Inhibition of ABCB1, ABCC2 and ABCG2

The potency of new tariquidar-like analogs to modulate the ABCG2 transporter was determined in the flow cytometric mitoxantrone efflux assay. In ABCG2-overexpressing MCF-7/Topo cells red fluorescent mitoxantrone is not accumulated but extruded by the efflux pump. Therefore, ABCG2 inhibitors can easily be recognized by the determination of intracellular mitoxantrone levels. Different concentrations of efflux pump inhibitors cause changes in the mitoxantrone efflux that can be measured by the relative fluorescence intensity of the cells. Similarly, in ABCB1 overexpressing Kb-V1 cells, the accumulation of the fluorescent ABCB1 substrate calcein can be quantified by the use of flow cytometric techniques or the use of microplate readers. In the presence of ABCB1 inhibitors higher intracellular calcein levels lead to increased relative fluorescence intensities of the cells. The modulation of ABCC2 was investigated by the use of ABCC2-overexpressing MDCK-cells that were incubated with chloromethylfluorescein diacetate (CMFDA) in the presence or absence of increasing concentrations of test compounds. After intracellular formation of the ABCC2 substrate glutathione methylfluorescein, (GMSF), the extent of intracellular fluorescence was monitored with a plate reader in a concentration-dependent manner. Known modulators of the 3 efflux pumps were investigated as reference compounds.

Visible precipitations in the culture medium (EMEM / 5 % FCS), used for the mitoxantrone efflux assay, occurred at a number of tariquidar analogs at final concentrations above 10 µM.

Therefore, unless otherwise stated, concentrations higher than 10 µM were not considered in the calculation of IC50 values. The data were used for the construction of concentration-response curves as depicted in Figures 4.2, 4.3, 4.4 and 4.5. For a summary of all data from the investigation of test compounds 1-26 and reference substances for inhibition of ABCG2, ABCB1 and ABCC2 cf. Table 4.1.

Modulation of the efflux pump ABCG2 89

concentration [µM]

0.001 0.01 0.1 1 10 100

ABCG2 inhibition [%]

0 20 40 60 80 100

concentration [µM]

0.001 0.01 0.1 1 10

ABCG2 inhibition [%]

0 20 40 60 80 100

Figure 4.2: Concentration-response curves for ABCG2 modulation by the reference compounds fumitremorgin C (circles), Ko143 (squares), elacridar (triangles) and tariquidar (inverted triangles).

Figure 4.3: Concentration-response curves for ABCG2 modulation by the new tariquidar analogs 1-7. Compound 1 (circles), 2 (squares), 3 (triangles), 4 (inverted triangles), 5 (diamonds), 6 (stars) and 7 (hexagons).

concentration [µM]

0.001 0.01 0.1 1 10

ABCG2 inhibition [%]

0 20 40 60 80 100

concentration [µM]

0.001 0.01 0.1 1 10

ABCG2 inhibition [%]

0 20 40 60 80 100

Figure 4.4: Concentration-response curves for ABCG2 modulation by the new tariquidar analogs 8-14. Compound 8 (circles), 9 (squares), 10 (triangles), 11 (inverted triangles), 12 (diamonds), 13 (stars) and 14 (hexagons).

Figure 4.5: Concentration-response curves for ABCG2 modulation by the new tariquidar analogs 15-18. Compound 15 (circles), 16 (squares), 17 (triangles), and 18 (inverted triangles).

Modulation of the efflux pump ABCG2 91

Table 4.1: Inhibition of ABCG2, ABCB1 and ABCC2 by compounds 1-26 and reference substances.

a Mean values ± SEM, calculated from 2-3 independent experiments. ^b Maximal inhibitory effects (%) are expressed as inhibition caused by the highest concentration of the compound tested (10 µM, if not otherwise stated in parentheses) relative to the inhibitory effect caused by 10 µM fumitremorgin C (100% inhibition). ^c Mean values ± SEM, calculated from n=2-5. ^d No effect up to a concentration of 10 µM; Fumitremorgin C: 0.8

% inhibition at 10 µM, 22 % inhibition at 100 µM; Ko143: 1.6 % inhibition at 10 µM. ^e No effect up to a concentration of 50 µM; ^f N= 1. ^g MK571: 3-(((3-(2-(7-chloroquinoline-2-yl)ethenyl)phenyl)((3-dimethylamino-3-oxopropyl)thio) methyl) thio) propanoic acid;

LTC4: leukotriene C4. ^h Values obtained by the Calcein-AM microtiter assay. ⁱ Values obtained by Hoechst 33342 assay (see 4.4.2). IC50 values were calculated using SIGMA PLOT 9.0, four parameter logistic curve fitting and GraphPad Prism 5.0, respectively.

The fumitremorgin C analog Ko143 is a highly potent ABCG2 modulator with a maximal inhibitory effect of 82% at a concentration of 10 µM, referred to the control fumitremorgin C.

The acridone carboxamide derivative elacridar strongly inhibits both transporters without a preference, whereas tariquidar was about equipotent with elacridar at ABCB1 but approximately 4 times less potent at ABCG2. Transporter inhibition data show that minimal structural changes at the benzamide core of tariquidar resulted in a drastic change from ABCB1 to ABCG2 modulation. The shift of the quinoline-3-carboxamido substituent from position 2 (tariquidar) to position 3 of the benzamide moiety (compound 9) proved to be the key to increase the selectivity for ABCG2 over ABCB1: Whereas the moderate inhibition of ABCG2 was maintained (9: IC50 858 nM, tariquidar 916 nM), the modulation of ABCB1 decreased by a factor of >75 (IC50 >17,000 vs. 223 nM for 9 and tariquidar, respectively).

Furthermore, all compounds with bicyclic hetarylcarboxamides in position 3 and an ester group in position 4 (compounds 1-4 and 17) were highly potent and selective ABCG2 modulators superior to compounds with monocyclic heteroaromatic moieties (5-7). The 2 most potent inhibitors were obtained with a quinoline-2-carboxamido (compound 2 with an IC50 of 60 ± 10 nM) and quinoline-3-carboxamido substituent (compound 1 with an IC50 of 119 ± 22 nM), respectively. Presumably, the ester group in these compounds contributes considerably to the interaction with ABCG2 since structural modifications in position 4 and 5 of the benzamide core (8-18) resulted in a decrease in potency. For instance, in the series of quinoline-2-carboxamides 15-18 the ester 17, an analog of compound 2, lacking the 2 methoxy groups at the tetrahydroisoquinoline moiety, is only slightly less potent (by a factor of 5) than 2 as an ABCG2 modulator, whereas all other structural variations (15, 16 and 18) caused a more pronounced decrease in activity. This is specifically true for the amide analogs of 2 (16 and 18) in terms of both, potency and maximal response. Additional information on structure activity relationships is gained from the data for the substructures and intermediates (19-26). With the exception of compound 23, bearing a nitro group in position 3 and the ester group in position 4, analogs containing an intact tetrahydroisoquinoline moiety but lacking the quinolinecarboxamide substituent (19-23) were almost inactive. In the building blocks 24-26, both the ester group and the quinoline-2-carboxamide substituent are retained, but the characteristic benzamide N-substituent is lacking or truncated. The carboxylic acid 24 turned out to be inactive, whereas the benzamides 25, 26, which are not negatively charged at physiological pH, showed a moderate inhibition of ABCG2. Thus, ABCG2 modulation appears to depend on an intact 3-(acylamino)benzoic amide moiety rather than on the substituent at the benzamide N-atom.

Modulation of the efflux pump ABCG2 93

Only one compound (23) missed selectivity for ABCG2 vs. ABCB1 (ABCB1 transporter assays performed by Dr. Christine Müller and Peter Höcherl). The lack of selectivity of 23, a moderate inhibitor of ABCG2 and ABCB1, seems to be due to the lack of a “selectivity-conferring” bulky residue in either position 2 or 3 of the benzamide. Furthermore, a set of representative ABCG2 inhibitors (1-11), turned out to be inactive (up to concentrations of 20 µM or 50 µM, respectively) when studied for modulation of ABCC2 (investigations performed by Anne Mahringer), underlining the high transporter selectivity of these compounds.

4.4.2 Fluorescence-based microtiter plate assays for the characterization