Targeting PP2A activates AMPK signaling to inhibit colorectal cancer cells

LB-100 is a novel PP2A inhibitor. Its activity in human colorectal cancer (CRC) cells was tested. The in vitro studies demonstrated that LB-100 inhibited survival and proliferation of both established CRC cells (HCT-116 and HT-29 lines) and primary human colon cancer cells. Further, LB-100 activated apoptosis and induced G1-S cell cycle arrest in CRC cells. LB-100 inhibited PP2A activity and activated AMPK signaling in CRC cells. AMPKα1 dominant negative mutation, shRNA-mediated knockdown or complete knockout (by CRISPR/Cas9 method) largely attenuated LB-100-induced AMPK activation and HCT-116 cytotoxicity. Notably, microRNA-17-92-mediated silence of PP2A (regulatory B subunit) also activated AMPK and induced HCT-116 cell death. Such effects were again largely attenuated by AMPKα mutation, silence or complete knockout. In vivo studies showed that intraperitoneal injection of LB-100 inhibited HCT-116 xenograft growth in nude mice. Its anti-tumor activity was largely compromised against HCT-116 tumors-derived from AMPKα1-knockout cells. We conclude that targeting PP2A by LB-100 and microRNA-17-92 activates AMPK signaling to inhibit CRC cells.


LB-100 inhibits survival and proliferation of CRC cells
In order to test the potential activity of LB-100 on human CRC cells, the established HCT-116 HCC cells [5] were cultured in complete medium, and were treated with LB-100 at different concentrations. Simple cell counting assay results in Figure 1A displayed that treatment with LB-100, at 2.5 μM and 10 μM, significantly inhibited HCT-116 cell proliferation. The number of HCT-116 cells was decreased following LB-100 treatment ( Figure  1A). LB-100 at 10 μM was more potent than 2.5 μM in inhibiting HCT-116 cell proliferation, displaying a concentration-dependent response ( Figure 1A). BrdU incorporation is a well-utilized marker of cell proliferation. Results in Figure 1B confirmed that LB-100 treatment (2.5 μM and 10 μM) significantly decreased BrdU ELISA (enzyme-linked immunosorbent assay) optic density (OD) in HCT-116 cells, confirming its anti-proliferative activity. To test cell viability, the routine MTT tetrazolium assay was performed. As demonstrated, LB-100 treatment (2.5 μM and 10 μM, 96 hours) largely inhibited survival ("MTT OD") of HCT-116 cells ( Figure 1C). Additionally, the number of viable HCT-116 cell colonies was also significantly decreased following LB-100 treatment (2.5 μM and 10 μM, renewed every two days) ( Figure 1D). Notably, the anti-survival activity by LB-100 was also concentration-dependent. At a relative low concentration (0.5 μM), LB-100 failed to decrease HCT-116 cell survival ( Figure 1C and 1D).
We also examined the potential effect of this novel PP2A inhibitor [13,30,31] in other CRC cells: including the established HT-29 cells and two lines of primary human colon cancer cells ("Con Can1/2"). BrdU ELISA proliferation assay results ( Figure 1E) and MTT viability assay results ( Figure 1F) displayed that treatment with LB-100 (10 μM) dramatically inhibited survival and proliferation of the tested CRC cells, as the BrdU ELISA OD and MTT OD were both decreased following the LB-100 treatment ( Figure 1E and 1F). Intriguingly, the very same LB-100 (10 μM) treatment failed to inhibit proliferation and survival of the primary human colon epithelial cells ("Epi") ( Figure  1E and 1F). BrdU ELISA OD and MTT OD were not significantly decreased after LB-100 treatment ( Figure  1E and 1F). Collectively, the results indicate that LB-100 inhibits survival and proliferation of human CRC cells.

LB-100 activates apoptosis in CRC cells
We next studied the potential activity of LB-100 on cell apoptosis. Various apoptosis assays were performed, including the DNA end labeling (TUNEL) assay, Annexin V FACS assay and the caspase-3 activity assay. As displayed, treatment with LB-100, at 2.5 μM and 10 μM, in HCT-116 cells significantly increased the percentage of TUNEL-stained cells (Figure 2A), Annexin V ratio ( Figure 2B) and the caspase-3 activity ( Figure 2C). These results indicated profound apoptosis activation in LB-100treated cells (Figure 2A-2C). LB-100-induced apoptosis was again concentration-dependent ( Figure 2A-2C), and it was yet in-effective when applied at a lower concentration (0. 5 μM, Figure 2A-2C). In order to block cell apoptosis, the caspase-based inhibitors were employed. As displayed, co-treatment with the caspase-3 inhibitor zDEVDfmk or the pan caspase inhibitor zVADfmk largely attenuated LB-100-induced viability (MTT OD) reduction in HCT-116 cells ( Figure 2D). The two caspase inhibitors almost blocked LB-100-induced apoptosis activation (Data not shown). These results suggest that LB-100 activates caspase-dependent apoptosis to kill the CRC cells. It should be noted that significant apoptosis activation (TUNEL assay) was also observed in LB-100 (10 μM)treated HT-29 cells and the primary human colon cancer cells ( Figure 2E). TUNEL ratio was yet unchanged in the colon epithelial cells with same LB-100 treatment, further suggesting its selective response in cancerous cells ( Figure  2E). These results demonstrate that LB-100 activates apoptosis in CRC cells.

LB-100 induces G1-S arrest in CRC cells
The activity of LB-100 on cell cycle progression was also tested. HCT-116 cells were treated with LB-100 (10 μM), cells were further cultured for additional 36 hours, and FACS assay was employed to test cell cycle distribution. The quantified results in Figure 3A demonstrated that, following the LB-100 treatment, the percentage of G1-phase cells was increased, but the percentages S-phase and G2-phase cells were both decreased. These results indicate G1-S arrest by LB-100 treatment in HCT-116 cells, which should favor proliferation-inhibition and apoptosis. The very similar G1-S arrest result was also obtained in the LB-100-treated primary human colon cancer cells ( Figure 3B). Thus, LB-100 induces G1-S arrest in CRC cells.

LB-100 administration activates AMPK signaling and inhibits HCT-116 tumor growth in nude mice
The LB-100's activity on CRC growth in vivo was tested. HCT-116 cells, with or without AMPKα1, were injected s.c. to the nude mice. HCT-116 xenografts were   Figure 6A displayed that daily intraperitoneally (i.p.) administration of LB-100 (5 mg/kg body weight) potently inhibited growth of HCT-116 tumors in the mice. The volume of LB-100-treated tumors was markedly lower than that of the vehicle control mice ( Figure 6A). The estimated daily tumor growth (in mm 3 per day), which was calculated by (tumor volume at day-35-tumor volume at day-0)/35, was also significantly decreased by LB-100 administration ( Figure 6B). Additionally, the average tumor weight at day-35 was dramatically lower in the LB-100-treated group ( Figure 6C). These results indicated that LB-100 i.p. administration inhibited HCT-116 tumor growth in nude mice. The mice body weights, which

DISCUSSION
AMPKα1 phosphorylation at Thr-172 is vital for AMPK activation [20,47,48]. Studies have been focusing on the upstream mechanisms of phosphorylation of AMPKα1 [26]. A number of potential AMPKα1 kinases have been identified. These kinases, including LKB1 (Liver kinase B1) [48] and CaMKK (calcium/ calmodulin-dependent protein kinase kinase) [49], directly phosphorylate AMPKα1 at Thr-172, leading to profound AMPK activation. On the other hand, the mechanisms of AMPKα1 de-phosphorylation are largely unknown. Several potential AMPKα1 phosphatase were proposed [50]. One key AMPKα1 phosphatase is PP2A [15,37]. In the current study, we show that targeted inhibition of PP2A by LB-100 activated AMPK signaling in CRC cells. Notably, AMPKα in-activation, silence or complete depletion via the described methods didn't result in total abolition of LB-100-induced cytotoxicity against HCT-116 cells (See Figure 5). These results suggest that other signalings, besides AMPK activation, could also be responsible for LB-100's actions in CRC cells. These results are not surprising, as PP2A could de-phosphorylate substrate kinases other than AMPK [10,51]. It should also be noted that miR-17-92-mediated silence of PP2A also activated AMPK and induced HCT-116 cell death. Such effects were again largely attenuated by AMPKα mutation, silence or complete knockout.
Dysregulation of mTOR signaling has been detected in CRC tissues [52,53]. Sustained and over-activation of mTOR in CRC is positively involved in a number of key oncogenic behaviors, including uncontrolled cancer cell survival, proliferation and migration, as well as chemoresistance and angiogenesis [52,53]. Thus, mTOR overactivation represents an important oncotarget of CRC treatment [3,4]. AMPK activation shall lead to mTORC1 in-activation, thus inhibiting CRC and other cancer cells. Tuberous sclerosis 2 (TSC2) phosphorylation and activation by activated AMPK would lead to downstream mTORC1 in-activation [54][55][56]. Meanwhile, AMPK is also shown to directly phosphorylate and in-activate Raptor, a key mTORC1 component, to shut down mTORC1 activation [57,58].
In the current study, we show that AMPK activation by LB-100 also inhibited mTORC1 activation in CRC cells. Phosphorylations of mTORC1 substrates, including p70S6K1 and 4E-BP1, were dramatically inhibited in LB-100-treated cells. Notably, LB-100-induced mTORC1 inhibition was almost completely reversed with AMPKα1 mutation, shRNA knockdown or complete knockout. Thus, we propose that blockage of PP2A by LB-100 activates AMPK to inhibit mTORC1 activation in CRC cells. This mechanism could a primary reason of CRC cell inhibition. These results might also explain the in-effectiveness of LB-100 against human colon epithelial cells, as the normal cells displayed extremely low level of mTORC1 activation [59].

Reagents
LB-100 was provided by Selleck (Beijing, China). The caspase-3 inhibitor zDEVDfmk and the pan caspase inhibitor zVADfmk were purchased from Sigma (Nanjing, China). All the antibodies of this study were purchased from Cell Signaling Technology (Beverly, MA). The tissue-culture reagents were provide by Gibco (Nanjing, China).

Culture of established cell lines
The established human CRC cell lines, HCT-116 and HT-29, were purchased from the iBS Cell Bank of Fudan University (Shanghai, China). Cells were cultured in DMEM medium, plus 6% FBS. DNA fingerprinting and profiling were performed to distinguish the cell line from possible cross-contamination.

Primary human colon cancer cells and colon epithelial cells
The protocols of isolation and culture of human colon cancer cells as well as the primary colon epithelial cells were described in detail previously [60,61]. In brief, the colon cancer issues and the surrounding normal colon epithelial tissues from patients (two male, 49 and 59 years old) were obtained, washed, and digested. The single cell suspensions were pelleted, and resuspended in the described medium plus growth factors for the primary human cells [61]. A total of two primary human colon cancer cell lines and one colon epithelial cell line were established. Written-informed consent was obtained from each patient. The protocols were approved by the IRB (Institutional Review Board) and Ethics Committee of all authors' institutions, and experiments were conducted according to the principles of Declaration of Helsinki.

Cell viability assay
The cell survival was always tested by the 3-[4,5-dimethylthylthiazol-2-yl]-2,5 diphenyltetrazolium bromide (MTT) (Sigma, Nanjing, China) assay according to the attached protocol [61]. The MTT OD of treatment cells was always expressed as the percentage of that of untreated control cells.

Colony formation assay
HCT-116 cells were initially plated at 5 × 10 3 cells per 10-cm tissue culture dish. LB-100-containing medium was renewed every two days. After 10 days of incubation, the remaining HCT-116 colonies were stained and manually counted.

Cell proliferation assay
The simple cell counting assay and BrdU incorporation ELISA assay were performed to test cell proliferation. The detailed protocols were described previously [4,62].

Apoptosis assays
Other apoptosis assays, including TUNEL assay and caspase-3 activity, were performed as previously described [4].

Cell cycle analysis
Following the applied LB-100 treatment, CRC cells were labeled with 10 μM EdU (Biyuntian, Nanjing, China) for 1 hour and were then fixed. Cells were then analyzed by flow cytometry (Beckman Coulter, Shanghai, China). Cell cycle distribution was analyzed with Kaluza software (Beckman Coulter).

Western blotting assay
Cells and tumor tissues were lysed by the commercial available RIPA lysis buffer (Biyuntian) plus Complete Protease Inhibitor Cocktail (Roche) and phosphatase inhibitors (Roche). Quantified lysate proteins (40 μg of each condition) were separated by the 7.5-10 % SDS-PAGE gels (Biyuntian), and were transferred onto the polyvinylidene fluoride (PVDF) membrane. The blot was then blocked, and were probed overnight with designated primary antibody. The corresponding horseradish peroxidase (HRP)-conjugated secondary antibody was then added. Super-Signal West Pico Chemiluminescent Substrates (Thermo Scientific, Shanghai, China) were added to visualize the targeted protein band (based on the molecular weight), under the X-ray film development.

PP2A phosphatase activity assay
After the designated LB-100 treatment, CRC cells were washed and lysed in the described RIPA lysis buffer. The supernatants of 50 μg of total cellular proteins were assayed with the PP2A Phosphatase Assay Kit (Millipore) by the attached protocol. The PP2A activity data were presented as percentage of relative PP2A activity compared with control.

AMPK activity assay
Following the LB-100 treatment, cell lysates were achieved. AMPK was firstly immunoprecipitated with anti-pan-AMPKα antibody. The AMPK activity was determined in kinase assay buffer (see previous study [64]) plus AMP-[γ-32 P] ATP mixture, and SAMS peptide (HMRSAMSGLHLVKRR) [64]. The reaction was terminated by spotting the reaction mixture on phosphocellulose paper (P81), which was extensively washed with 150 mM of phosphoric acid. The radioactivity was measured with scintillation counter. AMPK activity in the treatment group was always normalized to that of the untreated control group.

AMPKα1 shRNA
The lentiviral particles, with human AMPKα1 shRNA or scramble non-sense control shRNA, were obtained from Santa Cruz Biotech (Nanjing, China). The lentivirus was added to HCT-116 cells for 24 hours. Puromycin (3.0 μg/mL, Sigma) was then added, which helped to select the stable cells.

CRISPR/Cas9-mediated AMPKα1 knockout
The small guide RNA (sgRNA) targeting human AMPKα1 was through the Optimized Crispr Design application from Dr. Zhang's lab (http://crispr.mit. edu/), which was inserted into the lentiCRISPR plasmid (Addgene plasmid 49535, Shanghai, China). The plasmid was transfected to HCT-116 cells using the described protocol [46]. AMPKα1 expression in the stable cells was confirmed by Western blotting assay.

Xenograft tumor assay
BALB/c nude mice (17.8-18.5 grams, 4-5 week old, all female) were injected subcutaneously (s.c.) in the right flanks with 1 × 10 6 HCT-116 cells (per mouse). Animals were housed in temperature-and humidity-controlled cages, with free access to water and rodent food on a 12-h light/dark cycle. After a tumor volume of 80 mm 3 was reached, LB-100 were injected intraperitoneally (i.p.) daily for a total of 21 days. Control mice were injected with equal quantity of vehicle. Tumor volumes were recorded weekly, calculated via the following formula: π/6 × larger diameter × (smaller diameter) 2 . All animal studies were performed in accordance with the standards of ERB and IACUC of all authors' institutions.

Statistical analysis
The results were expressed as the mean ± standard deviation (SD). Ordinary one-way ANOVA test was employed for comparison between groups. p < 0.05 was considered as statistically significant.

CONCLUSION
The previous cancer studies have suggested that PP2A inhibition is likely to be most effective for cancer therapy when combined with traditional cytotoxic agents [14,31,32]. The results of this study show that PP2A inhibition by LB-100 or miR-17-92 may have significant anti-CRC cell activity in vitro and in vivo. LB-100 or miR-17-92 could be further tested as promising anti-CRC agents.

Author contributions
All authors carried out the experiments, participated in the design of the study and performed the statistical analysis, participated in its design and coordination and helped to draft the manuscript.

CONFLICTS OF INTEREST
The listed authors have no conflicts of interest.