High Norepinephrine State Induces Growth of Colorectal Cancer Cells via ADP-Ribosyltransferase 1 in Type 2 Diabetes Mellitus

Background : Patients with type 2 diabetes mellitus have a higher susceptibility for colorectal cancer and poorer prognosis, but the mechanism is still unknown. Here, we investigated the effect of ADP-ribosyltransferase 1 (ART1) on the growth of colorectal cancer in an animal model of diabetes with high norepinephrine status, as well as the potential mechanism. Methods : We evaluated the size and weight of transplanted CT26 cell tumors with different ART1 expression levels in a mouse model of diabetes, as well as the survival time. CCK8 and flow cytometry were used to evaluate the growth of CT26 cells in vitro . Western blot was performed to analyze differentially expressed proteins in the ART1-modulated pathway. Results : High levels of norepinephrine and ART1 favored the proliferation of CT26 cells in vitro and in vivo . Moreover, inhibition of norepinephrine-dependent proliferation was observed in ART1-silenced CT26 cells compared to those with normal ART1expression. Following reduction of the serum norepinephrine level by surgery, the size and weight of transplanted CT26 cell tumors was significantly reduced compared to non-operated and sham-operated mice. Furthermore, the expression of ART1, mTOR, STAT3, and p-AKT protein in the tumor tissue of diabetic mice was higher than in non-diabetic mice. Following reduction of the norepinephrine level by renal denervation (RD), expression of the proliferation-related proteins mTOR, STAT3, p-AKT protein decreased, but no change was seen for ART1 expression. At the same concentration of norepinephrine, ART1 induced the expression of p-AKT, mTOR, STAT3, CyclinD1 and c-myc in CT26 cells in vitro . Conclusions : We conclude that faster growth of colorectal cancer in high norepinephrine conditions requires the expression of ART1, and that high ART1 expression may be a novel target for the treatment of diabetes-associated colorectal cancer.


Introduction
Many studies to date have shown that type 2 diabetes mellitus (T2DM) is closely related to the occurrence and development of colorectal cancer (CRC).Compared with non-diabetic patients, T2DM patients have an increased risk of CRC and worse prognosis, although the specific mechanism is still unclear [1,2].
Insulin resistance (IR) is the pathophysiological basis of T2DM.IR activates the sympathetic nervous system (SNS), resulting in an increased level of circulating norepinephrine (NE) [3][4][5].Long-term activation of the SNS leads to the onset of metabolic syndrome and increases the risk of T2DM, thus further aggravating IR and causing a vicious cycle [3,[6][7][8].A large number of studies indicate that NE activates the adrenergic receptor which in turn induces the cAMP-PKA, AKT-mTOR, ERK-Mnk1 and other signaling pathways.This subsequently affects the expression and activity of downstream molecules such as STAT3, c-myc and MMP-2, thereby affecting the biological behavior of tumor cells.The action of NE can be reduced by down-regulating its expression, or when the activity of beta-adrenergic receptors on the surface of tumor cells is blocked by alpha/beta-receptor [9][10][11][12][13][14][15][16][17].Therefore, a high level of NE-induced activation of signaling pathways such as AKT/mTOR/STAT3 may be a potential mechanism for the higher risk of CRC observed in patients with T2DM.ADP-ribosyltransferase 1 (ART1) is an important mono-ADP ribose transferase that catalyzes posttranslational modification of proteins by transferring a mono-ADP ribose to the protein target.ART1 is believed to have an important role in a variety of cellular biological properties [16,17].We previously reported that changes in ART1 expression altered the phosphorylation level of AKT, as well as the activity and expression of mTOR, GSK-3 and c-myc.These subsequently affected the proliferation, invasion, metastasis, differentiation, angiogenesis and apoptosis of mouse CT26 CRC cells [18][19][20].The AKT signaling pathway is one of the most important pathways involved in cell survival and insulin signaling [21].Activation of AKT can promote the phosphorylation of mTOR, STAT3 and other downstream substrates to exert extensive biological effects, indicating that ART1 is associated with glucose-related metabolic diseases.To our knowledge, the fundamental relationship between ART1 and the growth of CRC in conditions of T2DM and high NE status has yet to be reported.
In the present study, we investigated whether ART1 influences tumor growth and cell proliferation in an animal model of T2DM and CRC with high levels of NE.In addition, we explored whether the AKT/mTOR/STAT3 signaling pathway and the downstream molecules CyclinD1 and c-myc were involved in the mechanism, and whether there is crosstalk between ART1 and high NE status.

Cell Lines and Animals
The CT26 cell line was obtained from Professor YuQuan Wei (Sichuan University, Chengdu, Sichuan, China).
All experimental procedures were approved by the Animal Experimentation Ethics Committee (Chongqing Medical University) and were in accordance with the National Health and Medical Research Council of China Guidelines on Animal Experimentation.Mycoplasma testing has been done for the cell lines used in the research.Cell lines used have been authenticated by STR.

CT26 Cell Survival Assays and Flow Cytometry Analysis
CCK8 (CCK8 kit, Key Gen Biotechnology, Nanjing, PR China) assay was used to evaluate the effect of ART1 on the proliferation of CT26 cells in different concentrations of NE (0, 0.2, 0.4, 0.6, 0.8, 1.0 µM) for 24 h, or 1.0 µM of NE for different incubation times (12 h, 24 h, 36 h, 48 h).The absorbances (optical densities) were recorded with a universal microplate reader (Bio-Tek) at 450 nm.Assays were repeated at least three times.
Flow cytometry (Becton Dickinson Company, Franklin Lakes, NJ, USA) was used to evaluate the cell cycle distribution in each group treated with 1.0 µM NE for 48 h.All experiments were performed at least three times.

Establishment of Diabetic Mouse Model
A diabetic mouse model was established by feeding Balb/c mice with a HFD for six weeks and injecting with 1% streptozotocin (50 mg/kg, STZ, Sigma Chemical Co, St Louis, MO, USA) intraperitoneally.Control Balb/c mice fed a NCD were injected intraperitoneally with saline.Diabetes was defined as a random glucose concentration in tail vein blood of ≥11.1 mmol/L (test strips, Advantage, Bayer, Contour TS) one week after STZ injection [25][26][27].

Subcutaneous Tumor Transplant Model
Each experimental group consisted of 12 mice.The colorectal carcinoma complicated with diabetes mellitus (CRCD) group (n = 48) was comprised of T2DM Balb/c mice fed the HFD.These were equally and randomly divided into the GFP-ART1 group (subcutaneously transplanted with GFP-ART1 CT26 cells), the GFP-shART1 group (GFP-shART1 CT26 cells), the non-transfection group (non-transfected CT26 cells), and GFP-Vector group (GFP-Vector CT26 cells).The colorectal carcinoma complicated with diabetes mellitus (CRCO) group (n = 48) was comprised of Balb/c mice without T2DM and fed with NCD.These were also equally and randomly divided into GFP-ART1, GFP-shART1, non-transfected and GFP-Vector groups and subcutaneously transplanted with the appropriate CT26 cells as described above.
The CT26 cell suspension (2 × 10 6 /mL, 200 µL) was subcutaneously injected into the lateral skin of the right armpit of each mouse [28].After 14 days, six mice were randomly selected from each group for sacrifice, and the weight and volume of the subcutaneous tumor was recorded.The survival time of the remaining mice in each group was recorded.Tumor volume was calculated according to the formula: volume = the maximum diameter × the most trails 2 × ½ [29].

Renal Denervation (RD) and Sham Operation
Another 12 diabetic Balb/c mice inoculated with GFP-ART1 CT26 cells were equally divided into three groups: left RD (LRD, n = 4), left sham operation (LSO, n = 4), and no operation (GFP-ART1 group, n = 4).RD or sham surgery was performed as described previously [30,31].In brief, mice were anaesthetized with an intraperitoneal ketamine injection (87 mg/kg).The kidneys were exposed by paravertebral line incision and the renal arteries and veins were isolated from connective tissue.After stripping the visible nerves, the vessels were painted for 2 minutes with a solution of 10% phenol in absolute ethanol.The muscular layers of the abdominal wall were sutured with absorp- tive material and the skin was closed by non-absorptive filament.Animals recovered from anesthesia within 10-20 minutes after the end of surgery.In the sham operation, the renal nerves were isolated but preserved.As described earlier [31], successful denervation was assessed by measuring the renal tissue content of catecholamines using an enzyme-linked immunosorbent assay (ELISA) KIT (Cloud-Clone, Houston, TX, USA).Completeness of denervation was assumed when the NE tissue content was <10% of the mean value observed in the sham-operated group [30].
The day after operation, all mice were inoculated with GFP-ART1 CT26 cells in the right axillary fossa.The mice were killed after two weeks and the kidneys and transplanted tumor were excised.These were weighed and the tumor volume measured, then stored frozen in liquid nitrogen.Blood was also obtained from each animal.

Western Blot Analyses
CT26 cells and Balb/c mice transplant tumors were placed in lysis buffer (Beyotime, Shanghai, China) and the proteins electrophoresed on SDSPAGE gel and transferred to PVDF membranes.The membranes were blocked with 5% nonfat dried milk and incubated overnight at 4 °C with primary antibodies against ART1 (Abgent, San Diego, CA, USA; 1:500 dilution), Akt kinase (Akt), phospho-Akt (p-AKT), STAT3, mTOR, CyclinD1, and c-myc (Cell Signaling Technology, MA, USA; 1:1000 dilution).They were then incubated with horseradish peroxidase-conjugated secondary antibody at a dilution of 1:1000 (ZSGBBIO, Beijing, China).Finally, the membranes were assessed using enhanced chemiluminescence reagents (Beyotime) and analyzed with Quantity One software (BioRad Laboratories, Hercules, CA, USA).Betaactin was used as a loading control for the western blotting experiments.

Evaluation of Serum Insulin and NE Levels
Serum levels of insulin and NE were measured using commercial enzyme-linked immunosorbent assay kits (Cloud-Clone) according to the manual.

Statistical Analysis
Data was analyzed using the SPSS 19.0 statistical software package (SPSS, Chicago, IL, USA).Numerical data were compared using the chi-square test.Binary logistic regression was used to analyze T2DM and ART1 expression in CRC patients.All values shown are the mean ± standard error of the mean (SEM).Unpaired Student t-test was used for two-group comparisons.One-way ANOVA was used for intra-group comparisons, followed by a least significant difference post hoc test to compare between groups.p < 0.05 was considered as statistically significant.

High Concentrations of NE Boost the Proliferation of CT26 Cells and Requires ART1
CT26 cells from GFP-ART1, non-transfection, GFP-Vector and GFP-ShART1 groups were induced for 24 h with 0.2, 0.4, 0.6, 0.8 and 1 µM NE concentrations.With increasing concentrations of NE, cell proliferation was observed to gradually increase in the CT26 cell groups expressing ATR1 (GFP-ART1, non-transfection and GFP-Vector groups), but not in ATR1-silenced CT26 cells (Fig. 1A).We further studied the time-dependence of NE on cell proliferation in the four experimental groups.Each group was treated with 1 µM NE for 12, 24, 36 and 48 h.Cell proliferation was found to increase progressively from 12 h to 36 h in the GFP-ART1, non-transfection and GFP-Vector groups, and then decreased after 48 h treat- ment.However, no significant increase with treatment time was observed in the GFP-ShART1 group (Fig. 1B).Flow cytometry showed a decreased percentage of G1 phase and increased S phase and proliferation indices (PI) in the GFP-ART1 group compared to the non-transfection and GFP-Vector groups.In contrast, the GFP-shART1 group had an increased percentage of G1 phase and decreased S phase and PI.Following NE induction, CT26 cells that expressed ATR1 (GFP-ART1, non-transfection and GFP-Vector groups) showed a decreased percentage of G1 phase cells and increased percentage of S phase and PI cells.In contrast, ATR1-silenced CT26 cells showed no obvious difference in cell cycle distribution or PI (Fig. 1C, Table 1).Together, these results suggest that a high concentration of NE and an appropriate treatment time can induce ART1dependent proliferation of CT26 cells.At the same NE concentration and treatment time, the proliferation of CT26 cells was highest in the GFP-ART1 group and lowest in the GFP-ShART1 group (Tables 2,3), thus demonstrating that ART1 boosts the proliferation of CT26 cells.

Establishment of Animal Model of Diabetes
A diabetic Balb/c mouse model was established by feeding a HFD and intraperitoneal injection with 1% STZ.Blood tests showed significantly higher glucose concentrations, insulin levels and NE levels in diabetic mice compared to mice fed an NCD.Diabetic mice also had higher weight (Fig. 2).These results demonstrate successful establishment of the Balb/c mouse diabetic model.
Diabetic mice were inoculated with CT26 cells subcutaneously into the right axillary fossa.Following inoculation with cells having the same ART1 expression level, diabetic mice lost more weight than non-diabetic mice (p < 0.01).However, the absolute value of body weight was still higher in the CRCD group compared to the CRCO group (21.82 ± 1.64 vs. 21.04 ± 1.08 g, respectively, p < 0.05) (Table 4).Moreover, both the volume and weight of transplanted tumors in the CRCD group were higher than in the CRCO group (p < 0.05).Mice injected with high ART1 expressing CT26 cells showed the most body weight loss (8.54 ± 0.33 g) and the largest and heaviest tumors (p < 0.01), whereas mice injected with ATR1-silenced CT26 cells had the least body weight loss (4.82 ± 0.55 g) and the smallest and lightest transplanted tumors (Fig. 3A,B; Tables 5,6).

ART1 Impacts the Growth of Transplanted Tumor Cells and the Survival Time of Both Diabetic and Non-Diabetic Mice
We also investigated the effect of ART1 expression and diabetes on the survival of mice with CT26 tumor cell transplants.Following inoculation with CT26 cells expressing the same level of ART1, the survival of CRCD mice was shorter than that of CRCO mice (p < 0.05) (Fig. 3C).Furthermore, the survival of both CRCD and CRCO mice was shortest in those injected with high ART1 expressing CT26 cells, and longest in those injected with ATR1-silenced CT26 cells (p < 0.01) (Table 7).

Renal Denervation Reduces the Level of Blood Glucose and the Growth of Tumor Cell Transplants with High ART1 Expression in CRCD Mice
Animal models of RD were successfully established, with the NE level reduced by >10% in the kidney and plasma of the operated group (LRD) (p < 0.01) [30,31].The blood glucose and plasma insulin levels of each group were also measured, together with the volume and weight of transplanted tumor.The results showed lower blood glucose and plasma insulin levels, and smaller and lighter tumors in the LRD group compared with the sham-operated (LSO) and non-operated groups (p < 0.01).No significant differences were observed between the non-operated and LSO groups (p > 0.05) (Fig. 4; Tables 8,9).

ART1 is Positively Associated with the Expression of p-AKT, mTOR, STAT3, cyclinD1 and c-myc in CT26 Cells
Induced by High NE GFP-ART1 CT26 cells showed the highest levels of protein expression for ART1, p-AKT, mTOR, STAT3, Cy-clinD1 and c-myc (p < 0.01).In contrast, GFP-shART1 CT26 cells showed the lowest expression of these protein markers (p < 0.01), suggesting that ART1 contributes to the up-regulation of p-AKT, mTOR, STAT3, CyclinD1 and c-myc in cells with high NE (Fig. 5A).

Expression of ART1, p-AKT, mTOR and STAT3 in Transplanted Tumors from CRCD and CRCO Mice, and Changes after Renal Denervation in the CRCD Group
The expression of ART1, mTOR and STAT3 was significantly higher in CRCD mice compared with CRCO mice (p < 0.01).Although there was no difference in the expression of AKT (p > 0.05), the expression of p-AKT was significantly higher in CRCD mice compared with CRCO mice (p < 0.01) (Fig. 5B).In vivo data also revealed that a high-NE environment boosted the expression of ATR1, as well as p-AKT, mTOR and STAT3.
To investigate the effect of NE on the ART1dependent proliferation of CT26 cells, GFP-ART1 CT26 cells were inoculated in the non-surgery group, LSO group (sham operation) and LRD group (renal denervation).No significant differences in ART1 and AKT protein expression were found between the three groups (p > 0.05).However, the expression of p-AKT (p < 0.01), mTOR (p < 0.05) and STAT3 (p < 0.05) proteins was clearly lower in the LRD group compared to the non-surgery and LSO groups (Fig. 5C).This suggests that NE deprivation leads to a consistent decrease in p-AKT, mTOR and STAT3 expression.Interestingly, however, the ART1 and AKT levels did not decrease, suggesting that even if NE affects the expression of proliferation-related proteins, the expression of ART1 itself is not dependent on the NE level.

Discussion
Epidemiological studies and bioinformatic analyses have found that T2DM is a risk factor for CRC, as well as being a prognostic factor for adverse survival outcomes in such patients [1,2].T2DM is a metabolic disease characterized by IR, hyperinsulinemia, chronic activation of the SNS, increased circulating NE levels, and a high NE status [3,[6][7][8].Increased NE levels have been reported to promote CRC liver metastasis, possibly through the beta-adrenergic receptor signaling pathway [32].Large case-control studies have also found that long-term use of NE antagonists or beta-receptor blocking agents can reduce the risk of malignant tumors, including CRC [14,15,33].Therefore, high levels of NE are likely to be related to CRC development in addition to T2DM.
In the present study, ART1 was found to be necessary for the proliferation of CRC cells in a high-NE environment accompanied with diabetes.Our further research revealed  that ART1 activated the AKT-mTOR pathway, resulting in increased STAT3 phosphorylation and ultimately promoting Cyclin D1 and c-myc expression and cell proliferation.
Our previous studies [18][19][20]22] showed that high ART1 expression boosted the malignant biological behavior of CRC, and that ADP-ribosylation could affect glucose metabolism via several pathway [34].However, so far there has been no research on the effect of ART1 on CRC associated with diabetes mellitus.Therefore, in the present study we investigated the effect of changes in ART1 expression on CRC associated with T2DM.Regardless of whether mice were fed a normal or HFD, elevated ART1 expression was observed more often in those with lymph node metastasis than those without.This confirms the results of our previous study and suggests that ART1 expression is related to the malignant behavior of CRC.Importantly, mice fed a HFD had higher expression of ART1 compared to mice fed a normal diet.These results suggest that ART1 expression may be related to CRC-associated disorders of glucose metabolism, but the mechanism underlying this association remains to be clarified.
In vivo experiments showed that diabetes was a positive factor for the growth of CRC tumor cell transplants when ART1 expression was constant.Furthermore, increased ART1 expression accelerated the growth of these tumors and shortened the survival time in both the diabetic and non-diabetic mouse models.Thus, both diabetes and ART1 encouraged the growth of xenografts.However, diabetic and non-diabetic mice still showed a difference in xenograft growth when ART1 was silenced, suggesting the diabetes-accelerated growth of xenografts was not due entirely to ART1.
The in vitro study conducted here also showed that CT26 cells with high ART1 expression showed consistently higher proliferation rates with increasing NE concentrations and induction times.However, no significant changes were detected in the proliferation of ART1-silenced CT26 cells following increases in NE concentration and induction time.These results highlight that NE-promoted proliferation of CRC CT26 cells is dependent on ART1.Although the diabetes-accelerated growth of xenografts was not totally dependent on ART1, we conclude from our findings that NE promotion of CRC cell proliferation is dependent on ART1.Phosphorylated STAT3 is linked to the activation of Akt/mTOR, with this pathway being involved in cell proliferation and growth.Results from both in vitro and in vivo experiments have shown that ART1 expression in a high NE environment is positively associated with the expression of p-AKT, mTOR and STAT3, but not AKT.Following LRDinduced reduction in the blood level of NE, the expression of p-AKT, mTOR and STAT3 protein in transplanted tumor cells decreased significantly, whereas no change was observed in the expression of ART1 and AKT protein.Moreover, the volume and weight of the transplanted tumor also decreased significantly, thus confirming that reduction of the NE level is an effective strategy for inhibiting xenograft growth.Consistent with our observations, Lima-Seolin et al. [35] also found increased activity of the AKT/mTOR/STAT3 signaling pathway in high NE conditions.Our results suggest that NE is not required for ART1 expression, but is required for ART1-induced activation of the AKT/mTOR/STAT3 signaling pathway.STAT3 can increase the transcription of CyclinD1 and c-myc, thus promoting the indefinite proliferation of cells.Our results indicated that expression of ART1 up-regulated the expression of CyclinD1 and c-myc proteins in conditions of high NE.Together with previous results, we hypothesize that ART1 up-regulates the expression of Cy-clinD1 and c-myc proteins in high NE conditions by activating the AKT/mTOR/STAT3 signal pathway, thereby promoting the proliferation and growth of CT26 cells.
A recent study used TaqMan OpenArray panels to analyze gene expression patterns in colon carcinoma samples from T2DM and non-diabetic patients [36].This revealed that T2DM impacts CRC mainly through five pathways, namely Wnt (wingless-type)/β-catenin, Hippo, TNF, PI3K/Akt, and platelet activation, thus partially confirming the results of the present study.

Limitation
In this present study we applied long term high-fat diet combined with low-dose STZ injection to simulate the insulin resistance state of T2DM to the greatest extent [37][38][39].In view of the tolerance of the mouse model to the subsequent two surgeries (renal sympathetic denervation and subcutaneous tumor transplantation), we chose Blab/c mice instead of nude mice for the experiment, and therefore accordingly we did not use human colorectal cancer cell lines.That is a limitation for our study.It's worth seeking a mouse model compatible with human CRC cells for further validation of our findings in next step of research.

Conclusions
In conclusion, we hypothesize that ART1 plays an important role in stimulating the proliferation of CRC cells under high NE conditions.The specific mechanism is likely to include activation of the AKT/mTOR/STAT3 pathway, thereby up-regulating the expression of Cyclin D1 and cmyc protein and subsequently promoting the growth of CRC.Although the molecular mechanisms require further clarification, our results may provide an explanation for the poorer prognosis of CRC patients with T2DM and high NE status compared to non-diabetic patients.Moreover, these results suggest that ART1 could be used as a therapeutic target in CRC patients with T2DM.

Fig. 1 .
Fig. 1. High concentrations of NE induce ART1-dependent proliferation in CT26 cells.(A) Proliferation of CT26 cells with different ATR1 expression levels after treatment with different concentrations of NE. (B) Proliferation of CT26 cells with different ATR1 expression levels and treated with 1 µmol/L NE for different times.(C) Cell cycle distribution of CT26 cells with different ATR1 expression levels and treated with or without NE.(a) GFP-ART1 CT26 cells treated with 1 µmol/L NE.(b) Non-transfected CT26 cells treated with 1 µmol/L NE.(c) GFP-Vector CT26 treated with 1 µmol/L NE.(d) GFP-shART1 CT26 cells treated with 1 µmol/L NE.(e) GFP-ART1 CT26 cells with no NE treatment.(f) Non-transfected CT26 cells with no NE treatment.(g) GFP-Vector CT26 cells with no NE treatment.(h) GFP-shART1 CT26 cells with no NE treatment.NE, norepinephrine.

Fig. 3 .
Fig. 3. ART1 increases the growth of transplanted tumor while reduce the survival time of both CRCD and CRCO mice.(A) shows the results for DM mice transplanted with CT26 cells expressing different levels of ART1 and the results for non-DM mice transplanted with CT26 cells expressing different levels of ART1 (GFP-ART1 vs. Non-transfected, GFP-shART1 vs. GFP-Vector, # p < 0.05 ## p < 0.01; GFP-ART1 vs GFP-shART1: ▲▲ p < 0.01), respectively.(B) Tumor size in DM mice and in non-DM mice transplanted with CT26 cells expressing different levels of ART1.(a) DM mice transplanted with GFP-ART1 CT26 cells.(b) DM mice transplanted with non-transfected CT26 cells.(c) DM mice transplanted with GFP-Vector CT26 cells.(d) DM mice transplanted with GFP-shART1 CT26 cells.(e) Non-DM mice transplanted with GFP-ART1 CT26 cells.(f) Non-DM mice transplanted with non-transfected CT26 cells.(g) Non-DM mice transplanted with GFP-Vector CT26 cells.(h) Non-DM mice transplanted with GFP-shART1 CT26 cells.(C) Survival curves for DM mice and non-DM mice transplanted with CT26 cells expressing different levels of ART1.

Fig. 4 .
Fig. 4. Tumor size in DM mice transplanted with CT26 cells expressing high levels of ART1 and surgical intervention to alter the serum NE level.(A) GFP-ART1 group.(B) LSO group.(C) LRD group.