Drug resistance is a major challenge in cancer chemotherapy.1
The over-expression of anti-apoptotic Bcl-2 proteins which protect cells from apoptosis is one mechanism for tumors to acquire drug resistance.2-4
Inhibiting anti-apoptotic Bcl-2 proteins, therefore, is one promising strategy to nullify drug resistance in cancer treatment, which has been proved valid through anti-sense approach.5
Mechanistically, anti-apoptotic Bcl-2 proteins protect cells from apoptosis by antagonizing pro-apoptotic Bcl-2 proteins through dimerization.6,7
Structural studies of a complex of Bcl-XL
protein and a pro-apoptotic peptide (Bak BH3) have revealed a hydrophobic cleft on Bcl-XL
protein as the binding pocket for the pro-apoptotic peptide8
and molecules binding to that hydrophobic cleft may overcome the protective effect of the Bcl-XL
These observations have stimulated the recent discovery of small-molecule based antagonists - a few of which are currently at various stages of clinical evaluation, such as gossypol, apogossypol, TW-37, and ABT-737.10-14
As Bcl-2 proteins form homo/hetero oligomers for their physiological functions, multivalent modulators may potentially possess unique functions that are not present in the monomeric modulators.15
In addition, it is well established that multivalency of ligands could enhance potency compared to the monomeric ones.16,17
To test these concepts in the Bcl-2 protein family, we developed a series of dimeric analogs of 3e
based on BH3I-1, a small-molecule Bcl-2 antagonist developed by Degterev et al.18
Several dimeric modulators demonstrated enhanced potency to inhibit the binding of a BH3-domain Bak peptide binding to the anti-apoptotic Bcl-2 proteins while dimerization through the carboxylic acid group on 3e
resulted in substantial loss of the activity. Further SAR studies confirmed that the carboxylic acid was critical for the antagonism of the small molecule against the Bcl-2 proteins. The enhanced potency of the dimeric modulators was further confirmed by its ability to induce cytochrome c release from isolated mitochondia. The cytochrome c release study also suggested that the dimeric modulator functionally mimics the monomeric antagonist.
The preparation of the dimeric modulators (3e-D1 – D4
) followed the general synthetic route outlined in (The preparation of the dimeric modulators 3e-D5 – D7
followed a similar synthetic route detailed in Supplementary Materials
Briefly, the syntheses of 3e-D1 – D4
started with L-phenylalanine, which first reacted with carbon disulfide and cyclized upon chloroacetate treatment to form the rhodanine compound in 45% yield. The rhodanine compound (2.2 equivalent) was then condensed with various dimeric aldehydes in refluxing toluene to afford compound 3e-D1 – D4
in 76-92% yield. These four dimeric analogs were evaluated in the anti-apoptotic Bcl-2 proteins binding assays.19
The binding was assessed for their activity to prevent the binding of Bak peptide binding to anti-apoptotic Bcl-2, Bcl-XL
, and Bcl-w. All of the four dimeric modulators demonstrated improved potency to prevent the binding of Bak peptide to the three anti-apoptotic Bcl-2 proteins tested (). The best candidate, 3e-D2
demonstrated 5 – 16 fold increase of potency compared to the monomeric 3e
. The optimal linker between the monomers is a four-atom ethylene glycolic liner – increase or decrease of linker length would result in decreased activity.
Scheme 1 General synthetic route for the syntheses of the bottom-linked dimeric modulators. Reagents and conditions: (a) CS2, NaOH; ClCH2COONa; 5.5M HCl, reflux, overnight; (b) 4,4′-[methylenebis(oxy)]bis Benzaldehyde (0.45 equivalent), NH4OAc in Toluene, (more ...)
The inhibitory potency of BH3I-1 based modulators against Bak peptide binding to anti-apoptotic Bcl-2 proteins.
Delighted by these results, we further explored the other dimerization of the monomeric 3e
. 3e-D5 – D7
were prepared by using L-phenylalanine based rhodanine, which first reacted with various ethylene glycol to form the dimeric rhodanines in 34 – 81% yield (Supplementary Materials
). The dimeric rhodanines were then condensed with 5-bromobenzaldehyde (2.2 equivalent) for 3e-D5 – D7
(82 – 95%). These three dimers, however, all demonstrated substantial decrease of potency to prevent Bak peptide binding to anti-apoptotic Bcl-2 proteins (). One potential cause for the loss of activity would be due to the mask of the carboxylic acid functional group on 3e
. To test this hypothesis, we synthesized 3e ethyl ester
and 3e ethyl ester dimer
by following the same procedure for the preparation of 3e19
except that the ethyl ester of L-phenylalanine was used instead of L-phenylalanine. As hypothesized, 3e ethyl ester
and 3e ethyl ester dimer
all demonstrated substantially weakened ability to prevent Bak peptide binding to the anti-apoptotic Bcl-2 proteins relative to 3e
(> 5 fold loss) and 3e-1D
(> 15 fold loss) respectively. 3e ethyl ester dimer
still revealed better potency than the monomeric 3e ethyl ester
. These SAR studies demonstrate the carboxylic acid functional group on 3e
is critical for its interaction with anti-apoptotic Bcl-2 proteins. In fact, quite a few of the reported small-molecule antagonists against anti-apoptotic Bcl-2 proteins have acidic functional groups, such as gossypol, apogossypol, ABT-737, TW-37, and the recent flavonoid compound developed by Wang et al (The acidic protons are highlighted in ). Wang et al also observed that masking the acidic functional group resulted in dramatic loss of the activity against anti-apoptotic Bcl-2 proteins.20,21
Our study herein, in combination with the other studies, suggest that an acidic functional group may be optimal for small-molecule antagonists against anti-apoptotic Bcl-2 proteins.
Structures of some reported small molecule antagonists against anti-apoptotic Bcl-2 proteins (the acidic protons are highlighted).
To explore whether the monomeric modulator and dimeric modulator may function differently, we evaluated 3e and 3e-D2 for their ability to induce cytochrome c release from mitochondria, a key feature for mitochondrial-mediated apoptotic cascade. Both candidates induced cytochrome c release from the isolated mitochondria (). The respective concentration to induce 50% cytochrome c release for 3e was about 69 μM while that for 3e-D2 was ~ 6 μM, correlating with their potency to antagonize the anti-apoptotic Bcl-2 proteins.
The potency of 3e and 3e-D2 in releasing cytochrome c from isolated mitochondria.
In summary, dimerization of monomeric antagonist 3e could enhance the potency to bind the anti-apoptotic Bcl-2 proteins. The dimeric modulator also had enhanced activity to induce cytochrome c release from mitochondria. These studies demonstrated the dimerization is an effective approach to enhance the potency of Bcl-2 antagonists. In addition, our SAR studies in combination with others suggest that an acidic functional group is critical for the antagonism of the small molecules against anti-apoptotic Bcl-2 proteins. Our current studies also suggest that the dimeric modulators have the same function as the monomeric antagonists.