Hyperglycemia and hyperinsulinemia stimulate cervical cancer cell proliferation in vitro and in vivo study


 Background: Diabetes mellitus and malignant tumor are the second and third causes of women death in Mexico. Hyperglycemia, insulin and insulin-like growth factor 1 are the main risk factors involved in cancer development in patient with diabetes. The aim of this study was to evaluate the effect of hyperglycemia and hyperinsulinemia over cell proliferation and tumor growth in cervical cancer. Methods: Cell proliferation, apoptosis and cell cycle of cervical cancer cell lines (HeLa, SiHa and CaSki) in presence of hyperglycemia and/or insulin were evaluated. Xenograft model for cervical cancer was done in diabetic female nu/nu mice; biochemical parameters, body weight, tumoral volume and cell doubling time were evaluated. Results: Hyperglycemia and hyperinsulinemia significantly increase cell proliferation and decreases apoptosis with no change in cell cycle. Insulin treatment increase tumor volume and diminish cell doubling time, this group also developed hyperinsulinemia and in Langerhans pancreatic islet hypertrophy; whereas, hyperglycemic groups show the same effects but in lesser degree than the insulin treated group. Conclusion: Glucose and insulin stimulates both, proliferation and tumoral growth of cervical cancer, so this should be a possible explanation for the low survival of diabetic patients with cervical cancer in compare to non-diabetic patients with cervical cancer.

tumor volume and diminish cell doubling time, this group also developed hyperinsulinemia and in Langerhans pancreatic islet hypertrophy; whereas, hyperglycemic groups show the same effects but in lesser degree than the insulin treated group.
Conclusion: Glucose and insulin stimulates both, proliferation and tumoral growth of cervical cancer, so this should be a possible explanation for the low survival of diabetic patients with cervical cancer in compare to non-diabetic patients with cervical cancer.

Background
Diabetes mellitus is a metabolic alteration of carbohydrates, lipids and proteins associated to insulin deficiency or resistance, that prevents correct glucose metabolism causing increased glucose levels [1]. King and coworkers [2], estimate that by the year 2025, diabetes prevalence will increase 27%. By the way, Wild and coworkers [3], estimated that for the year 2030 diabetic people will be duplicated in the most of countries of world, without consider obesity. Those  It has been pointed out an association between diabetes and cancer, there are relative risk associated with diabetes for hepatic, pancreatic, endometrial, colon, breast, bladder and lung cancer, although the relative risk varies between each type of cancer [5]; unfortunately, in our country the probability of having cancer and diabetes is becoming more frequent. Reportedly that patient with both pathologies: cancer and diabetes does not respond equal to cancer treatment and have higher mortality comparing with no diabetic patient [6,7]. A lot of studies have reported that growth factors increase metastasis and invasion properties of cancer cells (epidermal growth factor (EGF) and Insulin-like growth factor 1 (IGF-1)) [8][9][10][11]. In normal condition, in smooth muscle cells in presence of IGF-1 promotes protein synthesis but no cell proliferation, conversely, if the same smooth muscle cell culture with some stress condition as hyperglycemic and/or hypercholesterolemic, IGF-1 stimulates cellular proliferation [12,13]. On the other hand, insulin is 4 responsible for introduce glucose into the cell and subsequently be metabolized; both receptors IGF-1 and insulin belong to tyrosine kinase transmembrane family; this similarity allow that insulin can bind to IGF-1 receptor (IGF-1R) and also IGF-1 can bind to insulin receptor (IR), but with low affinity [14]. Therefore, it is possible that in cancer patients who already have diabetes and are treated with insulin, this insulin could have two effects: introduce glucose into the cancer cell or act as grow factor binding to IGF-1R and promoting proliferation, migration and invasion. IGF-1R is overexpressed in many cancer cell lines and types of tumors, including cervical cancer [11] so is probable that IGF-IR is activated in presence of insulin [15]. Therefore, our objective was to test if insulin may act as a growth factor in promoting tumoral growth. To evaluate the effect of high concentrations of glycemic and insulin over proliferation of cervical cancer cells, we developed the following experiments in vitro and in vivo.

Methods
HeLa, SiHa and CaSki cervical cells were obtained from American Type Culture Collection (ATCC), cells were maintained in Dulbecco's Modified Eagle´s Medium (DMEM) supplemented with 5% of fetal bovine serum (FBS), 37 o C and 5% of CO 2 .
HeLa cell line was also used in the "in vivo" studies. All experiments were prepared under sterility into a laminar flow hood. All compounds used were purchased from recognized commercial houses.

Effect of insulin on cell proliferation
Cell proliferation was evaluated by MTT (3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide; Sigma) and crystal violet assays. 1X10 3 cells/well was seed and expose with normoglycemic (5 mM) or hyperglycemic (25 mM) medium by 7 days alone or in combination with insulin (0, 0.01, 0.1 or 1 ug/mL). SiHa and CaSki cells were only expose to the highest concentration of insulin. All experiments were made by triplicate. Time of exposition was stablished as a result of previous experiments made in the laboratory and by reports in the literature using different cancer cell line. Insulin concentration was selected because high physiological levels has been reported as inductors of proliferative effects (up 10 nM) [16,17].

MTT Assay
The assay is based on the conversion of MTT to formazan, this compound is developed by viable cells with active metabolism. The measured absorbance is proportional to the number of viable cells. Briefly, 5 mg/mL MTT was dissolved in DMEM (1:9), the solution was added in each well and was incubated during 1 to 3 h.
Formazan dye was dissolved with 2-propanol acid, optical density was measured using an ELISA reader at λ= 570 nm. The results were expressed as the percentage of MTT reduction in relation to cero insulin concentration groups.

Apoptosis assay
Apoptosis was evaluated by flow cytometry. Briefly, once the cells were expose to hyperglycemia and/or insulin, they were washed with PBS and harvested with PBS-EDTA-Tripsin, followed by centrifugation and resuspended in DMEM, finally they were exposed to Guava Nexin Reagent, in the dark at room temperature for 20 minutes. 10,000 events were acquired in the Guava easyCyte equipment (Merck Millipore, Darmstadt, DE) per sample. Early and late apoptosis were determined, and the results were expressed as the total percentage of apoptosis. Experiments were carried out independently and by triplicate. Experiments were carried out independently and by triplicate.  [19][20][21]. This protocol was approved for the Instituto Nacional de Cancerología (011/045/IBI and CB714/11).

Experimental Design
Thirty-nine female mice (BALB/c (nu/nu) of 20-25 g body weight) were used. Random 7 way, one part of the mice was hyperglycemia induced using streptozotocin (100 mg/Kg/week) until the mice became hyperglycemic. Once the mice present hyperglycemia (≥ 125 mg/dL), then were inoculated subcutaneously into the back with 5X10 6

Biochemical quantifications
Glucose, cholesterol, triglycerides, creatinine and BUN were quantified in serum at the end of the experiments using a Beckman Coulter laboratory analyzer AU680 Chemistry System. Also, blood glucose concentration was measured with glucometer (Accu-Chek Performa, Roche) each week for hyperglycemia follow-up. Plasma insulin concentration was determined using an ELISA kit (Abnova).

Histological studies
Pancreas was perfused through pancreatic conduct catheter with saline solution 8 following by 10% formalin until fixation was completed. After that, the tissue was including in paraffin, sectioned at 3 µm thickness and stained with hematoxylineosin method. The slides were analyzed blindly. Langerhans islets by mm 2 were count in all groups.

Statistics analyses
All the results were present as mean ± SEM (standard error of the mean).

Results
To test our hypothesis, we first perform in vitro assays; we quantified percentage of cell proliferation with high glucose concentration (25 mM) with or without insulin (Table 1), we found that in Hela cells under hyperglycemia condition, glucose lead to significant increase of cell proliferation (12%) in compare to normoglycemic condition. Those results were observed with both MTT and crystal violet assay. After that, the cells were expose to different insulin concentration (0.01, 0.1 and 1 µg/mL); in Table 1 we show that in normoglycemic condition (glucose 5 mM), insulin did not modify cell proliferation, conversely, in hyperglycemic condition with high insulin concentration (1 µg /mL) this increased cell proliferation, more that 20%. Each experiment was the average of three independent experiments. a P ≤ 0.05 vs glucose 5 mM without insulin; b P ≤ 0.05 vs glucose 25 mM without insulin.
We also evaluated the effect of cell proliferation in SiHa and CaSki cell lines in hyperglycemia and hyperinsulinemia condition. In the Table 2, we showed that in both cell lines had a significant increase on cell proliferation in presence of hyperglycemia and this proliferation show a tendency to increases even more when insulin (1 µg /mL) was added. To elucidate a possible mechanism of action, we explored the apoptosis and cell cycle in the three cervical cancer cell lines. In Fig. 1 we show apoptosis percentage in Siha and CaSki cells. Clearly, we observed a decrease in the total percentage of apoptosis in the hyperglycemia group in compared to the control group with statistic significance. When insulin was added, this percentage decreases even more, although there was not significant difference respect to only hyperglycemia group. In the other hand, in HeLa cell line we did not find statistical difference when insulin was added, only we find a decreasing tendency. Whereas, any of three cells lines show change in cell cycle phases when the cells were treated with insulin or hyperglycemia (data no show).
Once we found that both conditions hyperglycemia and hyperinsulinemia promote the cell proliferation; we translate these assays to in vivo studies in order to observe if the results obtained in vitro are reproduced in vivo. Blood glucose levels were monitored during the twelve weeks, Fig. 2B shows the temporal course of blood glucose levels where it is clearly observed that animals administered with streptozotocin significantly increased serum glucose with respect to the control group. Systemic toxicity was evaluated recording body weight during all experiment (12 weeks), no difference was found in this parameter ( Fig. 2A).
Several biochemical parameters were quantified in order to know if the hyperglycemic status developed by the mice already have the metabolic alterations that occurs during diabetes in humans. Figure  In Fig. 4A we show a temporary course of tumoral volume of xenotransplants of HeLa cells. We found a significantly increases of tumor growth in all studied groups in comparison to the control group; in fact, insulin group, independently of hyperglycemia condition, shows the highest tumor growth, some mice of this group had to be sacrificed before to finish the treatment because of ethic guidelines does not allow us to work with such big tumor (Fig. 4B). We did not find significant changes in the tumoral volume in hyperglycemia groups with or without insulin, however both groups show an increase in tumoral volume, that can suggest that both insulin and glucose are activating signaling pathways that culminate in cell proliferation and/or decreases in apoptosis.
Those results correlate with a decrease of cell doubling time, while control group need approximately 20 days to duplicate the tumoral volume, insulin´s groups need only approximately 13.5 days (Fig. 4C).

Discussion
Current evidence demonstrates that hyperglycemia promote cell proliferation to triggers several direct and indirect pathways such as induction of epithelial mesenchymal transitions [24], high density in GLUT 1 receptors and used of lactate [25,26], increase leptin and pro-survival AKT/mTOR signaling [27], increase levels of inflammatory cytokines in blood circulation [28,29] and WNT/β-catenin signaling [30]. These studies suggest that high glucose levels have an influence in clinical course of cancer patients. In this study we have demonstrated in vitro that the combination of hyperglycemia and hyperinsulinemia act synergistically to enhance cell proliferation in cervical cancer cells and decrease apoptosis without cell cycle 12 changes.
Those results are similar to report by Masur and coworkers [31], they report that glucose and insulin increase proliferation cell and migratory activity in colon, mama, bladder and prostate cell lines through activation of phosphatidylinositol 3kinase (PI3K), protein kinase C alpha (PKCα) and human myosin light chain kinase (MLCK) proteins. However, in our study we did no find significant changes in the cell cycle, similar results were reported by Liu and coworkers in head and neck cells lines [30], unlike Masur et. al. [31] who found an increase number of transcripts of cyclin A1 and E, genes involved in the transition from G1 to S phase, nevertheless, they do not confirm these results with a whole cell, so additional assay will be done.
Both insulin and IGF-1 have similar structure and activate almost the same transduction pathway; IR has metabolic function, when insulin activate IR, adapter proteins that contain phosphor-tyrosine binding as insulin receptor substrate (IRS-1-4) are phosphorylated, this initiates an autocatalytic process that culminates with the activation of proteins that allow translocations of GLUT4 receptors to cellular membrane for transport of glucose inside the cell. While, IGF-1R has proliferation, grow and anti-apoptosis effects; once IGF-1R is activated by IGF-1, IRS and Shc (Src homology collagen) are phosphorylated, which subsequently interact with p85 and p110 subunit of PI3K protein who activated AKT and in turn promotes cell growth and protein synthesis. Both insulin and IGF-1 can bind of both of two receptors IR and IGF-1R (although their affinity to own receptor is higher). It has been reported that in pathological conditions with hyperinsulinemia, insulin can trigger proliferation pathway activated by IGF-1 in mama, colon, cervix, kidney and lung cancer cells lines [32][33][34][35][36][37]. In addition, IR and IGF-1R are closely linked stimulate cellular proliferation and induce metastasis in experimental models [29,38]. In this 13 study we found that, in vitro, both hyperinsulinemia and hyperglycemia increase cell proliferation and decrease apoptosis, whereas in vivo increases tumoral volume.
So, our results are related to what is reported in the literature.
By the way, Novosyadlyy et. al. [39] performed in vivo studies, where shows that high phosphorylation of insulin and IGF-1 receptors in mammalian tumor cell extracted from diabetic rats, suggesting that hyperinsulinemia also have an influence on IGF-1 pathway. Our study shows that in normoglycemic condition there are no changes in presences of insulin (Table 1), wherever, insulin plus hyperglycemia increases cell proliferation significantly (more than 25%), and the same effect was observed in cisplatin resistant cells (Table 2). Even though that change seems not so big, it is significant and enough to increases tumoral volume ( Fig. 4A) and cell doubling time (Fig. 4B). Similar results were show in smooth muscle cells (SMC), by Clemmons and coworkers [40]. These authors observed that the IGF-1 stimulates both migration and proliferation in SMC in presence of hyperglycemia through activation αVβ3 integrin and Shc phosphorylation. It is worth mentioning that IGF-1R is present in cervical cancer cell line [11].
In this study we only used insulin but not IGF-1, and the group administered with insulin showed an increase in serum insulin levels (Fig. 3C) and also, this result correlated with the increase in tumor volume (Fig. 4A), indirectly indicating the possible activation of the IGF-1 pathway. However, the diabetic groups treated or not with insulin, also showed an increase in tumor volume although the serum insulin levels were the same to the control group, which suggests that the effect is due to hyperglycemia condition more than insulin. According to these results, Villarreal-Garza and coworkers [41] reported that patients with breast cancer and hyperglycemia are associated with a poor outcome in contrast to normoglycemic 14 patients, conferring an elevated risk to die. This effect is probably due to the low chemotherapy response produced by hyperglycemia when induced high proliferation, so it will be necessary corroborate it.
Our results show an increase of insulin concentration in insulin treated group (Fig. 3C) and that result correlated with a hypertrophy in the Langerhans islets ( Fig. 3A). This effect supports the results reported by Gao and coworkers [42] finding that a high-fat diet impaired glucose homeostasis increasing insulin release and causing hypertrophy in the Langerhans islets, which in turn resulted in insulin resistance.
Insulin stimulates both, proliferation and tumoral growth of cervical cancer, and this could be one of the possible causes for what diabetic people with cervical cancer present a low response to chemotherapy in compare to normoglycemic patients.
This propose is similar that reported by the Liu group [30], where they found that the long cultivation with high-glucose environment promote that the head and neck cancer increase malignancy, regardless of the lymph node metastasis that occurs in this type of cancer.

Conclusion
The present study confirms that both hyperglycemia and hyperinsulinemia enhance cell proliferation and tumor growth in cervical cancer. So, clinicians must be permanently monitoring glucose levels to avoid hyperglycemia or hyperinsulinemia in decrement of cancer response. were used to acquisition of material and not had impact on the analysis, interpretation of data or the writing of the manuscript.

Abbreviations
Availability of data and materials: All the data generated or analyzed in this study are included in the manuscript or the supplementary information.
hepatic steatosis and hyperinsulinemia in outbred CD-1 mice. PLoS One. Tumoral volume. A) Temporary course of tumors from transplanted HeLa cells in nude mice u