Generación de Valor

The fluorescent ligand BODIPY-JTE2-3 121 was used in confocal experiments to demonstrate whether the ligand had the potential to highlight CB2 receptor membrane

binding. Detailed experimental conditions are included in section 2.5 but in brief known concentrations of fluorescent ligand were incubated with CHO cells

-8 -7 -6 -5 -4 0 20 40 60 80 100 120 JTE2-3 (62) Bodipy-JTE2-3 (121) log [ligand] % bound [ 3H] CP55, 940 -10 -9 -8 -7 -6 -5 0 20 40 60 80 100 120 JTE2-3 (62) Bodipy-JTE2-3 (121) log [ligand] % bound [ 3H] CP55, 940 A. CHO B. HEK

expressing human CB2 receptors and images were recorded at various time courses

throughout the experiment. To assess whether the fluorescence observed was due to specific CB2 binding, fresh CHO cells were pre-incubated with a blocking

concentration of a non-fluorescent cannabinoid ligand (i.e. 1 µM HU210 14) and the fluorescent ligand was subsequently introduced. Images collected from this

experiment were compared to those conducted without the HU210 14 present, to determine if any fluorescence observed was due to CB2 receptor binding. Various

concentrations of 121, ranging from 10 – 400 nM, were used in experiments. The fluorescence observed was weak even when laser power and pin-hole aperture was increased. Even at concentrations of 400 nM fluorescence from the CHO cells was negligible (results not shown). Background fluorescence from the CB2 transfected

CHO cells was negligible at the excitation wavelengths used (633 nm) (Result not shown).

4.6 Discussion

Synthesis to access JTE2-3 62 and its 2-aminoethyl derivatives 119, 120, and 121 was successfully carried out. The positioning of the 2-aminoethyl linker was guided by molecular modelling results which demonstrated that attachment of a linker at this position should not interfere with the binding of the ligand in the receptor binding pocket.

In order to complete the synthesis of the 2-aminoethyl derivatives of JTE2-3 62 it was necessary to synthesise a novel ortho substituted tyramine derivative 109. By careful selection of the thermo stable Cbz protecting group247_{, we were able to}

access 109 via a thermally induced [3,3]-sigmatropic Claisen rearrangement.

Oxidation of the olefin using two different oxidants (OsO4 with a NaIO4 cleavage, and

O3) gave the aldehyde 111 and alcohol 112 respectively. Repeated attempts to

generate the aldehyde 111 using ozonolysis and DMS failed to give the desired product. A literature report detailing this reaction on a similar allylic benzene system did not suggest that there were any problems during this step, however only brief experimental conditions were reported248. It was obvious that the ozonide of 110 could be formed, demonstrated by the successful reduction of this complex using NaBH4. Therefore, we concluded that the problems observed in our synthesis were

due to the DMS reduction step. All attempts to rectify this by altering the conditions of the DMS reduction did not resolve the problem and we had to use the alternative OsO4 oxidation to generate the aldehyde 111. Unfortunately reductive amination of

this aldehyde 111 failed to give the amine required. We assessed the use of NH3 and

NH4Cl, as a source of ammonium ion, and combinations of these together to form the

imine in-situ and reducing it to the amine using NaCNBH3. However we could not

detect the 1° amine in any of the attempted reactions. A Beilstein search revealed that other reported experiments using ammonia and aldehydes to give 1° amines often exhibited lower yields (the highest reported was 40%251) than the corresponding reaction used to yield 2° and 3° amines. This is presumably due to the fact that once the 1° amine has been formed it is immediately available to react with the aldehyde present and form 2° amines. To produce the amine 115 we therefore had to use a more convoluted route, involving activation of the alcohol to the mesylate and forming the corresponding azide 114. This was reduced under mild conditions using PPh3 and

water to the corresponding amine 115. Boc protection of the amine resulted in a compound that was quantified analytically and we calculated an overall yield from the olefin as a respectable 32% for 5 steps; which represents an average yield of 80% for each step. The subsequent chemical reactions leading to the two fluorescent

conjugates Dansyl-JTE2-3 120 and BODIPY-JTE2-3 121 utilised chemistry fully discussed in section 2.6. The unlabelled ligand JTE2-3 62 was synthesised to allow us to test the effect of conjugation to the fluorescent dye using a radioligand binding competition assay discussed below.

The affinity of JTE2-3 62 and BODIPY-JTE2-3 121 to the human CB2

receptor was assessed using radioligand competition displacement assays against the non-specific cannabinoid [3H] CP55, 940. The membrane homogenates used were made up from both CHO and HEK cells stably transfected with the human CB2

ligand. Concentrations of up to 32 µM of BODIPY-JTE2-3 121 failed to displace any of the [3H] CP55, 940 from the receptors. The Ki value of 3.11 µM obtained for the

unlabelled JTE2-3 62 was much higher than that reported in the literature (0.4 nM)126. This represented a difference in affinity to that quoted in the literature of a factor of 7500. JTE2-3 62 displayed less affinity to the human CB2 than was expected.

Adulterating JTE2-3 62 with the fluorescent dyes Dansyl and BODIPY effectively abolished any of the human CB2 receptor binding ability JTE2-3 62 possessed. This

was confirmed in a fluorescent confocal microscopy experiment, because no membrane binding was observed at concentrations of 400 nM.

The significantly lower affinity of JTE2-3 62 observed in our displacement experiments, when compared to the literature Ki value, was unexpected and was a

matter of concern. Initially we checked that the binding assay was generating meaningful data by performing a saturation radioligand binding experiment (using [3H] CP55, 940) on the CB2 HEK membranes which yielded a Kd of 5.1 nM (results

not shown), which was well within the reported range of 0.29-7.37 nM for this ligand51. The analytical data for the compound JTE2-3 62 was checked against that published in the original patent. 1H NMR and m.p. were identical for our synthesised compound and the literature values (section 8.1). 13_{C and ES-MS, which were not}

reported in the patent, were consistent with the structure of JTE2-3 62. Finally a fresh stock solution of JTE2-3 62 was made up to check for errors in ligand concentration. We obtained the same results with the new stock solution and this result forms part of the n = 4 analysis shown in B of Figure 4-3. We had ruled out all experimental error in the calculation of the Ki of JTE2-3 62 (3.11 µM) and yet could not explain the 7500

fold difference to that reported in the patent (0.4 nM)126_{. Differences between the two}

experimental protocols used to calculate the Ki of JTE2-3 62 were checked. The

radioactively labelled ligand used and the species of CB2 receptor tested against

differed between the two protocols. [3H] Win 55212-2 and rat CB2 receptors (derived

from rat spleenocytes) were used in the Japan Tobacco publication126, where we had used [3H] CP55, 940 and the human CB2 receptor expressed in HEK and CHO cells.

[3H] Win 55212-2 is not commercially available so we were unable to assess the effect of using a different radioligand. However we were able to test the affinity of JTE2-3 62 against the rat CB2 receptor and discovered that the difference in the

different Ki values obtained was due to the species of CB2 receptor used. A full

discussion of this work is reported in Chapter 5.

4.7 Conclusions

• We successfully synthesised a derivative of the CB2 selective ligand JTE2-3

62, incorporating an alkyl linker which we labelled with Dansyl and BODIPY® fluorescent dyes.

• Computational modelling demonstrated that the optimum position for attachment of an alkyl linker was on the phenol ring, as the phenol ring was located near to the top of the receptor binding pocket.

• Development of novel chemistry was established, enabling the synthesis of orthogonally protected 2-aminoethyl tyramine, required for the synthesis of the fluorescent conjugate.

• Binding data for the BODIPY labelled compound 121, demonstrated a loss of affinity for the CB2 receptor (Ki > 32 µM).

• Confocal microscopy experiments showed that BODIPY labelled compound 121 did not bind to human CB2 receptors expressed in live CHO cells.

• Binding data obtained for the unlabelled lead compound JTE2-3 62,

demonstrated it had, surprisingly, 7500 times less affinity to the CB2 receptor

compared to the Ki value reported in the original patent. Thorough analysis of

this anomaly suggested that the difference between the two results was due to the different species (rat and human) of receptor used in the binding assay.