NPC-26 kills human colorectal cancer cells via activating AMPK signaling

NPC-26 is novel mitochondrion-interfering compound. The current study tested its potential effect against colorectal cancer (CRC) cells. We demonstrated that NPC-26 induced potent anti-proliferative and cytotoxic activities against CRC cell lines (HCT-116, DLD-1 and HT-29). Activation of AMP-activated protein kinase (AMPK) signaling mediated NPC-26-induced CRC cell death. AMPKα1 shRNA knockdown or dominant negative mutation abolished NPC-26-induced AMPK activation and subsequent CRC cell death. NPC-26 disrupted mitochondrial function, causing mitochondrial permeability transition pore (mPTP) opening and reactive oxygen species (ROS) production. ROS scavengers (NAC or MnTBAP) and mPTP blockers (cyclosporin A or sanglifehrin A) blocked NPC-26-induced AMPK activation and attenuated CRC cell death. Significantly, intraperitoneal injection of NPC-26 potently inhibited HCT-116 tumor growth in severe combined immuno-deficient (SCID) mice. Yet, its anti-tumor activity was significantly weakened against AMPKα1-silenced HCT-116 tumors. Together, we conclude that NPC-26 kills CRC cells possibly via activating AMPK signaling.


INTRODUCTION
Colorectal cancer (CRC) is still a major malignancy in the world, which causes significant mortality each year [1][2][3]. Mitochondrion is vital for regulating cell signaling and survival [4,5]. Recent studies have developed a small-molecule mitochondrion-interfering compound, named NPC-26 [6]). It disturbs normal mitochondrial functions, causing mitochondrial permeability transition pore (mPTP) opening and reactive oxygen species (ROS) production, and eventually leading to cell death [6,7]. NPC-26 induces a conversion from elongated to punctate mitochondria, and provokes non-apoptotic cell death which is BAX-/BAK-independent [6]. Further studies have proposed that NPC26-induced cell death is dependent on activation of kinase signaling pathways [6]. To our best knowledge, the potential effect of this compound in CRC cells has not been tested thus far. More importantly, the underlying signaling mechanisms of NPC-26-induced cell death are still vague. Here, we suggest that NPC-26 kills human CRC cells possibly via activating AMP-activated protein kinase (AMPK) signaling.

NPC-26 is cytotoxic and anti-proliferative to cultured CRC cells
First, the potential effect of NPC-26 on cultured CRC cells was tested. As demonstrated, treatment with NPC-26 (for 72 hours) in HCT-116 cells dose-dependently inhibited cell survival, which was tested by the CCK-8 OD reduction ( Figure 1A). NPC-26, at over 1 μM, significantly decreased CCK-8 OD of HCT-116 cells ( Figure 1A). NPC-26's IC-50, or the concentration that inhibited 50% of cell survival, was 7.31±0.55 μM ( Figure 1A). NPC-26 also displayed a time-dependent response in inhibiting HCT-116 cell survival ( Figure 1B). It would require 48 hours for 10 μM of NPC-26 to exert a significant effect ( Figure 1B). When analyzing cell death, we showed that the number of trypan blue positive cells ("dead" cells) was significantly increased following 1-30 μM of NPC-26 treatment ( Figure 1C). Thus, these results indicate that NPC-26 is cytotoxic to cultured HCT-116 cells.
We also tested the potential role of NPC-26 on CRC cell proliferation. BrdU ELISA assay was performed, and results showed that NPC-26 dose-dependently decreased the BrdU ELISA OD in HCT-116 cells ( Figure 1D), indicating its anti-proliferative activity. Note that for the BrdU assay, cells were treated with NPC-26 for only 24 hours, when no significant cytotoxicity was yet noticed ( Figure 1B). Further studies showed that NPC-26 at 1-30 μM dramatically decreased the number of proliferative HCT-116 colonies, again confirming its anti-proliferative activity ( Figure 1E). Therefore, NPC-26 is also antiproliferative to HCT-116 cells.
CCK-8 assay was again utilized to test the potential activity of NPC-26 to other CRC cells. Results showed that 10 μM of NPC-26 significantly inhibited survival of two other established CRC cell lines: HT-29 and DLD-1 ( Figure 1F). On the other hand, same NPC-26 treatment (10 μM, 72 hours) failed to affect the survival of two normal colon epithelial cell lines: FHC and CCD-841 ( Figure 1F). Thus, it appears that NPC-26 is only cytotoxic to the cancerous cells, which is proposed by other studies [7]. 5-Flurouracil (5-FU) in a very common chemotherapeutic drug for CRC treatment [26]. We showed that 10 μM of NPC-26 was more potent than same concentration of 5-FU in inhibiting HCT-116 cell survival ( Figure 1G) and inducing cell death ( Figure 1H). The IC-50 for 5-FU (72 hours treatment) was 33.21±2.45 μM ( Figure 1G), which was also significantly higher than that of NPC-26 (7.31±0.55 μM).

AMPKα1 mutation inhibits NPC-26-induced killing of HCT-116 cells
The above shRNA results imply that AMPK activation mediates NPC-26-induced cytotoxicity against HCT-116 cells. To further support this hypothesis, a dominant negative AMPKα1 ("dn-AMPKα1", T172D) construct [27,31,32] was introduced to the HCT-116 cells. Via puromycin selection, two stable HCT-116 cell lines with this construct were established ["dn-AMPKα1 (L1/2)"]. Western blot assay results in Figure 3A confirmed expression of dn-AMPKα1 (Flag-tagged) in the stable cells. Significantly, dn-AMPKα1 expression almost completely blocked NPC-26-induced AMPK activation ( Figure 3A). As a result, NPC-26-induced HCT-116 cell death was also attenuated ( Figure 3B). Interestingly, treatment with NPC-26 (10 μM, 4h) induced significant AMPK activation in HT-29 cells ( Figure 3C), but not in the FHC colon epithelial cells ( Figure 3D). As a matter of fact, expression of total AMPKα1 and ACC was also extremely low in the epithelial cells ( Figure 3C and 3D). These results could at least in part explain why these epithelial cells were not killed by NPC-26 ( Figure 1F). The above AMPKα1 shRNA and mutation experiments were also repeated in HT-29 cells, and similar results were obtained (Data not shown). Together, these results suggest that NPC-26-induced killing of CRC cells requires AMPK activation.

NPC-26 disrupts mitochondrial function, causing AMPK activation
A very recent study by Dong et al., [7] demonstrated that NPC-26 disrupts normal mitochondrial function, leading to mPTP opening and ROS production in pancreatic cancer cells. Here, NPC-26 treatment (10 μM) in HCT-116 cells similarly induced mitochondrial depolarization (JC-10 intensity increase, indicating mPTP opening [7]) ( Figure 4A) and significant ROS production ( Figure 4B). Thus, normal mitochondrial functions could also be disrupted by NPC-26 in HCT-116 cells. To study the link between mitochondrial dysfunction and AMPK activation in NPC-26-treated cells, pharmacological strategy was applied. As demonstrated in Figure 4C, NPC-26-induced AMPK activation was largely inhibited with co-treatment of ROS scavengers (NAC and MnTBAP [33]) or mPTP blocker cyclosporin A (CsA) [34] and sanglifehrin A (SfA) [35]. Remarkably, above inhibitors also alleviated NPC-26-induced killing of HCT-116 cells ( Figure 4D and 4E). Treatment with these inhibitors alone didn't affect AMPK activation and HCT-116 cell survival/ death (Data not shown). These results suggest that NPC-26 induces mitochondrial dysfunction, which possibly leads to AMPK activation and subsequent cell death.

NPC-26 inhibits HCT-116 tumor growth in SCID mice
In order to test the anti-tumor activity of NPC-26 in vivo, the severe combined immuno-deficient (SCID) mice bearing HCT-116 xenograft tumor model was applied. Results in Figure 5A demonstrated that i.p. injection of NPC-26 (25 mg/kg body weight, daily) [7] dramatically suppressed growth of HCT-116 tumors with scramble control shRNA ("sc-shRNA", or control tumors). Significantly, NPC-26-induced anti-tumor activity in vivo was remarkably weakened against HCT-116 tumors with AMPKα1 shRNA ("Seq-1") ( Figure 5A), indicating that AMPK activation could also be required for NPC-26's actions in vivo. Indeed, when analyzing tumor tissue samples, we found that NPC-26 administration (12 hours after initial administration) induced significant AMPK activation, or p-AMPK/p-ACC, in control tumors ( Figure  5B, left panel), which was absent in tumors expressing AMPKα1 shRNA ( Figure 5B, left panel). Tumors with AMPKα1 shRNA also showed extremely low expression of total AMPKα1 ( Figure 5B, left panel). Proliferating cell nuclear antigen (PCNA) expression is a well-established marker of proliferation. Cyclin D1 is important for cell proliferation [36,37]. Here, we showed that PCNA and Cyclin D1 were both downregulated in NPC-26-treated control tumors, but not in AMPKα1-silenced tumors ( Figure 5B, right panel). IHC staining assay further confirmed AMPK activation [p-AMPKα1 (Thr-172) staining] by NPC-26 in control tumors, but not in the AMPKα1 shRNA-expressing tumors ( Figure 5C). Notably, expression of AMPKα1 shRNA alone didn't affect HCT-116 tumor growth in SCID mice. Results in Figure 5D demonstrated that the above NPC-26 administration didn't affect the body weight of experimental mice, indicating that the regimens were relatively safe [7].
Under certain circumstances, it has yet been proposed that AMPK activation could also be pro-survival [41,42]. The difference might be due to the intensity of AMPK activation. Low level of AMPK activation might promote cell survival, but intensified AMPK activation could promote cell death via regulating its downstream signalings (p53, mTOR inhibition and autophagy etc.). In fact, the activity of AMPK could increase over 100fold on phosphorylation of a conserved threonine residue (Thr-172) within the activation loop at α1 subunit [38,43]. In the current study, we showed that NPC-26 induced significant AMPKα1 phosphorylation at Thr-172, indicating a profound AMPK activation. Remarkably, AMPKα1 shRNA knockdown or Thr-172 dominant negative mutation not only abolished NPC-26-induced AMPK activation, but also attenuated CRC cell death. Thus, AMPK activation by NPC-26 is indeed pro-death in CRC cells. Notably, NPC-26 was non-cytotoxic to normal colon epithelial cells, where AMPK was also not significantly provoked. Further studies showed that NPC-26 disrupted mitochondrial function, causing mPTP opening and ROS production, which served as the upstream signal for AMPK activation. Remarkably, ROS scavengers (NAC or MnTBAP) and mPTP blockers (CsA or SfA) almost completely blocked NPC-26-induced AMPK activation.
It should be noted that mitochondria in cancer cells are structurally and functionally different from those in normal ("epithelial") cells, which are often highly-active in malignant cells to participate in metabolic reprogramming and cell activities [4,5]. Intriguingly, existing literatures have also reported that certain key mPTP components are up-regulated in cancer cells. For example, the ATP synthase c subunit was upregulated in human breast cancer cells [44]. VDAC-1 over-expression was also observed in several cancer cells [7,45]. Unique upregulation of mPTP components and high mitochondrial activity in cancer cells could explain why only cancer cells, but not the epithelial cells, were killed by NPC-26 treatment. As a matter of fact, we found that NPC-26 failed to induce ROS production, AMPK activation and significant cytotoxicity in two normal colon epithelial cell lines (FHC and CCD-841).  scramble control shRNA ("sc-shRNA") or AMPKα1 shRNA ("Seq1") were inoculated to the SCID mice to establish xenografted tumors. Mice were administrated with NPC-26 (25 mg/kg, daily, i.p.) or saline ("Vehicle") for a total of 30 days, tumor volume A. and mice body weight D. were recorded every 5 days. Twelve hours after initial NPC-26 administration, one tumor of each group was isolated, and expression of listed proteins was tested by Western blot assay B. or IHC staining assay C. * p <0.05 vs. "Vehicle". # p <0.05 vs. NPC-26 treatment of "sc-shRNA" tumors. Bar=100 μm (C).
The selective cytotoxicity of NPC-26 to cancerous cells has been reported early as well [7].
Intriguingly, AMPK blockage, via AMPKα1 shRNA knockdown or Thr-172 dominant negative mutation, didn't completely abolished NPC-26-mediated killing of CRC cells (Figure 2 and 3). Meanwhile, ROS scavengers (NAC or MnTBAP) as well as mPTP blockers (CsA and SfA) only alleviated, but didn't abolish NPC-26's cytotoxicity (Figure 4). It is possible that these interfering strategies didn't result in complete inhibition of the targeted pathways (AMPK, ROS and mPTP). It is more likely that other signalings besides AMPK may also contribute to NPC-26's actions in CRC cells. Therefore, further studies will be needed to explore the relationship between AMPK and these other pathways in mediating NPC-26's actions in CRC cells. It will also be important to further characterize the underlying mechanism of NPC-26induced AMPK activation. In summary, we propose that NPC-26 kills CRC cells possibly via activating AMPK signaling. NPC-26 might have translational value for the treatment of CRC.

Cell culture
HCT-116, DLD-1 and HT-29 CRC cell lines were provided by Dr. Lu's group [27,46]. Cells were maintained in DMEM medium (with 10% FBS). Two normal colon epithelial cell lines, FHC and CCD-841, were purchased from the Cell Bank of Shanghai Institute of Biological Science (Shanghai, China). The above epithelial cells were also cultivated in DMEM medium.

Cell survival assay
To test cell survival, cell counting kit-8 (CCK-8, Dojindo Laboratories, Kumamoto, Japan) assay kit was performed based on the attached manual. CCK-8 absorbance optic density (OD) was recorded at 450 nm.

Trypan blue staining assay
After treatment, trypan blue (0.2%)-stained cells was counted by the Countess automatic cell counter (Invitrogen, Shanghai, China). The cell death percentage (%) was calculated by the number of the trypan blue cells divided by the total cell number.

Colony formation assay
HCT-116 cells with applied NPC-26 treatment were re-suspended in DMEM medium containing 0.5% agar (Sigma), which were then plated onto a pre-solidified six-well plate. Afterwards, cells were cultured in NPC-26-containing medium for 8 days. The remaining colonies were counted manually.

Western blot assay
Cell or HCT-116 tumor tissue samples were incubated in lysis buffer described [27,46]. Quantified protein lysates (30 μg/lane) were separated by 8-10% of SDS-PAGE gels, and were transferred onto polyvinylidene difluoride (PVDF) membranes. The membranes were then blocked with 10% milk, and incubated with designated primary and secondary antibodies. The blots were then subjected to ECL detection. Indicated protein band was quantified of total gray via ImageJ software (NIH).

ROS detection
Following treatment, cells were stained with CellRox Orange Reagent (5 μM, Invitrogen) at 37°C for 30 min. ROS content was detected by the above fluorescence microplate reader (Titertek Fluoroscan, Germany).

Mice xenograft assay
The protocols using 6-8 week-old severe combined immunodeficient (SCID) mice (weighting 18-19g) were approved by the IACUC of all authors' institutions. Exponentially growing HCT116 cells (5*10 6 cells per mouse), expressing scramble control shRNA or AMPKα1 shRNA ("Seq1"), were subcutaneously (s.c.) injected into SCID mice. Within three weeks, the xenografted tumors were established around 100 mm 3 in volume. The SCID mice were then treated as described. Tumor volumes, recorded every 5 days, were calculated using the formula: (mm 3 ) = (A 2 × B)/2: A and B were the shortest and the longest diameter, respectively. Mice body weights were also recorded.

Statistical analysis
All values were expressed as the mean ± standard deviation (SD). A p-value, calculated by ANOVA, of less than 0.05 was considered statistically significant.

ACKNOWLEDGMENTS
This work is supported by Shanghai medical key subject construction project (ZK2015B15), Shanghai weak discipline construction plan (2016ZB0202) and Shanghai Minhang District science and technology research plan (2015MHZ039). The funders have no role in design, in the collection, analysis, and interpretation of data; in the writing of the manuscript; and in the decision to submit the manuscript for publication.

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