T. Zhou et al. / Bioorg. Med. Chem. Lett. 22 (2012) 3219–3222
3221
MDR using a P-gp over-expressed cancer cell line (KB-vin) was
evaluated by measuring the reversal of vincristine (VCR) resistance
in the presence of 4 lM of DCP or DCX analogs. Verapamil, a first
generation chemosensitizer and P-gp competitive inhibitor, was
used as the positive control. The results are shown in Table 1. In
general, the DCP and DCX analogs dramatically reversed the VCR
resistance in the KB-vin cells by lowering the GI50 values of VCR
under co-treatment conditions.
interesting that introducing bromide at R8 (compound 19) signifi-
cantly improved MDR reversal activity approximately two-fold
compared to compound 7, with a co-treatment GI50 value as low as
1.6 nM, indicating that R8 may be an important position for optimiz-
ing chemosensitization activity. Further investigation is needed to
verify and extend the preliminary SAR considerations noted above.
Data shown in Table 1 demonstrated a lack of concordance
between anti-HIV activity and MDR reversal potency. Some com-
pounds, such as compounds 1, 2, and 9, were potent anti-HIV
inhibitors, but weaker chemosensitizers. In contrast, compounds
5, 8, and 12 significantly reversed VCR resistance, but only weakly
inhibited anti-HIV replication. This finding is not too surprising,
since different targets are presumably involved. Nevertheless,
compounds 3, 4, 7, 11, and 17 exhibited significant polypharmacol-
ogy and, as such, are interesting dual function anti-HIV and chemo-
sensitization leads.
DCP analogs 3 and 4, with a methyl group substituted at R2,
afforded a 10-fold better chemosensitization than their counter-
parts 1 and 2, as shown in Table 1. (The GI50 of VCR is 2.9 or
3.9 nM in combination with compound 3 or 4, versus 30 or
20 nM in combination with compound 1 or 2.) This suggests that
R2-methyl in DCP is important for optimizing the activity. The
small alkyl group is also tolerated well at R1 in DCP. Both 3 and 4
exerted high activity significantly better than verapamil. Both
compounds also showed potent anti-HIV activity. Significantly,
compound 3 fully reversed the VCR resistance in KB-vin under
the co-treatment condition (with GI50 of 2.0 nM for VCR in the par-
ent drug-sensitive KB cell line).
Compounds 5–19 are DCX analogs with diverse substitutions on
the A and C rings. As shown in Table 1, most of the analogs exhib-
ited high chemosensitization activity, indicating that a planar ring
extension from chromone to xanthone is not detrimental. Com-
pounds 5–7 bearing different substitutions on R7 afforded the anti-
viral compounds with a wide spectrum of chemosensitization
activities. Compounds 5 and 7 with non-substitution or R7-meth-
oxy-substitution exhibited comparable effects to 3 and 4, with
co-treatment GI50s as low as 2.1 and 3.4 nM, respectively. In con-
trast, 6, an R7-hydroxy-DCX, failed to reverse the cytotoxicity of
It is known that the acquired VCR resistance in KB-vin is due to
the enhanced drug efflux via P-gp over-expression.12,13 Since the
addition of DCP and DCX reversed VCR’s cytotoxicity as determined
by phenotypic assay, we next investigated whether select com-
pounds were active using a P-gp functional assay. The effect of
DCP or DCX analogs on efflux of the P-gp substrate Doxorubicin
(Dox) was determined using an established fluorometric cell-based
assay. In this assay, inhibition of P-gp leads to the intracellular
accumulation of Dox, a fluorescent anthracycline cancer drug,
which serves as a quantitative measurement of P-gp inhibi-
tion.14–16 Three compounds with the most active chemosensitiza-
tion action (3, 4, and 7 in Table 1) were tested for their ability to
facilitate Dox influx in comparison to the positive control inhibitor,
verapamil. The results are shown in Fig. 2. All three compounds in-
duced significant intracellular accumulation of Dox under co-treat-
ment conditions, and all were more active than the positive control
compound, verapamil. The results suggest that chemosensitization
by DCP and DCX analogs can be attributed in part to the inhibition
of P-gp efflux activity in the KB-MDR-1 cell system. Further work is
needed to determine whether P-gp inhibition is the sole mecha-
nism of chemosensitization by these novel anti-HIV compounds.
In summary, nineteen pyranochromone derivatives (DCPs and
DCXs) with selective anti-HIV activity were tested using a pheno-
typic assay for chemosensitization activity. Most derivatives eval-
uated were better chemosensitizers than the control agent
verapamil. Seven of them (3, 5, 8, 11, 12, 17 and 19) fully reversed
VCR resistance under the co-treatment condition used. In general,
the SAR profiles of chemosensitization and anti-HIV activities were
not concordant but five compounds, 3, 4, 7, 11 and 17, showed sig-
nificant dual pharmacologic actions. Using a P-gp functional assay,
the inhibition of cancer drug efflux was demonstrated in co-treated
cells, suggesting a likely mechanism for the chemosensitization ac-
tion involves P-gp inhibition. Taken together, our results show that
several pyranochromone derivatives possessing potent anti-HIV
activity also exert substantial chemosensitization activity via P-
gp inhibition. In principle, this unique, dual-function behavior
would make these ideal candidates for exploring new combination
(HAART) therapies since they are a novel mechanistic class of re-
verse transcriptase inhibitor (RTI) that could sensitize and increase
efficacy of other anti-HIV drugs that are P-gp substrates. Work is
ongoing to better define the mechanism of action of these pyra-
nochromone derivatives and to establish pre-clinical anti-HIV
activity in combination regimens.
VCR at 4 lM. A possible reason is as follows: if the oxo group in
the B ring functions as an important H-bonding acceptor for the
interaction with the target protein, the R7-OH group may destroy
the interaction by forming intramolecular H-bonding. Compounds
8–18 contain diverse substitutions around the A-ring. Generally, an
alkyl, alkoxy, or fluoride substituent at R3 or R6 sustained or en-
hanced the chemosensitization activity compared with 5 and 7.
Compounds 8, 11, 12, and 16 fully reversed VCR resistance with
co-treatment GI50 values around 2–3 nM. Compounds with R4 or
R5 substituents (9, 10, 13, 14, 15 and 18) showed no significant
chemosensitizion activity, although some of them were good HIV
inhibitors, suggesting that chemosensitizion activity is sensitive
to the substitutions at R4 or R5. Extension/modification on the ring
system horizontally or increasing the molecular hindrance was
unfavorable for the interaction of the drug molecules with the tar-
get protein(s), such as P-gp. Compound 17, with a fluoro moiety at
R5, demonstrated a comparable chemosensitizion activity to the
non-F-substituted analog 7 and its positional isomer 16. This result
could be attributed to the isosteric relationship of F vs H. It is
Acknowledgements
This investigation was supported by grant AI-33066 from the
National Institute of Allergy and Infectious Disease (NIAID)
awarded to K.H.L. This study was also supported in part by Taiwan
Figure 2. Dose-dependent Doxorubicin accumulation.17