammonia in an attempt to develop a rapid extraction procedure that allowed subsequent analysis of the labeled metabolites. Metabolites were extracted from kidneys and brain in rats given in vivo injections of the radiotracers O-[2-11C]acetyl-Lcarnitine and N-[11C]methylpiperidyl benzilate, respectively. Only a minimal sample pretreatment of the tissue was necessary, i.e., cutting into 10-20 pieces and mixing with the drying agent Hydromatrix, before it was loaded into the extraction vessel. Extraction efficiency was measured for SFE at temperatures over the range of 70150 °C and a pressure of 400 bar. For O-[2-11C]acetylL-carnitine, 66% of the radioactivity was trapped in the collected fractions and 12% remained in the extraction vessel. For the more lipophilic N-[11C]methylpiperidyl benzilate, 93% of the activity was collected and less than 1% remained in the extraction vessel. Labeled metabolites were analyzed by LC and also, in the case, of O-[211C]acetyl-L-carnitine by LC/MS. The complete extraction procedure, from removal of the biological tissue until an extract was ready for analysis, was 25 min, corresponding to about one half-life of the radionuclide 11C. Compounds labeled with short-lived β+-emitting radionuclides (e.g. 11C t1/2 ) 20.3 min; 18F t1/2 ) 109.7 min) are used for in vivo and in vitro studies of physiological and biochemical processes. The distribution of the radioactivity representing the original tracer and its metabolites in the body may be followed by positron †
Department of Organic Chemistry. Uppsala University PET Centre, Subfemtomole Biorecognition Project. Department of Analytical Chemistry. | The Subfemtomole Biorecognition Project is a joint endeavor between Uppsala University and Research Development Corp. of Japan. ‡ §
S0003-2700(96)00786-X CCC: $14.00
© 1997 American Chemical Society
emission tomography (PET).1 In order to interpret the PET data correctly, it is often desirable to perform complementary small animal experiments, especially with respect to formation of metabolites and their contribution to the radioactivity as measured in tissue and blood. If labeled metabolites are to be identified, rapid extraction procedures are required due to the decay of the radionuclide. The use of supercritical fluid extraction (SFE) has been shown to offer a quick alternative compared to conventional liquid solvent extraction techniques.2 Some supercritical fluids are gases at ambient conditions and will vaporize upon trapping. This can be of benefit if trace amounts of analytes are to be extracted since no concentration step is necessary before analysis. These favorable aspects of SFE can be utilized and are of great benefit for extraction of 11C-labeled metabolites from biological tissues, where it is desired to use a minimum of sample pretreatment and produce extracts rapidly for analysis. The most commonly used supercritical fluid for SFE is carbon dioxide, and modifiers, such as methanol, are often added to increase the solubility of polar analytes.3 For certain applications a more polar supercritical fluid would be desirable. Most polar solvents have high critical temperatures that can cause degradation of the analytes, e.g., methanol (Tc ) 239.5 °C, Pc ) 80.9 bar) and water (Tc ) 374.1 °C, Pc ) 220.5 bar), although subcritical water could be another viable option. An attractive alternative is ammonia (Tc ) 132.5 °C, Pc ) 113.1 bar) due to its solvent strength and more favorable critical point. The use of supercritical ammonia has been limited due to practical difficulties associated with its corrosive nature. Careful considerations must be taken (1) Wagner, H. N., Jr.; Szabo, Z.; Buchanon, J. W. Principles of Nuclear Medicine, 2nd ed.; W. B. Saunders: Philadelphia, PA 1995. (2) Chester, T. L.; Pikston, J. D.; Raynie, D. E. Anal. Chem. 1994, 66, 106R130R. (3) (a) Lee, M. L., Markides, K. E. Eds. Analytical Supercritical Fluid Exraction and Chromatography, Chromatography Conference, Inc.: Provo, UT, 1990. (b) Paulaitis, M. E.; Krukonis, V. J.; Kurnik, R. T. Rev. Chem. Eng. 1983, 1, 179-250.
Analytical Chemistry, Vol. 69, No. 3, February 1, 1997 275
Chart 1. O-[2-11C]Acetyl-L-carnitine (1) and N-[11C]Methylpiperidyl Benzilate (2)
Figure 1. Schematic of the SFE system: (A) cryostat, -10 °C, (B) pump, (C) pulse dampener, (D) oven. V1 and V2 are three-port twoposition valves.
in choosing instrumentation and often modified versions of commercial instrumentation, must be used.4 In this paper, the use of supercritical ammonia for SFE of 11Clabeled metabolites from biological tissue is presented. The SFE was performed using a system designed for supercritical fluid synthesis and on-line preparative SFC in supercritical ammonia, which has been described previously.5 By replacing the reaction cell with an extraction vessel, both static and dynamic SFE could be performed at 20-150 °C and 9-400 bar in this system. Two different 11C-labeled tracers, O-[2-11C]acetyl-L-carnitine (1) and the more lipophilic N-[11C]methylpiperidyl benzilate (2) were used in the study. The tracers were injected into rats, and after a short distribution time, the rats were sacrificed and the organs with highest uptake of radioactivity were removed. From previous studies in monkeys, it was known that a high uptake of N-[11C]methylpiperidyl benzilate was obtained in the brain, but for O-[211C]acetyl-L-carnitine, the radioactivity distribution in the rat was investigated to identify a suitable organ. The extraction efficiency for static SFE with supercritical ammonia for the labeled compounds from rat tissue was studied by measuring the radioactivity distribution from the sample loaded in the extraction vessel and into the collected fractions. The trapped extracts were analyzed by LC, LC/MS, and radio techniques. EXPERIMENTAL SECTION General Information. O-[2-11C]Acetyl-L-carnitine6 (1) and 11 N-[ C]methylpiperidyl benzilate7 (2) were synthesized at the Uppsala University PET Centre according to published procedures. The radiochemical purity of the tracers was >98%, and they were obtained in a phosphate-buffered (pH 7) solution. Routine LC analyses were performed with a Beckman 126 gradient pump (Beckman Instruments, Inc., Fullerton, CA) and a Beckman 166 variable-wavelength UV detector in series with a β+-flow detector.8 Both detectors were connected via a Beckman AI 406 interface to the Beckman System Gold Chromatography software package for data processing. The columns used were (4) (a) Grolimund, K.; Jackson, W. P.; Joppich, M.; Nussbaum, W.; Anton, K.; Widmer, H. M. In Proceedings of the 7th International Symposium on Capillary Chromatography; Ishii, D., Jinno, K., Sandra P., Ed.; Gifu, Japan, May 11-14, 1986; pp 625-636. (b) Kuie, J. C.; Markides, K. E.; Lee, M. L. J. High Resolut. Chromatogr. Chromatogr. Commun. 1987, 10, 257-262. (5) (a) Jacobson, G. B.; Markides, K. E.; La˚ngstro¨m, B. Acta Chem Scand. 1994, 48, 428-433. (b) Jacobson, G. B.; Westerberg, G.; Markides, K. E.; La˚ngstro ¨m B. J. Am. Chem. Soc. 1996, 118, 6868-6872. (C) Jacobson, G. B.; Markides, K. E.; La˚ngstro ¨m, B. Acta Chem Scand., in press. (6) Jacobson, G. B.; Watanabe, Y.; Valind, S.; Kuratsune, H.; La˚ngstro ¨m, B. Nucl. Med Biol., in press. (7) Mulholland, G. K.; Kilbourne, M. R.; Sherman, P.; Carey, J. E.; Frey, K. A.; Koeppe, R. A.; Kuhl, D. E. Nucl. Med. Biol. 1995, 22, 13-17. (8) Modified version of the radio detector described in: B. La˚ngstro ¨m, Lundqvist, H. Radiochem. Radioanal. Lett. 1979, 41, 375-382.
276 Analytical Chemistry, Vol. 69, No. 3, February 1, 1997
(A) Beckman Ultrasil NH2 (4.6 × 250 mm; 10 µm) and (B) Sherisorb ODS1 (4.6 × 250 mm; 5 µm) with mobile phases (C) 0.01 M potassium dihydrogen phosphate (pH 4.6), (D) 0.05 M ammonium formate (pH 3.5), (E) methanol, and (F) acetonitrile/ water 50:7 (v/v). The LC/MS equipment consisted of Beckman 126 and 116 pumps, a CMA 240 autosampler (CMA Microdialysis, Stockholm, Sweden), and a VG Quattro triple-quadrupole mass spectrometer equipped with pneumatically assisted electrospray (3 kV cone voltage) (Fisons Instruments, Cheshire, UK). A postcolumn 1:100 split was used, with 1% of the total flow delivered to the electrospray probe and 99% to a Beckman 166 variable-wavelength UV detector followed by a Flow-Count β+-detector (Bioscan, Inc., Washington, DC). The columns used were (G) Kromasil C-18 (4.6 × 100 mm, 5 µm; Chromtech, Sweden) with a guard column (4.6 × 10 mm; Technosphere, HPLC Technology Ltd.) and (H) YMC-AL C-18 (4.6 × 100 mm, 5 µm; Chromtech), with mobile phases (I) water, (J) acetonitrile, (K) 100 mM acetic acid in water/ acetonitrile 1:1 (v/v), and (L) 75 mM acetic acid and 25 mM ammonia in water/acetonitrile 1:1 (v/v). Measurements of radioactivity in LC fractions and rat organs were made in a Na(Tl) well counter. The data were collected with a scaler (Newport P6000) and decay corrected. Animal Experiments. Male Sprague-Dawley rats, 10-12 weeks old and with weights of ∼350 g, were used. The studies were performed with permission from the Research Animal Ethical Committee of Uppsala, no. C184/95. A total of 1 mL of the 11Clabeled tracer in saline was injected into the tail vein of the unanesthesized rat. After predefined times, the rats were briefly anesthesized and sacrificed by decapitation, before the selected organs were removed. The radioactivity and weight of the organs were measured, and the SUVs (standard uptake value) were calculated as the organ radioactivity concentration divided by the ratio of total injected radioactivity and rat weight. SFE Procedure. A 2.2 mL (5 × 100 mm) stainless steel extraction vessel (Keystone Scientific Inc.) with 0.5 µm frits was used. The rat organs (half of a kidney or brain) were cut into ∼10-20 pieces (approximate dimensions 2 mm cubes) and mixed with an equal volume of Hydromatrix (purified diatomaceous earth; International Sorbent Technology, Wales, UK). A plug of Hydromatrix was loaded into the bottom of the extraction vessel before loading the sample mixture. The SFE was performed with a supercritical fluid system previously designed for synthesis and on-line SFC with supercritical ammonia, which has been described in detail elsewhere.5 A simplified scheme of the system, containing only the instrumentation involved in the extraction mode, is shown in Figure 1. The main parts of the system were a pump, Gilson Model 308, a pressure regulator, Model 821, an oven, Model 831, and a cryostat refrigerating circulator. Valves V1 and V2 were three-port two-position valves (A3C3WEHCY, Valco
Instruments Co. Inc., Houston, TX) located inside the oven. The tubing in the system was 1/16 in. o.d. 316 grade stainless steel. From the cryostat to valve V1 the inner diameter of the tubing was 0.127 mm and from V1 to the trapping solution, 0.25 mm. Careful considerations were taken in the choice of material in the system due to the corrosive nature of supercritical ammonia. Check valves made of poly(vinyl difluoride) (PVDF) and a Teflon (PTFE)-graphite piston seal in the pump head were used. The Valco valves (V1 and V2) contained rotor material of Valcon E which were resistant to ammonia at the pressures and temperatures used. The extraction vessel was placed inside the oven, which had been preheated to the desired temperature. With valve V2 closed, ammonia was pumped at 0.5 mL/min until the desired pressure had been obtained inside the extraction vessel. Valve V1 was then closed, and static SFE was performed for 5 min. Valves V1 and V2 were opened, and the extracts were trapped in 1 mL of water by positioning the outlet tubing into the trapping solvent, contained in a 20 mL glass vial fitted with a septum and a vent. The collected fractions were heated (∼40 °C) under reduced pressure for 2 min to remove most of the ammonia before further analysis. When working with supercritical ammonia, high-pressure and temperature-resistent instrumentation is required and the experiments should be performed in a well-vented area. Due to safety considerations, all SFE experiments were performed in a leadshielded hood. Body Distribution of O-[2-11C]Acetyl-L-carnitine. Five groups of rats, four rats in each group, were used in the experiment. The first four groups were injected with 10 MBq of 1 and were sacrificed after 1, 5, 10, or 30 min. The fifth group was injected with 500 MBq of 1 (the high-dose group) and were sacrificed after 5 min. The brain, heart, lungs, liver, kidneys, and a blood sample were removed, and the radioactivity and weight of the samples were measured. Extraction Efficiency. Four rats were injected with 15-36 MBq of 1. They were sacrificed after 5 min, and the kidneys were removed. Another four rats were injected with 10 MBq of 2 and after 5 min the brain was removed. The radioactivity and weight of the organs were measured prior to loading the SFE vessel. A total of nine static SFEs were performed with each tissue sample (two at 70 °C, three at 100 °C, and four at 150 °C), with a pressure of 400 bar. Five minute fractions were collected in 1 mL of water at room temperature. The radioactivity in the collected fractions and residues in the extraction vessel was measured and compared to the radioactivity loaded into the extraction vessel. A second set of static SFE was performed as above from a mixture of 350 µL of tracer solution and 2 mL of Hydromatrix. The extracts were collected in either 1 mL of water, as above, or in a 1:1 mixture of 30 mL of 0.5 M HCl/ethanol, cooled in an ice bath. These blank experiments were performed to compare the extraction efficiency and complete recovery with and without any tissue. The extracts were also analyzed by LC (see below for conditions) to verify whether the tracers were intact after exposure to supercritical ammonia or whether degradation products were produced. Metabolite Analysis. In different experiments, 15 rats were injected with ∼1 GBq of 1 and 2 rats with 1 GBq of 2. They were sacrificed 5 min after injection and the kidneys and brain removed.
Static SFE was performed for 3 min at 400 bar and 100 °C. During a 10 min heating period (100-140 °C), dynamic SFE was performed with a flow of 0.5 mL/min. A second 3 min static SFE was performed at 400 bar and 150 °C. All extracts were trapped in the same fraction. The total extraction time was 16 min. The metabolites from 1 were analyzed by LC on column A with mobile phases C/F of the following conditions: linear gradient 5-10 min 84-20% F, column temperature 40 °C, flow 2.0 mL/min, wavelength 230 nm. References of acetic acid, acetylcarnitine, and glutamic acid eluted at 3.8, 7.4, and 10.6 min, respectively. The metabolites from 2 were analyzed by LC on column B with mobile phases D/E 65:35 (v/v) under the following conditions: column temperature 40 °C, flow 2.0 mL/min, wavelength 254 nm. The retention time of 2 was 5.9 min. The eluent from the LC analysis were collected in 30 s intervals during 20 min. The radioactivity in the 40 fractions was measured in the well counter. The metabolites from 1 were also analyzed by LC/MS with positive electrospray (cone voltage 30 V) on column G with mobile phases I/J/K 90:7:3 (v/v/v) and on column H with mobile phases I/J/L 97:1:2 (v/v/v). RESULTS AND DISCUSSION In order to validate the efficiency of the SFE method of 11Clabeled metabolites with supercritical ammonia, two 11C-labeled tracers of different polarity were selected. The endogenous O-[211C]acetyl-L-carnitine functions as an acetyl group transporter, acting in the inner mitochondria membrane. It is of great interest in PET studies of, for example, the pathways of the tricarboxylic acid (TCA) cycle.6,9 N-[11C]Methylpiperidyl benzilate is a muscarinic antagonist and has been used in PET studies of the cholinergic system.10 The tracers were injected into rats, and the organ with highest uptake of radioactivity was selected to obtain enough radioactivity for further analysis of the extracted metabolites by LC and LC/ MS. The earlier experience of 2 permitted the brain to be selected for this tracer, but for 1 a body distribution study was performed to identify a suitable organ. The activity distribution of 1 related to time after injection was evaluated in several organs and in blood (Figure 2). The highest relative uptake was found in the kidneys. At 5 min after injection the relative SUV was 8.7. In all other cases, the SUV values were lower than 2. These results were compared to injecting 50 times higher radioactivity of 1 (high-dose group) with a distribution time of 5 min, to see whether a higher injected dose would effect its distribution in the body. The high-dose group gave a slightly lower SUV value in the kidneys compared to the 5 min group having a lower injected dose, but when the data were normalized to the radioactivity found in blood, no obvious change was observed. The kidneys were selected as a suitable organ with a tracer distribution time of 5 min, for the remainder of the study, due to the highest uptake of radioactivity. In the initial attempts to extract the tissue matrix alone, problems with blockage of the extraction cell frits were encountered. This problem occurred with both kidney and brain and (9) Yamaguti, K.; Kuratsune, H.; Watanabe, Y.; Takahashi, M.; Nakamoto, I.; Machii, T.; Jacobson, G. B.; Onoe, H.; Matsumura, K.; Valind, S.; La˚ngstro ¨m, B.; Kitani, T. Biochem. Biophys. Res. Commun., submitted. (10) Shinotoh, H.; Asahina, M.; Inoue, O.; Suhara, T.; Hirayama, K.; Tateno, Y. J. Neural Transm.: Parkinson’s Dis. Dementia Sect. 1994, 7, 35-46.
Analytical Chemistry, Vol. 69, No. 3, February 1, 1997
277
Figure 2. Uptake of O-[2-11C]acetyl-L-carnitine in rat tissues at different times after iv injection. Expressed as SUV.
Figure 3. LC chromatogram and radioactivity from N-[11C]methylpiperidyl benzilate extracts. UV: The UV signal from the sample spiked with N-methylpiperidyl benzilate (tr ) 5.9 min). Radio: The radioactivity in the eluate, collected every 30 s.
Table 1. Extraction Efficiency of Supercritical Ammonia for 11C Metabolitesa activity trapped (%)b 1 fractn 1 2 3 4 5 6 7 8 9
2
T (°C)
tissuec
blankd
tissuec
blankd
70 70 100 100 100 150 150 150 150
17 ( 13 1(1 13 ( 9 4(1 3(2 22 ( 11 3(2 2(1 1(1
75 5 1
