AT7867 Inhibits Human Colorectal Cancer Cells via AKT-Dependent and AKT-Independent Mechanisms

AKT is often hyper-activated in human colorectal cancers (CRC). This current study evaluated the potential anti-CRC activity by AT7867, a novel AKT and p70S6K1 (S6K1) dual inhibitor. We showed that AT7867 inhibited survival and proliferation of established (HT-29, HCT116 and DLD-1 lines) and primary human CRC cells. Meanwhile, it provoked caspase-dependent apoptosis in the CRC cells. Molecularly, AT7867 blocked AKT-S6K1 activation in CRC cells. Restoring AKT-S6K1 activation, via expression of a constitutively-active AKT1 (“ca-AKT1”), only partially attenuated AT7867-induced HT-29 cell death. Further studies demonstrated that AT7867 inhibited sphingosine kinase 1 (SphK1) activity to promote pro-apoptotic ceramide production in HT-29 cells. Such effects by AT7867 were independent of AKT inhibition. AT7867-indued ceramide production and subsequent HT-29 cell apoptosis were attenuated by co-treatment of sphingosine-1-phosphate (S1P), but were potentiated with the glucosylceramide synthase (GCS) inhibitor PDMP. In vivo, intraperitoneal injection of AT7867 inhibited HT-29 xenograft tumor growth in nude mice. AKT activation was also inhibited in AT7867-treated HT-29 tumors. Together, the preclinical results suggest that AT7867 inhibits CRC cells via AKT-dependent and -independent mechanisms.

AKT, or protein kinase B (PKB), is a serine/threonine kinase that lies downstream of phosphatidylinositol 3-kinase (PI3K) [7,8]. Over-expression and/or hyper-activity of AKT and a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 AKT-regulated signalings are often detected in human CRC [7,8]. AKT plays a pivotal role in a number of cellular behaviors, including cell growth, proliferation and metabolism as well as survival and apoptosis-resistance [7,8]. On the other hand, AKT inhibition, silence or loss-offunction mutation could lead to CRC cell death [9,10]. Therefore, different AKT inhibitors are being evaluated in both pre-clinical and clinical CRC studies [11,12,13,14]. Grimshaw et al. has recently developed a dual inhibitor of AKT and p70S6K1 ("S6K1"), named AT7867 [15]. This dual inhibitor was shown to block AKT-S6K1 activation and inhibit human tumor cell proliferation [15].Although the effect of AT7867 on human CRC viability was examined by Grimshaw et al [15], this effect remains to be fully characterized. Importantly, the mechanisms underlying AT7867-mediated anti-cancer activity are still illusive [15]. We are interested to know whether there are AKT-independent mechanisms also responsible for AT7867-mediated killing of cancer cells. Here, we provided evidences to suggest that sphingosine kinase 1 (SphK1) inhibition and subsequent ceramide production should also participate in AT7867-induced anti-CRC cell activity.

Primary culture of patient-derived colon cancer and epithelial cells
Fresh human colon cancer tissues and surrounding epithelial tissues were separately carefully. Tissues samples were then mechanically dissociated, filtered through a 70-μm strainer, and digested as previously reported [10]. Primary cells were then cultured in the described complete medium [10]. Two lines of primary colon cancer cells and one line of primary colon epithelial cells were established. Experiments and the protocols requiring clinical samples were approved by the Ethics Review Board (ERB) of Nanjing Medical University. The written-informed consent was obtained from each participant. A total of two colon cancer patients (Male, 56/66 years old) administrated in the First Affiliated Hospital of Nanjing Medical University (Nanjing, China) were enrolled. All investigations were conducted according to the principles expressed in the Declaration of Helsinki as well as national/international regulations.

Clonogenicity assay
As described [17], cells (5 × 10 4 per treatment) were suspended in agar-containing complete medium or plus AT7867 treatment, which were then added on top of a six-well plate. After 8 days, colonies were stained and manfully counted.

BrdU assay of proliferation
Cells with/out the AT7867 treatment were incubated with BrdU (10 μM). Cells were then fixed, and BrdU incorporation was determined by the BrdU ELISA kit (Roche Diagnostics) according to the attached protocol.

Trypan blue assay of cell death
As described [17], after applied treatment, the percentage of "dead" cells was calculated by the number of the trypan blue stained cells divided by the total cell number.

Quantification of apoptosis by ELISA
After applied treatment, the single strand DNA ("ssDNA") Cell Apoptosis ELISA Kit was applied to detected denatured DNA in ELISA format to reflect cell apoptosis [18].

Annexin V assay
The adherent and floating cells were collected and washed. Cells were then incubated in Annexin V solution (10 μg/mL, Invitrogen, Shanghai, China) for 15 minutes. Immediately prior to reading on a FACS Calibur flow cytometer (BD, Nanjing, China), 10 μg/mL of propidium iodide (Invitrogen) was added to the mix. Annexin V positive cells were gated as apoptotic cells.

TUNEL assay and caspase activity assay
The detailed protocols of TUNEL staining assay and caspase activity assay were described in detail in other studies [17,19].

Western blot assay
After treatment, both floating and adherent cells were collected and washed. Cells were then harvested using the RIPA buffer (Biyuntian, Nanjing, China). Aliquots of 30 μg lysates per sample were separated by SDS-PAGE and transferred to PVDF membranes (Millipore, Nanjing, China). The blots were blocked and incubated with designated primary and secondary antibodies. Targeted protein bands were visualized with ECL reagents and developed with Hyper-film (GE Healthcare, Shanghai, China). Results were quantified via the ImageJ software (NIH).

AKT1 shRNA knockdown
The two lentiviral AKT1 shRNAs ("-a/-b"), with non-overlapping sequences, were designed by Genepharm (Shanghai, China). The AKT1shRNA (10 μL/mL) or the scramble control shRNA (Santa Cruz Biotech, Nanjing, China) was added to cultured cells for 24 hours. Puromycin (5.0 μg/mL) was then included to select stable colonies for 4-6 passages. The AKT1 knockdown in the stable cells was verified by Western blot assay.
2.14. SphK1 activity assay and ceramide content assay Analyzing sphingosine kinase 1 (SphK1) activity and cellular ceramide content were described in detail in our previous study [17].

In vivo tumor studies
HT-29 cells were injected subcutaneous (s.c.) in the right flanks of female nude mice (from the Experimental Animal Center of Nanjing Medical University, Nanjing, China). Animals were randomized into three groups, and treatment was started with vehicle or AT7867 when established tumors were~100 mm 3 in average volume. Control mice received vehicle only (10% DMSO, 90% saline) and treatment mice received 10 or 50 mg/kg AT7867 intraperitoneally (i.p.). Tumor volumes and body weights were monitored every four days. The tumor size was calculated using the formula: V (volume) = 0.5328 × Long × Width × High (mm 3 ). For recording mouse body weight, the estimated tumor weight (tumor volumes × 1g/cm 3 ) was subtracted from total weight of each mouse. Mice were maintained under the following conditions: 12hour dark/12-hour light cycle, 24±2˚C temperatures, and 50±10% humidity. Mice were observed extremely carefully throughout the experimental period. The clinical signs of mice were recorded daily, and if the criteria of humane endpoints were met, animals were sacrificed. Humane endpoints were considered as rapid weight loss (>15%), abnormal changes in behavior and motion (social and eating behavior), tumor size greater than 2 cm 3 or skin problems (wounds or signs of inflammation). If animals reached these endpoints, they were euthanized by exsanguination under 2,2,2-tribromoethanol anesthesia (4 mg/10 g body weight, Sigma). All injections in this study were performed via the above anesthesia method [21]. The protocol was approved by the Nanjing Medical University's Institutional Animal Care and Use Committee (IACUC) and Ethics Review Board (ERB).

Statistical analysis
Data were normalized to control values of each assay, and were presented as mean ± standard deviation (SD). Statistics was analyzed by one-way ANOVA followed by a Scheffe's f-test using the SPSS 18.0 software (Chicago, IL). P< 0.05 is considered statistically significant.
The potential effect of AT7867 on other human CRC cells was also tested. As demonstrated, AT7867 (10 μM) was cytotoxic when added to the two other established CRC cell lines (HCT116 and DLD-1). It also decreased the percentage of viable patient-derived primary human CRC cells (two lines, "Pri1/2") ( Fig 1E). Remarkably, the AT7867 treatment (10 μM, 72 hours) was non-cytotoxic to the primary colon epithelia cells ("Epi", Fig 1E). Together, these results demonstrate that AT7867 is cytotoxic and anti-proliferative to cultured human CRC cells.

AT7867 provokes apoptosis in CRC cells
We next tested the potential effect of AT7867 on CRC cell apoptosis. Under many apoptotic stimuli, cytochrome c released from mitochondria will associate with procaspase-9 and Apaf-  1), primary human CRC cells (two lines, "Pri1/2") and the primary human colon epithelia cells ("Epi") were treated with/out a range of concentrations of AT7867 for applied time; MTT assay was performed to test viable cell percentage (A and E); Cell proliferation was tested by the clonogenicity assay (B) and the BrdU ELISA assay (C); Cell death was tested by the trypan blue staining assay (D). Data were shown as the mean (n = 5, within the same experiment) with the standard deviation (SD). * P < 0.05 vs. "C" (medium control) group. 1, causing caspase-9 activation [22,23]. The latter further activates caspase-3 and caspase-7 to initiate a caspase cascade, leading to mitochondrial/intrinsic apoptosis pathway activation [22,23]. On the other hand, activation of caspase-8 is the characteristic marker of extrinsic apoptosis pathway activation [22,23,24]. In the current study, we showed that AT7867 (at 1-25 μM) significantly increased the activity of caspase-3 and caspase-9, but not caspase-8, in HT-29 cells (Fig 2A), suggesting activation of intrinsic, but not extrinsic, apoptosis pathway. Furthermore, 1-25 μM of AT7867 also significantly increased the ssDNA ELISA optic density (OD) (Fig 2B). The percentages of TUNEL positive cells and Annexin V positive cells were also increased significantly following AT7867 (10 μM) treatment (Fig 2C). These results demonstrated that AT7867 promoted apoptosis in HT-29 cells.
To study the link between AKT inactivation and AT7867-induced actions, a constitutivelyactive AKT1 ("ca-AKT1", flag-tagged) [20] was introduced to HT-29 cells. Western blot assay results in Fig 3B confirmedca-AKT1 expression in the HT-29 cells. Remarkably, the ca-AKT1 completely restored AKT activation (pAKT/pGSK3β/pS6K1) in AT7867-treated HT-29 cells, even higher to the control level (Fig 3B, quantification). Yet, it only partially inhibited AT7867induced HT-29 cell death ( Fig 3C) and apoptosis ( Fig 3D). Therefore, restoring AKT-S6K1 signaling was not enough to completely rescue HT-29 cells from AT7867, indicating that AKTindependent mechanisms should also play a significant role in mediating AT7867's cytotoxicity. As a matter of fact, the cytotoxic effect of AT7867 on HT-29 cells was significantly greater than the effects of other known Akt inhibitors, such as perifosine[26], MK2206[27] and AKT inhibitor II (Fig 3E and 3F), although these other AKT inhibitors also blocked AKT-S6K1 activation (Data not shown).

AT7867 inhibits SphK1 to promote ceramide production in CRC cells
Above results suggest that AKT-independent mechanisms could also contribute to AT7867-induced cytotoxicity in CRC cells. Growing evidences have indicated that SphK1 is an important oncotarget of CRC [28,29]. Over-expression and/or sustained-activation of SphK1 contributes to cancer cell progression and apoptosis-resistance [30,31]. Inhibition, mutation or silence of SphK1 would increase cellular ceramide production to promote cell apoptosis [30,31]. Interestingly, a very recent study has shown that AKT inhibitor A-674563 could also inhibit SphK1 activity, independent of AKT inhibition [32]. Thus, SphK1 activity and ceramide production in AT7867-treated CRC cells were tested using the methods described previously [17]. Results in Fig 4A showed clearly that AT7867 dose-dependently decreased SphK1 activity in HT-29 cells. SphK1 protein expression was not altered following the AT7867 treatment (Data not shown). Consequently, the cellular ceramide level was significant increased (Fig 4B). Intriguingly, restoring AKT activation by expression of ca-AKT1 (See Fig 3) failed to affectAT7867-induced ceramide production (Fig 4C). Further, a comparable ceramide production by AT7867was noticed in AKT1-silenced HT-29 cells (Fig 4D). These results imply that AT7867-induced ceramide production apparently is independent of AKT inhibition.
On the other hand, other AKT inhibitors, perifosine, MK2206 and AKT inhibitor II, failed to inhibit SphK1 activity in HT-29 cells (Fig 4I). These results further suggest that SphK1 inhibition and ceramide production, independent of AKT inhibition, should also contribute to AT7867-induced cytotoxicity against CRC cells.

AT7867 inhibits HT-29 tumor growth in nude mice
In order to examine the effect of AT867 on tumour growth in vivo, HT-29 cells were injected s. c. to the nude mice to establish the xenograft tumors. Tumor growth curve results in Fig 5A  displayed that i.p. injection of AT7867 (10 and 50 mg/kg) significantly inhibited HT-29 tumor growth in nude mice. AT7867 at 50 mg/kg was more efficient than 10 mg/kg in inhibiting HT-29 tumor growth (Fig 5A). Estimated daily tumor growth results in Fig 5B further confirmed the significant anti-HT-29 tumor activity byAT7867. Notably, the mice body weight was not significantly affected by the AT7867 administration. Neither did we notice any signs of apparent toxicities. These results, consistent with reports from other studies [15], suggested that the AT7867 treatment regimens here were relatively safe to the mice. IHC staining assay results in Fig 5D demonstrated that pAKT level was significantly lower in the AT7867 (50 mg/kg)treated HT-29 tumors. Together, these results show that i.p. injection ofAT7867efficientlyinhibits HT-29 tumor growth in nude mice.

Discussion and Conclusions
Several mechanisms are responsible for the sustained activation of AKT in CRC and other malignancies, including constitutive activation of upstream receptor tyrosine kinases (i.e. EGFR), PIK3CA/PTEN mutations, AKT amplification and/or mutation [7,8]. These will lead to persistent growth of CRC cells [7,8]. Thus, pharmacological inhibition of AKT represents a rational approach for the treatment of CRC [7,8]. A number of AKT inhibitors of different mechanisms of action have recently been developed [7,8].In the current study, we showed that AT7867, a novel AKT and S6K1 dual inhibitor [15], inhibited survival and proliferation of established and primary human CRC cells. Meanwhile, AT7867 provoked caspase-dependent apoptosis in CRC cells. In vivo, AT7867 i.p. injection suppressed HT-29 xenograft tumor growth in nude mice. AT7867 blocked AKT-S6K1 signaling in CRC cells. However our results indicated that AKT-S6K1inhibition is unlikely the sole mechanism responsible for AT7867-mediated cytotoxicity in CRC cells. First, exogenous expression of ca-AKT1 completely restored AKT-S6K1 activation in AT7867-treated HT-29 cells, but only partially inhibited HT-29 cell death. Second, in AKT1-silenced HT-29 cells, AT7867 was also cytotoxic and pro-apoptotic. activity was tested (I). Data were shown as the mean (n = 5, within the same experiment) with SD. * P < 0.05 vs. "C" (medium control) group. # P < 0.05 vs. AT7867 only group. doi:10.1371/journal.pone.0169585.g004 In line with previous studies [15], we showed that AT7867 inhibited AKT and S6K1 activations in both established and primary human CRC cells. Besides, we here proposed a novel mechanism to explain its cytotoxicity against CRC cells. AT7867 inhibited SphK1 to promote pro-apoptotic ceramide production in CRC cells. This AKT-independent mechanism also participated in AT7867-induced anti-CRC cell activity.S1P, which inhibited AT7867-provoked ceramide production, also attenuated subsequent HT-29 cell death. Reversely, PDMPor C6 ceramide significantly augmented AT7867-induced HT-29 cell death.K6PC-5, a SphK1 activator, attenuated AT7867-induced HT-29 cell death (Fig 4). This should also explain why AT7867 was more efficient than other specific AKT inhibitors (perifosine, MK2206 and AKT inhibitor II) in killing CRC cells. As these other AKT inhibitors failed to affect SphK1 activity (Fig 4). It could also be the reason of in-effective of this compound in normal colon epithelial cells. Indeed, AKT and S6K1 expression/phosphorylation as well as SphK1 expression were significantly lower in these epithelial cells, as compared to HT-29 CRC cells (S1 Fig). Intriguingly, AT7867-induced ceramide production appears independent of AKT inhibition. As ca-AKT1 failed to inhibit AT7867-indued ceramide production. Meanwhile, in AKT1-silenced HT-29 cells, a compare ceramide production was still noticed. In summary,  HT-29 tumor-bearing nude mice (n = 10 per group) were administrated with AT7867 (10 or 50 mg/kg body weight, i.p., daily for 16 days) or vehicle control ("Veh"); Tumor volumes (A) and mice body weights (C) were recorded ever four days; Estimated daily tumor growth (mm 3 /day) was also shown (B). Three days after initial drug administration, pAKT (Ser-473) in "Veh" and AT7867 (50 mg/kg)-treated HT-29 tumors was tested by the IHC staining assay, representative images were presented (D). Data were shown as the mean with SD.* P < 0.05 vs. "Veh" group. # P < 0.05 vs. AT7867 only group. Bar = 75 μm (D). the results of this preclinical study suggest that AT7867 inhibits human CRC cells in vitro and in vivo. Both AKT-dependent and AKT-independent (i.e.SphK1 inhibition and ceramide production) mechanisms could be responsible for its superior actions against CRC cells.