Evaluation of Novel N¹-Methyl-2-phenylindol-3-ylglyoxylamides as a New Chemotype of 18 kDa Translocator Protein-Selective Ligand Suitable for the Development of Positron Emission Tomography Radioligands

Abstract

A novel series of N¹-methyl-(2-phenylindol-3-yl)glyoxylamides 19–31, designed in accordance with our previously reported pharmacophore/topological model, showed high affinity for the 18 kDa translocator protein (TSPO) and paved the way for developing a new radiolabeled probe. Thus ligand 31, N,N-di-n-propyl-(N¹-methyl-2-(4'-nitrophenyl)indol-3-yl)glyoxylamide, featuring the best combination of affinity and lipophilicity, was labeled with carbon-11 for evaluation with positron emission tomography (PET) in monkey. After intravenous injection, [¹¹C]31 entered brain to give a high proportion of TSPO-specific binding. These findings augur well for the future application of [¹¹C]31 in humans. Consequently, the binding of 31 to human TSPO was tested on samples of brain membranes from deceased subjects who through ethically-approved in vitro study had previously been established to be high-affinity binders (HABs), mixed-affinity binders (MABs) or low-affinity binders (LABs) for the known TSPO ligand, PBR28 (2). 31 showed high affinity for HABs, MABs and LABs. In conclusion, [¹¹C]31 represents a promising new chemotype for developing novel TSPO radioligands as biomarkers of neuroinflammation.

Introduction

Translocator protein (TSPO), formerly known as the peripheral benzodiazepine receptor or PBR, is an 18 kDa protein, which is mainly located at the contact sites between the outer and inner mitochondrial membranes. This receptor is a component of a trimeric complex with the 32 kDa voltage-dependent anion channel (VDAC) and the 30 kDa adenine nucleotide translocase (ANT) to constitute the mitochondrial permeability transition pore (MPTP).^1,2

TSPO is involved in a variety of biological processes, such as cholesterol transport, steroidogenesis, calcium homeostasis, lipid metabolism, mitochondrial oxidation, cell growth and differentiation, apoptosis induction, and regulation of immune functions.^2,3

Consistent with its key cellular functions, TSPO is expressed in many peripheral tissues, including liver, heart, kidney, lung, and the immune system, with higher levels appearing in steroid producing tissues.^1,2 In the central nervous system, TSPO normally exists at low level and is mainly located in glial and ependymal cells. Recent evidence suggests that TSPO on glial cells may regulate the biosynthesis of neurosteroids, leading to the hypothesis of a potential role for TSPO in the treatment of neuropathological conditions.^4–7

TSPO expression is up-regulated in several human pathologies, including gliomas and neurodegenerative disorders (Huntington's and Alzheimer's diseases), as well as in various forms of brain injury and inflammation. Under neuroinflammatory conditions, TSPO markedly increases in activated microglia.⁸ Changes in TSPO level have been found in patients affected by generalized anxiety, panic, post-traumatic stress, obsessive-compulsive disorders, and separation anxiety.^4,9,10

Consequently, TSPO has been suggested as a promising target for a number of therapeutic applications,¹¹ and also as a diagnostic marker for related disease progression, so prompting the development of specific fluorescent^12-14 or radiolabeled^15-17 ligands, as powerful tools to image and measure the expression levels of this protein in living cells, isolated tissues or living subjects. In particular, the use of a selective radioligand with positron emission tomography (PET) allows measurement of the relative density of a target protein in animal and human subjects in vivo. PET studies of TSPO could offer quantitative measures of inflammation, so providing valuable biomarker-type measurements during drug development and early clinical trials.^17,18

[¹¹C]PK 11195¹⁹ ([¹¹C]1; Figure 1), first as racemate²⁰ and later as R-enantiomer²¹, was the first radioligand to be used extensively in the PET imaging of neuroinflammation. However, this radioligand has several limitations, including a high level of non-specific binding, and a poor signal-to-noise ratio which complicate its quantification.^18,22 Therefore, many groups have been searching for TSPO ligands with improved performance for quantifying TSPO expression.^15-17 Recently, more-sensitive radioligands have been introduced for the study of human subjects, including [¹¹C]PBR28 ([¹¹C]2, Figure 1).^22–25 Recent studies with these new radioligands reveal heterogeneity with regard to the form of expression of TSPO in human subjects. For example, in the case of the radioligand [¹¹C]2, human subjects may be classified as high-affinity binders (HABs), mixed-affinity binders (MABs) or low-affinity binders (LABs).²⁶ The existence of such different populations, and in particular the more difficult to identify MABs, compromises the accuracy of quantitative comparisons between test and control cohorts. Consequently, there remains a need to find TSPO radioligands that bind with an equally high affinity in all human subjects. One strategy for meeting this need is to explore new chemotypes of TSPO ligand for PET radioligand development.

Figure 1.

Structures of some known TSPO ligands.

In recent years, N,N-dialkyl-(2-phenylindol-3-yl)glyoxylamides (I), Figure 2, have been described as a series of potent and selective TSPO ligands, with K_i values in the nanomolar or subnanomolar range; this series represents conformationally constrained analogues of the indoleacetamide derivative, FGIN-1 (5) (Figure 1).^27,28

Figure 2.

Ligand 1 and N,N-dialkyl-2-phenylindolylglyoxylamides I and II in the pharmacophore/receptor model of TSPO.²⁷,²⁸

Within this class, SAR findings were rationalized in the light of the most recently reported pharmacophore/receptor model made up of three lipophilic pockets (L₁, L₃, and L₄) and an H-bond donor group (Figure 2).^27,28 Specifically, the second carbonyl group of the oxalyl bridge engages an H-bond with the donor site H₁; the two lipophilic substituents on the amide nitrogen, R₁, and R₂ (linear or ramified alkyl, arylalkyl groups) interact hydrophobically with the L₃ or L₄ lipophilic pockets; the 2-phenyl moiety establishes a putative π-stacking interaction within the L₁ pocket (Figure 2).

The high affinities of these 2-phenylindolglyoxylamide derivatives I have recently permitted the development of new fluorescent probes useful for investigating the localization and the expression level of TSPO.^13,14

Pursuing our interest in this field, we evaluated the opportunity to develop a novel PET radioligand for TSPO from this chemotype. SAR data and our previously reported pharmacophore/receptor model hypothesis suggested that the indole NH should not be involved in the recognition of the receptor protein (Figure 2). In this view, we developed a small library of novel 2-phenylindol-3-ylglyoxylamide derivatives of general formula II (compounds 19–31), featuring a methyl group on the indole nitrogen (Figure 2), as predicted high-affinity TSPO ligands. These compounds, due to the presence of this methyl group, are amenable to labeling with the positron-emitter carbon-11 (t_1/2 = 20.4 min). Thus, [¹¹C]31 was synthesized and studied for its behaviour in monkey in vivo and for the characteristics of its binding to brain membranes collected from a cohort of human subjects.

Results and Discussion

Chemistry

The general procedure for the synthesis of the target N,N-dialkyl-1-methyl-2-phenylindol-3-ylglyoxylamides 19–31 is outlined in Scheme 1, and involved the treatment of previously described Ia–k ^27,28 or newly synthesized indoles (17 and 18) with sodium hydride and subsequent addition of an excess of methyl iodide, in dry DMF. All products were purified by flash chromatography (CHCl₃ as eluent). Their chemical-physical properties were determined (Table 1) and their structures were confirmed by IR, ¹H-NMR, and elemental analysis (Supporting Information). The new compounds, 17 and 18, were prepared essentially following procedures similar to those used to obtain products Ia–k,^27,28 as summarized in Scheme 1.

Scheme 1.

Synthesis of new N¹-methyl TSPO ligands 19-31.

Table 1.

Physical Properties of N,N-dialkylindolylglyoxylamide Derivatives 17-31.

17-31
Compd.	R1	R2	R3	R4	R5	Yield (%)	mp (°C)	formulaα
17	Me	Bn	H	H	H	72	98-100	C24H20N2O2
18	Me	Bn	F	H	H	89	72-74	C24H19FN2O2
19	nPr	nPr	H	Me	H	90	110-112	C23H26N2O2
20	nBu	nBu	H	Me	H	79	83-85	C25H30N2O2
21	nHex	nHex	H	Me	H	90	oil	C29H38N2O2
22	Et	nBu	H	Me	H	76	oil	C23H26N2O2
23	Et	Bn	H	Me	H	70	oil	C26H24N2O2
24	Me	nBu	H	Me	H	64	124-126	C22H24N2O2
25	Me	nPen	H	Me	H	70	126-128	C23H26N2O2
26	Me	Bn	H	Me	H	92	148-150	C25H22N2O2
27	nPr	nPr	H	Me	F	73	120-122	C23H25FN2O2
28	nPr	nPr	F	Me	H	83	132-134	C23H25FN2O2
29	nHex	nHex	F	Me	H	65	oil	C29H37FN2O2
30	Me	Bn	F	Me	H	70	138-140	C25H21FN2O2
31	nPr	nPr	NO2	Me	H	97	144-146	C23H25N3O4

^αElemental analyses for C, H, N were within ± 0.4% of the calculated values.

Binding Affinity on Rat Kidney

The binding affinity at TSPO of all the newly synthesized indolylglyoxylamides 17–31 was determined in rat kidney membranes by binding competition experiments against [³H]1 as radioligand.²⁷ Because of the well-established TSPO versus central benzodiazepine receptor (BzR) selectivity of N,N-dialkyl-2-phenylindol-3-ylglyoxylamides I,^27,28 a few randomly selected novel indole derivatives were evaluated for their BzR affinity, using membranes from rat brain tissues and [³H]flumazenil as the radioligand. The tested compounds showed no significant binding properties in this assay (data not shown).

The binding affinities of all the newly synthesized N¹-methylated derivatives 19–31, expressed as K_i values, are listed in Table 2, together with those of the standard TSPO ligands 1, 3, and 4 (Figure 1). The binding data of some of the parent compounds, Ia–k,^27,28 and of 17 and 18 are also included for comparison.

Table 2.

TSPO binding affinity of N,N-dialkylindolylglyoxylamide derivatives I a–k, and 17–31.

Ia–k, 17–31
Ligand	R1	R2	R3	R4	R5	Ki (nM)a
1						9.30 ± 0.50
3						23.0 ± 3.0
4						0.5–7
Ia b	nPr	nPr	H	H	H	12.2 ± 1.0
Ib b	nBu	nBu	H	H	H	7.50 ± 0.70
Ic b	nHex	nHex	H	H	H	1.40 ± 0.20
Id c	Et	nBu	H	H	H	12.6 ± 1.0
Ie c	Et	Bn	H	H	H	11.0 ± 1.0
If c	Me	nBu	H	H	H	53.3 ± 4.0
Ig c	Me	nPen	H	H	H	12.1 ± 1.0
17	Me	Bn	H	H	H	12.0 ± 1.0
Ih c	nPr	nPr	H	H	F	2.67 ± 0.48
Ii b	nPr	nPr	F	H	H	4.28 ± 0.32
Ij b	nHex	nHex	F	H	H	0.370 ± 0.13
18	Me	Bn	F	H	H	1.80 ± 0.10
Ik c	nPr	nPr	NO2	H	H	0.950 ± 0.10
19	nPr	nPr	H	Me	H	19.5 ± 1.5
20	nBu	nBu	H	Me	H	0.300 ± 0.050
21	nHex	nHex	H	Me	H	0.299 ± 0.050
22	Et	nBu	H	Me	H	60.2 ± 6.0
23	Et	Bn	H	Me	H	2.92 ± 0.30
24	Me	nBu	H	Me	H	190 ± 10
25	Me	nPen	H	Me	H	38.2 ± 4.0
26	Me	Bn	H	Me	H	20.9 ± 1.8
27	nPr	nPr	H	Me	F	23.7 ± 2.1
28	nPr	nPr	F	Me	H	19.8 ±2.0
29	nHex	nHex	F	Me	H	1.07±0.11
30	Me	Bn	F	Me	H	6.77 ± 0.50
31	nPr	nPr	NO2	Me	H	5.70 ± 0.45

^aThe concentration of tested compounds that inhibited [³H]1 binding to rat kidney mitochondrial membranes (IC₅₀) by 50% was determined with six concentration of the displacers, each performed in triplicate. K_i values are the mean ± SEM of three determination.

^bData taken from ref. 27.

^cData taken from ref. 28.

All the newly synthesized compounds were found to be potent TSPO ligands, with K_i values in the low nanomolar or subnanomolar range.

The insertion of the methyl group at the N¹ position did not significantly influence TSPO binding, as most of the derivatives 19–31 showed K_i values ranging from 0.3 to 60 nM, essentially comparable with those of their unsubstituted counterparts Ia-k, 17 and 18. The only exception was compound 24 which showed a low affinity (K_i 190 nM). Compounds 20 and 21 possessed the highest affinity in this series (K_i 0.30 nM) and a 25- and 4.6-gain in activity with respect to their desmethyl counterparts, Ib and Ic, respectively. These results agree with our previously reported pharmacophore/receptor model hypothesis, where the indole NH is not involved in the recognition of the receptor protein.

Radioligand Synthesis

The high affinity of the newly synthesized N¹-methylated-2-phenylindol-3-ylglyoxylamide derivatives 19–31 made them good candidates for the development of radioligands for the imaging of TSPO. The choice of the best candidate to be labeled with carbon-11 was guided by two main factors, namely adequate high affinity towards TSPO and suitably moderate lipophilicity for achieving adequate brain entry and low non-specific binding.^29–32 Based on these two parameters, we selected 31 as the ligand from the new series showing the best combination of affinity (K_i 5.7 nM) and lipophilicity (cLogP 3.95) to be labeled with carbon-11 and evaluated as a TSPO radioligand in vivo.

[¹¹C]31 was rapidly prepared by treating the N-desmethyl analog 1k with [¹¹C]methyl iodide and sodium hydride in DMF, and purified by HPLC preceding formulation for intravenous injection. [¹¹C]31 was obtained in high radiochemical purity (> 99%), chemical purity and specific radioactivity (typically > 4 Ci/μmol at end of synthesis) and was radiochemically stable. Adequate activities (> 25 mCi) were readily obtained for PET imaging in monkey.

PET Imaging of [¹¹C]31 in Monkeys

The ability of [¹¹C]31 to image brain TSPO in vivo was tested with PET in rhesus monkeys. After injection of no-carrier-added (NCA) [¹¹C]31 into monkeys, radioactivity entered the brain well with peak radioactivity uptake occurring in all TSPO-containing regions by about 40 min after injection (Figure 3A). Maximal uptake occurred in putamen (2.32 ± 0.72, SUV; n = 7) between 12 and 32 min after injection. This moderately high uptake is comparable to that achieved with other prominent TSPO radioligands, such as [¹¹C]2²³ and [¹⁸F]6³³ (Figure 1). Subsequent washout of radioactivity from all TSPO-containing regions was quite slow. In repeating the experiments in three monkeys in which TSPO receptors were preblocked by administration of 1 (5 mg/kg, i.v.), the kinetics of brain radioactivity was strikingly different. Radioactivity was rapidly taken into all examined brain regions to a higher level than in the baseline experiment, and then washed out rapidly to a low common level (Figure 3B). In a displacement experiment in another monkey, administration of 1 (5 mg/kg, i.v.) at 48 min after radioligand caused a rapid decline in radioactivity in all TSPO-containing regions to a low common level (Figure 3C). These data show the presence of a high proportion of specific binding of radioligand to brain TSPO in the baseline experiments and furthermore demonstrate the reversibility of this binding.

Figure 3.

Brain region time–activity curves in monkey after injection of [¹¹C]31 in A) a baseline (radioligand alone) experiment, B) a TSPO pre-block experiment in which the TSPO ligand 1 (5 mg/kg, i.v.) was given at 10 min before radioligand and C) a displacement experiment in which 1 (5 mg/kg, i.v.) was given at 48 min after radioligand. Key: Anterior cingulate (△), cerebellum (▽), hippocampus (◇), pre-frontal cortex (□), putamen (▲), and thalamus (×).

Average PET images of monkey brain acquired over 4–100 min after injection of [¹¹C]31 (Figure 4) displayed a high level of radioactivity, especially in putamen, cerebellum, and intermediate levels in cortical regions. Scans from the corresponding pre-block experiments showed a very uniform low distribution of radioactivity, consistent with absence of radioligand specific binding (Figure 4).

Figure 4.

Horizontal PET images of monkey brain at the level of the caudate acquired between 4 and 100 min after intravenous injection of monkey with [¹¹C]31under baseline and pre-block conditions, and corresponding MRI scan.

Emergence of Radiometabolites of [¹¹C]31 in Monkey Plasma In Vivo

Generally, PET radioligands are appreciably metabolized over the short time-span of a PET scanning session. Detailed quantitative analysis of radioligand behavior in vivo to derive important output measures such as binding potential or volume of distribution may require measurement of the free fraction of radioligand in plasma (f_P) and also a determination of the amount of unmetabolized radioligand in plasma over the full duration of a scanning session. Generally, PET radioligands which produce brain-penetrant radiometabolites are often difficult to quantify.^30,31 The plasma free fraction of [¹¹C]31 in monkey blood in vitro was low (0.71 ± 0.13%; n = 3), in accordance with the moderately high lipophilicity of 31 (cLogP = 3.95). After administration of [¹¹C]31 into monkey, both the total radioactivity in plasma and the parent radioligand component reduced rapidly (Figure 5). Recoveries of radioactivity from plasma into supernatant acetonitrile for HPLC analysis exceeded 94%. HPLC analyses of plasma from monkeys studied at baselinerevealed [¹¹C]31 eluting at 5.18 ± 1.41 min, and three less lipophilic radiometabolites [¹¹C]A–[¹¹C]C eluting at 2.02 ± 0.60, 2.79 ± 0.80 and 3.71 ± 0.95 min (n = 96), respectively. Monkey plasma radioactivity became composed equally of radiometabolites and [¹¹C]31 at 38.8 and 80.8 min in two baseline experiments and at 7.8 and 17.5 min in TSPO-pre-block experiments in the same monkeys, respectively (Figure 5). A greater availability of TSPO-free radioligand for metabolism in the pre-block experiments may account for the more rapid appearance of radiometabolites in the preblock experiments. The identities of the three radiometabolites are currently unknown, as is their ability to penetrate the blood-brain barrier. Their lower lipophilicity relative to the radioligand may however render them less brain-penetrant and relatively untroublesome to quantitative analysis.

Figure 5.

Emergence of three radiometabolites, [¹¹C]A–[¹¹C]C, of [¹¹C]31 in plasma of one monkey under baseline conditions.

The concentration of unchanged [¹¹C]31 in plasma (SUV) was appreciably higher in TSPO pre-block experiments than in baseline experiments, especially over the initial 10 min (Figure 6). Such increases have also been seen for other radioligands^23,33 and are attributed to the prevention of radioligand binding to abundant TSPO receptors in peripheral organs by the administered TSPO pre-blocking agent. In this regard, whole body PET scanning of a monkey before and after treatment with 1 clearly showed displacement of activity from TSPO-containing organs such as liver, lung, heart, kidney and spleen with a concomitant increase in radioactivity in bladder (see Supporting Information).

Figure 6.

Time-course of concentration of unmetabolized [¹¹C]31 in plasma after intravenous injection into monkey under baseline and pre-blocked conditions.

Binding of Ligand 31 to Human Brain Tissue

Ligand 31 showed similar high-affinity brain tissue binding in ‘high-affinity binders; HABs’ (K_i = 1.57 ± 1.03 nM, n = 4) and ‘mixed-affinity binders; MABs’ (K_i = 1.82 ± 1.09 nM, n = 4) and somewhat lower affinity in ‘low-affinity binders; LABs’ (K_i = 9.53 ± 6.25 nM, n = 5) (Figure 7). The differences in affinity towards LABs versus HABs or MABs are statistically significant (P < 0.05 for both HABs and MABs). However, these differences in K_i should be treated with some caution since they may reflect possible differences in the affinity of binding of the reference radioligand ([³H]1) towards HABs, MABs and LABs. Although, ligand 31 appears to show some sensitivity to LABs, sensitivity to MABs versus HABs is less clear from the small number of tests (and is statistically insignificant, P = 0.250). If as suggested, by Owen et al.³⁴ the MABs have about a 1:1 mixture of the high-affinity and low-affinity binding sites, then by the Cheng-Prusoff equation³⁵ the affinity expected for 31 in the MABs should be about two-fold lower than in the HABs. The data here, from only 4 HABs and 4 MABs, indicate they have similar affinities. Binding data in a larger group of subjects, preferably with tritiated 31 as radioligand, are needed to clarify whether true differences in binding affinities exist between HABs and MABs.

Figure 7.

Binding affinity (1/K_i) of 31 for brain tissue from human ‘high-affinity binders, HABs’, ‘mixed-affinity binders, MABs’ and ‘low-affinity binders, LABs’. Each point is from a single human subject.

Summary

In this study, a novel series of 2-phenylindol-3-ylglyoxylamides featuring a N¹-methyl substituent was developed. When tested on rat kidney membranes, most of the new compounds showed high affinity for TSPO, with K_i values comparable to those of their N¹-unsubstituted parent derivatives, in accordance with our previously reported pharmacophore/topological model hypothesis, where the indole NH is not involved in the recognition of the receptor protein. Thus, N¹-methyl-2-phenylindol-3-ylglyoxylamides constitute a new chemotype of high-affinity TSPO ligands with potential for development to produce effective radioligands for imaging brain TSPO with PET, of which [¹¹C]31 is the first example. [¹¹C]31 was easily prepared and readily entered monkey brain to give a high proportion of reversible specific binding to TSPO. As for other chemotypes,^26,34 such as the aryloxyanilides, exemplified by 2 and 6, and the linked bicyclics, exemplified by DPA713 (7) and PBR111 (8) (Figure 1),^26,34 this new chemotype shows different affinity for the two forms of brain TSPO found among human subjects. However, the difference in affinities is quite small compared to that of the aryloxyanilide 2, and indicates the possibility to find a new PET radioligand from this structural class with a truly insignificant difference that will be applicable in all human subjects.

Experimental Section

Materials and Methods

All reagents used were obtained from commercial sources (Sigma-Aldrich). All solvents were of an analytical grade. 2-Phenylindole (9) and 2-(4-fluorophenyl)indole (10) were from Sigma-Aldrich. 5-Fluoro-2-phenylindole (11), 2-(4-nitrophenyl)indole (12), 2-phenylindol-3-ylglyoxyl chloride (13), 2-(4-fluorophenyl)indol-3-ylglyoxylchloride (14), 5-fluoro-2-phenylindol-3-ylglyoxyl chloride (15), and 2-(4-nitrophenyl)indol-3-ylglyoxylchloride (16), were prepared as previously described.^27,28 1 and 3 were obtained from Sigma-Aldrich. Racemic [³H]1 (S.A., 84.8 Ci/mmol) and [³H]flumazenil (S.A., 83.4 Ci/mmol) were purchased from Perkin-Elmer Life Sciences.

Melting points were determined using a Reichert Köfler hot-stage apparatus and are uncorrected. Infrared spectra were recorded with a Nicolet-Avatar 360 FT-IR spectrophotometer in Nujol mulls. Routine nuclear magnetic resonance spectra were recorded in d₆-DMSO solution on a Gemini 200 spectrometer (Varian) operating at 200 MHz. Evaporation was performed in vacuo (rotary evaporator). Analytical TLC was carried out on 0.2 mm pre-coated silica gel aluminum sheets (60 F-254; Merck). Silica gel 60 (230-400 mesh) was used for column chromatography. Combustion analyses on target compounds were performed by our Analytical Laboratory in Pisa. All compounds showed ≥ 95% purity.

γ-Radioactivity from ¹¹C was measured using a calibrated dose calibrator (Atomlab 300; Biodex Medical Systems). Radioactivity measurements were corrected for physical decay. All radiochemistry was performed in lead-shielded hot-cells for personnel protection from radiation.

General Procedure for the Synthesis of N,N-Dialkyl-2-(phenylindol-3-yl)glyoxylamide Derivatives 17–18

Oxalyl chloride (0.65 mL, 7.5 mmol) was added dropwise, at 0 °C to a well-stirred mixture of the appropriate indole 9 or 10 (3.0 mmol) in freshly distilled diethyl ether (10 mL). The mixture was maintained at room temperature for 2–24 h (TLC analysis). The generated precipitate was collected and washed with portions of anhydrous diethyl ether to give the acyl chlorides 13 or 14^27,28 which were dried over P₂O₅ in vacuo, and then suspended in 50 mL of dry toluene. A solution of N-benzyl-N-methylamine (0.30 mL, 2.5 mmol) in dry toluene (5 mL) was then added dropwise, at 0 °C, followed by the addition of a solution of triethylamine (0.3 mL, 2.5 mmol) in the same solvent (5 mL). The reaction mixture was allowed to warm to room temperature, stirred for 2–24 h (TLC analysis) and then filtered. The collected precipitate was triturated with a saturated NaHCO₃ aqueous solution, washed with water, and collected again to give a first portion of crude product. The toluene solution was evaporated to dryness, and the residue was treated with saturated NaHCO₃ aqueous solution, washed with water, and collected to yield an additional amount of crude product. The quantities of amide derivatives obtained from the initial insoluble precipitate or from the toluene solution were variable, depending upon the solubility of the various compounds. Products 17 and 18 were purified by flash-chromatography (CHCl₃ as eluent). Yields and melting points of compounds 17 and 18 are listed in Table 1; spectral data are reported in the Supporting Information.

General Procedure for the Synthesis of N,N-Dialkyl-1-methyl-2-phenylindol-3-ylglyoxylamides Derivatives, 19–31

Sodium hydride (0.11 mmol, 50% dispersion in mineral oil) was added portion-wise, under a nitrogen atmosphere, to an ice-cooled solution of the appropriate N,N-dialkyl-2-phenylindol-3-ylglyoxylamide derivative Ia–k,^27,28 17, 18 (0.1 mmol) in dry DMF (3 ml). Once hydrogen evolution had ceased, an excess of methyl iodide (0.018 mL, 0.3 mmol) was quickly added at 0 °C. The reaction was maintained under stirring at room temperature until the disappearance of the starting material (12–24 h, TLC analysis). The solvent was eliminated under vacuum and the residue was triturated with crushed ice and collected by filtration. The products were finally purified by flash chromatography (CHCl₃ as eluent). Yields and melting points of compounds 19-31 are listed in Table 1; spectral data are reported in the Supporting Information.

Inhibition of [³H]1 binding to rat kidney mitochondrial membranes

For binding studies, crude mitochondrial membranes were incubated with [³H]1 (0.6 nM) in the presence of a test compound in the concentration range (0.05 nM–10 μM) in Tris-HCl buffer (50 mM, pH 7.4) as previously described.²⁷ For the active compounds, the IC₅₀ values were determined and K_i values were derived in accordance with the Cheng-Prusoff equation.³⁵

Inhibition of [³H]flumazenil binding to rat cerebral cortex membranes

Rat cerebral cortex membranes were prepared as previously described.³⁶ After differential centrifugation, the obtained crude membrane fraction was subjected to washing procedures to remove endogenous GABA.³⁷ The washed membranes were incubated with [³H]Ro15-1788 (0.4 nM) for 90 min at 0 °C in Tris–citrate buffer (50 mM 500 μL pH 7.4), as previously described.³⁸

Radiochemistry

Production of NCA [¹¹C]carbon dioxide

No-carrier-added (NCA) [¹¹C]carbon dioxide (~ 2.3 Ci) was produced with a PETtrace cyclotron (GE Medical Systems; Milwaukee, WI) according to the ¹⁴N(p,α)¹¹C reaction³⁹ by irradiation of nitrogen gas (initial pressure 160 psi; 75-mL volume) containing 1% oxygen with a proton beam (16.5 MeV, 45 μA) for 40 min.

Production of NCA [¹¹C]methyl iodide

NCA [¹¹C]methyl iodide was produced from NCA [¹¹C]carbon dioxide via reduction to [¹¹C]methane and then vapour phase iodination.⁴⁰ Thus, at the end of the proton irradiation [¹¹C]carbon dioxide was delivered to a PETtrace MeI Process Module (GE Medical Systems, Model: KAB301-31001) inside a hot-cell through stainless tubing (OD 1/8 in, ID 1/16 in) over 2 min and trapped on molecular sieve (13X) and reduced to [¹¹C]methane over nickel at 360 °C. The [¹¹C]methane was recirculated over iodine at 720 °C to generate [¹¹C]methyl iodide, which was trapped on Porapak Q held in the recirculation path.

Labeling of 31 with carbon-11

Compound 1k (0.5 mg) was added to a V-vial containing anhydrous DMSO (0.4 mL) containing finely ground KOH (4 mg). [¹¹C]Methyl Iodide was released from the Porapak trap at 180 °C into a stream of nitrogen and bubbled into the DMSO at 17 mL/min until the radioactivity in the vial maximized. The mixture was then heated at 80 °C for 5 min, after which the crude reaction mixture was diluted with water (500 μL). The product was injected onto a Gemini C18 column (5 μm, 10 mm × 250 mm; Phenomenex) eluted at 6 mL/min with a mobile phase of MeCN-0.1M-aq. HCOONH₄ initially composed of 50% MeCN and increased linearly to 80% MeCN over 10 min. Eluate was monitored for absorbance at 254 nm (Beckman 166 UV detector) and for γ-radiation with a semiconductor pin-diode detector (Bioscan). [¹¹C]31 (t_R =10 min) was collected and mobile phase and ammonium formate removed under high vacuum at 80 °C. Ethanol (1.5 mL) was added to dissolve the radioactive residue and the solution then diluted with saline (13.5 mL). Finally, this solution was sterilized by filtration through a sterile MP filter to provide [¹¹C]31 ready for intravenous injection, pending HPLC analysis.

[¹¹C]31 was analyzed for radiochemical purity, specific activity and chemical purity with HPLC on a Luna C18 column (5 μm, 4.6 × 250 mm) eluted with MeCN-0.1M aq. HCOONH₄ (60: 40 v/v) at 3.0 mL/min. Eluates were monitored for radioactivity (pin diode detector; Bioscan) and absorbance at 254 nm. Samples were injected alone, and then co-injected with the reference non-radioactive compound to check for co-elution. The identity of the product was further confirmed by LC/MS/MS of associated carrier on a Finnigan LCQ deca instrument.

Computation of cLogP

cLogP for 31 was computed with ChemDraw.

PET Experiments with [¹¹C]31 in Monkeys

All animals were handled in accordance with the Guide for the Care and Use of Laboratory Animals⁴¹ and the National Institute of Health Animal Care and Use Committee. For each scanning session, the subject monkey was immobilized with ketamine and maintained under anesthesia with 1–3% isoflurane in oxygen. An intravenous perfusion line filled with saline (0.9% w/v) was used for bolus injection of [¹¹C]31 of 99.9% radiochemical purity. Dynamic PET images of brain radioactivity were obtained for up to 100 min on an Advance (GE Medical Systems, WI) PET camera.

Decay-corrected time-activity curves (TACs) were obtained for irregular volumes of interest (VOIs) in brain, selected from pre-frontal cortex, temporal cortex, parietal cortex, occipital cortex, hippocampus, striatum, thalamus and cerebellum. Radioactivity levels in VOIs were normalized for injected dose and monkey weight and expression as a standardized uptake value (SUV):

$$\mathrm{SUV}=[(\%\text{injected dose per g tissue})\times \text{body weight in g}]∕100.$$

The following scans were performed among four monkeys. Baseline scans: All four male rhesus (Mucacca mulatto) monkeys (6.4–12.0 kg) received brain scans at baseline after bolus intravenous injection of [¹¹C]31 (4.76–6.70 mCi; dose of 31, 0.155-0.859 nmol/kg). TSPO pre-block scans: Two monkeys (11.1 and 12 kg) were scanned with [¹¹C]31 (6.10 and 6.40 mCi; dose of 31, 0.108 and 0.135 nmol/kg of 31) at 10 min after administration of the TSPO ligand 1 (5 mg/kg, iv). TSPO displacement scan: One monkey (12.82 kg) was scanned with [¹¹C]31 (5.60 mCi; dose of 31, 0.156 nmol/kg) with 1 (5 mg/kg, iv) administered over 2 min at 48 min from the start of the scan.

Stability of [¹¹C]31 in Monkey Whole Blood and Plasma in Vitro, and in Buffer

[¹¹C]31 was incubated for 45 min in whole monkey blood (1 mL). A sample (0.5 mL) was removed, added to acetonitrile (0.8 mL), and centrifuged. Then the supernatant liquid was analyzed by reverse phase radio-HPLC. The stability of [¹¹C]31 to incubation in sodium phosphate buffer (0.15 M, pH 7.4) for 2 h at room temperature was also assessed by reverse phase HPLC.

Plasma Protein Binding of [¹¹C]31.⁴²

The radioligand was added to pooled human plasma, placed at the top of an “Amicon” Centrifree filter unit (200 μL/unit) and filtered by ultra centrifugation at 5000 × g. Then all components of filter units were counted for radioactivity to allow calculation of radioligand plasma protein binding (f_P).

Emergence of Radiometabolites of [¹¹C]31 in Monkey Plasma In Vivo

During each of four PET scans (two baseline and two preblock), blood samples were drawn periodically from the monkey femoral artery and collected in heparin-treated Vacutainer tubes. The samples were centrifuged and the plasma separated. A sample of plasma (0.45 mL) was mixed with acetonitrile (0.72 mL) and centrifuged. The supernatant liquid was analyzed with radio-HPLC on a Novapak C18 column (4 μm; 8 × 100 mm) eluted at 2.0 mL/min with MeOH/H₂O/Et₂O (75: 25: 0.1 by vol.). The time-courses for percentages of radioactivity in plasma represented by [¹¹C]31 and its radiometabolites were calculated.

Human Brain Membrane Preparation

Human brain tissue from 13 donors was obtained from the UK National Multiple Sclerosis (MS) Brain Bank. Ethical approval for the study procedures was granted by the local Research Ethics Committee (Imperial College). Four of these donors had been classified previously, through ethically approved in vitro studies, as high-affinity binders (HABs) for the TSPO ligand 2, four as mixed-affinity binders (MABs) and five as low-affinity binders (LABs).²⁶ Donor tissue blocks were homogenized in 10 times weight for volume buffer (0.32 mM sucrose, 5 mM Tris-Base, 1 mM MgCl₂, pH 7.4, 4 °C). Homogenates were centrifuged (32,000 × g, 20 min, 4 °C) followed by removal of the supernatant. Pellets were resuspended in at least 10 times w/v (weight for volume) buffer (50 mM Tris-Base, 1 mM MgCl₂, pH 7.4, 4 °C) followed by two washes by centrifugation (32,000 × g, 20 min, 4 °C). Membranes were suspended in buffer (50 mM Tris-Base, 1 mM MgCl₂, pH 7.4, 4 °C) at a protein concentration of about 4 mg protein/mL and aliquots were stored at – 80 °C until use.

Human Brain Homogenate and Competition Binding Assays

Aliquots (about 250 mg protein/mL) of membrane suspension were prepared using assay buffer (50 mM Tris-Base, 140 mM NaCl, 1.5 mM MgCl₂, 5 mM KCl, 1.5 mM CaCl₂, pH 7.4, 37 °C) and incubated with [³H]1 (5 nM) and one of 11 concentrations of unlabeled 31 ranging from 0.03 nM to 10 μM at 37 °C in a final volume of 500 μL for 60 min. The specific binding component was defined by addition of unlabeled 1 (10 mmol/L). After incubation, assays were terminated by filtration through GF/B filters (Whatman, Maidstone, UK), followed by 3 × 1 mL washes with ice-cold wash buffer (50 mM Tris-Base, 1.4 mM MgCl₂, pH 7.4, 4 °C). GF/B filters were pre-incubated with 0.05% polyethyleneimine (60 min) before filtration. Scintillation fluid (3 mL/vial, Perkin Elmer Ultima Gold MV) was added and vials counted on a Perkin Elmer Tricarb 2900 liquid scintillation counter. Each point was performed in quadruplicate. K_i (nM) values were determined with GraphPad Prism 5.0 software (GraphPad Software Inc., La Jolla, CA, USA).

Abbreviations

ANT

adenine nucleotide translocase

BzR

central benzodiazepine receptor

HABs

high-affinity binders

LABs

low-affinity binders

MABs

mixed-affinity binders

MPTP

mitochondrial permeability transition pore

NCA

no-carrier-added

PET

positron emission tomography

PBR

peripheral benzodiazepine receptor

S.A.

specific radioactivity

TSPO

translocator protein (18 kDa)

VDAC

voltage dependent anion channel