Improved syntheses of P2X7 ligands based on substituted benzyl amide of pyroglutamic acid motif labelled with iodine-123 or iodine-125

P2X7 sites are emerging targets for molecular imaging research, notably in the context of neurodegeneration and inflammatory conditions. Therefore, we prepared a precursor for (radio)iodination of the P2X7 ligand VPGIS191. We then developed a radioiodination method with 123I for SPECT with a radiochemical yield of 71 ± 13% and 125I for autoradiography with a radiochemical yield of 85 ± 6% Autoradiography of [125I]VPGIS191 in mouse brain cryostat sections demonstrated approximately 36 nM binding affinity and Bmax of approximately 400 pmol/gram tissue for P2X7 binding sites. VPGIS191 (cis) had a two-fold lower affinity compared to its geometric trans-isomer TZ6019.


Introduction
P2X 7 (Fig. 1) is a transmembrane ligand-gated cation channel that is activated by extracellular adenosine triphosphate (ATP) and certain other adenosine nucleotides, thereby mediating cytokine release from cells of hematopoietic lineage such as microglia and macrophages in the brain, retina, and peripheral nervous system [1].Activation of P2X 7 receptors mediates host immune responses participating in the regulation of apoptosis and inflammation [2] in a variety of pathologies including Alzheimer's and Parkinson's diseases [3], cardiovascular disorders [4], and cancers [5].Due to their potential as markers or therapeutic targets in diverse disease conditions, P2X 7 receptors are an emerging topic for molecular imaging by positron emission tomography (PET) and single photon computer tomography (SPECT).However, clinical molecular imaging of P2X 7 receptors is in its infancy, in part due to the inadequate sensitivity of available radioligands [6].

In vitro autoradiography
We followed an autoradiographic procedure much as described by Kuhar et al. [22].Saline-perfused mouse brains were frozen by immersion to isopentane at − 40 °C, and stored at − 80 °C until use.A cerebral hemisphere was mounted in the sagittal orientation in a cryostat for cutting at − 20 °C into 20 μm-thick sections, which were thawmounted onto Superfrost glass slides.After air-drying and storage overnight at − 80 °C, the slides were preincubated in buffer (50 mM buffer: Tris-HCl, pH 7.4, 1 mM EDTA, 0.1% BSA) for 15 min.After removal of the excess buffer, the slides were incubated with the same buffer (1 mL) modified by addition of radiotracer [ 125 I]VPGIS191 (A m = 4.6 MBq/nmol) at final concentrations of 2, 10, 50 and 100 nM.Nonspecific binding was evaluated in consecutive brain sections by the addition of GSK1482160 to a final concentration of 10 µM.After one-hour incubation, the sections were washed by immersion in ice cold buffer (3 × 1 min) and finally dipped in distilled water (30 s, 5 °C) to remove buffer salts, followed by rapid drying under an air stream.The dried slides were then exposed to a phosphor storage screen for 1 − 6 min together with slides bearing dried drops of buffer of known concentrations of The imaging screens were read using a Cyclone phosphor imager (Perkin-Elmer, USA), and the brain binding results (total, non-specific, and specific) converted into units of pmol/gram tissue (wet weight).

Data analysis
Data from autoradiography experiments were analyzed using Microsoft Excel (Microsoft, Redmond, WA, USA).The saturation binding parameters (B max and K d ) were determined using linear regression of the Scatchard plot.
The preparation of the (E)-stereoisomer of substituted lactam 2 was as previously described [8], through N-alkylation in DMF with tributyl-(3-chloropropenyl)stannane in the presence of lithium bis(trimethylsilyl)amide (LHMDS) as a base.In our case, we used a bromine-substituted analog as the alkylation reagent (Scheme 1).Reaction and isolation of the product were unproblematic, with 68% yield, compared to 88% as reported for the (E)-stereoisomer [8].We suppose that our replacement of chlorine for bromine accounts for the lower yield.
We carried out the syntheses of the non-radioactive chromatographic VPGIS191 standard by two different methods, using either 82.2 µmol (Method I) or 10 nmol of precursor (Method II) (Scheme 1).Method I followed the procedure of Jin et al. [8], but replacing CHCl 3 -MeOH (20:1-10:1, v/v) for hexane-EtOAc (4:1, v/v) as solvent in the silica-gel column purification.We obtained the product in 54% yield, as compared to the 90% reported by Jin et al. [8].In the slightly modified Method II, we stopped the reaction with sodium ascorbate instead of sodium thiosulfate, with reaction time of 5 min instead of 15 min.We developed a new procedure for the SPE purification of the radiotracer, using 10% ethanol in water for SPE washing and 60% ethanol to elute the product, and had optimized the SPE purification for Bond Elute Plexa and alternately for Strata Phenomenex cartridges.Despite these measures, the product contained up to 0.01% tin-containing precursor according to HPLC-UV (Suppl.Mat.Fig. S1).We estimate a 100-fold margin of safety for the maximum possible carryover of tin from the synthesis, were this method eventually applied for human SPECT studies.We also tested the Sep pack C 18 cartridge for SPE purification, which proved to retain the product even in 95% ethanol.Therefore, we used the optimized SPE procedure above for the radiosyntheses with 125 I and 123 I.

In vitro autoradiography of [ 125 I]VPGIS191 in mouse brain tissue
We performed in vitro autoradiography of [ 125 I]VPGIS191 in mouse brain sagittal cryostat sections (Fig. 3), following an established procedure with optimized washing conditions by Kuhar et al. [22] as modified in [8].The specific binding was invariably less than 50% of total binding, which disfavored quantitation by Scatchard analysis.Nonetheless, we were able in one trial to obtain estimates of K d = 36 nM and B max = 402 pmol/g, which closely matched corresponding affinity for [ 123 I]TZ6019 in HEK cells (K d = 19 nM) [8].In what seems to be the first quantitation of P2X 7 sites in rodent brain cryostat sections, we find a rather high B max , comparable to that seen for [ 3 H]PK11195 binding at microglial TSPO sites [23].Despite this high B max , the low affinity and high non-specific binding of [ 125 I]VPGIS191 disfavored its use in quantitative autoradiography, and may likewise predict low signal-to-background in SPECT studies with this ligand (Fig. 4).

Conclusions
Several structural classes of radioligands for the P2X 7 receptor intended for PET or SPECT imaging have affinities in the range 1-20 nM (Table 1); among these, only the pyroglutamic acid derivative [ 123 I]TZ6019 is intended for SPECT imaging had 20 nM affinity in HEK cells [8], which may not suffice for visualization on P2X 7 sites in living brain.As an alternative to [ 123 I]TZ6019, we prepared its geometric isomer VPGIS191 as reference standard, and also labelled with iodine-123 for SPECT, or iodine-125 for autoradiography in vitro.Unlike the literature synthesis for TZ6019, our synthesis of VPGIS191 employs a commercially available intermediate, namely (Z)-3-(tributylstannyl)prop-2-en-1-ol.We obtained the precursor in 68% yield, the reference standard VPGIS191 VPGIS191 in 52% yield, 123 I labelled radiotracer in 71 ± 13% radiochemical yield, and 125 I labelled radiotracer in 85 ± 6% radiochemical yield.Despite its high C]F-GSK1482160, [ 123 I]TZ6019 and [ 18 F]IUR-1602, the pyroglutamic acid-based ligands have high affinity for P2X 7 .Notably, [ 11 C]GSK1482160 had 1.2 ± 0.1 nM affinity (K d ) [15] versus 19.3 ± 2.8 nM for [ 123 I] TZ6019 [8].These lead compounds for PET and SPECT

Fig. 1
Fig. 1 Cartoon depicting the cellular locations of P2X 7 receptors and their predicted involvement in various pathologies with an inflammatory component.Created with BioRender.com

Fig. 4
Fig. 4 Scatchard analysis of the saturable binding of [ 125 I]VPGIS191 in mouse brain sagittal cryostat sections.Each point indicates the mean of four determinations at 2, 10, 50, and 100 nM [ 125 I] VPGIS191.