Enhancing Mammalian Target of Rapamycin (mTOR)–Targeted Cancer Therapy by Preventing mTOR/Raptor Inhibition-Initiated, mTOR/Rictor-Independent Akt Activation

The mammalian target of rapamycin (mTOR), a phosphatidylinositol 3-kinase (PI3K)-related serine/threonine kinase, plays a central role in regulating cell growth, proliferation, and survival, in part by regulation of translation initiation, through interactions with other proteins such as raptor [forming mTOR complex 1 (mTORC1)] and rictor [forming mTOR complex 2 (mTORC2); refs. 1–3]. The best-characterized downstream effectors of mTORC1 are the 70-kDa ribosomal S6 kinase (p70S6K) and the eukaryotic translation initiation factor 4E–binding protein 1 (4E-BP1; ref. 1). In response to mitogenic stimuli or nutrient availability, mTORC1 is activated (4), leading to phosphorylation of p70S6K and 4E-BP1, and the subsequent enhanced translation of mRNAs that are critical for cell cycle progression and proliferation (1).

PI3K/Akt signaling represents a major cell survival pathway. Its activation has long been associated with malignant transformation and apoptotic resistance (5, 6). It is generally thought that mTOR (i.e., mTORC1) functions downstream of the PI3K/Akt pathway and is phosphorylated (or activated) in response to stimuli that activate the PI3K/Akt pathway (1, 7). However, the recent discovery of mTORC2 as an Akt Ser⁴⁷³ kinase also places mTOR upstream of Akt. Although mTORC2 is thought to be insensitive to rapamycin, it has been shown that prolonged rapamycin exposure inhibits mTORC2 assembly and Akt activity in certain types of cancer cells (8). We and others have shown that mTOR inhibitors activate Akt while suppressing mTORC1 signaling in different types of cancer cell lines and clinical human tumor samples (9–11). Currently, it is unclear how mTOR inhibitors activate Akt survival signaling.

mTOR signaling has recently emerged as an attractive therapeutic target for cancer therapy (1, 12). The potential applications of mTOR inhibitors for treating various types of cancer have been actively studied both preclinically and clinically. In the United States, several phase II or III trials are ongoing that test the effects of mTOR inhibitors on various cancers (1, 13, 14). A recent study has shown encouraging results that the mTOR inhibitor CCI-779 improved overall survival among patients with metastatic renal cell carcinoma (15).

In addition to the intrinsic resistance of cancer cells to mTOR inhibition by rapamycin, cancer cells can acquire resistance to rapamycin (16). Therefore, understanding the mechanisms by which cells become resistant to mTOR inhibitors such as rapamycin has long been an interesting subject and may eventually guide the development of successful mTOR-targeted cancer therapy by avoiding or overcoming cell resistance to mTOR inhibition.

The current study aimed at showing the relationship between mTORC2 and mTORC1 inhibition-induced Akt activation, and particularly the biological significance of Akt activation in mTOR-targeted cancer therapy.

Reagents. Rapamycin and LY294002 were purchased from LC Laboratories. RAD001 powder, formulated RAD001, and matched placebo control were provided by Novartis Pharmaceuticals Corp. The powders of these agents were dissolved in DMSO at a concentration of 20 mmol/L, and aliquots were stored at −80°C. Stock solutions were diluted to the desired final concentrations with growth medium just before use. Rabbit polyclonal antibodies against Akt, mTOR, raptor, phosphorylated Akt (p-Akt; S473), p-Akt (T308), phosphorylated p70S6K (p-p70S6K; T389), and phosphorylated S6 (p-S6; S235/S236), respectively, were purchased from Cell Signaling Technology, Inc. Rabbit polyclonal anti-actin and mouse monoclonal anti-tubulin antibodies were purchased from Sigma Chemical Co. Rabbit polyclonal anti–glyceraldehyde-3-phosphate dehydrogenase antibody was purchased from Trevigen, Inc. Goat polyclonal mTOR (FRAP; N-19) and rabbit polyclonal rictor (BL2178) antibodies were purchased from Santa Cruz Biotechnology, Inc. and Bethyl Laboratories, Inc., respectively.

Cell lines and cell treatment. Human lung cancer cell lines and PC-3 cells were purchased from the American Type Culture Collection. The U937 cell line was provided by Dr. Yongkui Jing (Mount Sinai School of Medicine, New York, NY). These cell lines were grown as described previously (9). The rapamycin-resistant A549 cell line (A549-RR) was established by exposing the rapamycin-sensitive A549 parental cells (A549-P) to gradually increased concentrations of rapamycin from the initial 1 nmol/L to the final 20 μmol/L over a 6-mo period. A549-RR cells were routinely cultured in complete medium containing 1 μmol/L rapamycin. All treatments with mTOR inhibitors were done in the respective media containing 5% fetal bovine serum.

Western blot analysis. The procedures for preparation of whole-cell protein lysates and for Western blotting were described previously (9, 17).

Immunoprecipitation. mTOR complexes were immunoprecipitated with goat polyclonal mTOR (FRAP; N-19) antibody according to the same procedure described previously (18, 19). At the end, the samples containing an equal amount (20–50 μg) of whole-cell protein lysates and immunoprecipitates from 0.5 to 1 mg cell lysates captured with protein A-Sepharose were analyzed by Western blotting.

Gene knockdown by small interfering RNA. Control (nonsilencing) small interfering RNA (siRNA), raptor siRNA that targets 5′-AAGGCTAGTCTGTTTCGAAAT-3′, and rictor siRNAs that target 5′-AAGCAGCCTTGAACTGTTTAA-3′ (rictor-1) and 5′-AAACTTGTGAAGAATCGTATC-3′ (rictor-2), respectively, were described previously (19) and synthesized from Qiagen. The transfection of siRNA was conducted in a 12- or 24-well plate using Lipofectamine 2000 (Invitrogen) following the manufacturer's instructions. Forty-eight hours after transfection, the cells were treated with DMSO and rapamycin for the given times and then subjected to detection of the given proteins by Western blot analysis. In addition, we also used lentiviral raptor, rictor, and scramble short hairpin RNAs (shRNA) described previously (19), which were ordered from Addgene, Inc., and used for stably knocking down raptor and rictor as described previously (19).

Growth inhibition assay. Cell number was estimated by the sulforhodamine B (SRB) assay and the growth inhibition was calculated as previously described (20).

Colony formation assay. Cells (single-cell suspension) were plated in 12-well plates at a density of 250 per well. On the second day, cells were treated with the given agents. Every 3 d, the medium was replaced with fresh medium containing the corresponding concentrations of the agents. After a 10-d treatment, the medium was removed and cell colonies were stained with crystal violet (0.1% in 20% methanol) and counted. Pictures were also taken using a digital camera to record the result.

Cell cycle analysis. The procedure for analysis of cell cycle by flow cytometry was described previously (21).

Lung cancer xenografts and treatments. Animal experiments were approved by the Institutional Animal Care and Use Committee of Emory University. Four- to 6-wk-old (∼20 g of body weight) female athymic (nu/nu) mice were ordered from Taconic and housed under pathogen-free conditions in microisolator cages with laboratory chow and water ad libitum. A549 cells at 5 × 10⁶ in serum-free medium were injected s.c. into the flank region of nude mice. When tumors reached certain size ranges (100–500 mm³), the mice were randomized into four groups (n = 6/group) according to tumor volumes and body weights for the following treatments: vehicle control, formulated RAD001 (4 mg/kg/d, oral gavage) and LY294002 in DMSO (25 mg/kg/d, i.p.), and the combination of RAD001 and LY294002. Tumor volumes were measured using caliper measurements once every 2 d and calculated with the formula V = π(length × width²)/6. After a 14-d treatment, the mice were sacrificed with CO₂. The tumors were then removed, weighed, and frozen in liquid nitrogen or fixed with formalin. Certain portions of tumor tissues from each tumor were homogenized in protein lysis buffer for preparation of whole-cell protein lysates as described previously (20). Western blotting results were quantitated using Kodak Image Station 2000R (Eastman Kodak Co.).

Immunohistochemistry. Immunohistochemical analysis on formalin-fixed, paraffin-embedded nude mouse xenograft tissues was done using Cytomation EnVision+ Dual Link System-HPR (DAB+; Dako North America, Inc.) following the standard manufacturer's protocol. The primary antibody against p-Akt (S473) was purchased from R&D Systems and used as 1:500 dilution. The incubation time for the primary antibody was overnight at 4°C.

Statistical analysis. The statistical significance of differences in p-Akt and p-S6 levels and tumor sizes between two groups was analyzed with two-sided unpaired Student's t tests when the variances were equal or with Welch's corrected t test when the variances were not equal by use of GraphPad InStat 3 software (GraphPad Software). Data were examined as suggested by the same software to verify that the assumptions for use of the t tests held. Results were considered to be statistically significant at P < 0.05.

Effects of prolonged treatment with mTOR inhibitors on Akt phosphorylation are dose dependent. We and others previously showed that rapamycin induces a rapid and sustained increase in Akt phosphorylation in several types of cancer cells, including lung, breast, and prostate cancer cells (9, 10). However, two recent studies have shown that prolonged treatment with mTOR inhibitors decreases Akt phosphorylation in certain cancer cell lines (e.g., PC-3 and U937; refs. 8, 22). In this study, we further examined the effects of RAD001 in comparison with rapamycin on Akt phosphorylation in a group of lung cancer cell lines after a prolonged treatment. Both RAD001 and rapamycin at 10 nmol/L increased p-Akt levels while inhibiting p70S6K phosphorylation in all of the cell lines after a 24-h treatment (Fig. 1A). We also treated H157 and A549 lung cancer cells with 1 nmol/L RAD001 or rapamycin for a prolonged period of time from 24 to 96 h and then harvested the cells for analysis of Akt phosphorylation. As shown in Fig. 1B, p-Akt levels remained elevated at all the tested times over the prolonged period of time even when decreased p-p70S6K levels returned at 96 h (i.e., RAD001 at 96 h). This result clearly shows that mTOR inhibitors induce a sustained Akt activation in the tested cell lines. We noted that p-p70S6K levels recovered at 96 h after treatment with RAD001 but not with rapamycin (Fig. 1B). Because we treated cells only once, it is likely that rapamycin may have a longer half-life in cell culture than RAD001, resulting in better efficacy than RAD001 in inhibiting mTOR signaling.

Moreover, we examined the effects of prolonged treatment with rapamycin or RAD001 on Akt phosphorylation in two cell lines (i.e., PC-3 and U937), in which Akt phosphorylation was decreased by prolonged treatment with rapamycin (8), in a more detailed way. Previous studies used 100 nmol/L rapamycin (8) or >1,000 nmol/L CCI-779 (22), which decreased p-Akt levels after a 24-h treatment. In our study, we could repeat this result after both 24- and 48-h treatments with 100 nmol/L rapamycin in PC-3 cells. However, when the concentration of rapamycin was reduced to 1 nmol/L, we consistently observed an increase in Akt phosphorylation at both 24- and 48-h treatments. Similar results were also obtained from cells treated with RAD001 (Fig. 1C). In U937 cells, prolonged treatment with either 1 nmol/L rapamycin or RAD001 clearly increased the levels of p-Akt, although at 10 or 100 nmol/L they decreased p-Akt levels (Fig. 1D). Similar results with RAD001 were also observed in Jurkat cells (Fig. 1D). We noted that both rapamycin and RAD001 at 1 nmol/L sufficiently inhibited mTORC1 signaling evidenced by reduction of p-S6 or p-p70S6K levels (Fig. 1C and D). Thus, the effects of prolonged treatment with mTOR inhibitors on Akt phosphorylation are clearly dose dependent in these cell lines. We also noted that both rapamycin and RAD001 at 1 to 100 nmol/L increased Akt phosphorylation at Thr³⁰⁸ in a dose-dependent manner in PC-3 cells (see Supplementary Fig. S1), suggesting that mTOR inhibitors also activate phosphoinositide-dependent protein kinase 1 kinase. We noted that our data here on Akt phosphorylation at Thr³⁰⁸ by rapamycin or RAD001 in PC-3 cells are different from previous report that rapamycin at 100 nmol/L slightly decreased Akt phosphorylation at Thr³⁰⁸ after a 24-h treatment (8). The reason for this inconsistency is not clear but may be due to the different ways the cells were treated by us and other investigators.

Rapamycin increases Akt phosphorylation accompanied with inhibition of the assembly of mTORC2. We were interested in the effects of rapamycin on the assembly of mTORC2 under the conditions that Akt phosphorylation is increased. To this end, we immunoprecipitated mTOR complexes from rapamycin-treated cell lysates using a mTOR-specific antibody and then detected raptor and rictor, respectively, in these immunoprecipitates by Western blotting. In the tested cell lines exposed to 10 nmol/L rapamycin for 24 h, the amounts of raptor and particularly rictor in mTOR complexes were substantially reduced, indicating that both mTORC1 and mTORC2 were inhibited in cells exposed to rapamycin, although the levels of p-Akt remained elevated in these cell lines (Fig. 2A). Moreover, we detected mTORC2 in PC-3 cells after a prolonged treatment with rapamycin at either 1 nmol/L (which increases p-Akt levels) or 100 nmol/L (which decreases p-Akt levels) as we presented in Fig. 1C. Rapamycin at both 1 and 100 nmol/L effectively decreased the levels of rictor in mTOR complexes precipitated by a mTOR antibody (i.e., inhibition of mTORC2 assembly), albeit with differential effects on alteration of Akt phosphorylation. These results clearly indicate that rapamycin inhibits mTORC2 assembly regardless of its differential effects on regulation of Akt phosphorylation.

mTOR inhibitor-induced Akt activation is secondary to mTORC1 inhibition and cannot be abrogated by inhibition of mTORC2. To dissect the roles of mTORC1 and mTORC2 in mTOR inhibitor-induced Akt phosphorylation, we knocked down raptor and rictor expression, which would result in disruption of mTORC1 and mTORC2, respectively. In both Calu-1 and H157 cells, raptor knockdown alone increased p-Akt levels as did rapamycin without altering the levels of p-p70S6K (Fig. 2C,, lanes 3 and 11), indicating that disruption of mTORC1 activates Akt. On treatment with rapamycin, p-Akt levels were even further increased (lane 4 versus lane 3 and lane 12 versus lane 11), likely due to additional inhibition of the activity of the residual mTORC1. Silencing of rictor using two different siRNAs slightly decreased basal levels of p-Akt (lanes 5, 7, 13, and 15). However, rapamycin still increased p-Akt levels in these cells (lanes 5–8 and lanes 13–16). Similar results were also generated from H157 cells exposed to rapamycin for 24 h, in which raptor and rictor were stably silenced using lentiviral raptor and rictor shRNAs, respectively. Under such conditions, stable silencing of raptor did reduce basal levels of p-p70S6K (Fig. 2D). Collectively, these results indicate that rapamycin-mediated increase in Akt phosphorylation is secondary to mTORC1 inhibition independent of mTORC2. Because transient knockdown of raptor in our system did not apparently decrease p-p70S6K but substantially increased p-Akt levels, these results also suggest that p-Akt is more susceptible than p-p70S6K to modulation by mTOR inhibition, suggesting that mTOR inhibition-induced Akt phosphorylation is unlikely a secondary event to p70S6K inhibition.

The rapamycin-resistant cell line exhibits increased levels of p-Akt with disrupted mTORC2. To further show the effect of long-term mTOR inhibitor exposure on Akt activity, we established a rapamycin-resistant cell line named A549-RR by exposing rapamycin-sensitive A549 cells (A549-P) to gradually increased concentrations of rapamycin from the initial 1 nmol/L to the final 20 μmol/L over a 6-month period. A549-RR cells were resistant not only to rapamycin but also to RAD001 (Fig. 3A) and were at least 10,000-fold more resistant to either rapamycin or RAD001 than A549-P cells by comparing their IC₅₀s. The A549-RR cell line had a comparable growth rate to that of A549-P (Fig. 3B). To maintain the acquired resistance to rapamycin, we routinely cultured A549-RR cells in complete medium containing 1 μmol/L rapamycin. Twenty-four hours before each experiment, rapamycin was withdrawn from the medium. We observed that A549-RR cells had much higher basal levels of p-Akt than A549-P cells; these high levels of p-Akt were not increased further by either rapamycin or RAD001 (Fig. 3C). In A549-P cells, rapamycin at either 1 nmol/L or 1 μmol/L increased p-Akt levels. The total levels of Akt in both A549-P and A549-RR cell lines were not altered (Fig. 3C,, bottom). Both glycogen synthase kinase 3β (GSK3β) and FOXO3a are well-known substrates of Akt. The basal levels of phosphorylated GSK3β (p-GSK3β) but not phosphorylated FOXO3a were accordingly elevated in A549-RR cells compared with those in A549-P cells (Fig. 3C). We noted that p-p70S6K levels were not decreased by rapamycin or RAD001 in A549-RR cells, although the p-S6 levels were slightly decreased by high concentration (i.e., 1 μmol/L) of rapamycin or RAD001 (Fig. 3C). There results indicate that A549-RR cells lose responses to mTOR inhibitor-mediated inhibition of mTORC1-p70S6K signaling while exhibiting increased levels of p-Akt.

It has been suggested that down-regulation of 4E-BP1 is associated with rapamycin resistance (23). Therefore, we compared the levels of 4E-BP1 and its phosphorylation between A549-P and A549-RR cell lines. As presented in Fig. 3C, we did not find an obvious difference in basal levels of 4E-BP1 between A549-P and A549-RR cell lines. The expression levels of 4E-BP1 were not altered by mTOR inhibitors in both cell lines. We found that both cell lines had comparable levels of phosphorylated 4E-BP1 (p-4E-BP1). p-4E-BP1 levels were reduced by both low (1 nmol/L) and high (1 μmol/L) concentrations of rapamycin or RAD001 in A549-P cells but not in A549-RR cells, except for the high dose (1 μmol/L) of rapamycin. These results suggest that 4E-BP1 levels cannot account for cell resistance to mTOR inhibitors in our system.

Following these studies, we determined whether the assembly of mTOR complexes was altered in A549-RR cells. Therefore, we compared the levels of mTORC1 and mTORC2 between A549-P and A549-RR cells. The total levels of mTOR, raptor, and rictor in cell lysates were not altered in A549-RR cells; however, the amounts of raptor and rictor in mTOR complexes precipitated by a mTOR antibody were strikingly decreased (Fig. 3D), indicating that both mTORC1 and mTORC2 were inhibited in A549-RR cells. Under such circumstances, the levels of p-Akt (S473), p-Akt (T308), and p-GSK3β (S9) were elevated in cell lysates from A549-RR cells compared with those from A549-P cells (Fig. 3D), indicating that A549-RR cells have increased Akt activity, albeit with disrupted mTORC2.

Sustained Akt activation is associated with development of cell resistance to mTOR inhibitors. We were interested in the biological significance of sustained Akt activation in mTOR-targeted cancer therapy. To this end, we took advantage of the rapamycin-resistant cell line (i.e., A549-RR) that has elevated levels of p-Akt as described above. We first determined whether the acquired rapamycin resistance in A549-RR cells was reversible. To do so, we cultured A549-RR cells in rapamycin-free complete medium for up to 5 months and monitored cell responses to mTOR inhibitors and p-Akt levels at 1-month intervals. At 2 months after rapamycin withdrawal, the cell line, which was named A549-RR2W, was slightly more sensitive than A549-RR cells to either rapamycin or RAD001 (Fig. 4A). Even at 3 or 4 months after rapamycin withdrawal, the cells (i.e., A549-RR3W and A549-RR4W) were still partially resistant to mTOR inhibitors, although their sensitivities to rapamycin or RAD001 were increased compared with A549-RR2W cells (data not shown). After a 5-month withdrawal of rapamycin, the cell line, which was named A549-RR5W, was as sensitive as A549-P cells to both rapamycin and RAD001 (Fig. 4B), indicating a complete restoration of rapamycin sensitivity. Collectively, these results indicate that the acquired rapamycin resistance in A549 cells is reversible, although it sustains for over 5 months.

Accordingly, we examined basal p-Akt levels and their modulation by mTOR inhibitors in rapamycin-resistant cell lines during rapamycin withdrawal. After a 2-month withdrawal of rapamycin, we found that the basal levels of p-Akt in A549-RR2W cells were still much higher than that in A549-P cells and were only increased by high concentrations of rapamycin or RAD001 (i.e., 1 μmol/L; Fig. 4C). The basal levels of p-p70S6K in A549-RR2W and A549-P cells were comparable and could be effectively inhibited by both rapamycin and RAD001. Similarly, the p-S6 levels in A549-RR2W and A549-P cells were also comparable and inhibited by mTOR inhibitors (Fig. 4C). After 5-month withdrawal of rapamycin when cell sensitivity to rapamycin is fully restored, we noted that p-Akt levels in A549-RR5W cells were as low as those in A549-P cells (Fig. 4D). On treatment with rapamycin or RAD001, p-Akt levels were substantially increased in A549-RR5W cells as was observed in A549-P cells (Fig. 4D). As we already showed in A549-RR2W cells, p-p70S6K levels in A549-RR5W cells were comparable with those in A549-P cells and could be effectively decreased by rapamycin or RAD001 (Fig. 4D). Collectively, our results clearly indicate that sustained Akt activation (i.e., increase in p-Akt levels) during mTOR-targeted cancer therapy is associated with cell resistance to mTOR inhibitors.

To further show this association, we examined whether enforced reduction of p-Akt levels by Akt siRNA alter cell sensitivity to rapamycin. To this end, we decreased p-Akt levels by knocking down the levels of total Akt using Akt siRNA and then examined its effect on cell sensitivity to rapamycin. As presented in Supplementary Fig. S2, silencing of Akt by Akt siRNA substantially reduced the levels of p-Akt (Supplementary Fig. S2A). Accordingly, these cells were much more sensitive than control siRNA-transfected cells to rapamycin (Supplementary Fig. S2B), indicating that enforced reduction of p-Akt levels restores cell sensitivity to rapamycin. Thus, these results further support the notion that sustained increase in p-Akt levels is associated with the development of cell resistance to mTOR inhibitors.

The rapamycin-resistant cell line retains sensitivity to PI3K inhibitors. Because of the increased levels of p-Akt in A549-RR cells, we determined whether A549-RR cells were cross-resistant to PI3K inhibitors. A549-RR cells responded as well as A549-P cells to either LY294002 or wortmannin in terms of a 3-day monolayer culture assay (Supplementary Fig. S3A). By a long-term (10-day) colony formation assay, we found that LY294002 effectively inhibited the growth of both A549-P and A549-RR cells (Supplementary Fig. S3B). At the tested concentrations of up to 15 μmol/L, LY294002 failed to induce apoptosis in either A549-P or A549-RR cells by examining cell morphologic changes and analysis of sub-G₁ populations (data not shown). However, LY294002 induced G₁ arrest in both A549-P and A549-RR cells with comparable potencies (Supplementary Fig. S3C). Moreover, we compared the effects of LY294002 on p-p70S6K and p-Akt in A549-P and A549-RR cells and found that LY294002 effectively decreased the levels of not only p-p70S6K and p-S6 but also p-Akt in both cell lines, although A549-RR cells had very high basal levels of p-Akt (Supplementary Fig. S3D). Collectively, these results indicate that A549-RR cells do not exhibit cross-resistance to PI3K inhibitors.

Cotargeting mTOR and PI3K/Akt signaling augments inhibition of tumor growth. Given that sustained Akt activation is associated with development of cell resistance to mTOR inhibitors, whereas mTOR inhibitor-induced Akt activation was suggested to be PI3K dependent (9), it was plausible to speculate that blockage of mTOR inhibitor-induced Akt activation by a PI3K inhibitor would enhance the anticancer efficacy of the mTOR inhibitors and prevent development of cell resistance to mTOR inhibitors. Thus, we examined the effects of RAD001 combined with LY294002 on the growth of lung cancer cells in cell culture. The RAD001 and LY294002 combination exhibited growth-inhibitory effects that are greater than that caused by each single agent in a 3-day monolayer culture (Fig. 5A). In the long-term colony formation assay, we obtained similar results. This combination worked better than either single agent in decreasing colony size and number (Fig. 5B).

Furthermore, we tested the effects of the combination of RAD001 and LY294002 on the growth of lung cancer xenografts in nude mice. In agreement with the results in cell cultures, the combination of RAD001 and LY294002 exhibited a significantly greater effect than RAD001 or LY294002 alone in inhibiting the growth of A549 xenografts (P < 0.001; Fig. 5C). During the 2-week period of treatment, the tumor sizes in mice receiving both RAD001 and LY294002 were smaller in comparison with other groups receiving either vehicle or single-agent treatment (Fig. 5C), indicating an effective anticancer efficacy for the combination treatment. In a H460 xenograft model, we began treatments with relatively larger tumors (in average 300–400 mm³). Both RAD001 and LY294002 alone failed to achieve significant effects on inhibiting the growth of tumors; however, the combination of RAD001 and LY294002 significantly inhibited the growth of H460 xenografts compared with control (P < 0.05 or 0.01; Fig. 5D). Collectively, these results clearly show that cotargeting mTOR and PI3K/Akt signaling exhibits enhanced anticancer efficacy.

Cotargeting mTOR and PI3K/Akt signaling enhances inhibition of mTORC1 signaling while preventing Akt phosphorylation in vivo. We also determined whether continuous RAD001 treatment in cancer xenograft models led to an increase in Akt phosphorylation as we observed in cell cultures. By Western blot analysis, we detected p-Akt levels in tumors exposed to RAD001 for 14 days and found that p-Akt levels were significantly increased (P < 0.05) in the RAD001-treated group compared with the vehicle control group in both A549 and H460 xenografts (Fig. 6A). As expected, p-Akt levels in tumors exposed to the combination of RAD001 and LY294002 were not increased (Fig. 6A). Immunohistochemical analysis of p-Akt in H460 xenografts also showed that p-Akt levels were increased in RAD001-treated tumors but not in tumors exposed to the combination treatment of RAD001 and LY294002 (Fig. 6C). Thus, these results clearly indicate that continuous treatment of lung tumors with a mTOR inhibitor in nude mice leads to an increase in Akt phosphorylation and this increase can be abrogated by inclusion of a PI3K inhibitor.

Moreover, we determined whether the presence of LY294002 affected the inhibitory effect of RAD001 on mTORC1 signaling in tumor tissues. As presented in Fig. 6B, RAD001 alone significantly decreased the levels of p-S6 (P < 0.001), indicating that RAD001 indeed inhibits mTORC1 signaling; however, the presence of LY294002 further reduced the levels of p-S6, which were significantly lower than those in tumors exposed to RAD001 alone (P < 0.05 or 0.01). Thus, these results indicate that cotreatment of tumors with a mTOR inhibitor (e.g., RAD001) and a PI3K inhibitor (e.g., LY294002) not only blocks RAD001-induced Akt phosphorylation but also exhibits an enhanced effect on inhibiting mTORC1 signaling.

In the current study, we further showed that prolonged treatment with either rapamycin or RAD001 increased p-Akt levels in several human lung cancer cell lines (Fig. 1). A549-RR cells, which were routinely cultured in the presence of 1 μmol/L rapamycin, still exhibited increased levels of p-Akt compared with the parental A549 cells (Fig. 3). Moreover, we detected significantly increased levels of p-Akt in lung cancer xenografts exposed to RAD001 for 14 days (Fig. 6). In current studies, we used 1 or 10 nmol/L of rapamycin or RAD001, which is lower than concentrations (100 or >1,000 nmol/L) used in other studies, showing that prolonged treatment with a mTOR inhibitor decreases p-Akt levels (8, 22). At 100 nmol/L (or even at 10 nmol/L), both rapamycin and RAD001 indeed decreased p-Akt levels after a 24- or 48-h treatment in PC-3, U937, and Jurkat cells as reported (8, 22). However, both rapamycin and RAD001 at 1 nmol/L consistently increased p-Akt levels even after a 48-h exposure in these cell lines (Fig. 1). Thus, it seems that there are two types of cancer cells: one type exhibits increased levels of p-Akt after a prolonged treatment with a mTOR inhibitor regardless of concentrations (e.g., A549 and H157 cells), whereas another type shows dose-dependent alterations in p-Akt levels after prolonged treatment with a mTOR inhibitor (e.g., PC-3 and U937 cells). In the latter cell type, low doses (e.g., 1 nmol/L) of mTOR inhibitors, which sufficiently block mTORC1 signaling (9), clearly increase p-Akt levels.

It has been suggested that mTORC2 is rapamycin insensitive (18), although it can be inhibited by prolonged rapamycin treatment (8). It has been suggested that an equilibrium may exist between mTORC1 and mTORC2 complexes (7). Therefore, it is possible that inhibition of mTORC1 by a mTOR inhibitor somehow shifts the equilibrium to favor or facilitate formation and activation of mTORC2, leading to increase in Akt phosphorylation. In our study, we found that a prolonged treatment with rapamycin (i.e., 24 h) inhibited not only mTORC1 but also mTORC2 (e.g., H157 and A549) with increased Akt phosphorylation in all three lung cancer cell lines (Fig. 2A). In rapamycin-resistant A549-RR cells where p-Akt levels were increased, the assembly of both mTORC1 and mTORC2 was also clearly inhibited (Fig. 3D). Thus, our results clearly indicate that p-Akt levels can be increased under the condition that mTORC2 activity is inhibited.

Although mTORC2 has been recently shown to be an Akt Ser⁴⁷³ kinase (19), our results indicate that mTOR inhibitor-induced Akt phosphorylation is unlikely to be mediated by mTORC2 because it is inhibited during mTOR inhibitor treatment. This notion is further supported by our findings that disruption of mTORC2 by knocking down rictor did not block rapamycin-induced Akt phosphorylation (Fig. 2). In agreement with previous findings that raptor knockdown increases Akt phosphorylation (19), we also observed that inhibition of mTORC1 by silencing raptor was sufficient to increase Akt levels in our cell lines tested. These results indicate that mTOR inhibitor-induced Akt activation is the consequence of mTORC1 inhibition. Collectively, we conclude that mTOR inhibitors induce Akt activation through a mTORC1-dependent mechanism independent of mTORC2.

It is well documented that PI3K/Akt represents a major survival pathway that is often associated with resistance to cancer therapy (24–26). The biological significance of mTOR inhibitor-induced Akt activation in mTOR-targeted cancer therapy is unclear. In our study, we observed that p-Akt levels were drastically increased in the rapamycin-resistant cell line (A549-RR). Moreover, when the selective pressure (i.e., rapamycin) was removed, the acquired high levels of p-Akt remained for a long period of time and were tightly associated with cell resistance to mTOR inhibitors. When the sensitivity of rapamycin-resistant (A549-RR) cells to mTOR inhibitors was fully restored after a 5-month removal of rapamycin, p-Akt levels dropped to normal levels comparable with those in rapamycin-sensitive parental cells (A549-P; Fig. 4). Additionally, enforced reduced p-Akt levels by silencing total Akt levels with Akt siRNA increase cell sensitivity to rapamycin (see Supplementary Fig. S2). Thus, our results suggest a critical role of Akt activation in the development of cell resistance to mTOR inhibitors. Although we suggest the association between sustained Akt activation and development of acquired resistance to mTOR inhibitors, the mechanistic insights into how sustained Akt activation negatively regulates the efficacies of the mTOR inhibitors are still unclear and need further investigation.

PI3K/Akt works upstream of mTORC1 and regulates mTORC1 activity. Therefore, inhibition of PI3K/Akt signaling using PI3K inhibitors should affect mTORC1 activity as well. Moreover, mTOR is a PI3K-related serine/threonine kinase, and its activity can be directly inhibited by the PI3K inhibitors LY294002 and wortmannin (27, 28). Thus, it has been proposed that PI3K inhibitors may share similar signaling pathways with rapamycin, such as mTOR/p70S6K, to exert their biological function (27). If PI3K inhibitors suppress cell growth solely through inhibition of mTOR signaling, cells resistant to rapamycin should be cross-resistant to PI3K inhibitors as was seen with RAD001. In our study, LY294002 or wortmannin was equally effective in inhibiting the growth of A549-P and A549-RR cells. Moreover, LY294002 induced G₁ arrest in both A549-P and A549-RR cells with comparable potencies. We also found that LY294002 effectively decreased the levels of p-p70S6K, p-S6, and p-Akt in both A549-P and A549-RR cells (see Supplementary Fig. S3). Together, these results indicate that rapamycin resistance does not interfere with the action of PI3K inhibitors, suggesting that mTOR and PI3K inhibitors exert their biological functions through different mechanisms or PI3K inhibitors suppress cell growth through other mechanisms in addition to inhibition of mTOR signaling.

Rapamycin resistance is an important subject of mTOR-targeted cancer therapy in the clinic. Our finding that rapamycin-resistant cells retain sensitivity to PI3K inhibitors has important clinical implications. To overcome or avoid cell resistance to mTOR inhibitors during mTOR-targeted cancer therapy, combination of a mTOR inhibitor with a PI3K inhibitor or intermittent use of a PI3K inhibitor and a mTOR inhibitor may be good approaches. Indeed, our results clearly show that RAD001 combined with LY294002 exhibited enhanced inhibitory effects on the growth of human lung cancer cells in cell cultures (Fig. 5). Importantly, the RAD001 and LY294002 combination worked better than each single agent alone in inhibiting the growth of human lung cancer xenografts in nude mice (Fig. 5), indicating an enhanced anticancer activity in vivo. As expected, treatment of xenografts with RAD001 increased p-Akt levels, which could be abrogated by cotreatment with LY294002. Besides, we found that RAD001 plus LY294002 also exerted an enhanced effect on reduction of p-S6 levels, indicating that inhibition of PI3K/Akt enhances the effect of the mTOR inhibitor on inhibition of mTORC1 signaling (Fig. 6). Collectively, our results validate the strategy for cancer therapy by cotargeting mTOR and PI3K/Akt signaling and warrant clinical evaluation of this strategy for cancer therapy.

No potential conflicts of interest were disclosed.

Grant support: NIH grants RO1 CA118450-01 (S-Y. Sun) and PO1 CA116676-01 (Project 1 to F.R. Khuri and S-Y. Sun), Georgia Cancer Coalition Distinguished Cancer Scholar award (S-Y. Sun), Department of Defense IMPACT award W81XWH-05-0027 (Project 5 to F.R. Khuri and S-Y. Sun), and BATTLE award W81XWH-06-1-0303 (Project 4 to F.R. Khuri and S-Y. Sun). S-Y. Sun, H. Fu, and F.R. Khuri are Georgia Cancer Coalition Distinguished Cancer Scholars.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Dr. Yongkui Jing for kindly providing cell lines and Dr. H. Elrod in our laboratory for editing of the manuscript.