Synthesis, Radiolabeling, and Biological ... - ACS Publications

May 18, 2016 - ... Vladimir-Prelog-Weg 1-5/10,. Zurich 8093, Switzerland ...... were approved by the Veterinary Office of the Canton Zurich, Zurich,. ...
2 downloads 0 Views 2MB Size
Subscriber access provided by University of Sussex Library

Article

Synthesis, Radiolabelling and Biological Evaluation of 5-Hydroxy-2[18F]Fluoroalkyl-Tryptophan Analogues as Potential PET Radiotracers for Tumor Imaging. Aristeidis Chiotellis, Adrienne Müller Herde, Simon Leonard Rössler, Ante Brekalo, Erika Gedeonova, Linjing Mu, Claudia Keller, Roger Schibli, Stefanie D. Krämer, and Simon M. Ametamey J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b00057 • Publication Date (Web): 18 May 2016 Downloaded from http://pubs.acs.org on May 19, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of Medicinal Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 63

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

Synthesis, Radiolabelling and Biological Evaluation of 5-Hydroxy-2-[18F]Fluoroalkyl-Tryptophan Analogues as Potential PET Radiotracers for Tumor Imaging. Aristeidis Chiotellis,*,† Adrienne Müller Herde,† Simon L. Rössler,† Ante Brekalo,† Erika Gedeonova,† Linjing Mu,‡ Claudia Keller,† Roger Schibli,† Stefanie D. Krämer,† Simon M. Ametamey†

† Center for Radiopharmaceutical Sciences ETH-PSI-USZ, Institute of Pharmaceutical Sciences ETH, Zurich, Switzerland ‡ Center for Radiopharmaceutical Sciences ETH-PSI-USZ, Department of Nuclear Medicine, University Hospital Zurich, Zurich, Switzerland

ACS Paragon Plus Environment

1

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 63

ABSTRACT

Aiming at developing mechanism-based amino acid synthesized two

18

18

F-PET tracers for tumor imaging, we

F-labelled analogues of 5-hydroxy-L-[β-11C]tryptophan ([11C]5HTP) whose

excellent in vivo performance in neuroendocrine tumors is mainly attributed to its decarboxylation by aromatic amino acid decarboxylase (AADC), an enzyme overexpressed in these malignancies. Reference compounds and precursors were synthesized following multistep synthetic approaches. Radiosynthesis of tracers was accomplished in good radiochemical yields (15-39%), high specific activities (45-95 GBq/µmol) and excellent radiochemical purities. In vitro cell uptake was sodium-independent and was inhibited ≥95% by 2-amino-2norbornanecarboxylic acid (BCH) and ~30% by arginine. PET imaging in mice revealed distinctly high tumor/background ratios for both tracers, outperforming the well-established O(2-[18F]fluoroethyl)tyrosine ([18F]FET) tracer in a head-to-head comparison. Biological evaluation revealed that the in vivo performance is most probably independent of any interaction with AADC. Nevertheless, the excellent tumor visualization qualifies the new tracers as interesting probes for tumor imaging worthy for further investigation.

ACS Paragon Plus Environment

2

Page 3 of 63

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

INTRODUCTION

In the past two decades, positron emission tomography (PET) has evolved into a powerful tool for the detection, staging and response to therapy monitoring of many cancerous lesions. Often used in conjunction with anatomic imaging (e.g. magnetic resonance imaging, MRI), PET is capable of characterizing tumors based on biochemical changes at the molecular level which allows higher sensitivity and specificity compared to anatomic imaging modalities alone. PET tracers developed for tumor imaging are designed to take advantage of the metabolic irregularities associated with several hallmarks of cancer. The uncontrolled cell proliferation, which represents the essence of neoplastic disease, involves corresponding metabolic adjustments in order to fuel cell growth and division.1 One such alteration is that tumor cells exhibit up-regulated amino acid metabolism so as to satisfy their increased rate of protein synthesis2 and/or to use them as precursors of biogenic amines or as energy source3. Therefore, it is no surprise that amino acids labelled with PET radionuclides have been used successfully for clinical oncologic imaging of brain tumors and some peripheral cancers like prostate, lung, head and neck cancer as well as neuroendocrine tumors.4 Transport of amino acids across cell membranes is mediated by membrane-associated carrier proteins. In order to keep up with their high need for amino acids, cancer cells overexpress certain amino acid transporters.5 A number of studies have implicated increased system L expression in tumor cells. The Na+ independent amino acid transport system L (LAT) is the major route for transporting large branched and aromatic amino acids. There are four subtypes of LAT identified, designated LAT1 to LAT4. LAT1 has been extensively documented to be up-regulated in a wide range of tumor cell lines, primary human cancers and metastases6,

7

while LAT2 to LAT4 show a more restricted

ACS Paragon Plus Environment

3

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 63

expression pattern in cancer cells. Additionally, upregulation of LAT1 correlates with cell proliferation and angiogenesis and it is also involved in cell signaling through the mTOR pathway that regulates cell growth and division.8 The heavy involvement of the system L transporters in cancer fueled, as expected, the development of many amino acid based tracers aiming to image LAT transport and in particular LAT1 activity. The uptake of several clinically useful

PET

tracers

such

as

L-[11C-methyl]methionine

([11C]MET),9

O-(2-

[18F]fluoroethyl)tyrosine [18F]FET,10 L-3,4-dihydroxy-6-[18F]fluorophenylalanine ([18F]FDOPA) and 5-hydroxy-L-[β-11C]tryptophan ([11C]5HTP)11 has been shown to proceed, at least to some extent, via system L transport. Regarding brain cancer imaging, tracer substrates of LAT1 have the additional advantage that the transporter is abundant at the blood-brain barrier, allowing tracer equilibration across the blood-brain barrier. [11C]5HTP and [18F]FDOPA are two amino acid tracers that are being used successfully for the imaging of neuroendocrine tumors (NETs).12,

13

NETs are a group of rare malignancies

whose small size and slow growth make their detection challenging. The success of these tracers derives not only from the fact that they are efficiently taken up by LAT1. NETs possess the unique property of synthesis, storage and secretion of biogenic amines by following the Amine Precursor Uptake and Decarboxylation (APUD) mechanism.14 The catecholaminergic/ serotonergic metabolic pathways, expressed in these tumors, contain many different enzymatic steps, from biosynthesis of the related neurotransmitters (through decarboxylation of the corresponding amino acids), to translocation into secretory vesicles or degradation by monoamine oxidases (MAO).13,

15

The efficient retention of [11C]5HTP and [18F]FDOPA in

NETs is considered to be the result of uptake (LAT1), decarboxylation (Aromatic Amino Acid Decarboxylase, AADC) and granular storage by vesicular monoamine transporters (VMAT). The

ACS Paragon Plus Environment

4

Page 5 of 63

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

latter process prevents enzymatic break down in the cytoplasm from MAO that leads to irreversible trapping of the tracers in the tumors.11 Despite the fact that both tracers are being used for the detection of NETs, it has been sporadically reported that [11C]5HTP has superior biological properties.13, 16 Nevertheless, the difficult radiosynthesis of [11C]5HTP which involves a complicated multi-step reaction including a chemoenzymatic step and the short physical halflife of [11C]carbon (20.3 min) are major obstacles that hamper its more widespread clinical use.17 As a result, [11C]5HTP is available only to a few facilities with an on-site cyclotron.

Figure 1. Chemical structures of [11C]5HTP and [18F]FDOPA, two clinically utilized tracers for endocrine tumor imaging and 18F-labelled tryptophan analogues. The numbering of the atoms of the tryptophan core and side chain is denoted for 1H and 13C NMR chemical shift assignments. Aiming at finding an 18F-labelled analogue of [11C]5HTP which ideally would exhibit equally good in vivo properties and at the same time allow a facile radiosynthetic procedure, our

ACS Paragon Plus Environment

5

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

laboratory has been focusing on the development of various

Page 6 of 63

18

F-labelled tryptophan-based

radiotracers. Ideally, the new tracer should be efficiently taken up by LAT1 and be a good substrate for AADC. We previously reported the radiosynthesis and biological evaluation of 2[18F]fluoroethoxy-DL-tryptophan analogues that have a 18F-bearing side chain at either position 4-, 5-, 6- or 7- of the indole core.18,

19

We have also synthesized and evaluated tryptophan

derivatives bearing a 3-[18F]fluoropropyl side chain at positions 2- and 5- of the tryptophan ring20 (Figure 1). All the tracers showed efficient uptake via the LAT1/2 system but none was further metabolized in AADC-expressing xenografts. Despite the fact that AADC is non-specific and able to recognize and decarboxylate a broad spectrum of amino acids including tryptophan,21 our findings suggest that the previous tryptophan tracers are not substrates for AADC as only the parent molecule was detected 1h post-injection. It seems that AADC has strict demands with regard to steric or electronic properties of the aromatic substrates. For example, methylation of the 5-OH position of 5HTP and methylation of the 3-OH group of L-3,4-dihydroxyphenylalanine leads to derivatives that completely abolish the capacity to be decarboxylated by AADC.21,

22

Moreover, while tryptophan is a fairly good substrate for AADC, 5-methyl and 6-methyl tryptophan show no activity.21 Our results confirm these observations. Also, 5HTP is five-fold a better substrate for decarboxylation than is tryptophan21 which could explain why [11C]tryptophan showed low uptake in carcinoid liver metastases and was outperformed by [11C]5HTP.16 In this respect, focusing on the development of

18

F-labelled derivatives of 5HTP

rather than tryptophan seems a more attractive option. The aforementioned data indicate that in order for a tryptophan-based structure to have optimal properties for AADC recognition, it should keep the following prerequisites: a) have an unmodified 5-OH group on the indole core and b) avoid steric hindrance in the near vicinity of

ACS Paragon Plus Environment

6

Page 7 of 63

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

the 5-OH position. In this work, we describe the synthesis, radiolabelling and biological evaluation of two new 5HTP analogues in which the 5-OH position remains intact while the 18Flabel is attached via an alkyl side chain of varying lengths (n=2, 3) at 2-position of the indole core, which is the most remote site from the 5OH group. These derivatives are namely 5hydroxy-2-(3-[18F]fluoropropyl)-DL-tryptophan (5OH-2-[18F]FPTRP, [18F]10) and 5-hydroxy-2(2-[18F]fluoroethyl)-DL-tryptophan (5OH-2-[18F]FETRP, [18F]22). A shorter side chain would mean that the new tracer is structurally more close to 5HTP and thus have better chances of exhibiting similar biological properties. Efforts to prepare a more structurally compound close to 5HTP i. e. 5-hydroxy-2-([18F]fluoromethyl)-DL-tryptophan (5OH-2-[18F]FMTRP, n=1) were not successful due to unexpected synthetic complications which are briefly discussed. For the in vitro and in vivo evaluations, we chose the following cell lines: small cell lung cancer NCI-H69, prostate cancer PC-3 and glioma brain cancer C6. The NCI-H69 cell line displays high AADC activity23 while the PC-3 and C6 cell line express low levels of AADC at the messenger RNA level.24, 25 LAT1 levels are increased in all these cell lines.26-28

ACS Paragon Plus Environment

7

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 63

RESULTS AND DISCUSSION

Chemistry

5OH-2FPTRP was synthesized from commercially available (R,S)-5HTP 1 as depicted in Scheme 1. We used racemic (R,S)-5HTP as we wanted to avoid synthetic complications and find out as early as possible whether the new analogues exhibit any remarkable biological action. As in

the

case

of

our

previously

synthesized

2-fluoropropyl-tryptophan

(2-FPTRP),20

functionalization at the 2-position of the indole core was accomplished by applying Danishefsky’s procedure using allyltributyltin as the nucleophile.29 This would install the allyl functionality at the desired position which is necessary to construct the 3-hydroxypropyl side chain via hydroboration-oxidation. Based on our good experience using a phthalyl protecting group for this sensitive alkylation step, we decided to proceed in the same manner with the new analogues; install first a phthalyl group which would later be substituted with a Boc group so as to facilitate the radiolabelling procedure.

ACS Paragon Plus Environment

8

Page 9 of 63

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

Scheme 1. Synthesis of 5OH-2FPTRP (10) Reaction of 5HTP with phthalic anhydride afforded 2 in 98% yield. The phenolic alcohol and the carboxyl group were then protected simultaneously as tert-butyl ether and tert-butyl ester, respectively by reacting 2 with tert-butyl 2,2,2-trichloroacetimidate.30, 31 The procedure afforded the desired bis-protected compound 3a in 45% yield while the mono-protected compound 3b (with the phenol group free) was isolated in 25% yield. Compound 3a was then subjected to 2allylation following Danishefsky’s methodology which adequately afforded 4. The phthalyl group was then cleaved using methylhydrazine20 and the crude amine was protected with a Boc group by treatment with Boc anhydride to afford intermediate 5 (84%, over two steps). To ensure a smooth hydroboration, the pyrrolic nitrogen was additionally Boc protected providing the fully

ACS Paragon Plus Environment

9

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 63

tert-butyl protected intermediate 6. Hydroboration with 9-BBN and subsequent oxidation with H2O2 under basic conditions afforded alcohol 7 which was then mesylated to furnish precursor 8 in quantitative yield. Subsequent fluorination with CsF in t-BuOH afforded 9 in excellent yield (93%) and final cleavage of all the tert-butyl protecting groups with TFA in the presence of scavengers (to avoid alkylation on the electron rich tryptophan core) gave reference compound 5OH-2FPTRP (10) in 64% yield after preparative HPLC purification. The synthesis of the 5OH-2FETRP derivative is shown in Scheme 2. A literature procedure had reported on the installation of an indolic 2-malonyl substituent followed by decarboxylation to provide the corresponding desired ester.32 Selective reduction of the methyl ester over the tbutyl ester would then provide the desired 2-hydroxyethyl side chain functionality for further modifications.

ACS Paragon Plus Environment

10

Page 11 of 63

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

Scheme 2. Synthesis of 5OH-2FETRP For this approach, the mono-tert butyl protected compound 3b, isolated earlier as a byproduct in the synthesis of 10, was first benzylated on the phenolic alcohol to provide 11 in excellent yield. The phthalyl group was then substituted with di-Bn in a two-step procedure affording 12 in 66% overall yield. Treatment of 12 with tert-butyl hypochlorite (t-BuOCl) followed by the addition of freshly prepared lithium dimethylmalonate and ZnCl2 as acidic promoter afforded intermediate 13 in excellent yield (96%). Subsequent Krapcho decarboxylation efficiently provided methyl ester 14. Before cleaving the Bn-protecting groups and fully protecting with Boc groups we decided at this point to reduce the methylester to the corresponding alcohol to avoid possible interference of the bulky neighboring Boc protecting groups. Reaction of 14 with DIBAL-H afforded the corresponding alcohol 15 (91%) and the hydroxyl functionality was subsequently protected as an acetate group yielding intermediate 16. All benzyl groups were then cleaved via hydrogenation and subsequently the amine nitrogen was Boc-protected to provide 17 in 73% yield over two steps. The pyrrole and phenol moieties were finally protected with Boc groups affording the fully tert-butyl protected compound 18. The following cleavage of the acetate group proved more challenging than expected. All standard methods tested including transesterification in methanol using catalytic amounts of KCN33, hydrolysis with the use of K2CO3/MeOH or LiOH in MeOH/THF/water readily promoted the migration of the Boc-group from the pyrrolic nitrogen to the OH group of the hydroxyethyl side chain. However, reduction of the acetate using DIBAL-H proceeded smoothly yielding the desired compound 19. The crude product was not isolated but used directly in the succeeding mesylation reaction. A small amount of the crude product was purified by flash column chromatography to afford pure alcohol 19 which was used for collecting analytical data.

ACS Paragon Plus Environment

11

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 63

It is worth mentioning that, 19 is stable under storage at -25 °C and did not spontaneously react to give the migration adduct. Mesylation of crude 19 afforded precursor 20 in 68% yield over 2 steps. Subsequent fluorination using the same conditions as in the synthesis of 10 (CsF, t-BuOH) provided fluorinated compound 21 albeit in 34 % yield. This rather modest yield might be attributed to increased steric hindrance of the neighboring Boc groups due to the shorter length of the alkyl chain. Finally, cleavage of Boc groups with TFA/scavengers afforded reference compound 22 (5OH-2FETRP) in 60% yield after semipreparative HPLC purification. For the synthesis of the 5OH-2FMTRP (Scheme 3), we envisioned that the desired hydroxymethyl functionality could derive from the reduction of 2-formyl-5-hydroxytryptophan, obtained by the SeO2 oxidation of a suitably protected 1,2,3,4-tetrahydro-b-carboline-3carboxylic acid as previously reported.34, 35 We additionally aimed at fully protecting the amino nitrogen with two electron withdrawing Boc groups to prevent potential intramolecular attack on the electrophilic carbon of the precursor, to form back the rigid 6-membered β-carboline ring. Despite all precautions, mesylation of 31 afforded quantitatively a very polar product, which after isolation was found to be the quaternary ammonium salt 32. Given the synthetic challenges we encountered and also the fact that 5OH-2FMTRP could potentially undergo in vivo defluorination as reported for similar aromatic fluoromethyl substrates,36, 37 we decided at this point not to pursue its synthesis any further.

ACS Paragon Plus Environment

12

Page 13 of 63

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

Scheme 3. Synthetic attempt towards 5OH-2FMTRP

ACS Paragon Plus Environment

13

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 63

Radiochemistry

Both [18F]10 and [18F]22 were prepared in analogy to a procedure previously reported in our group.20 The method involves a two-step reaction sequence consisting of nucleophilic fluorination of the corresponding mesylate precursors 8 and 20 followed by cleavage of the protecting groups (Scheme 4). Nucleophilic substitution was achieved by heating the corresponding mesylate precursors with [18F]TBAF in acetonitrile at 100 °C for 10 min. Deprotection was accomplished in the same reaction vial under acidic conditions using 4M HCl . For [18F]10 a temperature of 110°C was necessary to efficiently cleave the more resilient t-butyl ether protecting group whereas for [18F]22, a temperature of 100 °C was sufficient. Higher temperatures favored the formation of unknown impurities.

Scheme 4. General radiosynthetic scheme towards 5OH-2-[18F]ALKTRPs Purification of both target compounds was accomplished by semi-preparative HPLC. The product peaks were collected and neutralized to pH ≈ 6 using either 10% sodium bicarbonate or via an ion exchange resin for sodium free formulations. Sodium ascorbate or ascorbic acid was added to prevent radiolysis, which was observed for almost all of our previously synthesized tryptophan tracers. Chemical and radiochemical purities of both HPLC purified tracers were examined with analytical HPLC and were always found to be greater than 95%. The products were stable for up to 4 hours in their formulation. The average decay corrected radiochemical

ACS Paragon Plus Environment

14

Page 15 of 63

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

yields for [18F]10 and [18F]22 were 39% (n=12) and 15% (n=6) respectively. The lower yield obtained for [18F]22 originates from the low [18F] incorporation during the first fluorination step. This result is not surprising given that similar observations were made during the corresponding non-radioactive fluorination step. The specific activities for both tracers were in the range of 45 95 GBq/µmol at the end of synthesis and radiochemical purity was always ≥95% for both tracers. The average synthesis time was approximately 70 min from end of bombardment (EOB).

In vitro cell uptake and enzyme studies Figure 2 shows the uptake of the two tracers by NCI-H69 tumor cells under various conditions. In the absence of an inhibitor, both tracers showed an initial high temperature-independent uptake within the first 5 minutes reaching more than 50% of the maximal uptake. Up to 60 min, radioactivity uptake increased further at 37°C but not at 4°C. Sodium had no substantial influence on the uptake of [18F]10, excluding major involvement of a Na+-dependent transporter. The LAT1/2 and ATB0,+ inhibitor 2-amino-2-norbornanecarboxylic acid (BCH) completely inhibited uptake of both tracers at 37 and 4 °C (Figure 2A). Arginine (Arg), a substrate for cationic amino acid transporters (CAT), b0,+ and ATB0,+ inhibited the temperature-dependent fraction of [18F]10 uptake as concluded from the superposition of the uptake at 4°C in the absence of Arg and uptake at 37°C in the presence of Arg (Figure 2B). As expected, addition of histidine (His) and lysine (Lys) to Arg completely inhibited uptake since His is a substrate of LAT1/2. Based on these results, we hypothesize that [18F]10 is taken up by LAT1/2 and possibly b0,+ and/or ATB0,+. Both LAT1 and ATB0,+ are upregulated in various cancer types.38 [18F]22 showed a similar uptake pattern with respect to temperature-dependency and inhibition by BCH compared to its analogue with the longer alkyl chain (Figure 2C).

ACS Paragon Plus Environment

15

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 63

Figure 2. Time-dependent uptake of [18F]10 (A, B) and [18F]22 (C) in NCI-H69 cells at 37 °C and

4

°C.

(A)

Influence

of

sodium

(Earle

balanced

salt

solution

(EBSS)

vs

HEPES/TRIS/cholineCl) and 10 mM BCH, respectively, as indicated. (B) Influence of 10 mM Arg, a mix of 10 mM Arg, His, Lys or 10 mM BCH in sodium-containing buffer as indicated. (C) [18F]22 uptake with and without BCH in EBSS buffer. For experiments with n = 2 error bars indicate individual values, for n > 2 error bars indicate standard deviations. In A and B, results from incubations with BCH at 4°C are not shown for clarity. Uptake was < 1.2% of added radioactivity. In order to investigate whether the target compounds are substrates for AADC, both 5HTP analogues 10 and 22 were incubated with recombinant human AADC. The enzyme preparation generated free amines in the absence of an added substrate, presumably by decarboxylation of available amino acids in the enzyme preparation. Incubation with 0.5 mM of the AADC substrate L-DOPA increased the free amine concentration by at least 30% while both 5HTP derivatives reduced the basal activity to less than 25% (10, 2 independent experiments; 22, 1 experiment;

ACS Paragon Plus Environment

16

Page 17 of 63

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

Figure 3). The generation of free amines from L-DOPA was reduced in samples containing both L-DOPA and 10 or 22 at equal concentrations of 0.5 mM (1 experiment each). The reduction was about 20% for 10 and 50% for 22 (Figure 3). These results indicate that both 10 and 22 have an inhibitory effect on the enzyme, with Ki values in the range or higher than KM of LDOPA (since decarboxylation was reduced by 50% or less at equimolar concentration of LDOPA and test compound). It is worth mentioning that previously synthesized tryptophan derivatives 2-FPTRP20 and 6-fluoroethoxy-tryptophan (6-FEHTP)18 did not show inhibitory effect on AADC when used in the same assay and they were not substrates for decarboxylation either (data not shown). The development of tracers which are (irreversible) inhibitors of AADC was suggested as an alternative strategy to the APUD concept for NET imaging with PET. The respective tracers afluoromethyl-6-[18F]fluoro-m-tyrosine (FM-6FmT)39 and 6[18F]fluoro-m-tyrosine (FMT)40 have been reported in the literature as inhibitors of AADC although a detailed biological evaluation has not yet been published. Our in vivo PET studies (see below) excluded AADC interaction as a mechanism of tumor uptake. We therefore did not further investigate the details of AADC inhibition, such as concentration dependence or stereoselectivity. According to early investigations on the enzyme activity,41 preference for one or the other enantiomer of amino acid substrates and inhibitors can be expected.

ACS Paragon Plus Environment

17

Journal of Medicinal Chemistry

1.8 Test compound 1.6 Test compound + LDOPA 1.4

Fold baseline activity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 63

1.2 1.0 0.8 0.6 0.4 0.2 0.0 Baseline

5OH-2FPTRP

5OH-2FETRP

Figure 3. Inhibition of AADC by 5OH-2FPTRP and 5OH-2FETRP. The enzyme preparation and cofactor were incubated together with vehicle (water, baseline) or one of the amino acids at 0.5 mM concentration, with (grey) or without (black) addition of 0.5 mM L-DOPA. Shown is the amount of amine production normalized to the baseline experiment without L-DOPA. A value < 1 indicates inhibition of the baseline activity of the enzyme. Test compound, baseline (water), 10 or 22, as indicated. n = 2 (baseline with and without L-DOPA and 10 without L-DOPA) or 1 (other conditions). Error bars indicate the individual values for conditions with n=2. Indoleamine 2,3-dioxygenase (IDO) has been identified as a novel target in cancer therapy since tryptophan depletion caused by IDO expressing tumors facilitates their escape from immune surveillance by inhibiting effector T cells.42 In this respect, probes imaging IDO activity would be of great interest for PET. We investigated thus whether the new tracers are substrates of IDO. While the positive control TRP was oxygenated at a specific activity of > 0.5 nmol/min per µg protein (single experiment), none of the amino acids 10, 22, 2-FPTRP, 6-FEHTP or LDOPA was oxygenated under the same experimental conditions (data not shown).

Small animal PET with xenograft-bearing mice

ACS Paragon Plus Environment

18

Page 19 of 63

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

Dynamic PET scans with either NCI-H69, PC3 or C6 xenografted-mice showed marked accumulation of [18F]10, [18F]22, and [18F]FET in the xenografts up to 150 min post injection. The uptake ratios between xenograft and reference region (muscle) for all radiotracers were highest between 60 and 75 min post injection and Figure 4 shows representative PET images superimposed on CT. For a head-to-head comparison, each xenograft-bearing animal was scanned consecutively with [18F]10, [18F]22 and [18F]FET on different days. As estimated from the PET images in Figure 4, radiotracer accumulation in all three xenografts were in the similar range for [18F]10 and [18F]22. A higher xenograft-to-reference ratio was found for both 5HTP analogues compared to [18F]FET.

ACS Paragon Plus Environment

19

Journal of Medicinal Chemistry

Page 20 of 63

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment

20

Page 21 of 63

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

Figure 4. PET/CT images of NCI-H69, PC3 and C6 xenograft-bearing mice after intravenous injection of 10-21 MBq [18F]10, [18F]22 or [18F]FET. PET images are averaged from 60 to 75 min post tracer injection, when xenograft –to-reference ratios were highest. Cross hairs indicate xenografts. Note: The same animal was scanned with all three radiotracers. SUV scales are adjusted to provide similar color (blue) for background tissue in the images, allowing a direct comparison of the uptake ratio between xenografts and healthy tissues. Maximal (max) SUV for NCI-H69 xenografts are 1 ([18F]10 and [18F]22), 3.2 ([18F]FET); for PC3 xenografts 0.7 ([18F]10), 0.9 ([18F]22), 3,4 ([18F]FET); for C6 xenograft 0.5 ([18F]10), 0.6 ([18F]22) and 3.6 ([18F]FET). MIP: maximal intensity projection. Uptake analysis revealed similar standardized uptake values (SUVs) for [18F]10 and [18F]22 and the values were ~0.6 for NCI-H69 xenografts, ~0.5 for PC3 and ~0.4 for C6 (Figure 5A). For comparison, [18F]FDOPA reached average SUVs of 1.6 and 1.4 in NCI-H69 and PC3 xenografts, respectively, when AADC was inhibited with the AADC inhibitor S-carbidopa. SUV in PC3 xenografts was 0.6 without S-carbidopa pre-administration.19 [11C]5HTP in BON xenografts showed SUV values of approximately 0.2.11 In the current study, [18F]FET showed the highest uptake values of ~1.6, ~1.4, and ~1.6 for NCI-H69, PC3, and C6 xenografts, respectively. The new tryptophan analogues however, showed significantly lower background (muscle) radioactivity than [18F]FET. This resulted in NCI-H69-to-reference ratios of 4.3±0.9 for [18F]10, 3.9±0.2 for [18F]22 and 1.8±0.1 for [18F]FET (Figure 5B). Similar SUV ratios were obtained for PC3-xenografts and values were 4.2±0.6 for [18F]10, 4.2±0.4 for [18F]22 and 1.8±0.2 for [18F]FET. In general, SUV ratios were significantly higher for both tryptophan analogues than [18F]FET. C6 xenografts-to-reference SUV ratios were lower than for the other two xenografts and amounted to 2.7±0.3 for [18F]10 and 3.2±1.1 for [18F]22. SUV ratio for

ACS Paragon Plus Environment

21

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 63

[18F]FET in C6 xenografts was 1.9±0.2 and similar to the other [18F]FET uptake ratios. [18F]10 and [18F]22 reached higher xenograft/reference ratios than all our previous tryptophan derivatives, of which 6-[18F]FEHTP18 (Figure 1) performed best with a SUV ratio of 2.6±0.2 in NCI-H69 xenograft-bearing mice. The AADC substrate [18F]FDOPA reached average SUV ratios of ≤ 2 in NCI-H69 (with S-carbidopa) and PC3 (with or without S-carbidopa) xenograftbearing mice in our previous study.19 The brain/muscle, kidney/muscle and bone/muscle ratios of [18F]10, [18F]22 and [18F]FET were determined in the 60-75 min scans of the NCI-H69 xenograft-bearing mice. Brain/muscle ratios were 0.96±0.09 and 0.94±0.07 for [18F]10 and [18F]22, respectively, and 0.71±0.06 for [18F]FET (significantly lower). The ratios close to unity would be in agreement with equilibration across the blood-brain barrier by LAT143, 44 and are a promising feature towards the development of the tracers for the imaging of brain cancer. Kidney/muscle ratios were significantly higher for [18F]10 and [18F]22 (7.2±1.4 and 12.1±2.2) than for [18F]FET (1.4±0.2). The high radioactivity of [18F]10 and [18F]22 in the kidneys, in particular in the pelvis (not shown) indicates renal excretion in the absence of re-absorption, as shown for [18F]fluorodeoxysorbitol recently.45 The high radioactivity accumulation most probably results from re-absorption of water after filtration leading to an increase in tracer concentration in the urine. [18F]FET was discussed as a substrate of LAT2, which is the predominant isoform of the LAT transport systems in kidney, where it re-absorbs its substrates from the proximal tubules.46, 47 Re-absorption by LAT2 would explain the low accumulation of [18F]FET in kidney and the higher general radioactivity in all tissues after [18F]FET administration compared to [18F]10 and [18F]22.

ACS Paragon Plus Environment

22

Page 23 of 63

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

Bone/muscle ratios were significantly higher for [18F]10 and [18F]22 than [18F]FET. The ratios between shoulder joints and muscle were 1.4±0.13 and 2.0±0.5 for [18F]10 and [18F]22 and 1.0±0.1 for [18F]FET. This may indicate minor defluorination of the two tryptophan analogues. In this study we compared racemic mixtures of [18F]10 and [18F]22 with enantiomerically pure (S)-[18F]FET. Taking into account that the stereochemistry of radiolabeled amino acids can influence the rate and selectivity of amino acid transport,48 we cannot exclude the possibility of the new tracers exhibiting different transport kinetics and/or tumor uptake between the two isomers. The LAT system mainly transports L-amino acids but it lacks marked stereoselectivity since it can transport certain (R)-amino acids (e.g. (R)-phenylalanine, (R)-leucine and (R)methionine) but not (R)-tryptophan.6 It is possible that optically pure (S)-[18F]10 or (S)-[18F]22 could show even better biological performance in particular if the (R) isomer in the racemic mixtures is not efficiently recognized by the involved transporter(s). In order to investigate whether the high uptake of both tracers could be a consequence of their binding to AADC as inhibitors (or substrates), we performed PET experiments after pretreatment of the mice with S-carbidopa that would act as a competitor at the binding site. Results showed that S-carbidopa had no significant effect on tracer accumulation in tumor or healthy tissue (Figure 5C). This result, in combination with the results obtained from the in vitro AADC assay and the ex vivo radiometabolite studies, suggest that tumor uptake of the new 5HTP derivatives is most likely independent of any interaction with AADC or IDO.

ACS Paragon Plus Environment

23

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 63

Figure 5. Uptake of [18F]10, [18F]22 or [18F]FET in NCI-H69, PC3 or C6 xenografts and reference region (muscle) expressed as (A) SUVs and (B) SUV ratios of xenograft to reference region. (C) SUV of [18F]10, [18F]22 or [18F]FET in NCI-H69 xenograft and reference tissue under baseline conditions (vehicle) and AADC inhibition (25 mg/kg S-carbidopa). SUVs and SUV ratios represent PET data averaged between 60 to 75 min post tracer injections. For each group, n = 4. SUV of [18F]FET relative to SUV of the other two radiotracers: ** p