EMPLEO

As observed in the crystal structure of 7.21 co-crystallised in VIM-2 (Figure

82), it is believed that adding an electron withdrawing group onto the phenyl

ring attached to the ‘war head’ group would increase the binding affinity. This is due to increased face to face π-stacking interactions with the electron rich Tyr-67 residue. A structure activity relationship study was conducted to confirm this hypothesis (Figure 86).

Figure 86: Location of SAR variation for thiol series Thiol

The phenyl group of compound 7.21 was also substituted for a bromine (7.41) to see if this had negative effects on the binding affinity due to the loss of the π-stacking interactions. Two examples were included which contained 5 membered rings (7.36 and 7.38) to see what effect changing from a 6 membered ring had on the binding affinity. Scheme 28 shows the synthetic route to the thiol based compounds.

Scheme 28: Synthetic route to compounds 7.28-7.41 based around thiol compound 7.21 Due to the observed dihedral angle in the crystal structure of compound 7.21 co-crystallised in VIM-2. Substituents were placed in the ortho position of the phenyl ring attached to the ‘war head’ group (Position X, Figure 86) to explore the role of the dihedral angle of the biphenyl ring system to investigate if changing the dihedral angle effects the binding affinity (7.28 and 7.33). Substituents were also placed in the meta and para positions of the phenyl group, attached to the ‘war head’ group (Position X, Figure 86), to investigate if placing substituents around the ring changed the binding affinity (7.28, 7.29 and 7.30).

Compounds containing strong electron withdrawing groups, such as nitro groups and carboxylic acids, gave very poor yields for the Suzuki coupling reaction (<30%). The compounds also proved be very water soluble leading to problems with isolation of the iodine containing intermediate compounds following opening of the lactone with TMSI. A number of impurities from this step were then carried through the introduction of the thiol and hydrolysis steps lead to significantly impure final thiol compounds. Although purification

attempts through column chromatography and recrystallization were made, no compounds containing electron withdrawing groups could be obtained in high enough purity for biological testing. There was therefore no compounds containing electron withdrawing groups, such as nitro or carboxylic acids, that were tested against the MBL enzymes.

Generally, there is no clear trend from the SAR study, and it has been observed that changing the electron density on the phenyl ring of 7.21 has no real effect on the potency of the molecules (Table 14).

It is also observed that putting groups into the ortho position (7.28, 7.33 and

7.40) on the phenyl ring does not affect the binding affinity of the compounds

to the IMP-1 enzyme implying the dihedral angle of the biphenyl unit does not have a large contribution to the binding affinity. However, In placing groups in the ortho position does lower the IC50 of the compounds against

the VIM-2 enzyme. This result suggests that the increase in the dihedral angle of the biphenyl unit is favourable for binding to the VIM-2 enzyme. Substitution of the phenyl ring for five membered rings (7.36 and 7.38) shows no significant change in the binding affinity of the inhibitors. This implies that the π-stacking interaction is not a major component of the binding affinity. Substitution of the phenyl ring for a bromine shows no change to the binding affinity for VIM-2. However, for IMP-1 there is a significant increase in the IC50 indicating that a bromine is not tolerated as

well within this enzyme.

Placing groups onto the phenyl ring, attached to the ‘war head’ group, in the meta position lowers the IC50 of the inhibitor compounds towards the VIM-2

enzyme. In general increasing the hydrophobicity of the inhibitors causes them to exhibit stronger binding to the VIM-2 enzyme (7.33 and 7.35). The same effect is not observed for the IMP-1 enzyme where placing groups into the meta position weakens the binding affinity. Compound 7.30 exhibits the best IC50 against IMP-1 and contains a chloro group in the para position. This is potentially rationalised by the more enclosed active site seen in the IMP-1 enzyme compared to the VIM-2 enzyme (see Section 2.4.2.1)

Table 14: Biological evaluation results of thiol series SAR study Compound Number X RA [%], 100 µM (VIM-2) IC50 VIM-2 (μM) IC50 IMP-1 (μM) 7.21 < 1 0.234 ± 0.044 0.232 ± 0.043 7.28 < 1 0.180 ± 0.098 0.309 ± 0.067 7.29 < 1 0.072 ± 0.059 0.147 ± 0.057 7.30 < 1 0.109 ± 0.039 0.069 ± 0.052 7.31 < 1 0.207 ± 0.069 0.397 ± 0.089 7.32 < 1 0.231 ± 0.030 0.307 ± 0.094 7.33 < 1 0.045 ± 0.031 0.378 ± 0.062 7.34 < 1 0.459 ± 0.073 1.723 ± 0.047 7.35 < 1 0.071 ± 0.031 0.220 ± 0.035 7.36 < 1 0.365 ± 0.023 0.393 ± 0.035 7.37 < 1 N.A. N.A 7.38 1.25 0.174 ± 0.127 0.189 ± 0.049 7.39 < 1 0.184 ± 0.165 0.323 ± 0.037 7.40 < 1 0.158 ± 0.089 0.480 ± 0.045 7.41 < 1 0.170 ± 0.018 0.804 ± 0.030

The thiolactone (7.64) was also synthesised and biologically evaluated. This could investigate if the thiolactone ring could be hydrolysed under the conditions of the assay to give inhibitor 7.21. The idea being that this method could be used to produce a pro-drug to help with the administration of the compound with the thiol ‘war head’ masked until hydrolysed by the enzyme. To so synthesise 7.64, compound 7.21 was refluxed in TFA to afford thiolactone 7.64 in good yield (99%) (Scheme 29).

Scheme 29: Synthesis of thiolactone 7.64

Additionally, compound (7.67) with the thiol group directly on the aromatic ring and a spacer to the carboxylic acid moiety, was also prepared and screened to probe the importance of the spatial arrangement of the groups. To synthesise 7.67, boronic acid 7.65 was oxidised to thiolactone 7.66 with hydrogen peroxide in good yield (97%). Thiolactone 7.66 was then hydrolysed using base to yield compound 7.67 in good yield (93%) (Scheme

30).

Scheme 30: Synthesis of compound 7.67

Neither of these compounds (7.64 and 7.67) gave significant inhibitory activity in the presence of the NDM-1 enzyme. The thiolactone 7.64 gave a residual activity (RA) of 90% at 100 µM against NDM-1, which is significantly higher than the RA of 7.21 which is < 1% at 100 µM against NDM-1. No IC50

was determined for 7.64 against the MBLs as the RA was above 30%. The observed RA implies that 7.64 is not being hydrolysed rapidly by the NDM-1

enzyme. The results however do not confirm if hydrolysis is occurring slowly or not at all. Further biological evaluation would be required to establish this. Compound 7.67 has a RA of 92% at 100 µM against NDM-1, which is significantly higher than the RA of 7.21 of < 1% at 100 µM against NDM-1. No IC50 was determined 7.67 against the MBLs as the RA was above 30%.

This shows that the spatial arrangement of the thiol in relation to the benzene and carboxylic acid is important to the observed binding affinity of

7.21.