Small Prosthetic Groups in 18F-Radiochemistry:
`Useful Auxiliaries for the Design of 18F-PET
`Tracers
`Ralf Schirrmacher, PhD,* Björn Wängler, PhD,† Justin Bailey, PhD,* Vadim Bernard-Gauthier, PhD,*
`Esther Schirrmacher, PhD,* and Carmen Wängler, PhD‡
`
`Prosthetic group (PG) applications in 18F-radiochemistry play a pivotal role among current 18F-labeling
`techniques for the development and availability of 18F-labeled imaging probes for PET (Wahl, 2002)
`(1). The introduction and popularization of PGs in the mid-80s by pioneers in 18F-radiochemistry
`has profoundly changed the landscape of available tracers for PET and has led to a multitude of
`new imaging agents based on simple and efficiently synthesized PGs. Because of the chemical
`nature of anionic 18F− (apart from electrophilic low specific activity 18F-fluorine), radiochemistry
`before the introduction of PGs was limited to simple nucleophilic substitutions of leaving group
`containing precursor molecules. These precursors were not always available, and some target
`compounds were either hard to synthesize or not obtainable at all. Even with the advent of re-
`cently introduced “late-stage fluorination” techniques for the 18F-fluorination of deactivated aromatic
`systems, PGs will continue to play a central role in 18F-radiochemistry because of their robust
`and almost universal usability. The importance of PGs in radiochemistry is shown by its current
`significance in tracer development and exemplified by an overview of selected methodologies for
`PG attachment to PET tracer molecules. Especially, click-chemistry approaches to PG conjuga-
`tion, while furthering the historical evolution of PGs in PET tracer design, play a most influential
`role in modern PG utilization. All earlier and recent multifaceted approaches in PG development
`have significantly enriched the contingent of modern 18F-radiochemistry procedures and will con-
`tinue to do so.
`Semin Nucl Med 47:474–492 © 2017 Elsevier Inc. All rights reserved.
`
`Introduction
`B efore discussing a prosthetic group (PG)’s purpose in
`
`radiochemistry,1 its definition needs to be clarified. The
`use of the term “prosthetic group” has been traditionally em-
`ployed in enzymology and defined as a cofactor that is “either
`tightly or loosely bound to the enzyme. If tightly con-
`nected, the cofactor is referred to as a prosthetic group.”2 How
`did this definition migrate into the field of radiochemistry and
`what is the significance of it in this particular context? It should
`
`*Medical Isotope and Cyclotron Facility, Cross Cancer Institute, University
`of Alberta, Alberta, Canada.
`†Molecular Imaging and Radiochemistry, Department of Clinical Radiology
`and Nuclear Medicine, Medical Faculty Mannheim of Heidelberg
`University, Germany.
`‡Biomedical Chemistry, Department of Clinical Radiology and Nuclear
`Medicine, Medical Faculty Mannheim of Heidelberg University, Germany.
`Address reprint requests to Ralf Schirrmacher, PhD, Medical Isotope and
`Cyclotron Facility, Cross Cancer Institute, University of Alberta, AB,
`Canada. E-mail: schirrma@ualberta.ca
`
`474
`
`http://dx.doi.org/10.1053/j.semnuclmed.2017.07.001
`0001-2998/© 2017 Elsevier Inc. All rights reserved.
`
`be recalled that for the development of a new radiotracer,
`usually (in almost all cases), a nonradioactive lead com-
`pound devised by medicinal chemistry is taken as a chemical
`template or scaffold and translated into a radiotracer for in
`vivo imaging by introducing a radiolabel. This general strat-
`egy holds true for tracers intended for PET, where so-called
`organic radionuclides such as carbon-11 and fluorine-18 (18F)
`are commonly used, and where the size of compounds varies
`between small tracers for brain imaging, and peptide- and
`protein-based probes for cancer imaging.3-8 If the original lead
`structure does not contain, for example, fluorine, the radio-
`chemist has to add fluorine at a position in the molecule that
`most likely does not interfere with its binding properties to
`the target. The smaller the change in the molecular struc-
`ture of the lead compound, the less probable the binding
`behavior to the target structure will change. Therefore, ra-
`diochemists and also medicinal chemists (addition of fluorine
`to a molecule can have very positive effects on its bio-
`availability and stability in vivo9) often simply exchange a
`
`Petitioner GE Healthcare – Ex. 1019, p. 474
`
`

`

`Smallprostheticgroupsin18F-radiochemistry
`
`475
`
`hydrogen atom in a lead with a fluorine atom based on the
`similar van der Waals radii of hydrogen and fluorine. Un-
`fortunately, it is not always possible to introduce radioactive
`18F at the same position in a molecule where medicinal chem-
`ists are able to add stable fluorine in their multistep synthetic
`schemes. This ostensible contradiction can be explained easily
`by the different requirements for nonradioactive and radio-
`active fluorine in preparative chemistry. Although 18F is still
`fluorine in chemical terms, the radioactive nature of 18F in-
`troduces the problem of radioactive decay to the synthetic
`equation. In general, multistep syntheses up to this day are
`not desirable in 18F radiochemistry because such laborious
`and lengthy procedures will never yield the desired radio-
`tracer in high radiochemical yields (RCYs) and high specific
`activities, not to mention the cumbersome purification of many
`intermediates on the way toward the final product.
`
`What Is a Prosthetic Group?
`These ramifications are still the reasons why PGs have been in-
`vented and used in tracer development ever since. In principle, a
`PG overcomes the inability to radiolabel certain compounds with
`18F in 1-2 steps via nucleophilic 18F− substitution of a leaving group
`in a timely manner. Comparable with carbon-11 methylations using
`11C-methyliodide, small reactive 18F bearing labeling synthons were
`envisioned as analogous labeling reagents to engage nucleophilic moi-
`eties such as amino, hydroxy, carboxy, and thiol groups. This strategy
`made it possible to easily transfer a 11C-methylation to an equiva-
`lent 18F-fluoro methylation or ethylation (or alkylation in general),
`taking advantage of the already existing labeling precursor for
`11C-labeling. This concept has been recently extended to a multi-
`tude of chemical 18F-bioconjugations based on click chemistry,
`significantly supporting the use of PG applications in radiotracer de-
`velopment despite the development of novel metal catalyzed late-
`stage 18F-fluorination techniques that have substantially enriched
`existing 18F-labeling techniques.10-14 Referring back to the above-
`introduced definition of a PG in enzymology, a PG in
`radiopharmaceutical chemistry is a serviceable auxiliary that can easily
`
`be attached to a precursor molecule in analogy to existing 11C-labeling
`techniques (alkylation, acylation, amination, etc) or click-chemistry
`methodologies (triazole-, dihydropyrazine-, oxime formations, etc).
`However, the use of PGs for tracer development carries an impor-
`tant limitation. Usually, a first-in-kind radiotracer (especially for small
`compounds of low molecular weight) is directly derived from the
`lead structure that has no functionalities that could be easily modi-
`fied to introduce 18F by a PG. As a result, in most cases, a methyl
`group that is usually abundantly available in most small organic com-
`pounds is substituted for a 11C-methyl group. This strategy yields
`a radiotracer that is in every regard exactly the same as the nonra-
`dioactive lead compound. In contrast, the structural exchange of a
`methyl group with a higher 18F-labeled homolog such as an
`18F-fluoroethyl group or even a closely related 18F-fluoromethyl group
`can significantly alter the biological and chemical properties of the
`resulting molecule. It is the responsibility of the researcher to prove
`that this very minor chemical-radiochemical alteration in struc-
`ture does not compromise important parameters such as toxicity,
`metabolism, and binding affinity to the target. This of course adds
`a new layer to the development process of 18F-radiopharmaceuticals,
`a noticeable detriment in comparison with an unaltered 11C-tracer
`that has most often already been tested for toxicity, binding affin-
`ity, and metabolism in the course of commercial drug development.
`This handicap, however, is offset by the advantages of 18F over 11C
`in terms of nuclide properties, and longer half-life, which allows for
`a prolonged duration of PET measurement and thus a higher quality
`of corresponding PET images.
`Many of the abovementioned concepts of 18F-PG introduction
`can be illustrated by the synthesis of 6-O-(2-[18F]fluoroethyl)-6-O-
`desmethyl diprenorphine, a PET tracer for opioid in vivo imaging,
`where the original methyl group in 6 position is substituted by an
`18F-fluoroethyl moiety (Fig. 1).
`The methoxy group can be easily converted into a hydroxy group
`(the desmethyl precursor) that can be labeled either by [11C]methyl
`iodide or 2-[18F]fluoroethyl tosylate ([18F]FETos).15 This particular
`position is the most suited to make small changes to the structure
`of diprenorphine without compromising its binding to the opioid
`receptors. Another group had previously chosen the N-17 posi-
`tion to replace the cyclopropyl methyl group with an 18F-fluoro ethyl
`
`unfavorable derivatization site
`
`18F
`
`n=1-2
`N
`
`18F-PG labeling
`at N-17 position
`
`OH
`
`N-17
`
`18F-PG labeling
`at O-6 position
`
`OH
`
`N
`
`HO
`
`O
`
`OCH3
`
`HO
`
`O
`
`O-6CH3
`
`HO
`
`O
`
`O
`
`OH
`
`18F
`
`reduced binding to opioid receptors
`
`favorable derivatization site
`
`preserved binding to opioid receptors
`
`Figure 1 Derivatization sites of the opioid receptor ligand diprenorhine. Structural changes of diprenorphine at the O-6 position (right) yield
`preserved binding to the receptor, whereas changes at the N-17 position (left) lead to a reduced binding to the receptor.
`
`Petitioner GE Healthcare – Ex. 1019, p. 475
`
`

`

`476
`
`R.Schirrmacheretal.
`
`and -propyl moiety, but the in vivo evaluation of the resulting com-
`pounds revealed reduced binding in opioid receptor-rich regions
`in comparison with the 11C-original.16 Fortunately, the O-6 derivatized
`18F variant of diprenorphine overcomes the shortcoming of the origi-
`nal 11C-labeled compound of a very limited PET scan duration and
`even enables shipment of the compound to PET centers without a
`cyclotron. This review does not claim to summarize the entire lit-
`erature with regard to reported PGs over the last 30 years but will
`rather explain conceptual strategies behind PG utilization using de-
`monstrative examples.
`
`The Early Days of Prosthetic
`Group Chemistry
`Early research was reported in the late 1970s and early 1980s,
`when the synthesis of halo-18F-fluoromethanes was de-
`scribed in the context of hot-atom recoil labeling where the
`precursor substrate was directly added to the neon gas target
`and bombarded with 14-MeV deuterons.17 The “hot” 18F−
`formed via the 20Ne(d,α)18F reaction could react directly
`with in-target components to form the corresponding
`halo-18F-fluoromethanes. Unfortunately, the high energy trans-
`fer rates within a target during bombardment led to an increased
`formation of radioactive impurities impairing the produc-
`tion of higher activities. The irradiation of halo-alkanes in the
`presence of 18F generates halo-18F-fluoro alkanes, which were
`characterized but not used for any kind of follow-up label-
`ing. Another experimental study with more detailed information
`but still limited practical applicability for preparative
`18F-radiochemistry was reported by Root and Manning.18 At
`the same time, 18F-fluorinated surface oxidized silver wool as
`a curious source of 18F was used for the synthesis of
`halo-18F-fluoro alkanes. The authors did not have the so-
`phisticated labeling utilities at their disposal, as modern
`radiochemists and consequently the reaction mechanism using
`silver wool is still not well understood.19 Another early con-
`tribution from Coenen and coworkers in 1985 fully
`characterized for the first time 1-bromo-[18F]fluoromethane
`with the intent to use this synthon analogously to [11C]methyl
`iodide.20 The labeling conditions to 18F-fluorinate the pre-
`cursor dibromomethane via nucleophilic substitution were
`different from previous approaches. Almost at the same time,
`Dae and coworkers reported on a mechanistic study of
`halofluorination of olefins that paved the way towards other
`investigations.21 The full automation of the synthesis of
`bromo-18F-fluoromethane was achieved by automated dis-
`tillation of the reaction mixture over 3 Sep-Pak Plus silica
`cartridges to remove the dibromomethane precursor and
`unreacted 18F−.22 The introduction of the aminopolyether
`Kryptofix222 and potassium carbonate system as well as the
`tetrabutylammonium 18F-fluoride complex23,24 constituted im-
`portant improvements enhancing the nucleophilicity of the
`18F−anion, via complexation of the potassium cation through
`Kryptofix222, leaving the 18F− un-solvatized and thus “naked”
`or through the formation of highly soluble nBu4N[18F]F (due
`to the phase transfer capabilities of nBu4N+), respectively. The
`implementation of the Kryptofix222/K[18F]F labeling system
`
`marked a turning point in 18F-radiochemistry. It is still an im-
`portant pillar of modern 18F-labeling methodology and had
`a catalytic effect on the further development of PGs and ra-
`diotracers in general. Previously, hard to label compounds
`suddenly became available through this innovation, invigo-
`rating and intensifying the research on 18F-labeled compounds
`and making 18F the most important diagnostic radionuclide
`today.25 Besides bromo- and iodo-[18F]fluoromethane, the more
`reactive [18F]fluoromethyl triflate was introduced in 2002 via
`online conversion of bromo-[18F]fluoromethane using silver
`triflate in a gas-solid phase reaction.26 The year 1986 was most
`memorable with regard to PG development and PG applica-
`tion in tracer research. Several pioneering researchers in
`radiochemistry such as Kilbourn, Katzenellenbogen, Welch,
`Shiue, Wolf, Barrio, Coenen, and Stöcklin presented their early
`or continuing research on 18F-fluorinated alkyl tosylates,
`-mesylates, and -halides at the Sixth International Sympo-
`sium on Radiopharmaceutical Chemistry in Boston,
`Massachusetts, USA, initializing a new important chapter in
`radiochemistry that still continues to have an effect on tracer
`conception. It cannot be overstated that putting forward the
`Kryptofix222/K2CO3 labeling system for anionic 18F-fluorination
`sustainably invigorated and accelerated progress in 18F-tracer
`development. Although some new methods have completely
`omitted toxic Kryptofix222 from the synthesis and also omitted
`the azeotropic drying procedure,27 this 18F-nucleophilicity-enhancing
`auxiliary is still in frequent use. Shiue and Wolf for example applied
`this labeling system in the synthesis of higher 18F-fluoroalkylated
`analogs of spiroperidol (spiperone 4, Fig. 2) (at the same confer-
`ence there were several contributions from different groups featuring
`the synthesis of 18F-N-alkyl labeled spiroperidols 5, Fig. 2), an im-
`portant ligand to investigate dopamine receptors in humans.23 The
`motivation to use PG chemistry was clearly motivated by the dif-
`ficult multistep synthesis of the original compound, making a routine
`application for human PET imaging unlikely. The obtained
`N-[18F]fluoroalkyl spiroperidol derivatives 5, despite being struc-
`turallyslightlydifferentfromthelead,couldbesynthesizedconsistently
`from their corresponding 18F-fluoroalkyl halides in radiochemical
`yields of up to 60% (decay corrected) (Fig. 2). Generally for con-
`venience’s sake, a 1-step fluorination is always preferable and the
`chosen route if a labeling precursor is available. However, if the syn-
`thesis of a direct 1-step labeling precursor is cumbersome or the
`direct 1-step labeling results in the formation of many side-
`products because of harsh labeling conditions, the PG labeling
`approach is preferred for reasons of more efficient synthesis and faster
`and immediate evaluation of the tracer. The pioneering work from
`all those early groups provided the necessary innovative momen-
`tum to advance radiochemistry to the next level and set the state for
`numerousapplications,furtherdevelopments,and,finally,clickchem-
`istry in PG application.28-31
`
`2-[18F]fluoroethyl tosylate: The PG
`Workhorse
`If there is one PG that almost every radiochemist has at least
`used once in his or her professional career, it would most prob-
`ably be [18F]FETos, first introduced by Block et al at the
`
`Petitioner GE Healthcare – Ex. 1019, p. 476
`
`

`

`Smallprostheticgroupsin18F-radiochemistry
`
`477
`
`HN
`
`O
`
`N
`
`N
`
`O
`
`18F
`
`O
`
`n
`
`N
`
`n=1-3
`
`N
`
`N
`
`O
`
`F
`spiperone
`
`F
`18F-fluoroalkyl-spiperone
`
`X
`
`OTf
`n
`n=1, X=Br
`n=2, X=Br
`n=2, X=I
`n=3, X=Br
`
`nBu4N[18F]
`
`X
`
`18F
`n
`
`Chi et al. (28)
`
`TosO
`
`OTos
`n
`
`K[18F]F/Kryptofix 222
`
`18F
`
`TosO
`
`Block et al. (29)
`
`K[18F]F/Kryptofix 222
`Br,I
`
`18F
`n
`
`Shiue et al. (23)
`
`n=1
`
`Br
`n
`
`I
`
`n n
`
`=1-2
`
`Br
`
`I
`
`Figure 2 Synthesis of various 18F-fluoro alkylation PGs for the synthesis of an 18F-derivative of the D2 receptor ligand spiperone.
`
`Forschungszentrum Juelich, Germany, in 1987.32 The
`easy and reliable synthesis of [18F]FETos, its simple and con-
`venient purification, its perfect balance between reactivity
`and stability in solution, and its unprecedented number of
`labeling applications attesting to its usefulness make it the
`most stand-out PG to date (originally cited from Block et al:
`“This compound appears optimal as fluoroalkylation agent
`with respect to size, stability and ease of its preparation”)32
`(Fig. 2). Countless applications of 18F-FETos have been
`published in the literature and it would be beyond the scope
`of a review to just count them all. Although many other similar
`2-[18F]fluoroethyl sulfonates (and triflates cf. above) have
`been the subject of systematic investigations33 (eg, for the
`synthesis of the dopamine reuptake ligand [18F]FECNT 7
`from precursor 6, Fig. 3), [18F]-FETos remains the most
`prominent labeling synthon among 18F-fluoro-ethylating agents
`targeting amino and deprotonated hydroxy functions. [18F]FETos
`functions as an 18F-surogate PG for [11C]-methyl iodide34
`although bromo-[18F]fluoromethyl or [18F]trifluoromethyl
`synthons structurally more closely resemble [11C]methyl iodide
`in terms of spatial and steric demand. A drawback of these
`smaller PGs is their more intricate preparation (see above).
`Especially for less-experienced radiochemists, the synthesis
`of 1-bromo-18F-fluoro methane bears some preparative
`pitfalls, which can translate into low radiochemical yields
`or an impure PG. Furthermore, 18F-trifluoromethyl-chemistry,
`despite its high synthetic potential, has only recently been
`introduced and has not been widely applied yet.35 In stark
`contrast, the synthesis of [18F]FETos is almost foolproof. The
`precursor ethylene 1,2-ditosylate is commercially available,
`constitutes a non-volatile solid (unlike dibromo- or diiodo
`ethane, which are liquids but also greenhouse gases and
`should be used with care) and has a high molecular weight
`
`allowing accurate, easy, and convenient weighing control.
`The 18F-fluorination of ethylene 1,2-ditosylate reliably
`yields the PG between 50% and 80% RCY. The [18F]FETos is
`easily separated by High Performance Liquid Chromatogra-
`phy (HPLC) from its more lipophilic precursor, and final
`workup is equally simple via solid phase extraction. Even pro-
`cedures without HPLC separation have been reported.36 As
`for many PGs, one drawback of [18F]FETos should not be dis-
`regarded, namely the fact that the exchange of a 11C-methyl
`group with a 18F-fluoro ethyl group means a change in mo-
`lecular structure. Albeit small, this structural dissimilarity
`necessitates that any resulting radiotracer has to be tested for
`toxicity and change in biological behavior irrespective of whether
`these tests have already been performed for the original
`11C-tracer that is most often structurally equal to a fully evalu-
`ated pharmaceutical drug. [18F]FETos has been challenged over
`time by similar PGs such as bromo-18F-fluoro ethane.37,38 The
`efficient alkali iodide promoted 18F-fluoroethylation of
`p-anisidine 8, a very unreactive substrate for alkylations; using
`both 18F-FETos and bromo-18F-fluoroethane demonstrates the
`usefulness of both PGs for radiolabeling (Fig. 3). Several
`radioligands of relevance in nuclear medicine were labeled
`applying this methodology, for example, 18F-fluoroethyl-choline
`9 (Fig. 3). Although strategies have been reported to
`routinely prepare bromo-18F-fluoroethane by various meth-
`odologies, 18F-FETos seems continuously to assert itself as the
`most applied PG in the history of PG applications. An inter-
`esting metabolically stable alternative to 18F-FETos has been
`proposed in 2013 where a 18F-fluorocyclobutyl PG serves as
`an 18F-labeling agent 10 to provide an 18F-labeled amino acid
`11 (Fig. 3). The labeling yields using this synthon unfortu-
`nately did not compare favorably with other 18F-alkylating
`PGs despite its higher in vivo metabolic stability.39
`
`Petitioner GE Healthcare – Ex. 1019, p. 477
`
`

`

`478
`
`R.Schirrmacheretal.
`
`18F
`
`N
`
`COOMe
`
`Cl
`
`leaving group
`
`X
`
`Y
`
`O O
`
`O S
`
`18F
`
`X=Y=H
`X=Br, Y=H
`X=NO2, Y=H
`X=Y=Br
`X=Me, Y=H : tosylate moiety
`
`COOMe
`
`Cl
`
`HN
`
`18F
`
`70-80%
`
`no reaction
`
`HN
`
`low reactivity
`NH2
`
`18F
`
`18F
`
`OTos/Br
`
`OTos/Br
`
`NaI or KI or CsI
`
`MeO
`
`OH
`
`18F
`
`OTos/Br
`
`18F
`
`NaI or KI or CsI
`
`OH
`
`OTos-/Br-
`
`N
`
`[18F]FECH
`
`18F-/Kryptofix 222
`acetonitrile
`
`TosO
`
`OTos
`
`+tyrosine
`base
`
`18F
`
`H2N
`
`COOH
`
`O
`
`18F
`
`MeO
`
`N
`
`TosO
`
`Figure 3 Application of several different 18F-fluoroalkylating PGs for the 18F-fluoroalkylation of different precursor mol-
`ecules such as the dopamine re-uptake radioligand 18F-FECNT 7, 18F-FECH 9, and the tyrosine derivative 11.
`
`18F-Active Esters: Unspecific but
`Highly Reactive Labeling Agents
`In spite of [18F]FETos’s usefulness in labeling organic mol-
`ecules of low to medium molecular weight (<1000 u), it is
`less suited to react with complex molecules such as pep-
`tides and proteins for diverse reasons. First of all, [18F]FETos
`is not “super” reactive (OTos is a nucleofuge of medium re-
`activity) and requires some energy (eg, higher reaction
`temperatures) to react efficiently and in high radiochemical
`yields. Proteins in particular do not tolerate high tempera-
`tures and denature quickly. Peptides, if not fully side-chain
`protected, can easily yield more than just 1 18F-fluoroalkylation
`product, demanding laborious purification via HPLC.
`[18F]FETos has never found widespread application for peptide
`(or other larger compounds such as oligonucleotides) label-
`ing, and conditions for labeling have not been optimized
`because there are far better PGs for that purpose. Highly re-
`active chemically activated esters, radiolabeled with 18F, have
`early on been envisioned as reactive secondary labeling pre-
`cursors for peptide and protein labeling under preferably
`aqueous conditions because water perfectly dissolves both.
`Direct 18F-labeling approaches in water were not available at
`that time and were only introduced later in 2005-2006 using
`unconventional radiochemistries based on B-18F and Si-18F
`
`formation.40 The most prominent example of an activated ester-
`based PG is N-succinimidyl-4-[18F]fluorobenzoate ([18F]SFB)
`introduced by Vaidyanathan and Zalutsky in the 1990s.41,42
`Its publication has set a field-wide search in motion for new
`and even more efficient active ester-based PGs despite the
`report of significant radioactive side-product formation. The
`initial synthetic procedure (Fig. 4) where 4-formyl-N,N,N-
`trimethyl anilinium trifluoromethane sulfonate 12 is
`radiolabeled with 18F− to yield 4-[18F]fluorobenzaldehyde 13,
`which is oxidized into 4-[18F]fluoro benzoic acid 14 and finally
`converted in situ into [18F]SFB 15, has been improved, modi-
`fied, and automatized to make it more convenient and
`guarantee high RCYs for routine applications. Depending on
`the setup and method used, the overall synthesis can take
`approximately 2 hours, which is decidedly at the higher end
`of the preferable time line for PG syntheses. Automation of
`this usually 3-step preparation includes the initial labeling
`of the triflate salt of ethyl 4-(N,N,N-trimethylammonium)benzoate,
`its ester hydrolysis, and final in situ active ester formation. It was
`always a challenge for radiochemists to automatize the 18F-SFB
`synthesis to make its use more attractive for routine radiochemis-
`try. The introduction of other ester moieties such as tert-butyl ester
`into the labeling precursor 16 that is easily cleavable after
`18F-fluorination via trifluoroacetic acid yielding 18F-fluorobenzoic
`acid 14, has improved the original procedure for [18F]SFB synthesis
`
`Petitioner GE Healthcare – Ex. 1019, p. 478
`
`

`

`479
`
`O
`
`N
`
`O
`
`O O
`
`Smallprostheticgroupsin18F-radiochemistry
`
`Vaidyanathan et al. (41, 42)
`
`18F
`
`O O
`
`NHS, DCC, THF
`active ester
`H
`formation
`
`KMnO4, NaOH
`
`18F
`
`O H
`
`18F-/Kryptofix 222
`DMSO
`
`18F
`
`O H
`
`CF3SO3-
`
`(H3C)3N
`
`Wüst et al (43)
`
`active ester
`formation
`
`TSTU
`
`O O
`
`H
`
`1. 18F-/Kryptofix 222
`DMSO
`2. TFA
`
`18F
`
`O O
`
`CF3SO3-
`
`(H3C)3N
`
`18F-SFB reaction with a simple di-peptide
`
`18F
`
`R
`
`O
`
`NH2
`
`R'
`[18F]SFB
`
`HN
`
`R'
`
`O
`
`NH
`
`HOOC
`
`H2O
`
`R
`
`O
`
`NH
`
`HOOC
`
`Figure 4 Synthesis of [18F]SFB 15 and its application in radiolabeling of a dipeptide (box).
`
`(Fig. 4).43 Besides technical advancements in [18F]SFB synthesis
`such as a recently reported 3-step 1-pot synthesis for automation,44
`other active esters were devised to challenge [18F]SFB with regard
`to ease of preparation, synthesis time, and labeling efficiency. The
`most desirable synthetic route toward an 18F-active ester would be
`a single-step fluorination. One complication on the way toward a
`1-step labeling procedure is caused by the inherent lability of
`the active ester moiety. This strategy was tried very early in 1988
`and proved to be unsuccessful at that time.45 Although desirable
`for the intended labeling of proteins, this chemical reactiveness
`hinders the 1-step 18F-fluorination of an active ester precursor.
`Even if possible, the reactiveness of the active ester moiety leads
`to side products that have to be removed before the subsequent
`labeling with this PG. The synthesis of N-succinimidyl
`4-([18F]fluoromethyl)benzoate ([18F]SFMB) 18 by Lang and Eckelman
`was the first example of a 1-step active ester labeling that yielded
`the reactive PG in 18% RCY at the end of a 30- to 35-minute
`synthesis demonstrating impressively the reduction in synthesis
`time in comparison with the originally published 18F-SFB synthesis.46
`The relatively low yields of the 18F-fluorination of precursor 17
`were explained through experimentally confirmed formation of
`radioactive side products 19 and products of active ester hydroly-
`sis 20-21 (Fig. 5). [18F]Fluoromethyl benzoate synthesis was further
`improved by changing the leaving group of the precursor mol-
`ecule from Tos to Nos (4-nitobenzene sulfonate) as demonstrated
`in a 1-step labeling reaction of peptides.47,48 In 2010, Olberg et al
`and Malik et al reported on the 1-step synthesis of the 18F-labeled
`nicotinic acid derived active ester [18F]Fpy-TFP 23, where the
`2,3,5,6-tetrafluorophenyl active ester group survived the labeling
`of the trimethylammonium moiety at the 6-position of nicotinic
`acid 22.49,50 The 18F-fluorination surprisingly takes place at room
`temperature (Fig. 5). This PG 23 has been used in the synthesis of
`a PSMA targeting ligand. The purification of the PG was conve-
`niently achieved by simple Sep-Pak cartridge separation, omitting
`the need for an HPLC system. The yields of the labeling reaction
`were surprisingly high with 60%-70% at just 40°C, and the con-
`
`secutive reaction with several peptides (eg, arginylglycylaspartic
`acid (RGD)) was almost quantitative at relatively low peptide con-
`centrations. The recently introduced non-canonical labeling
`approaches based on B-18F51,52 and Si-18F40,53 that are character-
`ized by an easy 1-step introduction of 18F into small as well as
`complex organic compounds at very mild temperatures, led to
`the introduction of exotic PGs such as [18F]SiFB 24,54,55 which is
`structurally very similar to [18F]SFB. One caveat, however, is the
`higher lipophilicity of 18F-SiFB in comparison with [18F]SFB, lim-
`iting its use to protein labeling only, which can sufficiently compensate
`for the higher lipophilic character of the introduced silicon fluo-
`ride acceptor (SiFA) group.55 However, the synthesis of [18F]24 is
`outstandingly convenient, relying on the concept of isotopic ex-
`change and one simple cartridge purification step only as a result
`of the chemical identity of precursor and labeled product (Fig. 5).
`Another PG bearing an active ester group has been published in
`2013, tackling the high lipophilicity of the transferred 18F-fluoro
`benzyl group that can sometimes cause trouble when attached to
`small peptides.56 Although not based on a convenient 1-step reac-
`tion, this sulfonated hetero-bifunctional 18F-crosslinking PG 28 is
`based on the ring opening of a 1,3-propane sultone 25 with 18F−
`yielding 26, its deprotection to compound 27, and its final con-
`version into an N-hydroxysuccinimidyl ester 28 (Fig. 6). The
`18F-ring opening yielded only low amounts of product under
`conventional 18F-labeling conditions (acetonitrile, Kryptofix222,
`base) but works exceptionally well in tertiary alcohols (up to 90%
`radiochemical yield) at 90°C, a new labeling paradigm introduced
`by Kim et al.57 The sultone ring opening leads to the liberation of
`a sulfonic acid group, which imparts hydrophilicity to the PG.
`The PG 28 has only been used so far in the synthesis of an
`18F-labeled cyanine dye, and its usefulness for peptide and protein
`labeling has still to be proven. Shortly after, the same group applied
`a sultone-based labeling approach featuring a bis-sultone cross-
`linker to the synthesis of a peptide without implication of an
`active ester moiety.58 Another well-tolerated and synthetically proven
`18F-active ester was introduced by Guhlke et al in 1994,59 and
`
`Petitioner GE Healthcare – Ex. 1019, p. 479
`
`

`

`480
`
`R.Schirrmacheretal.
`
`isotopic exchange labeling
`
`F
`Si
`
`SiFB
`
`O O
`
`O
`
`N
`
`O
`
`Kryptofix 222
`acetonitrile
`RT
`
`18F
`Si
`
`[18F]SiFB
`
`O O
`
`O
`
`N
`
`O
`
`18F-
`
`TfO-
`
`(H3C)3N
`
`O
`
`F
`
`O
`
`F
`
`F
`
`F
`
`N
`
`activated ester
`activated carbon
`
`Kryptofix 222
`acetonitrile
`RT
`
`F
`
`F
`
`F
`
`O
`
`F
`
`O
`
`18F
`
`N
`
`[18F]Fpy-TFP
`
`18F-
`
`activated ester
`
`NO
`
`O
`
`b)
`
`OO
`
`a)
`
`b)
`
`TosO
`leaving group
`
`a)
`
`TosO
`
`hydrolysis
`
`18F
`
`O
`
`18F
`
`NO
`
`O
`
`OO
`
`[18F]SFMB
`
`18F
`
`hydrolysis
`
`18F
`Si
`
`high lipohilicity
`
`SiFA group
`
`HO
`
`O
`
`OO
`
`H
`
`18F
`
`Figure 5 Synthesis of different active ester PGs (18, 22, and 24) and possible side reactions (hydrolysis of 18 to 20 and 19 to 21).
`
`O
`
`OtBu
`
`Kryptofix 222/18F-
`Cs2CO3
`R-OH, 90ºC
`
`O
`S
`O O
`
`O
`
`O
`
`O
`
`N
`
`O
`
`O
`
`in situ formation
`of active ester
`
`18F
`
`HO
`SO O
`
`O
`
`TSTU, DIEA
`acetonitrile, 50ºC, 5 min
`
`18F
`
`HO
`SO O
`
`18F
`
`HO
`SO
`O
`
`O
`
`OtBu
`
`deprotection
`HCl, 80ºC
`
`O
`
`OH
`
`O
`
`O
`
`OTos/Br
`OR
`
`H3C
`
`O
`
`H3C
`
`O
`
`N
`
`O
`
`O
`
`Kryptofix 222/18F-
`K2CO3
`acetonitrile
`
`18F
`
`H3C
`
`OR
`
`O
`
`N
`
`OH-
`
`N,N'-disuccinimidyl carbonate
`
`18F
`
`H3C
`
`O-
`
`O
`
`O
`R= various ester groups
`18F
`
`Figure 6 Synthesis of active esters PGs 28 (sultone based) and N-succinimidyl-2-[18F]fluoropropionate 32.
`
`Petitioner GE Healthcare – Ex. 1019, p. 480
`
`

`

`Smallprostheticgroupsin18F-radiochemistry
`
`481
`
`used in the synthesis of 18F-peptides.60-62 The small PG
`N-succinimidyl-2-[18F]fluoropropionate 32 was used in the
`labeling of RGD dimer peptides, confirming its usefulness for
`peptide-based PET tracer development.63 The synthesis of 32 starts
`from either the bromo or the tosyl precursor 29, which is 18F-labeled
`under standard conditions, yielding ester compound 30. The ester
`moiety is cleaved by a base, and the resulting 18F-acid 31 is con-
`verted into the final active ester PG 32 (Fig. 6). A simplified
`1-step synthesis for 32 has been added recently to the portfolio of
`active ester PGs.64
`
`Click Chemistry: Modern Tools for
`PG Development
`Besides recent metal-mediated late-stage 18F-fluorination (see
`above), the facilitation and perfecting of click chemistry for
`18F-radiolabeling (and other radioactive tags) has been seri-
`ously changing radiochemistry over the last 2 decades after
`K. Barry Sharpless introduced the terminology to organic chem-
`istry in 2001.65-67 Click chemistry applications in all fields of
`radiochemistry have extensively been reviewed previously.68-73
`Especially, 18F-PG development has tremendously benefit-
`ted from the concept of click chemistry. The chemical efficiency
`of 18F-PG bioconjugation has been maximized through the
`utilization of a wide contingent of click-chemistry reactions
`such as oxime formation, Michael addition (eg, thiol-
`maleimide conjugation), 1,3-dipolar Huisgen cycloaddition
`(eg, (Cu(I) assisted as well as catalyst-free triazole forma-
`tion), and more recently Diels-Alder-like reactions (eg, tetrazine-
`dienophile r