Generation of alpelisib-resistant cells in breast cancer cell lines with PIK3CA mutations
To determine the ability of alpelisib to induce apoptosis in human breast cancer cells, we generated cell lines of alpelisib-resistant (BT474 and MDA-MB-361), which were treated with alpelisib. The alpelisib-resistant cell lines (BT474-AR and MDA-MB-361-AR) exhibited enhanced viability in the presence of 1 μM alpelisib (Figure 1A and S1A), and the dose-response curve showed a rightward shift (Figure 1B and S1B). Concurrently, cell cycle arrest due to alpelisib was attenuated in the chronically exposed cells; compared to the parental cells these cells has higher cells amount in the S phase (Figure 1C and S1C).
We next treated BT474-AR, MDA-MB-361-AR, and their parental cells with alpelisib for 24 hours and then detected the cleavage of the caspase. Our results showed that, compared to the control treatment, 1 μM alpelisib treatment efficiently enhanced the amount of annexin V-positive cells (Figure 1D and S1D). Furthermore, it can also raise the levels of cleaved forms of caspase-3 and caspase-8 (Figure 1E and S1E) in the drug-sensitive parental cells. However, alpelisib treatment didn’t rise the annexin V-positive cell percentage or induce the cleavage of caspase-8 and caspase-3 in the alpelisib-resistant cells. In summary, these data suggest that alpelisib treatment prompts apoptosis in drug-sensitive breast cancer cells with mutated PIK3CA.
Alpelisib modulates Bim and Mcl-1 levels and suppresses AKT-dependent Bim induction and Mcl-1 phosphorylation in breast cancer cell lines with mutated PIK3CA
For understanding alpelisib-mediated apoptotic induction of mutated PIK3CA in drug-sensitive breast cancer cell lines, we investigated whether alpelisib affects the protein level that regulates the Bcl-2 family proteins. Alpelisib treatment with dose of 10-1000 nmol/L didn’t adjust the expression levels of Bcl-2 and Bax in the parental cell lines. However, alpelisib treatment decreased Mcl-1 levels, increased BimEL expression, and reduced p-AKT levels in these cell lines (Figure 1F and S1F). Within 6 hours Mcl-1 reduced and elevation of Bim elevated. after alpelisib treatment (Figure 1G and S1G). Nevertheless, these alterations didn’t occur or minimally occurred in the resistant cell lines, even when they were exposed to 1 μM alpelisib (Figure 1H and S1H). Thus, alpelisib modulates Mcl-1 and Bim levels by suppressing AKT in drug-sensitive PIK3CA-mutant breast cancer cell lines.
FoxO3a mediates alpelisib-induced Bim expression in breast cancer cells with mutated PIK3CA
Next, we demonstrate that whether alpelisib affects levels of Mcl-1 and Bim and study the mechanism through which this occurs in drug-sensitive PIK3CA-mutant breast cancer cells. In both BT474 and MDA-MB-361 cells, alpelisib treatment substantially showed the induction of Bim mRNA expression compared with the control treatment (Figure 2A and S2A). Next, we determine the mechanism of Bim induction in response to alpelisib. Transcription factors includingE2F1(31), Egr-1(32) and c-Myc(33) are not involved due to unchanged expression following alpelisib treatment (Figure S2B). In a previous study, the induction of Bim could be regulated by the transcription factor FoxO3a in T and mast cells after cytokine deprivation. Interestingly, we found that siRNA depletion of FoxO3a in BT474 cells abrogated Bim induction by alpelisib, indicating that FoxO3a is needed for Bim induction after kinase inhibition (Figure 2B). Furthermore, the ChIP assay indicated that FoxO3a is directly recruited to the Bim promoter. The binding of FoxO3a to the Bim promoter was enhanced statistically following treatment with alpelisib (Figure 2C). Our findings also indicated that alpelisib prompted dephosphorylation of FoxO3a (Figure 2D), which stalls its consequent transactivation and nuclear translocation. Conversely, active AKT transfection constitutively induced phosphorylation of FoxO3a (Figure 2E), and inhibited Bim induction by alpelisib (Figure 2E). Consequently, AKT inhibition cause FoxO3a-mediated Bim induction following alpelisib treatment.
Alpelisib modulates Mcl-1 levels through AKT/GSK-3β-dependent degradation in drug-sensitive breast cancer cells with mutated PIK3CA
Next, we investigated the mechanism by which alpelisib induces Mcl-1 depletion. Our findings indicated that alpelisib did not affect the mRNA level of Mcl-1 (Figure 2F and S2C). We also found speedy degradation of Mcl-1 in alpelisib-treated BT474 or MDA-MB-361 cells than in cells mixed with DMSO (Figure 2G and S2D). Unswervingly, the occurrence of MG132, a proteasome inhibitor, enhanced Mcl-1 basal amount and inhibited the alpelisib-induced reduction in Mcl-1 (Figure 2H and S2E), indicating that the protein degradation depended on the ubiquitin/proteasome system.
In alpelisib-treated cells, phosphorylation of Mcl-1 at Ser159 was observed (Figure 2I). GSK3β phosphorylates Mcl-1 to promote its degradation and is inhibited by AKT through Ser9 phosphorylation. Our results showed that alpelisib prevented Ser9 phosphorylation of GSK3β through analyzing by upstream kinase signaling (Figure S2F). Treating cells with the SB216763(GSK3β inhibitor) inhibits alpelisib-induced phosphorylation and degradation of Mcl-1 (Figure 2J). These data indicate that alpelisib-mediated phosphorylation of Mcl-1 is dependent on AKT/GSK3β, which facilitates proteasomal degradation.
Modulation of Bim and Mcl-1 levels is an important mechanism responsible for alpelisib-mediated apoptotic induction in drug-sensitive breast cancer cells with mutated PIK3CA
In order to evaluate whether Bim rise can induce apoptosis by alpelisib, we utilized Bim siRNAs to knockdown Bim expression and then analyzed its effect upon alpelisib's capability to cause apoptosis in BT474. Bim siRNA effectively prevented the Bim expression triggered by alpelisib, as found through Western blotting (Figure 3A). Thereafter, we showed decreased amounts of cleaved caspase-3 in transfected cells with siRNA to Bim than that in transfected cells with siRNA control (Figure 3A). Following this finding, we also found considerably less apoptotic cells in BT474 cells transfected with Bim siRNA than in siRNA-transfected cells under regulation (Figure 3B). Consequently, Bim elevation blockade prevents the apoptosis that alpelisib induces. Consistent with our previous results, which reported that MDA-MB-361 cells in which Bim expression using Bim siRNA was knocked down (Figure S3A and S3B).
Next, we study if a reduction in Mcl-1 levels has a role in alpelisib-induced apoptosis in drug-sensitive breast cancer cells with mutated PIK3CA. For this, we expressed ectopic Mcl-1 (S121A/E125A/S159A/T163A; 4A) in both BT474 and MDA-MB-361 cell lines and then checked its effects on alpelisib-induced apoptosis. We found that alpelisib significantly induced caspase cleavage of -3 and -8 (Figure 3C and S3C) and enhanced the percentage of apoptotic cells (Figure 3D and S3D) in vector-control cell lines. However, it minimally or doesn’t affect BT474/Mcl-1 (4A) or MDA-MB-361/Mcl-1 (4A) cells. These findings reveal that the expression of Mcl-1 shields cancer cells from alpelisib-induced apoptosis.
AKT inhibitors restore Bim elevation and Mcl-1 reduction in alpelisib-resistant PIK3CA-mutant breast cancer cell lines
Present results indicate key roles for Mcl-1 and Bim modulation in controlling alpelisib-mediated apoptotic induction in PIK3CA-mutant breast cancer cell lines. Alpelisib 's inability to simulate these protein levels in BT474-AR and MDA-MB-361-AR cells to acquired alpelisib resistance prompted us to hypothesize whether there is a mechanism that boosts Bim levels with simultaneous Mcl-1 suppression to reduce the threshold for alpelisib-resistant cells to respond to alpelisib, leading to apoptotic cell death. Provided that AKT controls Bim induction and the phosphorylation and degradation of Mcl-1, we considered that AKT suppression contributes to Bim induction and lessening of Mcl-1 in PIK3CA-mutant alpelisib-resistant breast cancer cells. In the alpelisib-resistant cell lines, BT474-AR and MDA-MB-361-AR, alpelisib was not able to suppress AKT, increase Bim levels, and reduced Mcl-1 levels (Figure 4A-4D). These finding additionally verify the significance of Bim levels in defining cell sensitivity to alpelisib and subduing AKT signaling and modulating Mcl-1. Two different AKT inhibitors, MK-2206 and perifosine, individually or in combination with alpelisib, effectively suppressed AKT phosphorylation, increased Bim expression, and reduced Mcl-1 expression in the mentioned resistant cell lines (Figure 4A-4D). Our results conﬁrm that AKT inhibitors suppress AKT phosphorylation, reduce Mcl-1, and increase Bim in alpelisib-resistant breast cancer cell lines with mutated PIK3CA.
AKT inhibitors effectively overcome the acquired resistance of PIK3CA-mutant breast cancer cells to alpelisib
Next, we evaluate if AKT inhibition indeed restored the sensitivity of alpelisib-resistant breast cancer cells to alpelisib. MK-2206 and perifosine (AKT inhibitors) at the determined concentration ranges didn’t suppress the growth of BT474-AR cells, nor every week. As envisioned, this cell line was alpelisib-resistant. The combined effect of alpelisib for each of the AKT inhibitors however very efficiently prevents the development of BT474-AR cells (Figure 5A-5D and S4A-S4D). The CIs were well below 0.5, suggesting highly synergistic growth inhibitory effects. In long-term colony formation testing, we observed that the combining alpelisib with either of the two AKT inhibitors abolished colony formation, while either agent itself inhibited the development of BT474-AR cells in colony mildly. (Figure 5E-5F and S4E-S4H). Taken together, these data suggest that AKT inhibitors restore the sensitivity of alpelisib-resistant cells to alpelisib to efficiently stunned alpelisib resistance.
Alpelisib in combination with an AKT inhibitor effectively induces the apoptosis of alpelisib-resistant breast cancer cells
Additionally, we evaluated the effects of alpelisib in combination with an AKT inhibitor on apoptotic induction in alpelisib-resistant breast cancer cells. Alpelisib or perifosine alone didn’t increase or tend to enhance the percentage of apoptotic cells (Figure 5G) and the cleavage of caspase-8 and caspase-3 (Figure 5H) in BT474-AR and MDA-MB-361-AR cells. Nevertheless, their combination improved the percentage of apoptotic cells (Figure 5G) and the cleavage of caspase-8 and caspase-3 (Figure 5H). The same results were observed with MK-2206 treatment (Figure 5I-5J). Taken together, these data show that a combination of alpelisib with an AKT inhibitor effectively induces apoptosis in alpelisib-resistant breast cancer cells.
Alpelisib in combination with perifosine effectively inhibits the growth of alpelisib-resistant breast cancer xenografts in a mouse model
After the promising therapeutic activity of alpelisib in combination with the AKT inhibitor in alpelisib-resistant breast cancer cells, next, we evaluate its efficiency in xenograft tumor mice model alpelisib-resistant. As shown in vitro, both alpelisib and perifosine alone had no impact or marginal effects on the development of BT474-AR tumors while treated regularly for approximately 3 weeks. On the other hand, when they were combined, they markedly eliminated the growth of these xenografts under the conditions tested (Figure 6A and 6B) without any toxicity determined from the mean body weight observed at an indicted time interval (Figure 6C). We have repeated the MDA-MB-361-AR xenograft experiment with an intermittent treatment schedule of 1 week on treatment, 1 week off treatment, 1 week on treatment. Similar results for possible increased toxicity were observed as shown in Figure S6A-S6C. It is, therefore, obvious that the combination therapy can successfully inhibit the growth of alpelisib-resistant tumors in vivo with no obvious toxicity.
In xenograft tumors of mice undergoing the specified therapies, we further measured p-AKT, Mcl-1, and Bim rates. In consistence with in vitro observations, perifosine alone significantly raises the levels of Bim and decrease levels of Mcl-1 and p-AKT, while alpelisib alone didn’t alter concentrations of the proteins. The combination of alpelisib and perifosine was not extra effective than perifosine alone at rising Bim levels but signiﬁcantly reduced Mcl-1 levels beyond the level observed with the single agent (Figure 6D). Furthermore, the alpelisib and perifosine combination induced more apoptosis in BT474-AR tumors than single agents alone (Figure 6E). Therefore, we found that in vivo regulation of Mcl-1 and Bim rates by the combination of alpelisib and the AKT inhibitor.