Utility of 1,3,4,6-tetra-O-acetyl-2-deoxy-2-[18F] fluoroglucopyranoside for no-carrier-added18F-glycosylation of amino acids

JOURNAL OF LABELLED COMPOUNDS AND RADIOPHARMACEUTICALS

J Label Compd Radiopharm 2005; 48: 701–719.

Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jlcr.963

Research Article

Utility of 1,3,4,6-tetra-O-acetyl-2-deoxy-2-[¹⁸F]ﬂuoro-

glucopyranoside for no-carrier-added F-glycosylation

of amino acids

18

Simone Maschauer¹, Monika Pischetsrieder², Torsten Kuwert¹

and Olaf Prante^1,*

¹Laboratory of Molecular Imaging, Clinic of Nuclear Medicine, Friedrich-Alexander

University, Schwabachanlage 6, D-91054 Erlangen, Germany

²Institute of Pharmacy and Food Chemistry, Friedrich-Alexander University,

Schuhstrasse 19, D-91052 Erlangen, Germany

Summary

A radiochemical method for the ¹⁸F-glycosylation of amino acid side chains was

developed starting from peracetylated 2-deoxy-2-[¹⁸F]ﬂuoroglucopyranoside (TA-

[¹⁸F]FDG). O-(2-deoxy-2-[¹⁸F]ﬂuoro-d-glucopyranosyl)-l-serine and the correspond-

ing threonyl compound were obtained in a radiochemical yield of 25% and 12%

(related to [¹⁸F]ﬂuoride), respectively, after Zemple

´

n deprotection within a total

reaction time of 90 min. The anomeric conﬁguration of the corresponding ¹⁹F-

substituted compounds revealed preferential a-stereoselectivity. The ¹⁸F-glycosylation

method using TA-[¹⁸F]FDG is compatible with the short half-life of ﬂuorine-18 and

combines glycosylation and ¹⁸F-labelling of a target compound within a single

reaction step. TA-[¹⁸F]FDG is a promising ¹⁸F-labelled prosthetic group and could be

adapted to ¹⁸F-labelling of bioactive peptides to study their pharmacokinetics using

positron emission tomography (PET). Copyright # 2005 John Wiley & Sons, Ltd.

Key Words: ¹⁸F-glycosylation; F-18; glycosidic linkage; positron-emission-tomo-

graphy; PET

Introduction

The development of ¹⁸F-labelling methods adopted to proteins and bioactive

peptides for diagnostic imaging by positron emission tomography (PET) has

*Correspondence to: Olaf Prante, Friedrich-Alexander University Erlangen-Nurnberg, Clinic of Nuclear

¨

Medicine, Laboratory of Molecular Imaging, Schwabachanlage 6, D-91054 Erlangen, Germany. E-mail:

olaf.prante@nuklear.imed.uni-erlangen.de

Contract/grant sponsor: Deutsche Forschungsgemeinschaft; contract/grant number: GRK750

Contract/grant sponsor: GE Healthcare Bio-Sciences; contract/grant number: GEH-BS/NUK-03

Copyright # 2005 John Wiley & Sons, Ltd.

Received 19 April 2005

Revised 5 May 2005

Accepted 7 May 2005

702

S. MASCHAUER ET AL.

gained enormous interest in the ﬁeld of radiopharmaceutical sciences.¹

Particularly small radioactive labelled peptides were optimized for their

pharmacological properties and show remarkable advantages over large

proteins or monoclonal antibodies due to their higher uptake in target tissue

and improved blood clearance. The commonly used strategy to label peptides

still relies on the use of ¹⁸F-labelled prosthetic groups targeting the N-terminus

or lysine side chains. Up to now, N-succinimidyl-4-[¹⁸F]ﬂuorobenzoate

([¹⁸F]SFB) and 4-nitrophenyl 2-[¹⁸F]ﬂuoropropionate are amongst the most

eﬀective and suitable ¹⁸F-acylation agents as indicated by their prevalent

application in a large number of radiosyntheses for ¹⁸F-labelled peptide-based

imaging agents (i.e. [¹⁸F](d-Phe1)octreotide,²[¹⁸F]SAA-RGD,³[¹⁸F]NT(8-13)⁴

or [¹⁸F](Nle⁴, d-Phe⁷)-a-MSH).⁵Alternative ¹⁸F-labelling methods for

peptides include the use of ¹⁸F-labelled aldehydes,⁶photochemical conjuga-

tion⁷and ¹⁸F-alkylation.⁸However, these laborious strategies require multi-

ple-step syntheses and often suﬀer from low yields and the necessity of

protecting groups to prevent side-reactions. Furthermore, complicated and

time consuming tracer puriﬁcation steps are unfavorable for routine large-

scale synthesis of short-lived ¹⁸F-labelled radiopharmaceuticals.

Thus, additional and improved radioﬂuorination techniques are urgently

required in the ﬁeld of molecular imaging. Such developments include the

approach toward direct ¹⁸F-substitution of hydroxy groups in peptides,⁹the

development of a chemo-enzymatic ¹⁸F-glycosylation method¹⁰and, more

recently, the use of [¹⁸F]ﬂuorothiols for chemoselective labelling of peptides.¹¹

Furthermore, Poethko et al. developed a convenient and high yielding ¹⁸F-

labelling method for peptides by oxime conjugation of [¹⁸F]ﬂuorobenzalde-

hyde with unprotected aminooxy peptides providing an important improve-

ment in the search for improved ¹⁸F-labelling methodologies.¹²

In addition, it has been shown that glycosylation of peptides often improves

their biokinetics. In the case of RGD peptides, which are antagonists of a_vb₃

integrins, the methodology of glycosylation prior to ¹⁸F-acylation has

successfully led to the development of novel PET-radiopharmaceuticals for

the imaging of tumor angiogenesis.³Moreover, glycosylation of peptides has

been shown to enhance bioavailability and improve blood brain barrier (BBB)

permeability and in vivo clearance properties in a large number of studies.^13–16

Glycosylation strategies in carbohydrate chemistry are many including

syntheses of glycopeptide building blocks using peracetylated glycosides and

BF₃ꢀ Et₂O or glycosyl bromides and AgOTf or HgBr₂.^17–22In comparison to

these methods, a suitable ¹⁸F-glycosylation method should apply a rapidly

accessible glycosyl donor compound. Advantageously, the peracetylated

intermediate of the well-known radiosynthesis of 2-deoxy-2-[¹⁸F]ﬂuoro-

glucose ([¹⁸F]FDG), which represents the most frequently used PET-radio-

pharmaceutical for tumor imaging in nuclear medicine, can be produced in

Copyright # 2005 John Wiley & Sons, Ltd.

J Label Compd Radiopharm 2005; 48: 701–719

F-18 GLYCOSYLATION

703

highest radiochemical yields starting from [¹⁸F]ﬂuoride and the precursor

1,3,4,6-tetra-O-acetyl-2-O-triﬂuormethanesulfonyl-b-d-mannopyranose.²³

A

glycosylation reaction that made use of peracetylated [¹⁸F]FDG to obtain

N-glycosylated 2-nitroimidazole was reported by Patt et al.²⁴

The challenge of this project was the development of a radiochemical ¹⁸F-

glycosylation method that combines both, O-glycosylation and ¹⁸F-labelling

of the target compound. In this study we report the optimization of a no-

carrier-added ¹⁸F-glycosylation method using 1,3,4,6-tetra-O-acetyl-2-deoxy-

2-[¹⁸F]ﬂuoro-d-glucopyranose (TA-[¹⁸F]FDG, [¹⁸F]1) as ¹⁸F-labelled glycosyl

donor. In order to investigate the general feasibility of this ¹⁸F-glycosylation

strategy, we used the Fmoc-protected amino acids serine and threonine as

model glycosyl acceptors. We herein also report the stereochemical outcome of

glycosylation reactions of 2-deoxy-2-ﬂuoro glycosides and the mechanistic

implications of these results.

Results and discussion

Based on the [¹⁸F]FDG-synthesis by Hamacher and others,²³the ¹⁸F-labelled

glycosyl donor [¹⁸F]1 was obtained in a radiochemical yield (RCY) of 92% by

aminopolyether (Kryptoﬁx¹2.2.2) supported nucleophilic ¹⁸F-for-OTf sub-

stitution on the corresponding tetra-O-acetylated mannopyranoside (Scheme

1). For the subsequent reaction steps the crude reaction mixture of [¹⁸F]1 was

passed through a LiChrolut¹Si-cartridge with CH₃CN and submitted to

semipreparative reversed-phase HPLC for rapid radiochemical isolation of

[¹⁸F]1. Solid phase extraction (SPE) on a LiChrolut¹RP18-cartridge and

elution with CH₃CN followed by evaporation was performed in order to

Scheme 1. ¹⁸F-glycosylation of serine or threonine using TA-[¹⁸F]FDG ([¹⁸F]1).

.

(i) BF₃Et₂O, Fmoc-Ser-OH or Fmoc-Thr-OH, CH₃CN, T=808C, 5 min; (ii)

NaOMe/MeOH, T=608C, 10 min

Copyright # 2005 John Wiley & Sons, Ltd.

J Label Compd Radiopharm 2005; 48: 701–719

704

S. MASCHAUER ET AL.

provide [¹⁸F]1 for subsequent ¹⁸F-glycosylation of the model compounds

Fmoc-Ser and Fmoc-Thr (Scheme 1, step 2). The reaction parameters for the

¹⁸F-glycosylation key step were investigated with respect to the choice of

reaction solvent, reaction time, Lewis acid promotor (Table 1) and the

concentration of Fmoc-Ser (Table 2).

As shown in Table 1, CH₃CN was convenient for O-glycosylation (Scheme

1, step 2), whereas CH₂Cl₂turned out to be less suitable. However, the ¹⁸F-

glycosylation of Fmoc-protected lysine was hindered due to a solubility

problem in CH₃CN (data not shown).

Patt et al. suggested SnCl₄and Hg(CN)₂as promoters in the radiosynthesis

of an adduct of [¹⁸F]FDG and 2-nitroimidazole.²⁴This N-glycosylation used

Table 1. Radiochemical yields (RCY) for the ¹⁸F-glycosylation of Fmoc-Ser/Thr using

[¹⁸F]1 as glycosyl donor

Product

Solvent

Lewis acid^b

T (8C)

RCY (%)^a

[¹⁸F]2

[¹⁸F]3

CH₃CN

CH₂Cl₂

C₂H₅Cl₂

Et₂O

BF₃ꢀ Et₂O

BF₃ꢀ Et₂O, Hg(CN)₂

Cu(OTf)₂

SnCl₄

rt

40

60

70

80

120

60

reﬂux

60

reﬂux

60

0

32

35

48

0

47

0

11

0

30

9

SnBr₄

BF₃ꢀ Et₂O

SnCl₄

CH₃CN

Cu(OTf)₂

60

0

^at=10 min, 8 mM Fmoc-Ser/Thr, V=400 ml.

^b80 mmol BF₃; 40 mmol Hg(CN)₂; 20 mmol Cu(OTf)₂; 40 mmol SnBr₄; 85 mmol SnCl₄.

Table 2. Dependence of the radiochemical yield on

the concentration of precursor for the ¹⁸F-glycosyla-

tion of Fmoc-Ser using TA-[¹⁸F]FDG ([¹⁸F]1)

Concentration of Fmoc-Ser (mM)

RCY (%)^a

0.5

2

5

10

25

50

14 ꢁ 2

26 ꢁ 3

31 ꢁ 2

32 ꢁ 4

42 ꢁ 4

41 ꢁ 4

^an=3, t=10 min, 0.2 M BF₃ꢀ Et₂O, T=608C, V=200 ml

(CH₃CN).

Copyright # 2005 John Wiley & Sons, Ltd.

J Label Compd Radiopharm 2005; 48: 701–719

F-18 GLYCOSYLATION

705

large amounts of SnCl₄and catalytic amounts of Hg(CN)₂to achieve the N-

glycosylated product in surprisingly high radiochemical yields (80%) within

60 min. We observed only a moderate RCY of 11% for the O-linked

tetraacetylated [¹⁸F]FDG adduct [¹⁸F]2 when using SnCl₄(Table 1). In

contrast, our experiments revealed BF₃ꢀ Et₂O as the Lewis acid promoter of

choice, so that a maximum RCY of 48% ([¹⁸F]2) and 32% ([¹⁸F]3) was reached

in a much shorter reaction time (8–10 min) at 808C. A prolonged reaction time

did not signiﬁcantly lead to enhanced RCYs as determined by radio-HPLC.

As suggested by Patt et al.²⁴we also used Hg(CN)₂as a typical Helferich-

promotor for glycosylation reactions. We observed a moderate increase of

radiochemical yield of [¹⁸F]2 under the same reaction conditions as without

Hg(CN)₂(Table 1). However, we excluded Hg(CN)₂from the reaction mixture

due to its toxic properties which could cause problems when applying this

procedure to the production of radiopharmaceuticals for human use. Yamada

et al. used copper(II) triﬂate as an activator for peracetylated carbohydrates in

an approach for a one-pot strategy of oligosaccharide synthesis.²⁵In our

attempts to utilize copper(II) triﬂate as a promoter we were not successful

(Table 1), possibly due to the necessity to avoid moisture entirely. The

copper(II) triﬂate used in this study was dried with P₂O₅under reduced

pressure, which may not be suﬃcient for subsequent reactions under no-

carrier-added conditions.

When using higher reaction temperatures (>808C) clearly a more

accelerated degradation of [¹⁸F]2 and [¹⁸F]3 was observed, accompanied by

hydrolytic cleavage of the glycosidic bond, whereas at room temperature no

¹⁸F-glycosylated product was detected (Table 1).

Thus, the use of BF₃as a Lewis acid promotor and CH₃CN at 808C turned

out to be the optimum reaction. The concentration of Fmoc-protected amino

acid necessary for a reliable RCY of [¹⁸F]2 was determined to be 5–10 mM

(Table 2). As indicated in Table 2, a further increase in amino acid

concentration did not result in distinct increased RCYs for the ¹⁸F-

glycosylation key step. A precursor concentration of 5–10 mM is suitable for

peptide labelling, since these substrates are usually available in suﬃcient

amounts, whereas an application of this radiosynthetic approach to the ¹⁸F-

glycosylation of rare (or expensive) compounds could be problematic.

Interestingly, Zemple

´

n deacylation^26,27of [¹⁸F]2 or [¹⁸F]3 also led to

cleavage of the Fmoc-protecting group ([¹⁸F]6, [¹⁸F]7) under no-carrier-added

reaction conditions. This observation depended largely upon reaction time.

For compounds [¹⁸F]4 or [¹⁸F]5 a maximum RCY of about 40% (related to

[¹⁸F]2/[¹⁸F]3) was observed after 2 min as determined by radio-HPLC. After

20 min deprotection was complete and [¹⁸F]6 or [¹⁸F]7 were obtained in 85 and

40% RCY, respectively. The major by-product was 2-deoxy-2-[¹⁸F]ﬂuoroglu-

cose ([¹⁸F]FDG), especially in the case of the threonyl compound (50% after

Copyright # 2005 John Wiley & Sons, Ltd.

J Label Compd Radiopharm 2005; 48: 701–719

706

S. MASCHAUER ET AL.

20 min, radio TLC (SiO₂, CH₃CN:H₂O, 9:1 (v/v)), that was easily separated by

the subsequent cation exchange cartridge.

Optimized conditions and SPE techniques led to an experimental procedure

consistent with the use of TA-[¹⁸F]FDG ([¹⁸F]1) as a ¹⁸F-labelled glycosyl

donor. The total synthesis time was 90 min, the overall radiochemical yield was

about 25% ([¹⁸F]6) and 12% ([¹⁸F]7). Clearly, this two-step procedure

including two HPLC separation steps could be diﬃcult to adjust for

automation in large-scale radiopharmaceutical production for human use,

but for application in prior animal studies this ¹⁸F-glycosylation method

should be eminently practical. Therefore, we successfully utilized this

laboratory ¹⁸F-glycosylation procedure to obtain [¹⁸F]6 as a model compound

to test the stability of the glycosidic linkage in human serum in vitro. Not

surprisingly and as a proof of principle, [¹⁸F]6 was found to be stable in human

serum at 378C for up to 45 min (>98%, radio-TLC).

¹⁸F-labelled radioactive compounds were identiﬁed by characterization of

their corresponding ¹⁹F-substituted analogs. Scheme 2 depicts the syntheses of

the glycosyl donors 1,3,4,6-tetra-O-acetyl-2-deoxy-2-ﬂuoro-b-d-glucopyrano-

side (1) and the corresponding a-bromide (8), which were obtained following

literature procedures.^28–32With these glycosyl donors in hand, we turned our

attention toward the glycosylation reaction of Fmoc-Ser-OH and Fmoc-Thr-

OH according to Elofsson’s general procedure.³³Activation of the anomeric

b-acetyl of 1 (1 equiv) by BF₃ꢀ Et₂O (3 equiv) to a solution of Fmoc-Ser or

Fmoc-Thr (1 equiv) in dry CH₃CN provided low yields of 2 or 3, respectively

(Scheme 3). Interestingly, the main reaction was 3-O-acylation at the amino

acid side chain (48% (11), 43% (12), Scheme 3) as similarly observed by

Scheme 2. (i) DAST, ꢂ408C ! rt; (ii) HBr/AcOH, CH₂Cl₂

Scheme 3. Glycosylation method using boron triﬂuoride. (i) Fmoc-Ser-OH (for

.

11 and 2) or Fmoc-Thr-OH (for 12 and 3), BF₃Et₂O, CH₃CN, rt, 24 h

Copyright # 2005 John Wiley & Sons, Ltd.

J Label Compd Radiopharm 2005; 48: 701–719

F-18 GLYCOSYLATION

707

Salvador et al. using pentaacetylated glycosyl donors, although to a minor

extent.¹⁸In contrast to the reaction under no-carrier-added conditions

(Scheme 1), this observation indicated preferential orthoester formation as

an intermediate in the glycosylation reaction of 2-deoxy-2-ﬂuoroglucosyl

donors in the presence of BF₃, since orthoesters have been shown to rearrange

with O-acetylation of the glycosyl acceptor under Lewis acid catalysis.³⁴

In order to improve the yield of 2, we used a-bromide 8 in a silver triﬂate

mediated glycosylation, whereas 3 was advantageously obtained by treatment

of Fmoc-Thr-OH with 1 in the presence of TMSOTf (Scheme 4). Compounds

2 and 3 were isolated by preparative reversed-phase HPLC in 7–8% yield. As

the major by-products occurring during these coupling reactions, compounds

13 and 14 were isolated due to glycoester formation (Table 3, entry 7, 8). This

observation was also described by Salvador et al.¹⁸and noticed by Seitz et al.³⁵

for the synthesis of O-glycopeptides. Subsequent Zemplen deacetylation of 2

´

and 3 aﬀorded the Fmoc-protected compounds 4 and 5, respectively, as

analyzed by LC/MS. The deprotected glycosylated amino acids 6 and 7 were

obtained by starting from the fully protected amino acid derivatives Z-Ser-

OBn and Z-Thr-OBn using glycosyl donor 8. AgOTf-mediated glycosylation

provided 9 and 10 in improved yields, which were then deprotected by

hydrogenolytic benzyl removal followed by subsequent Zemple

´

n reaction

resulting in an anomeric mixture of 6 and 7, respectively.

The anomeric conﬁguration of glycosylated intermediates and product

compounds were analyzed by ¹H-NMR and ¹⁹F-NMR spectroscopy, since the

³J_F,1coupling constant is an excellent tool to distinguish between a (³J_F,1is

negligible) and b (³J_F,1ꢃ 2.5 Hz) conﬁguration at C-1 in 2-deoxy-2-ﬂuoro

glycosides. A summary of analytical data that were used to ensure identity of

the corresponding ¹⁸F-labelled analogs is presented in Table 3. Although the 2-

deoxy-2-ﬂuoro glucopyranosyl donors 1 and 8 used in this study were opposite

in anomeric conﬁguration, the TMSOTf-mediated glycosylation of Fmoc-Thr

Scheme 4. Glycosylation method using silver triﬂate. (i) AgOTf, CH₂Cl₂, rt, 2 h;

For compounds 6 or 7: (ii) 9 or 10, Pd black, HCOONH₄, methanol, rt, 15 min

and (iii) NaOMe/MeOH, rt, 2 h. For compounds 4 or 5: (iii) 2 or 3, NaOMe/

MeOH, rt, 24 h (4) or 4d (5). (* Treatment of Fmoc-Thr-OH with 1 and

TMSOTf advantageously provided 3 in a a:b ratio of >97:3; see Experimental)

Copyright # 2005 John Wiley & Sons, Ltd.

J Label Compd Radiopharm 2005; 48: 701–719

708

S. MASCHAUER ET AL.

Table 3. Analytical data of the various glycosylated compounds including a/b anomeric

1

ratios determined by H- and ¹⁹F-NMR spectroscopy

a

Entry

compound

yield (%)

R_f

k⁰

¹H-NMR

chemical

shift of

a/b^z

H-1 (ppm)

1

2

33

88

0.70^b

0.77^b

4.5

6.2

5.75 (b)^f

pure b

pure a

n.d.

3

4

5

6

8

0.20^d

0.18^d

0.71^b

0.86^c

10.3

11.2

13.8

14.7

5.17 (a),

4.94 (b)^g

66/34

>97/3

97/3

7

5.24 (a)^g

35

43

4.91 (a),

4.75 (b)^f

4.95 (a),

4.70 (b)^f

97/3

7

8

5

n.d.

10.8

12.0

6.30 (a),

6.05 (b)^g

70/30

80/20

40

6.31 (a),

6.08 (b)^g

Copyright # 2005 John Wiley & Sons, Ltd.

J Label Compd Radiopharm 2005; 48: 701–719

F-18 GLYCOSYLATION

709

Table 3. continued

a

Entry

compound

yield (%)

R_f

k⁰

¹H-NMR

chemical

shift of

a/b ^z

H-1 (ppm)

9

11

0.14^e

0.2

5.06 (a),

4.68 (b)^h

66/34

67/33

10

7

0.20^e

5.13 (a)^h

^aKromasil C8, 250 ꢄ 4.6, 1.5 ml/min, 40–100% CH₃CN in water (0.1% TFA) in 50 min.

^bEthyl acetate/n-hexane 1:1.

^cCH₂Cl₂/MeOH 95:5.

^dEthyl acetate/n-hexane 7:3.

^eCH₃CN/9.5mM tetrabutylammonium hydroxide 8:2.

^fIn CDCl₃.

^gIn DMSO-d₆.

^hIn D₂O.

ⁱdetermined by ¹⁹F-NMR.

using 1 (Table 3, entry 4) and also the silver-promoted glycosylation reactions

(Scheme 4, Table 3, entry 3, 5, 6) proceeded with preferential a-stereo-

selectivity. Thus, the product distribution obtained was consistent with an

S_N1-type reaction mechanism due to the thermodynamic eﬀect (anomeric

eﬀect) in tetrahydropyrans which generally favors axial linkage of electro-

negative functional groups to C-1.³⁶In contrast to 2-deoxy-2-iodo glycosyl

donors used for selective b-glycosylation,³⁷we conclude that glycosylation

reactions employing 2-deoxy-2-ﬂuoro glucopyranosyl donors 1 or 8 are

stereoelectronically controlled and proceed via an ﬂuoroglycosyl oxocarbo-

nium intermediate with excellent a-stereoselectivities. However, reaction

conditions for the syntheses of standard compounds were not in accordance

with those of the ¹⁸F-labelled analogs [¹⁸F]2 and [¹⁸F]3, so that further studies

are required to gain precise information about the a/b ratio under no-carrier-

added reaction conditions. Noteworthy, the decrease of a/b ratio for

compounds 6 and 7 after Zemplen deprotection (Table 3, entry 9, 10) could

´

not be attributed to epimerization of amino acid stereocenters or b-

elimination, since these side-reactions were excluded by a systematic study

of Sjolin et al. on a model compound.³⁸It is tempting to speculate that

¨

anomerization occurred during the work-up procedure for deprotected

anomeric mixtures of 6 and 7, respectively.

Copyright # 2005 John Wiley & Sons, Ltd.

J Label Compd Radiopharm 2005; 48: 701–719

710

S. MASCHAUER ET AL.

Experimental

General

All chemicals and reagents were of analytical grade and obtained from

commercial sources. [¹⁸F]Fluoride was obtained from PET Net GmbH

(Erlangen, Germany). Solid phase cartridges (Merck LiChrolut¹Si 200 mg,

LiChrolut¹RP-18 100 mg and LiChrolut¹SCX 100 mg) were purchased from

VWR International. Thin layer chromatography (TLC) was carried out on

silica gel-coated aluminum plates (Alugram¹, Sil G/UV₂₅₄, Macherey Nagel);

for radio-TLC plastic sheets (Polygram¹, Sil G/UV₂₅₄, Macherey Nagel) were

used. Compounds were visualized by UV light (254 nm), charring with

anisaldehyde solution (anisaldehyde/sulfuric acid/acetic acid, 1:2:97 (v/v/v)) or

charring with ninhydrin reagent (1% ninhydrin in ethanol). Analytical HPLC

was performed on the following system: HPLC Hewlett Packard (HP 1100)

with a quarternary pump and variable wavelength detector (HP 1100) and

radio-HPLC-detector D505TR (Canberra Packard). Computer analysis of the

HPLC data was performed using FLO-One software (Canberra Packard).

Preparative HPLC was performed on a Knauer system (HPLC pump 64, LC

photometer at 254 nm) using columns and conditions as speciﬁed in the text.

Electronic autoradiography (Instant Imager^TM, Canberra Packard) was used

to analyze radio-TLC data. NMR spectra were recorded on a Varian Gemini-

300 spectrometer, operating at 300.18 MHz for ¹H and 282.41 MHz for ¹⁹F by

Deutero GmbH (Kastellaun, Germany). ESI-MS and LC-MS analysis were

performed on an Agilent 1100 Series analytic HPLC system with a VWL

detector, coupled to a Bruker esquire 2000 mass spectrometer with electron

spray ionization. 1,3,4,6-tetra-O-acetyl-b-d-mannopyranose,^28,291,3,4,6-tetra-

O-acetyl-2-O-triﬂuormethanesulfonyl-b-d-mannopyranose³⁰and 3,4,6-tri-O-

acetyl-2-deoxy-2-ﬂuoro-a-d-glucopyranosyl bromide³¹(8) were synthesized

following literature procedures. ¹⁹F-NMR of 8 conﬁrmed the pure a-anomer

(¹⁹F-NMR (CDCl₃): d ꢂ189.27 (dd, J_F,2=49.4 Hz, J_F,3=11.0 Hz)).

Radiosynthesis of 1,3,4,6-tetra-O-acetyl-2-deoxy-2-ﬂuoro-d-glucopyranose (½¹⁸Fꢅ1)

For the preparation of [¹⁸F]1 a QMA-cartridge with [¹⁸F]ﬂuoride was eluted

with a solution of 15 mg Kryptoﬁx¹2.2.2/15 ml 1 M K₂CO₃in 1 ml acetonitrile/

water (8:2). The solution was evaporated using a stream of nitrogen at 858C

and co-evaporated to dryness with CH₃CN (2 ꢄ 200 ml). According to the

FDG-synthesis by Hamacher et al.,²³9 mg precursor (1,3,4,6-tetra-O-acetyl-2-

O-triﬂuormethanesulfonyl-b-d-mannopyranose) in 450 ml anhydrous acetoni-

trile was added. This mixture was stirred for 5 min at 858C and passed through

a SiO₂-cartridge (Merck, 200 mg). The radiochemical yield of TA-[¹⁸F]FDG

was 93% as determined by analytical HPLC from a sample withdrawn from

the reaction mixture. After elution with 1.0 ml acetonitrile and evaporation of

Copyright # 2005 John Wiley & Sons, Ltd.

J Label Compd Radiopharm 2005; 48: 701–719

F-18 GLYCOSYLATION

711

the solvent the residue was taken up in 500 ml acetonitrile/water (30:70) and

submitted to semipreparative HPLC (Kromasil C8, 125ꢄ 8, 3.5 ml/min,

acetonitrile/ water 30:70). The product fraction was diluted 1:10 with water

and ﬁxed on a C18-cartridge (Merck, 100 mg), dried in a nitrogen stream and

eluted with 1 ml acetonitrile in a reaction vessel. Starting from 370 MBq

[¹⁸F]F^ꢂthis procedure yielded 250 MBq [¹⁸F]1 in a reaction time of 30 min.

¹⁸F-glycosylation procedure

Radiosyntheses of N^a-(9-ﬂuorenylmethoxycarbonyl)-O-(3,4,6-tri-O-acetyl-2-

deoxy-2-[¹⁸F]ﬂuoro-d-glucopyranosyl)-l-serine (½¹⁸Fꢅ2) and N^a-(9-ﬂuorenyl-

methoxycarbonyl)-O-(3,4,6-tri-O-acetyl-2-deoxy-2-[¹⁸F]ﬂuoro-d-glucopyrano-

syl)-l-threonine (½¹⁸Fꢅ3)

Fmoc-protected serine or threonine (10 mM) in 200 ml anhydrous CH₃CN and

5 ml BF₃etherate were added to a reaction vessel containing dry [¹⁸F]1 at 808C.

After the reaction time of 5 min the mixture was diluted with H₂O 1:10 and

passed through a C18-cartridge (Merck, 100 mg). The cartridge was washed

with CH₃CN/H₂O (30:70) to remove hydrolytic by-products, dried and eluted

with 1.0 ml CH₃CN. The solvent was evaporated and the residue (diluted in

500 ml acetonitrile/water (40:60)) was transferred to the semipreparative

HPLC-system. The radioactive products ([¹⁸F]2 or [¹⁸F]3) were isolated by

gradient reversed-phase HPLC (Kromasil C8, 250 ꢄ 8 mm, 4 ml/min, 40–100%

CH₃CN in H₂O (0.1% TFA) within 50 min) in a radiochemical yield of 32%

for the threonine derivative or 48% for the serine derivative. The product

fraction was diluted with water (1:10), ﬁxed on a C18-cartridge (Merck,

100 mg), dried in a stream of nitrogen and eluted with 1 ml acetonitrile in a

reaction vessel. Starting from 250 MBq [¹⁸F]1 this procedure yielded 85 MBq

[¹⁸F]2 or 55 MBq [¹⁸F]3 within 30 min.

Optimization of the ¹⁸F-glycosylation procedure

The ¹⁸F-glycosylation procedure was optimized by repeating the reaction with

varying parameters as indicated in Table 1.

Radiosynthesis of O-(2-deoxy-2-[¹⁸F]ﬂuoro-d-glucopyranosyl)-l-serine (½¹⁸Fꢅ6)

The solution of [¹⁸F]2 in CH₃CN was evaporated to dryness as described

before. Subsequently, 50 mM NaOMe in dry methanol was added and the

mixture was stirred 20 min at 608C leading to deprotection of the glucose unit

and of the Fmoc-protecting group to aﬀord [¹⁸F]6. The radiochemical yield of

this cleavage was 85%. [¹⁸F]6 was ﬁxed on a SCX-cartridge (100 mg) and

eluted with 600 ml PBS. Starting from 85 MBq [¹⁸F]2 this procedure yielded

55 MBq [¹⁸F]6 in 27 min. The total synthesis time (starting from [¹⁸F]F^ꢂ) was

about 90 min, the overall radiochemical yield was about 25% for [¹⁸F]6.

Copyright # 2005 John Wiley & Sons, Ltd.

J Label Compd Radiopharm 2005; 48: 701–719

712

S. MASCHAUER ET AL.

Radiosynthesis of O-(2-deoxy-2-[¹⁸F]ﬂuoro-d-glucopyranosyl)-l-threonine (½¹⁸Fꢅ7)

The solution of [¹⁸F]2 in CH₃CN was evaporated. Subsequently, 20 mM

NaOMe in dry methanol was added and the mixture was stirred 20 min at

608C leading to deprotection of the glucose unit and of the Fmoc-protecting

group to aﬀord [¹⁸F]7. The radiochemical yield of this cleavage was 40%.

1; 3; 4; 6-Tetra-O-acetyl-2-deoxy-2-ﬂuoro-b-d-glucopyranose (1)

1 was synthesized by ﬂuorination of 1,3,4,6-tetra-O-acetyl-b-d-mannopyr-

anose (3.45 g, 10.35 mmol) using diethylamino sulfur triﬂuoride (DAST)

according to the method described by Card.³²The crude product was puriﬁed

by column chromatography (silica gel, ethyl acetate/n-hexane 7:3) to give 33%

1

(1.2 g, 3.4 mmol) after crystallization in diethylether. H- and ¹⁹F-NMR were

consistent with the literature data.^39,40

N^a-(Phenylmethoxycarbonyl)-O-(3,4,6-tri-O-acetyl-2-deoxy-2-ﬂuoro-d-gluco-

pyranosyl)-l-serine phenylmethyl ester (9)

138 mg (539 mmol) silver triﬂate were added to a cooled (08C) solution of

200 mg (539 mmol) 8 and 177 mg (539 mmol) N^a-(phenylmethoxycarbonyl)-l-

serine phenylmethyl ester (Bachem, Germany) in 2 ml CH₂Cl₂. The cooling

bath was removed and the suspension was stirred at room temperature. The

reaction was monitored by TLC. After 2 h the suspension was ﬁltered through

Celite¹, washed with saturated NaHCO₃and water, dried with Na₂SO₄and

concentrated. The residue was puriﬁed by column chromatography (silica gel,

ethyl acetate/n-hexane 1:1) to give 35% (117 mg , 189 mmol) of 9. TLC (ethyl

acetate/n-hexane 1:1): R_f= 0.71. HPLC (Kromasil C8, 250 ꢄ 4.6, 1.5 ml/min,

40–100% CH₃CN in water (0.1% TFA) in 50 min): k⁰=13.8. ¹H-NMR

(CDCl₃): d 7.40-7.23 (m, 10H, CH-benzyl), 5.84 (d, 1H, NH, J=8.1 Hz), 5.39

(dt, 1H, H-3, J_3,F=11.6 Hz, J_3,4=9.5 Hz), 5.21-5.08 (m, 4H, 2 ꢄ CH₂-benzyl),

4.92 (t(dd), 1H, H-4, J_4,5=9.5 Hz), 4.91 (d, 1H (a), H-1, J_1,2=3.6 Hz), 4.75

(dd, 1H, H-1 (b), J_1,2=9.6 Hz, J_1,F=4.0 Hz), 4.59 (dt, a-CH (Ser)), 4.40 (ddd,

1H, H-2, J_2,F=48.6 Hz, J_2,3=9.5 Hz), 4.20-4.09 (m, 2H, H-6a, H-6b), 4.02-

3.97 (m, 2H, b-CH₂(Ser)), 3.95-3.87 (ddd, 1H, H-5), 2.06 (s, 3H, OAc), 2.04 (s,

3H, OAc), 2.01 (s, 3H, OAc). ¹⁹F-NMR (CDCl₃): d ꢂ201.93 (dd,

J_F,2=48.6 Hz, J_F,3=11.6 Hz, a-anomer), ꢂ200.16 (ddd, J_F,2=50.1 Hz,

J_F,3=14.6 Hz, J_F,1=2.6 Hz, b-anomer); a/b= 97/3.

N^a-(Phenylmethoxycarbonyl)-O-(3,4,6-tri-O-acetyl-2-deoxy-2-ﬂuoro-d-gluco-

pyranosyl)-l-threonine phenylmethyl ester (10)

A mixture of 200 mg (539 mmol) 8 and 185 mg (539 mmol) N^a-(phenylmethox-

ycarbonyl)-l-threonine phenylmethyl ester in 2 ml CH₂Cl₂was treated as

described for the preparation of glycoside 9. The residue was eluted from a

Copyright # 2005 John Wiley & Sons, Ltd.

J Label Compd Radiopharm 2005; 48: 701–719

F-18 GLYCOSYLATION

713

column of silica gel with CH₂Cl₂/methanol (95:5) to give 43% (147 mg,

108 mmol) of 10. TLC (CH₂Cl₂/methanol 95:5): R_f= 0.86. HPLC (Kromasil

C8, 250 ꢄ 4.6, 1.5 ml/min, 40–100% CH₃CN in water (0.1% TFA) in 50 min):

1

k⁰=14.7. H-NMR (CDCl₃): d 7.36-7.25 (m, 10H, CH-benzyl), 5.51 (d, 1H,

NH, J=9.6 Hz), 5.37 (dt, 1H, H-3, J_3,F=12.9 Hz, J_3,4=9.6 Hz), 5.17-5.05 (m,

4H, 2 ꢄ CH₂-benzyl), 4.95 (d, 1H, H-1 (a), J_1,2=3.6 Hz), 4.91 (t(dd), 1H, H-4,

J_4,5=9.6 Hz), 4.70 (dd, 1H, H-1 (b), J_1,2=9.5 Hz, J_1,F=3.9 Hz), 4.47 (dd, a-

CH (Thr), J_H,NH=9.6 Hz, J_H,CH=2.2 Hz), 4.38 (m, 1H, b-CH (Thr)), 4.32

(ddd, 1H, H-2,

J_2,F=49.3 Hz, J_2,3=9.6 Hz), 4.19 (dd, 1H, H-6a,

J_6a,6b=12.4 Hz, J_6a,5=5.4 Hz), 4.02 (dd, 1H, H-6b, J_6b,5=2.2 Hz), 4.01 (ddd,

1H, H-5), 2.05 (s, 3H, OAc), 2.04 (s, 3H, OAc), 2.02 (s, 3H, OAc), 1.34 (d, 3H,

CH₃(Thr), J=6.7 Hz). ¹⁹F-NMR (CDCl₃): d ꢂ200.68 (dd, J_F,2=49.3 Hz,

J

_F,3=12.9 Hz, a-anomer), ꢂ200.78 (ddd, J_F,2=52.2 Hz, J_F,3=14.6 Hz,

J

_F,1=3.3 Hz, b-anomer); a/b 97/3. ESI-MS: 633.9 [M+H]⁺.

N^a-(9-Fluorenylmethoxycarbonyl)-O-(3,4,6-tri-O-acetyl-2-deoxy-2-ﬂuoro-d-

glucopyranosyl)-l-serine (2)

Method A: Glycosylation method using AgOTf. A mixture of 100 mg

(270 mmol) 8 and 88 mg (270 mmol) N^a-(9-ﬂuorenylmethoxycarbonyl)-l-serine

in 1 ml CH₂Cl₂was treated as described for the preparation of glycoside 9. The

residue was isolated by preparative HPLC (Kromasil C8, 250 ꢄ 20, 15 ml/min,

CH₃CN/H₂O 1:1 (0.1% TFA)) to give 8% (13 mg, 21 mmol) of 2. TLC (ethyl

acetate/n-hexane 7:3): R_f= 0.20. HPLC (Kromasil C8, 250 ꢄ 4.6, 1.5 ml/min,

40–100% CH₃CN in water (0.1% TFA) in 50 min): k⁰=10.3. ¹H-NMR

(DMSO-d₆): d 7.88-7.24 (m, 8H, CH-Fmoc), 5.38 (dt, 1H, H-3, J_3,F=12.5 Hz,

J_3,4=9.6 Hz), 5.17 (d, 1H, H-1 (a), J_1,2=3.6), 4.94 (dd, 1H, H-1 (b),

J_1,2=7.4 Hz, J_1,F=2.1 Hz), 4.87 (t(dd), 1H, H-4, J_4,5=9.6 Hz), 4.67 (ddd,

1H, H-2, J_2,F=48.5 Hz, J_2,3=9.6 Hz), 4.34-4.19 (m, 4H, a-CH (Ser), CH₂-

Fmoc, CH-Fmoc), 4.15-4.05 (m, 2H, H-6a, H-6b), 4.04-3.92 (m, 3H, H-5, b-

CH₂(Ser)), 2.00 (s, 3H, OAc), 1.96 (s, 3H, OAc), 1.95 (s, 3H, OAc). ¹⁹F-NMR

(DMSO-d₆): d ꢂ201.1 (dd, J_F,2=48.5 Hz, J_F,3=12.5 Hz, a-anomer), ꢂ199.6

(ddd, J_F,2=51.2 Hz, J_F,3=14.3 Hz, J_F,152.5 Hz, b-anomer); a/b=66/34. MS

(APCI): m/z=617.9 [M+H]⁺, ESI-MS: 639.9 [M+Na], 635.0 [M+H₂O]⁺,

615.9 [MꢂH]^ꢂ.

As a major by-product 3,4,6-tri-O-acetyl-2-deoxy-2-ﬂuoro-d-glucopyrano-

syl N^a-(9-ﬂuorenylmethoxycarbonyl)-l-serinoate (13) was characterized as

follows: HPLC (Kromasil C8, 250 ꢄ 4.6, 1.5 ml/min, 40–100% CH₃CN in

water (0.1% TFA) in 50 min): k⁰=10.8. ¹H-NMR (DMSO-d₆): d 7.88-7.24 (m,

8H, CH-Fmoc), 6.30 (d, 1H, H-1 (a), J_1,2=3.8 Hz), 6.05 (dd, 1H, H-1 (b),

J_1,2=7.9 Hz, J_1,F=3.3 Hz), 5.40 (dt, 1H, H-3, J_3,F=12.2 Hz, J_3,4=10.1 Hz),

4.96 (t(dd), 1 H, H-4, J_4,5=10.1 Hz), 4.93 (ddd, 1 H, H-2, J_2,F=48.6 Hz,

J_2,3=10.1 Hz), 4.34-4.19 (m, 4H, a-CH (Ser), CH₂-Fmoc, CH-Fmoc),

J Label Compd Radiopharm 2005; 48: 701–719

714

S. MASCHAUER ET AL.

4.15-4.05 (m, 2H, H-6a, H-6b), 4.04-3.92 (m, 3H, H-5, b-CH₂(Ser)), 2.00

(s, 3H, OAc), 1.96 (s, 3H, OAc), 1.95 (s, 3H, OAc). ¹⁹F-NMR (DMSO-d₆):

d ꢂ202.3 (dd, J_F,2=48.6 Hz, J_F,3=12.2 Hz, a-anomer), ꢂ201.2 (ddd,

J_F,2=51.9 Hz, J_F,3=12.0 Hz, J_F,153.5 Hz, b-anomer); a/b=70/30. MS

(APCI): m/z = 617.9 [M+H]⁺, ESI-MS: 635.0 [M+H₂O]⁺, 615.9 [MꢂH]^ꢂ.

Method B: Glycosylation method using BF₃. 108 ml (0.86 mmol) BF₃ꢀ Et₂O

were added to a solution of 100 mg (285 mmol) 1 and 93 mg (285 mmol)

N^a-(9-ﬂuorenylmethoxycarbonyl)-l-serine in 2 ml dry acetonitrile. The

mixture was stirred at room temperature and the reaction was monitored

by gradient HPLC (Kromasil C8, 250 ꢄ 4.6, 1.5 ml/min, 40–100% CH₃CN

in water (0.1% TFA) in 50 min). After 24 h the solution was washed with

saturated NaHCO₃and water, dried with Na₂SO₄and concentrated. The

residue was puriﬁed by preparative HPLC (Kromasil C8, 250 ꢄ 20, 15 ml/min,

CH₃CN/H₂O 1:1 (0.1% TFA)) to give only 51% (52 mg, 53 mmol) of 2.

HPLC (Kromasil C8, 250 ꢄ 4.6, 1.5 ml/min, 40–100% CH₃CN in water (0.1%

TFA) in 50 min): k⁰=10.3. The main reaction was O-acylation at the amino

acid side chain yielding 48% of N^a-(9-ﬂuorenylmethoxycarbonyl)-O-acetyl-l-

1

serine (11). This product was conﬁrmed by H-NMR and HPLC (Kromasil

C8, 250 ꢄ 4.6, 1.5 ml/min, 40–100% CH₃CN in water (0.1% TFA) in 50 min):

k⁰=6.6) as compared to the commercially available reference (Bachem No.:

1

B-1010). H-NMR (DMSO-d₆): d 13.0 (1H, COOH), 7.9-7.3 (m, 9H, NH,

CH-Fmoc), 4.4-4.2 (m, 6H, a-CH, CH₂-Fmoc, CH-Fmoc, b-CH₂), 2.0

(s, 3H, OAc).

N^a-(9-Fluorenylmethoxycarbonyl)-O-(3,4,6-tri-O-acetyl-2-deoxy-2-ﬂuoro-d-

glucopyranosyl)-l-threonine (3)

Method A: Glycosylation method using AgOTf. A mixture of 100 mg

(270 mmol) 8 and 92 mg (270 mmol) N^a-(9-ﬂuorenylmethoxycarbonyl)-l-threo-

nine in 2 ml CH₂Cl₂was treated as described for the preparation of glycoside

9. The residue was isolated by preparative HPLC (Kromasil C8, 250 ꢄ 20,

15 ml/min, CH₃CN/H₂O 60:40 (0.1% TFA)) to give only 51% (52 mg,

53 mmol) of 3.

As a major by-product (25%) 3,4,6-tri-O-acetyl-2-deoxy-2-ﬂuoro-d-gluco-

pyranosyl N^a-(9-ﬂuorenylmethoxycarbonyl)-l-threoninoate (14) was charac-

terized as follows: HPLC (Kromasil C8, 250 ꢄ 4.6, 1.5 ml/min, 40–100%

1

CH₃CN in water (0.1% TFA) in 50 min): k⁰=12.0. H-NMR (DMSO-d₆): d

7.88-7.25 (m, 8H, CH-Fmoc), 6.31 (d, 1H, H-1 (a), J_1,2=3.9 Hz), 6.08 (dd, 1H,

H-1 (b), J_1,2=8.1 Hz, J_1,F=3.9 Hz), 5.47 (dt, 1H, H-3, J_3,F=12.4 Hz,

J_3,4=9.5 Hz), 4.96 (t(dd), 1H, H-4, J_4,5=9.9 Hz), 4.93 (ddd, 1H, H-2 (a),

J_2,F=47.6 Hz, J_2,3=9.7 Hz), 4.55 (ddd, 1H, H-2 (b), J_2,F=50.8 Hz,

J_2,3=9.0 Hz, J_2,1=8.0 Hz), 4.32-3.95 (m, 8H, a-CH (Thr), CH₂-Fmoc,

J Label Compd Radiopharm 2005; 48: 701–719

F-18 GLYCOSYLATION

715

CH-Fmoc, b-CH (Thr), H-6a, H-6b, H-5), 2.02 (s, 3H, OAc), 1.96 (s, 3H, OAc),

1.93 (s, 3H, OAc), 1.14 (d, 3H, CH₃(Thr), J=6.3 Hz). ¹⁹F-NMR (DMSO-d₆):

d ꢂ202.4 (dd, J_F,2=48.2 Hz, J_F,3=12.6 Hz, a-anomer), ꢂ201.3 (ddd,

J_F,2=51.4 Hz, J_F,3=14.6 Hz, J_F,1=3.5 Hz b-anomer); a/b= 80/20. ESI-MS:

649.0 [M+H₂O]⁺.

Method B: Glycosylation method using BF₃. 108 ml (855 mmol) BF₃ꢀ Et₂O were

added to a solution of 100 mg (285 mmol) 1 and 97 mg (285 mmol) N^a-(9-

ﬂuorenyl-methoxycarbonyl)-l-threonine in 2 ml dry acetonitrile. The mixture

was stirred at room temperature and the reaction was monitored by gradient

HPLC (Kromasil C8, 250 ꢄ 4.6, 1.5 ml/min, 40–100% CH₃CN in water (0.1%

TFA) in 50 min). After 24 h the solution was washed with saturated NaHCO₃

and water, dried with Na₂SO₄and concentrated. The residue was puriﬁed by

preparative HPLC (Kromasil C8, 250 ꢄ 20, 15 ml/min, CH₃CN/H₂O 1:1

(0.1% TFA)) to give only 51% (52 mg, 53 mmol) of 3.

The main reaction was O-acylation at the amino acid side chain yielding 43%

of N^a-(9-ﬂuorenylmethoxycarbonyl)-O-acetyl-l-threonine (12). HPLC: (Kro-

masil C8, 250 ꢄ 4.6, 1.5 ml/min, 40–100% CH₃CN in water (0.1% TFA) in

1

50 min): k⁰=6.6). H-NMR (DMSO-d₆): d 12.8 (1H, COOH), 7.9-7.3 (m, 9 H,

NH, CH-Fmoc), 5.2 (m, 1H, b-CH₂), 4.4-4.2 (m, 4H, a-CH, CH₂-Fmoc, CH-

Fmoc), 2.0 (s, 3H, OAc), 1.2 (d, 3H, CH₃). LC-MS (APCI): 406.0 [M+Na]⁺.

Method C: Glycosylation method using TMSOTf. 250 ml (2.2 mmol) TMSOTf

were added to a solution of 33 mg (94 mmol) 1 and 36 mg (105 mmol) N^a-(9-

ﬂuorenylmethoxycarbonyl)-l-threonine in 3 ml dry CH₂Cl₂. The mixture was

stirred at room temperature and the reaction was monitored by gradient

HPLC (Kromasil C8, 250 ꢄ 4.6, 1.5 ml/min, 40–100% CH₃CN in water (0.1%

TFA) in 50 min). After 2 days the solution was washed with saturated

NaHCO₃and water, dried with Na₂SO₄and concentrated. The residue was

puriﬁed by semipreparative HPLC (Kromasil C8, 125 ꢄ 8, 4 ml/min, 40–100%

CH₃CN in water (0.1% TFA) in 50 min) to give 7% (4,5 mg, 7 mmol) of 3. TLC

(ethyl acetate/n-hexane 7:3): R_f=0.18. HPLC (Kromasil C8, 250 ꢄ 4.6, 1.5 ml/

1

min, 40–100% CH₃CN in water (0.1% TFA) in 50 min): k⁰=11.2. H-NMR

(DMSO-d₆): d 7.92-7.26 (m, 8H, CH-Fmoc), 5.46 (dt, 1H, H-3, J_3,F=11.8 Hz,

J_3,4=9.6 Hz), 5.24 (d, 1H, H-1(a), J_1,2=3.8 Hz), 4.84 (t(dd), 1H, H-4,

J_4,5=9.8 Hz), 4.61 (ddd, 1H, H-2, J_2,F=49.0, J_2,3=9,7 Hz), 4.31-3.97 (m,

8H, a-CH (Thr), CH₂-Fmoc, CH-Fmoc, b-CH (Thr), H-6a, H-6b, H-5), 2.00

(s, 3H, OAc), 1.99 (s, 3H, OAc), 1.97 (s, 3H, OAc), 1.26 (d, 3H, CH₃(Thr),

J=6.3 Hz). ¹⁹F-NMR (DMSO-d₆):

d

ꢂ200.07 (dd, J_F,2=48.8 Hz,

J

_F,3=12.3 Hz, a-anomer); a/b>97/3. ESI-MS: m/z=631.9 [M+H]⁺, 649.0

[M+H₂O]⁺.

J Label Compd Radiopharm 2005; 48: 701–719

716

S. MASCHAUER ET AL.

N^a-(9-Fluorenylmethoxycarbonyl)-O-(2-deoxy-2-ﬂuoro-d-glucopyranosyl)-l-

serine (4)

Deacetylation of 2 was performed following a procedure ﬁrstly described by

Zemple

´

n.²⁶Brieﬂy, 13 mg (21 mmol) of 2 were stirred at room temperature in a

solution of 25 mM NaOMe in anhydrous methanol (1 ml) for 24 h. The

mixture was neutralized with Amberlite¹IR-120 (H⁺) and separated by

semipreparative HPLC (Kromasil C8, 125 ꢄ 8, 4 ml/min, 40–100% CH₃CN in

water (0.1% TFA) in 50 min) to yield 40% (3.9 mg, 8.4 mmol) of 4. HPLC

(Kromasil C8, 250 ꢄ 4.6, 1.5 ml/min, 40–100% CH₃CN in water (0.1% TFA)

in 50 min): k⁰=2.6. ESI-MS: m/z=514.0 [M+Na]⁺.

N^a-(9-Fluorenylmethoxycarbonyl)-O-(2-deoxy-2-ﬂuoro-d-glucopyranosyl)-l-

threonine (5)

34 mg (54 mmol) 3 were stirred at room temperature in a solution of 6 mM

NaOMe in anhydrous methanol (3 ml) for 4 days. The mixture was neutralized

with Amberlite¹IR-120 (H⁺) and separated by semipreparative HPLC

(Kromasil C8, 125 ꢄ 8, 4 ml/min, 40–100% CH₃CN in water (0.1% TFA) in

50 min) to yield 10% (2.7 mg, 5.4 mmol) of 5. HPLC (Kromasil C8, 250 ꢄ 4.6,

1.5 ml/min, 40–100% CH₃CN in water (0.1% TFA) in 50 min): k⁰=4.2. ESI-

MS: m/z=505.9 [M+H]⁺.

O-(2-Deoxy-2-ﬂuoro-d-glucopyranosyl)-l-serine (6)

124 mg (3.8 mmol) ammonium formate and 78 mg (0.73 mmol) palladium

black were added to a solution of 117 mg (189 mmol) N^a-(phenylmethox-

ycarbonyl)-O-(3,4,6-tri-O-acetyl-2-deoxy-2-ﬂuoro-d-glucopyranosyl)-l-serine

phenylmethyl ester (9) in 8 ml methanol under a nitrogen atmosphere. The

suspension was stirred at room temperature for 15 min, ﬁltered through

Celite¹and concentrated. The residue was dissolved in 15 ml 50 mM NaOMe

in methanol for 2 h. The mixture was neutralized with Amberlite¹IR-120

(H⁺) and concentrated. Demineralization and isolation was realized by

passing the reaction through a BioGel¹P2 column (Biorad) and afterwards a

PD-10 Sephadex¹G-25 M column (Pharmacia). The aqueous solution was

lyophilized giving 11% (5.6 mg, 21 mmol) of 6. TLC (CH₃CN/9.5 mM

tetrabutylammonium hydroxide 8:2): R_f= 0.14. HPLC (Kromasil C8,

250 ꢄ 4.6, 1.5 ml/min, 40–100% CH₃CN in water (0.1% TFA) in 50 min):

k⁰=0.2. ¹H-NMR (D₂O): d 8.35 (s, 2H, NH₂), 5.06 (d, 1H, H-1 (a),

J_1,2=3.8 Hz), 4.68 (dd, 1H, H-1 (b), J_1,2=7.4 Hz, J_1,F=3.0 Hz), 4.33(ddd, 1H,

H-2, J_2,F=48.0 Hz, J_2,3=9.6 Hz), 4.10 (dd, 1H, H-6a, J_6a,6b=10.5 Hz,

J_6a,5=3.1 Hz), 3.94-3.35 (m, 6H, H-3, H-6b, H-5, H-4, CH₂(Ser)). ¹⁹F-

NMR (D₂O): d ꢂ196.13 (ddd, J_F,2=50.8 Hz, J_F,3=15.9 Hz, J_F,1=1.7 Hz, b-

anomer), ꢂ197.5 (dd, J_F,2=48.0 Hz, J_F,3=13.6 Hz, a-anomer); a/b=66/34.

J Label Compd Radiopharm 2005; 48: 701–719

F-18 GLYCOSYLATION

717

O-(2-Deoxy-2-ﬂuoro-d-glucopyranosyl)-l-threonine (7)

147 mg

(108 mmol)

N^a-(phenylmethoxycarbonyl)-O-(3,4,6-tri-O-acetyl-2-

deoxy-2-ﬂuoro-d-glucopyranosyl)-l-threonine phenylmethyl ester (10) were

treated as described for the preparation of 6 yielding 7% (2 mg, 7.5 mmol) of 7.

TLC (CH₃CN/9.5 mM tetrabutyl-ammonium hydroxide 8:2): R_f= 0.20.

HPLC (Kromasil C8, 250 ꢄ 4.6, 1.5 ml/min, 40–100% CH₃CN in water

(0.1% TFA) in 50 min): k⁰=0.2. ¹H-NMR (D₂O): d 8.35 (s, 2H, NH₂), 5.13 (d,

1H, H-1 (a), J_1,2=3.9 Hz), 4.28 (ddd, 1H, H-2, J_2,F=48.5 Hz, J_2,3=9.3 Hz),

3.94-3.30 (m, 6H, H-3, H-6a, H-6b, H-5, H-4, CH (Thr)), 1.30 (d, 3H, CH₃

(Thr), J=6.3 Hz). ¹⁹F-NMR (D₂O): d ꢂ195.76 (ddd, J_F,2=49.1 Hz,

J_F,3=14.3 Hz, J_F,1=2.6 Hz, b-anomer), ꢂ196.00 (dd, J_F,2=47.4 Hz,

J_F,3=12.8 Hz, a-anomer); a/b= 67/33.

Metabolic stability of ½¹⁸Fꢅ6 in human serum

200 ml of [¹⁸F]6 were dissolved in 2 ml of human serum and incubated at 378C.

Aliquots (100 ml) were taken at various time intervals (5–45 min) and quenched

in 500 ml methanol/CH₂Cl₂(1:1). The samples were centrifuged and the

supernatants analyzed by radio-TLC using CH₃CN/9.5 mM tetrabutylammo-

nium hydroxide (8:2) as eluant.

Conclusion

In summary, we developed a two-step ¹⁸F-glycosylation method based on the

commonly available intermediate 1,3,4,6-tetra-O-acetyl-2-deoxy-2-[¹⁸F]ﬂuoro-

glucopyranose ([¹⁸F]1), which occurs as an intermediate in the [¹⁸F]FDG

synthesis. This prosthetic group was successfully coupled to the side chains of

Fmoc-protected serine and threonine by the use of boron triﬂuoride as

promotor. The ¹⁸F-glycosylation method aﬀorded a radiochemical yield of

about 25% for the serinyl compound within a total synthesis time of 90 min.

Thus, [¹⁸F]1 is a promising ¹⁸F-labelled prosthetic group suitable for ¹⁸F-

glycosylation of amino acid side chains and could be adapted for ¹⁸F-labelling

of bioactive peptides to study their pharmacokinetics using positron emission

tomography (PET). Further studies have to investigate the scope of this

methodology particularly with regard to its potential to provide not only a

method for ¹⁸F-glycosylation but also the opportunity to improve the

biokinetics of PET-radiophamaceuticals.

Acknowledgements

The authors gratefully acknowledge ﬁnancial support by Amersham

Healthcare Deutschland GmbH. Special thanks are also due to Dr J. Einsiedel

and Dr S. Lober (Department of Medicinal Chemistry, Friedrich-Alexander

¨

University, Erlangen) for their expert help in ESI-MS and LC/MS analysis.

J Label Compd Radiopharm 2005; 48: 701–719

718

S. MASCHAUER ET AL.

We also gratefully acknowledge Wilhelm Hamkens and the cyclotron staﬀ

(PET Net GmbH, Erlangen) for producing and delivering [¹⁸F]ﬂuoride. This

project was supported by a grant from the Deutsche Forschungsgemeinschaft

(DFG, GRK 750).

References

1. Okarvi SM. Eur J Nucl Med 2001; 28: 929–938.

2. Wester HJ, Brockmann J, Rosch F, Wutz W, Herzog H, Smith-Jones P, Stolz B,

¨

Bruns C, Stocklin G. Nucl Med Biol 1997; 24: 275–286.

¨

3. Haubner R, Wester HJ, Weber WA, Mang C, Ziegler SI, Goodman SL,

Senekowitsch-Schmidtke R, Kessler H, Schwaiger M. Cancer Res 2001; 61:

1781–1785.

4. Bergmann R, Scheunemann M, Heichert C, Mading P, Wittrisch H, Kretzschmar

¨

M, Rodig H, Tourwe D, Iterbeke K, Chavatte K, Zips D, Reubi JC, Johannsen B.

´

Nucl Med Biol 2002; 29: 61–72.

5. Vaidyanathan G, Zalutsky MR. Nucl Med Biol 1997; 24: 171–178.

6. Herman LW, Fischman AJ, Tompkins RG, Hanson RN, Byon C, Strauss HW,

Elmaleh DR. Nucl Med Biol 1994; 21: 1005–1010.

7. Wester HJ, Hamacher K, Stocklin G. Nucl Med Biol 1996; 23: 365–372.

¨

8. Kilbourn MR, Dence CS, Welch MJ, Mathias CJ. J Nucl Med 1994; 28: 462–470.

9. Jelinski M, Hamacher K, Coenen HH. J Label Compd Radiopharm 2001; 44

(Suppl. 1): S151–S153.

10. Prante O, Hamacher K, Coenen HH. J Label Compd Radiopharm 1999; 42 (Suppl.

1): S111–S112.

11. Glaser M, Karlsen H, Solbakken M, Arukwe J, Brady F, Luthra SK, Cuthbertson

A. Bioconjug Chem 2004; 15: 1447–1453.

12. Poethko T, Schottelius M, Thumshirn G, Hersel U, Herz M, Henriksen G,

Kessler H, Schwaiger M, Wester HJ. J Nucl Med 2004; 45: 892–902.

13. Polt R, Porreca F, Szabo LZ, Bilsky EJ, Davis P, Abbruscato TJ, Davis TP,

Horvath R, Yamamura HI, Hruby V. Proc Natl Acad Sci USA 1994; 91:

7114–7118.

14. Egleton RD, Mitchell SA, Huber JD, Palian MM, Polt R, Davis TP. J Pharmacol

Exp Ther 2001; 299: 967–972.

15. Egleton RD, Mitchell SA, Huber JD, Janders J, Stropova D, Polt R, Yamamura

HI, Hruby V, Davis TP. Brain Res 2000; 881: 37–46.

16. Tomatis R, Marastoni M, Balboni G, Guerrini R, Capasso A, Sorrentino L,

Santagada V, Caliendo G, Lazarus LH, Salvadori S. J Med Chem 1997; 40:

2948–2952.

17. Nicolaou KC, Mitchell HJ. Angew Chem Int Ed 2001; 40: 1576–1624.

18. Salvador LA, Elofsson M, Kihlberg J. Tetrahedron 1995; 51: 5643–5656.

19. Gangadhar BP, Jois SDS, Balasubramaniam A. Tetrahedron Lett 2004; 45:

355–358.

20. Jensen KJ, Meldal M, Bock K. J Chem Soc, Perkin Trans 1 1993; 17: 2119–2129.

J Label Compd Radiopharm 2005; 48: 701–719

F-18 GLYCOSYLATION

719

21. Polt R, Szabo L, Treiberg J, Li Y, Hruby V. J Am Chem Soc 1992; 114:

10249–10258.

22. Carvalho I, Scheuerl SL, Kartha KPR, Field RA. Carbohydr Res 2003; 338:

1039–1043.

23. Hamacher K, Coenen HH, Stocklin G. J Nucl Med 1986; 27: 235–238.

¨

24. Patt M, Sorger D, Scheunemann M, Stocklin G. Appl Radiat Isot 2002; 57:

¨

705–712.

25. Yamada H, Hayashi T. Carbohydr Res 2002; 337: 581–585.

26. Zemplen G, Pascu E. Ber Dtsch Chem Ges 1929; 62: 1613–1614.

´

27. Agoston K, Dobo A, Rako J, Kerekgyarto J, Szurmai Z. Carbohydr Res 2001;

330: 183–190.

28. Deferrari JO, Gros EG, Mastronardi IO. Carbohydr Res 1967; 4: 432–434.

29. Kovac P. Carbohydr Res 1986; 153: 168–170.

30. Hamacher K. Carbohydr Res 1984; 128: 291–295.

31. Kovac P, Yeh HJC, Glaudemans CPJ. Carbohydr Res 1987; 169: 23–34.

32. Card PJ. J Org Chem 1983; 48: 393–395.

33. Elofsson M, Walse B, Kihlberg J. Tetrahedron Lett 1991; 51: 7613–7616.

34. Banoub J, Bundle DR. Can J Chem 1979; 57: 2091–2097.

35. Seitz O, Wong CH. J Am Chem Soc 1997; 119: 8766–8776.

36. Marzabadi CH, Franck RW. Tetrahedron 2000; 56: 8385–8417.

37. Chong PY, Roush PR. Org Lett 2002; 4: 4523–4526.

38. Sjolin P, Elofsson M, Kihlberg J. J Org Chem 1996; 61: 560–565.

¨

39. Binkley RW, Ambrose MG, Hehemann DG. J Carbohyd Chem 1987; 6: 203–219.

40. Adamson J, Foster AB, Hall DL, Johnson RN, Hesse RH. Carbohydr Res 1970;

15: 351–359.

J Label Compd Radiopharm 2005; 48: 701–719

Article Doi

Utility of 1,3,4,6-tetra-O-acetyl-2-deoxy-2-[18F] fluoroglucopyranoside for no-carrier-added18F-glycosylation of amino acids

DOI: 10.1002/jlcr.963

Source and publish data:

Authors:

Article abstract of DOI:10.1002/jlcr.963

Full text of DOI:10.1002/jlcr.963

Products guided by the article

R&D Labs maybe for 909784-85-6

Relevant to this article

Hot Product