Rapid synthesis of 11C-labeled FK506 for positron emission tomography

Article abstract of DOI:10.1016/S0040-4039(02)02705-3

The present study describes a rapid synthesis method for labeled [¹¹C]FK506 for positron emission tomography (PET). A one-pot reaction from [¹¹C]CH₃I, involving a Wittig reaction as the key carbon-carbon bond formation was developed. The chemical process was accomplished using a designed, fully automated synthetic apparatus, and an injectable solution of [¹¹C]FK506 was obtained in only 34 min from [¹¹C]CH₃I. The decay-corrected radiochemical yield based on [¹¹C]CH₃I was 11.9%, and the specific activity was 39.8 GBq/μmol.

Full text of DOI:10.1016/S0040-4039(02)02705-3

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 641–644
11
Rapid synthesis of C-labeled FK506 for positron emission
tomography
Yoshihiro Murakami,a,b,* Akio Kuroda,c Kazuhiko Osodaa,b and Shintaro Nishimuraa,b
aAdvanced Technology Platform Research Laboratory, Fujisawa Pharmaceutical Co. Ltd, 5-2-3, Tokodai, Tsukuba,
Ibaraki 300-2698, Japan
bThe Medical and Pharmacological Research Center Foundation, Wo32, Inoyama-cho, Hakui, Ishikawa 925-0613, Japan
cMedicinal Chemistry Research Laboratories, Fujisawa Pharmaceutical Co. Ltd, 5-2-3, Tokodai, Tsukuba,
Ibaraki 300-2698, Japan
Received 7 October 2002; revised 26 November 2002; accepted 28 November 2002
Dedicated to Dr. Kazuo Sakane, former Director of Medicinal Chemistry, Fujisawa Pharmaceutical Co. Ltd., who died on October 1st, 2001
Abstract—The present study describes a rapid synthesis method for labeled [11C]FK506 for positron emission tomography (PET).
A one-pot reaction from [11C]CH3I, involving a Wittig reaction as the key carbonꢀcarbon bond formation was developed. The
chemical process was accomplished using a designed, fully automated synthetic apparatus, and an injectable solution of
[11C]FK506 was obtained in only 34 min from [11C]CH3I. The decay-corrected radiochemical yield based on [11C]CH3I was 11.9%,
and the specific activity was 39.8 GBq/mmol. © 2003 Elsevier Science Ltd. All rights reserved.
Positron emission tomography (PET) is an advanced
biofunctional imaging technology that can be used for
clinical diagnosis, and uses very short half-life radioiso-
topes (ca. 11C: 20.2 min, 15O: 2 min, respectively). Since
the distribution of radio-tracer in the human body can
be detected by PET, we believe that PET studies can be
useful for pharmacokinetic studies of new drug candi-
dates using the corresponding labeled drug. In general,
however, synthesis of the labeled drug has a number of
technical issues which need to be addressed. First, a
very rapid synthesis process from the organic reaction
to the purification process is required. Second, the
synthetic operation must be performed with an auto-
mated synthesis apparatus to avoid radiation exposure,
and lastly, the reaction scale is very small, usually
nano- or pico-mole scale, so that quality control is
difficult. In this paper, we describe a rapid synthesis
method for 11C-labeled FK5061–3 (tacrolimus; Fujisawa
Pharmaceutical Co. Ltd., Osaka, Japan) using our in
house developed automated apparatus.
neuroprotective effects of FK506 were also found and
demonstrated in a non-human primate model of
stroke.4–8 For development of FK506 as a new drug
candidate for central nervous system disease, brain PET
of FK506 could serve as a source of accurate pharma-
cokinetic data a key requirement, because FK506 must
be regulated within a strict therapeutic range in a
clinical setting.
On the other hand, FK506 is a good target for total
synthesis and many research groups have attempted
total synthesis because of its complicated chemical
structure. Jones et al.9 and Nakatsuka et al.10 succeeded
in 1989 and 1990, respectively, however, these chemical
processes are not applicable for the synthesis of 11C-
labeled FK506 since they are multi-step and time con-
suming. As a result, we needed to develop a new and
short chemical process that ideally would involve few
synthetic operations.
As the first step, we investigated the appropriate label-
ing position for FK506. There are four possible posi-
tions for labeling by [11C]CH3I, which is a popular
labeling reagent in PET, and is made from the 14N(p,
a)11C nuclear reaction with a cyclotron, followed by
reduction with LiAlH4 and reaction with HI, as shown
in Scheme 1. Methylation of the hydroxyl groups by
CH3I were examined but suitable reaction conditions
The immunosuppressive agent FK506 is widely used for
prevention of rejection in transplantation and for treat-
ment of atopic dermatitis worldwide. Recently, the
Keywords: PET; 11C; FK506; Wittig reaction.
* Corresponding author. Tel.: +81-767-26-3311; fax: +81-767-26-3314;
e-mail: murakami@mprcf.or.jp
0040-4039/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)02705-3

642
Y. Murakami et al. / Tetrahedron Letters 44 (2003) 641–644
Scheme 1.
could not be obtained, thus our attention turned to the
synthesis of 11C-labeled alkenes from [11C]CH3I. 11C-
Labeled alkenes have been synthesized from [11C]CH3I
yielding 11C-labeled terminal alkenes,11,12 hence, we
decided to label the allyl position of FK506 using a
Wittig reaction as summarized in Scheme 2.
protected precursor proceeded in moderate yield. After
optimization of this key chemical process by cold syn-
thesis on a micro scale, the rapid synthetic method
shown in Scheme 2 was discovered. The reaction of
precursor 2† with tri-p-tolylphosphine and epichloro-
hydrin in o-dichlorobenzene at 140°C gave compound 4
in only 3 min.
Strong bases are commonly used for proton abstraction
of the phosphonium salt in the Wittig reaction, whereas
FK506 is unstable under basic conditions. Therefore,
neutral or acidic conditions used were seemed most
suitable for the process.
In our experiment, the tri-p-tolylphosphine reagent pro-
vided good yields in comparison with the corresponding
triphenylphosphonium salt. A rapid synthesis method
of the tri-p-tolylphosphonium reagent was also devel-
oped for thus PET study.
Kihlberg et al.11 used epichlorohydrin for rapid Wittig
reaction, instead of using strong base. Thus, we tried
Wittig reaction of aldehyde 113 under such neutral
conditions. However, these conditions were not suit-
able, and FK506 was not obtained. A trialkylsilyl
group was then chosen as a protecting group for two
hydroxy groups of the precursor,14 because deprotec-
tion is known to occur under mild acidic conditions. In
a preliminary cold experiment, Wittig reaction with the
Concerning the final deprotection step, the previous
report14 described that the trialkylsilyloxy groups were
cleaved by HF in CH3CN at room temperature. How-
ever, we found that reaction of triethylsilyloxy com-
pound 4 or t-butyldimethysilyloxy compound 5 in a
† Triethylsilyloxy compound 2 was prepared by the same procedure
as the t-butyldimethylsilyloxy compound 3.
Scheme 2. Synthetic route of 11C labelled FK506 from CH3I.

Y. Murakami et al. / Tetrahedron Letters 44 (2003) 641–644
643
mixture of 47% aqueous HF and CH3CN gave the final
compound in 3 min at rt and 12 min at 50°C, respec-
tively. In this way, a one-pot synthetic route of 11C-
labeled FK506 from CH3I was completed.
test runs using this apparatus. As a result, these investi-
gations brought to the shorter synthesis time of about
only 30 min. The automated labeled synthesis using
above synthesis apparatus was examined as shown in
Scheme 3.
This result encouraged us to carry out the hot synthetic
experiment on a pico mole scale, however it was found
that purification of [11C]FK506 from the reaction mix-
ture by preparative HPLC was difficult because of
tailing of the chromatographic peak of P(p-tol)3 and
o-dichlorobenzene, which overlapped the peak for
FK506. Thus, two additional processes were added to
the synthesis as follows: (1) o-dichlorobenzene in the
reaction mixture was distilled away under reduced pres-
sure before the deprotection step: (2) in the deprotec-
tion step, excess P(p-tol)3 in the reaction mixture was
oxidized with CBr4 to tris-4-methylphenylphosphine
oxide which has a very different HPLC retention time
in comparison to that of P(p-tol)3.
Every chemical step proceeded smoothly and
[11C]FK506 was obtained, however, the yield was only
3–5% from [11C]CH3I, whereas HPLC analysis of the
reaction mixture indicated 20–30% yield. It was found
that adsorption of [11C]FK506 occurred in the prepara-
tive HPLC silica gel column because the amount was
very small. Thus, two low adsorptive columns (Shodex
Rspak DE-LG (10×50 mm and SUPELCO TPR-100
(10×250 mm, 5 mm) were connected instead of the silica
gel column and were eluted with H2O (CH3CN:THF=
6:1)=35:65 (flow rate: 3.8 ml/min, column temperature:
60°C) to give a reasonable yield of the fraction contain-
ing [11C]FK506 which was collected and solvent
removed under reduced pressure. The residue was dis-
solved in 33% ethanol–saline solution and passed
On the basis of these results, a fully automated syn-
thetic apparatus for 11C-labeled FK506 was developed
as shown in Figure 1.
through
a sterile filter (Millipore GV) to give
[11C]FK506 as an injectable solution for PET study.
The total synthesis time was 44 min after 11C-labeled
carbon dioxide formation. The radioactivity was 0.99
GBq (end of synthesis), radiochemical purity was
higher than 99%, the specific activity was 39.8 GBq/
mmol.
This apparatus was constructed based on previous
reports,15,16 and was compact to minimize the loss in
solution transfer and synthesis time. The optimization
of the each reaction was accomplished by many cold
Figure 1. Diagram of the synthesis system; V1–26=solenoid valves; TCA1–2=3-way cock actuator; 1 and 30–33=reservoirs;
2=syringe pump; 3=filter; 4=vial; 5=waste bottle; 6=solvent bottle; 7=HPLC pump; 8=6-way valve; 9, 17 and 28=radio
detectors; 10=photosensor; 11=glass vessel; 12=HPLC column; 13=column oven; 14, 15 and 29=thermosensors; 16=evapora-
tor; 18 and 26=heaters.

644
Y. Murakami et al. / Tetrahedron Letters 44 (2003) 641–644
Scheme 3. Rapid synthesis of 11C-labeled FK506: (a) o-dichlorobenzene, 150°C, 3 min; (b) 2, o-dichlorobenzene, epichlorohydrin,
150°C, 3 min; (c) concentration of reaction mixture; (d) 47% aq. HF, CH3CN, CBr4, rt, 3 min.
In conclusion, we have succeeded in developing a rapid
automated synthesis method of [11C]FK506 using Wit-
tig reaction as a key reaction and a good quality
injectable solution of [11C]FK506 for PET study was
obtained our automated synthetic apparatus. This work
represents one of the most complicated 11C-labeled
tracer syntheses reported to date. We are hopeful that
the PET study using [11C]FK506 will provide useful
pharmacokinetic data for new drug research and
development.
5. Sharkey, J.; Crawford, J. H.; Butcher, S. P.; Marston, H.
M. Stroke 1996, 27, 2282–2286.
6. Butcher, S. P.; Henshall, D. C.; Teramura, Y.; Iwasaki,
K.; Sharkey, J. J. Neurosci. 1997, 17, 6939–6946.
7. Takamatsu, H.; Kondo, K.; Ikeda, Y.; Umemura, K.
Eur. J. Pharmacol. 1998, 362, 137–142.
8. Takamatsu, H.; Tsukada, H.; Noda, A.; Kakiuchi, T.;
Nishiyama, S.; Nishimura, S.; Umemura, K. J. Nucl.
Med. 2001, 42, 1833–1840.
9. (a) Jones, T. K.; Mills, S. G.; Reamer, R. A.; Askin, D.;
Desmond, R.; Volante, R. P.; Shinkai, I. J. Am. Chem.
Soc. 1989, 111, 1157–1159; (b) Jones, T. K.; Reamer, R.
A.; Desmond, R.; Mills, S. G. J. Am. Chem. Soc. 1990,
112, 2998–3017.
Acknowledgements
10. Nakatsuka, M.; Ragan, J. A.; Sammakia, T.; Smith, D.
B.; Uehling, D. E.; Schreiber, S. L. J. Am. Chem. Soc.
1990, 112, 5583–5601.
11. Kihlberg, T.; Gullberg, P.; La˚ngstro¨m, B. J. Label.
Compd. Radiopharm. 1990, 10, 1115–1120.
We wish to thank Mr. Shigeo Hayashi of NKK corpo-
ration for support in preparing [11C]iodomethane.
Fruitful discussions with Dr. Takashi Manabe and Dr.
Keizo Yoshida, Fujisawa Pharmaceutical Co. Ltd, are
also greatly appreciated.
12. Grierson, J. R.; Biskupiak, J. E.; Link, J. M.; Hodge, P.
Appl. Radiat. Isot. 1993, 44, 1449–1458.
13. Patent, WO9200313.
References
14. Organ, M. J.; Holmes, M. A.; Pisano, J. M.; Staruch, M.
J.; Wyvratt, M. J.; Dumont, F. J.; Sinclair, P. J. Bioorg.
Med. Chem. Lett. 1993, 3, 657–662.
15. Nishimura, S.; Yajima, K.; Harada, N.; Ogawa, Y.;
Hayashi, N. J. Autom. Chem. 1994, 16, 195–204.
16. Murakami, Y.; Nishimura, S.; Noda, A.; Harada, N.;
Tsukada, H. J. Label. Compd. Radiopharm. 1997, 39,
S1–3.
1. Starzl, T. E.; Todo, S.; Fung, J.; Demetris, A. J.;
Venkataramman, R.; Jain, A. Lancet 1989, 2, 1000–1004.
2. Fung, J.; Abu-Elmagd, K.; Jain, A.; Gordon, R.; Tzakis,
A.; Todo, S.; Takaya, S.; Alessiani, M.; Demetris, A.;
Bronster, O.; et al. Transplant Proc. 1991, 23, 2977–2983.
3. Klintmalm, G. B. Clin. Transplant. 1994, 8, 207–210.
4. Sharkey, J.; Butcher, S. P. Nature 1994, 371, 336–339.