NMR IN BIOMEDICINE, VOL. 6 , 27-31 (1993)

In Vivo 19F Nuclear Magnetic Resonance of a Monofluorinated Neuroleptic in the Rat?

Ulrich Giinther and Klaus Albert* Institut fur Organische Chemie der Universitlt, Auf der Morgenstelle 18, D-7400 Tiibingen, Germany

In viuo 'q NMR measurements of the fluorinated neuroleptic melperone [4'-fluoro-4-(4-methyIpiperidino)- butyrophenone hydrochloride] in the rat brain were performed using a geometrically optimized surface coil at 4.7T. It was possible for the first time to detect a signal of a monofluorinated neuroleptic drug with a time resolution of 30 min after i.p. application. The kinetic time course of the investigated neuroleptic melperone was recorded over 6 h and showed that the half-life in the rat brain is 4.3 h. The total amount of drug and its metabolites in the brain was estimated to be 50 p ~ . the chemical shift of the '% NMR signal shows the same upfield shift relative to that in aqueous solution as has been reported for trifluorinated neuroleptic drugs.

INTRODUCTION

Neuroleptic drugs are indispensible for the therapy of psychiatric diseases. Unfortunately it is not possible to define a general therapeutic index for any of these drugs. This is due to the fact that the neuroleptics pass the endothelia/glia (blood/brain) barrier at unpredic- table rates which seriously impairs clinical application and gives a poor correlation between plasma and brain levels of the neuroleptic. I9F in v i m NMR spectroscopy using surface coils is a very valuable method for non- invasive detection and quantification of fluorinated drugs and their metabolites in living organisms. This has been demonstrated by an increasing number of studies. 'L'' Our former investigations dealt with neuro- leptics carrying CF3 substituents, in the course of which it was possible to determine the half-life ( t I l2 ) of trifluo- perazine in the rat brain.'-4 Here we report on the first in vivo observation of a monofluorinated neuroleptic drug in the brain of a rat. Melperone has been selected for this study because it is rapidly metabolized after injection so that kinetic studies are facilitated. The structure of melperone is shown in Fig. 1.

NMR investigations can be performed with the naturally occurring fluorine isotope I9F which is dis- tinguished by a high sensitivity for NMR experiments (83% relative to 'H) and very low background levels in biological tissues.

Positron emission tomography (PET) is an alterna- tive technique which is superior to NMR in sensitivity.'* However, the applicability of PET is basically limited by the fact that it requires labelling with positron- emitting nuclei with a relatively short tl12 ( 18F: tI lz = 110 min; "C: f I l 2 = 20 min). Long-term kinetic stu- dies would require unacceptably high radiation expo- sure so that pharmacokinetic examinations are limited in length. The synthesis, purification and application of the labelled drugs must be performed within two or three fIl2 values, requiring considerable investment in

t Dedicated to Professor Ernst Bayer on the occasion of his 65th birthday.

Author to whom correspondence should be addressed. Abbreviations used: PET, positron emission tomography.

F 0: - cH,- cH,- cH,- N 3 CH,

Figure 1. Chemical structure of melperone.

technical facilities such as an on-site cyclotron. Finally NMR offers additional chemical information in chemi- cal shifts and linewidths.

EXPERIMENTAL

Drugs and chemicals

Melperone (as the hydrochloride salt) was donated by Nordmark (Uetersen, Germany). All other chemicals were obtained from Merck (Darmstadt, Germany)

* TURBINE

C O L

Figure 2. Experimental set-up for in vivo NMR measurements: (1) inhalation pump; (2) pC0, measuring instrument; (3) pulse- oxymeter sensor.

0952-3480/93/010027-05 $07.50 0 1993 by John Wiley & Sons, Ltd.

Received 2 December I991 Accepted (revised) 9 March 1992

28 U. GUNTHER AND K . ALBERT

l " " l " " I " " I " " 1 " ' -103.85 -103.90 -103.95 -104.00 -104.05

PPM

Figure 3. High-resolution 19F NMR spectrum of melperone.

In uitro NMR

All NMR spectra were recorded using a Bruker MSL 200 spectrometer (Bruker, Rheinstetten, Germany) with a 4.7 T magnet (150 mm vertical bore) equipped with an Aspect 3000 computer. High-resolution "F NMR spectra were recorded with a 10mm high- resolution probehead at 188.3 MHz. Samples contained a coaxial insert with a reference solution of trifluoro- ethanol in D 2 0 for calibration of chemical shifts and for 'H lock. TI determinations were performed with an inversion-recovery sequence.

In uiuo NMR

All studies using surface coils were performed with a Bruker in uiuo probehead with exchangeable surface coils. Five different surface coils with different dia- meters (13-19 mm) and geometries (circular and oval) were tested. For each of these surface coils a small spherical phantom of 4 mm diameter was positioned at

three different distances from the coil and the duration of the pulses was increased from 2 to 50 ps in order to determine the individual flip angle distributions. The parameters for the in uiuo experiments were selected so that the maximum amount of signal came out of the centre of the skull of the rat lying directly in front of the coil. For a two-turn spiral type surface coil with 19 rnm OD this can be either achieved by a pulse duration of 20 ps and full relaxation of the ''F nuclei or by a pulse duration of 15 ps and a relaxation delay of 500 ps describing an Ernst angle (56") for the 19F nuclei in the brain of the rat (0 = arccos exp( -TR/Tl)]. The latter localization method implies a flip angle of > 150" close to the coil so that signal contributions from tissues overlying the skull cannot contribute significantly to the total amount of the signal. The flip angle distribution and the absolute sensitivity was less satisfactory for all surface coils with a smaller diameter. An oval two-turn spiral type coil (14 X 22 mm) gave comparable results.

The probehead contained a plexiglass animal holder which could be moved vertically relative to the surface coil. After the rat was anaesthetized, it was placed in

IN VZVO "F NMR STUDIES 29

the probe in a horizontal position; its legs and head were fixed with adhesive tape so that its calvarium was lying directly above the surface coil (Fig. 2). The probehead was then slowly rotated into the vertical position and then placed in the magnet. The static magnetic field homogeneity was optimized with the probehead tuned to 'H by shimming on the 'H FID from tissue water until a linewidth of 15-25 Hz at half- height was achieved.

The calibration of chemical shifts was performed following a procedure proposed in Griin et al.,' which uses the resonance frequency of free water as a refer- ence for the chemical shift. After shimming, the B, magnetic field was adjusted to place the resonance signal of tissue water at a frequency offset of k 2-3 Hz from the basic spectrometer frequency (200.13 Hz). After the surface coil was retuned to the resonance frequency of I9F (188.3 MHz), the acquisition of the in uiuo spectrum was performed. Subsequently the same process was repeated with a test tube (2 cm long, 5 mm OD) containing an aqueous solution of trifluoroethanol whereby the field was readjusted in order to put the water signal on resonance. The reproducibility of this procedure was checked using different phantoms indi- cating that the accuracy of the resulting chemical shift values was in the range of kO.1 ppm.

Data acquisition was performed using a standard pulse sequence with phase cycling for quadrature detec- tion. The in uivo spectra were acquired with a pulse

b) Phantom in rat probehead

Linewidth 23 Hz a>

High resolution

~ ~ ~ ~ ~ I - - ~ I ' ' ~ ~ , ' . . T I " ' . I . -95 -100 -105 -110 -11s PPr

Figure 4. Comparison of the melperone signal: (a) in aqueous solution under high-resolution conditions; (b) in aqueous solu- tion in the in vivo probehead; (c) in vivo in the brain of a living rat.

width of 15 ys, a TR of 500 ms, and a spectral width of 15 kHz. Before Fourier transformation the FID data were treated with an exponential line broadening of 80 Hz.

Treatment of animals

Male Wistar rats weighing 200-250g were used. A single dose of 50 mg/kg of melperone dissolved in 1.5 mL of saline was injected intraperitoneally. In order to minimize the time delay between the injection of the drug and the beginning of the investigation, the animals were first anaesthetized with urethane (200% aqueous solution, 4-6 mL/kg) and fixed in the probe- head. After the field homogeneity had been adjusted the probe was removed from the magnet for drug administration. Subsequently the acquisition of data could be started within 20 s.

In a second experiment a rat was anaesthetized with halothane. For this purpose a Siemens Elma Servo Ventilator 900C was adopted for the conditions of the NMR experiment. The arrangement for this procedure is shown in Fig. 2. In order to monitor some physiologi- cal parameters the end expiratory pCOz (partial pres- sure) was controlled using a DATEX C02 monitor (Normokap, Helsinki, Finland). In addition the pulse of the animal was permanently monitored with a Hewlett Packard pulse-oxymeter sensor which was fixed to the tail of the rat.

High performance liquid chromatography

For quantitative determination of the concentrations of melperone, the weighed brains were homogenized after addition of a 50 mM phosphate buffer. The brain homo- genates were centrifuged at 105 000 g (Beckman ultra- centrifuge). The pellets were washed and centrifuged again. The combined supernatants were used for HPLC investigations without further purification. HPLC was performed on a nucleosil C18 column (250 x 4.6 mm). Referencing was achieved using melperone solutions with various concentrations as an external standard. The mobile phase consisted of acetonitrile and water (20530 v/v), adjusted to pH 2.2 with phosphate buffer. The pump rate was lmL/min, the HPLC pump employed was a Bruker LC 31. The signals were detected at a wavelength of 254 nm using a UV detector (Hitachi UV 655 A). Peaks were identified by compari- son with the retention times of the reference solutions which were also used for quantification.

RESULTS AND DISCUSSION

In vitro NMR experiments

Figure 3 shows the high-resolution I9F spectrum of rnelperone in D 2 0 indicating that the fluorine reso- nance is split by proton coupling. The 35 coupling constant of 8.8 Hz is caused by the protons in positions 2 and 6, resulting in a triplet which is again split by the protons in positions 3 and 5 with a 4J coupling constant

30 U. GUNTHER AND K. ALBERT

F Br I I

I I F - C- C-H

F C1

Halothane

Melperone

I - I . I - r . I . I - I ' I ' - T -50 -60 -70 -80 -90 -100 -110 -120

PPY

Figure 5. In vivo ''F NMR spectrum in the brain of a rat under halothane narcosis after application of 50 mg/kg of melperone.

of 5.5 Hz. The chemical shift of the center line lies at -103.9 ppm (relative to trifluorethanol at -76.0 ppm). Relaxation measurements of aqueous melperone solu- tions yielded a T, = 3600 ms which was reduced in brain

t,,* = 4.3 h

time [min]

Figure 6. Semi-logarithrnic plot of the concentration of melper- one a t a dose of 50 mg/kg.

homogenate to 865 ms. The detection limit of 19F NMR spectroscopy was examined with melperone solutions of different concentrations. The minimum concentra- tion necessary in a volume of 2 mL is 10 VM for detec- tion within 30 min.

In uitro and in uiuo NMR experiments

Figure 4(c) shows the NMR spectrum of melperone in the brain of a rat. The dosage for this experiment was 50mg/kg body wt; the rat was anaesthetized with urethane. After this spectrum was obtained the kinetic time course for melperone concentration was moni- tored over 4 h with a time resolution of 30 min for each spectrum (Fig. 5) . The signal intensity decayed expo- nentially with a t l l2 of 4.3 h.

The NMR signal represents a sum of specifically and non-specifically bound melperone. After a single dose application of melperone the contribution of metabo- lites to the NMR signal can be neglected. This has been proved by HPLC examinations of brain homogenates.

I N VIVO "F NMR STUDIES 31

The linewidth of the signal is at least 230 Hz (Fig. 4). An analogous line-broadening effect had been observed for other neuroleptic drugs such as fluphena- zine and trifluoperazine. 1-6 It is likely to be a direct consequence of an immobilization of the drug in the brain by binding to receptors and membranes. A simi- lar line broadening has been observed for 5-fluorouracil in blood plasma due to binding to plasma proteins.8

Another characteristic of the I9F in uiuo NMR signal of neuroleptics is the fact that it is shifted upfield relative to the signal in aqueous solution. The chemical shift of the in uiuo resonance of melperone lies between -104.5 ppm and -104.7 ppm, the signal of an aqueous solution has a chemical shift of - 103.9 ppm (relative to trifluoroethanol with 6 = -76.0 pprn). A high-field shift of the same amount has been observed for trifluori- nated neurolept ic~.~ This cannot be explained by local pH effects in the brain because the 19F chemical shift of melperone is independent of pH. However, it is well- known that the chemical shift of 19F nuclei changes significantly with the solvent. This gives rise to the assumption that the drug must be located in a non- aqueous binding site in the brain.

If a halothane narcosis is applied two halothane signals with chemical shifts of -75.3 and -79.0 pprn are observed (Fig. 6). The first signal (-75.3 ppm) can be related to the halothane in the brain of the animal which is dissolved in membranes and undergoes chemi- cal shift changes in the hydrophobic environment in the brain. The signal at -79.0ppm comes from gaseous halothane in the probe and is a very precise chemical shift reference. The origin of this signal was proved by Litt et ~ 2 1 . ' ~ by blowing air into the probe in order to

remove this signal. Comparing examinations showed that the signals of gaseous halothane have exactly the same chemical shift of -79.0 ppm.

By comparison of the absolute I9F NMR signal inten- sity for the in uiuo experiment vs an in uitro phantom with a volume similar to that of the rat brain (2 mL), it was possible to give a rough estimation of the concen- tration of the drug in the brain. The phantom sample was doped with CuSO,, in order to reduce its T I to a value in the order of 850ms which had been deter- mined for brain homogenate. According to this estima- tion, which did not consider the different linewidths in the phantom (ca 70 Hz) and the rat brain homogenate (230 Hz), the concentration of the drug 1 h after appli- cation is in the order of 50 p ~ .

This agrees well with the results of HPLC investi- gations which indicate that the maximum melperone concentration (ca 1 h after application of the drug) is 45 p ~ . In conclusion, the present data demonstrate that I9F in uiuo NMR might become a valuable tool for the determination of drug levels of >10 p ~ . At a dosage of 50 mg/kg of the monofluorinated neuroleptic drug mel- perone, it is possible to detect a I9F NMR signal within 30min in the brain of a living rat and to follow its kinetic course for 240 min.

Acknowledgements

We thank Nordmark for financial support of this work. We are indebted to Dr B. Kottler of the Institute for Anaesthesiology of the University of Tiibingen for his instructions concerning anaesthesia of rats and to Dr M. Bartels of the Institute of Psychiatry of the University of Tiibingen for his help in handling the animals.

REFERENCES

1.

2.

3.

4.

5.

Albert, K., Bartels, M., Kruppa, G., Mann, K., Schroth, G., Tabarelli, S. and Zabel. M. Untersuchung fluorsubstituierter Psychopharmaka mittels NMR. In Die NMR Tornographie, Klinischer Einsatz und Wirtschaftlichkeit, ed. by P. Pfannenstiel and M. Meves, pp. 34-36. G. Thieme Verlag, Stuttgart (1984). Albert, K., Rembold, H., Kruppa, G., Bayer, E., Bartels, M. and Schmalzing, G. In vivo "F nuclear magnetic resonance spectroscopy of trifluorinated neuroleptics in the rat. NMR Biomed. 3, 120-123 (1990). Bartels, M., Albert, K., Kruppa, G., Mann, K., Schroth, G., Tabarelli, S. and Zabel, M. Fluorinated psychopharmacolo- gical agents: noninvasive observation by fluorine-19 nuclear magnetic resonance. Psychiat. Res. 18, 197-201 (1 986). Bartels, M., Gunther, U., Albert, K., Mann, K., Schuff, N. and Stuckstedte, H. "F-magnetic resonance spectroscopy of neuroleptics: the first in vivo pharmakokinetics of trifluoper- azine in the rat brain and the first in vivo spectrum of fluphenazine in the human brain. Biol. Psychiatry 30, 656- 662 (1991). Komoroski, R. A., Newton, J. E. 0.. Karson, C., Cardwell, D., Jagannathan, N. R. and Sprigg, J. Detection of psychoactive drugs in vivo in human brain using "F NMR spectroscopy. 8th Annual Meeting of the Society of Magnetic Resonance in Medicine, Vol. 1, p. 444 (1989).

6. Komoroski, R. A., Newton, J., Karson, C., Walker, E., Kardwell, 0. and Ramaprasad, R. In vivo NMR spectroscopy of psychoactive drugs in humans. Magn. Reson. Irnag. 7, suppl. 1, 32 (1989).

7. Grun, 6. R., Berger, U., Oberdorfer, F., Hull, W. E., Ostertag, H., Friedrich, E., Lehmann, J. and Keppler, D. Metabolism

8.

9.

10.

11.

12.

13.

and actions of 2-deoxy-2-fluoro-o-galactose in vivo. Eur. J. Biochem. 190,l l-19 (1990). Hull, W. E., Post, R. E., Hermann, R., Britsch, 6. and Kung, W. Metabolites of 5-fluorouracil in plasma and urine, as moni- tored by "F nuclear magnetic resonance spectroscopy, for patients receiving chemotherapy with or without metho- trexate pretreatment. Cancer Res. 48, 1680-1688 (1988). Hull, W. E., Port, R., Osswald, H., Kunz, W., Juretschke, H. P. and Schuff, N. In vivo "F-NMR study of 5-fluorouracil metabolism in liver and implanted tumours of the mouse. 5th Annual Meeting of the Society of Magnetic Resonance in Medicine, vol. 3, pp. 594-595 (1986). Joseph,A., Davenport, C., Kwock, L., Burt, C. T. and London, R. E. Fluorine-19 NMR studies of tumour-bearing rats treated with difluoromethyl ornithine. Magn. Reson. Med. 4,

Wolf, W., Albright, M. J., Silver, M. S., Weber, H., Reichardt, U. and Sauer R. Fluorine-19 NMR spectroscopic studies of the metabolism of 5-fluorouracil in the liver of patients undergoing chemotherapy. Magn. Reson. Irnag. 5, 165-1 69 (1987). Buchsbaum, M. S . Positron emission tomography in schi- zophrenia. In Psychopharrnacology: the Third Generation of Progress, ed. by H. L. Meltzer, pp. 783-792. Raven Press, New York (1987). Litt, L., Gonzales-Mendez, R., James, T. L., Sessler, D. I., Mills, P., Chew, W. M., Moseley, M., Pereira, B., Severinghaus, J. W. and Hamilton, J. K. An in vivo study of halothane uptake and elimination in the rat brain with fluorine magnetic resonance spectroscopy. Anesthesiology

137-143 (1987).

67, 161-168 (1987).