Impact of AKT1 polymorphism on DNA damage, BTG2 expression, and risk of colorectal cancer development

This case-control study involved a total of 397 individuals including both colorectal cancer patients (CRC; n = 200) and population-based controls (n = 197). CRC patients (n = 200) from both sex, having age ≤ 60 and with documentary evidence of pathologically confirmed adenocarcinoma of colorectal cancer were included in this study. Subjects with mixed ethnic background, comorbidity, and patients who developed CRC at the age of above 60 years at the diagnosis were excluded. Determination of tumor stages and types were done by experienced pathologist at Institute of radiation and nuclear medicine (IRNUM), Peshawar. All patients and their guardians were informed about the nature of the study and important information of patients such as age, sex, ethnicity, medical records, pathology reports, drug history, family history, tumor size, tumor location and lymph node status etc. were obtained on a pre-designed proforma. Colorectal cancer risk factors such as taking red meat, vegetables, fibers, fruits and cooking choices and smoking history were also obtained. For control blood samples were collected from healthy individuals (n = 197) who had no sign of present or previous malignancy and no indication of CRC or nor any family history of cancer and had no blood relation with patients. Selection of control group of healthy donors was done on the basis of sex, age, smoking history and habits, residential, occupational and food intake. Informed consent of all the enrolled subjects was obtained on a questionnaire. The ethical approval was obtained from the institutional ethical board at Department of Biotechnology, University of Peshawar, Pakistan. Blood samples were collected both from colorectal cancer patients and controls at IRNUM Peshawar in 5 mL EDTA tubes and were stored at -20oC till further analysis.

DNA was extracted using DNA extraction kit (GeneJET Genomic DNA Purification Kit, Thermo Scientific, USA) and was quantified using UV-visible spectrophotometer (752 PC, China). Akt single nucleotide polymorphism was determined using polymerase chain reaction (PCR, Multigene Optimax, Labnet International, USA). PCR was performed in a 20 μL reaction mixture using allele specific primers. The sequences of primers and amplification condition are given in Table 1. The AA (379 bp), AG (245 and 379 bp) and GG (245 bp) genotypes were visualized with ethidium bromide and identified on agarose gel (2%) using UV transilluminator (Wealtec, USA).

Lymphocytes were isolated from fresh blood as described previously.21 Briefly, blood containing EDTA was mixed with phosphate buffer saline (PBS; Ca⁺² and Mg⁺² free) and layered over 2 mL ficoll / lymphocytes separation medium (LSMTM1077; Catalog Number: HiSep LSMTM 1077-LS001) in a 15 mL falcon tube. The mixture was centrifuged (2000 RPM for 30 min) that led to the formation of four distinct layers; the upper plasma layer, the second buffy coat layer containing lymphocyte and monocyte, the third ficoll layer (LSM) and the bottom layer of RBCs and cell debris. The buffy coat was isolated, mixed with 1 Ml PBS and centrifuged at 1500 RPM for 10 min. The pellet containing isolated lymphocytes washed with PBS, gently suspended in 1 ml PBS and used in subsequent experiments.

DNA damage in lymphocytes was assessed using comet assay, (also called single cell gel electrophoresis assay), as described previously.22 Briefly, cells were fixed in ethanol for 20 min, then hydrated in distilled water for 30 min followed by staining. The slides were washed with cold distilled water and mounted with the cover glass. For scoring of DNA comets, 100 stained nuclei were selected randomly from each group under the fluorescent microscope at 200x magnification and images were recorded. Total comet score was calculated as described previously.21

The isolated lymphocytes were lysed in a buffer containing Tris (50 mM, pH 7.4, Nacl (150 mM), EDTA (1.0 mM), phenylmethylsulphonyl fluoride (1.0 mM), aprotinin (1.0 μg/ml), leupeptin (1.0 μg/ml), NaF (1.0 mM, 1.0 mM) sodium orthovanadate, sodium deoxycholate (0.25 %) and Nonidet P-40 (1.0%). The extracted proteins were quantified and electrophoresed on SDS-PAGE, transferred onto a nitrocellulose membrane using immunoblotting kit. Membrane was incubated with anti-Akt, anti-pAkt and anti-tubulin and proteins were detected using immunoblotting detection kit Ab Signal™ (AbClon, Seoul, Republic of Korea). Antibodies for AKT and pAKT were purchased from Cell Signaling Technology while α-tubulin from Santa Cruz Biotechnology. α-Tubulin was used as a loading control.

Total RNA was extracted from purified PBMCs using Trizol reagent. RNA was reverse transcribed to cDNA using reverse transcription kit (Invitrogen). BTG2 expression was determined using conventional PCR followed by agarose gel electrophoresis. GAPDH was used as an endogenous control. The primer sequences and amplification conditions for BTG2 and GAPDH are given in Table 1.

Data was analysed using Minitab® 17 and was presented as Mean ± SD. Odds ratio (OR), 95% confidence interval (CI) were used to find out the association between AKT1 single nucleotide polymorphism and CRC risk. P ≤ 0.05 was considered as statically significant.

Overall 397 subjects including 200 colorectal cancer patients (cases) and 197 healthy individuals (control) were enrolled in this study and genotyping of AKT1 rs1130233 was performed to evaluate the status of rs1130233 polymorphism in case and control groups. Representative images of wild type GG, heterozygous mutatnt AG and homozygous mutatnt AA alleles for both control and CRC patients are given in Figure 1. The data has been presentated in Table 3. Among 200 colorectal cancer patients, 60 (30.0%) had wild type GG genotype, 120 (60.0%) had heterozygous mutatnt AG genotype while remaining 20 (10.0%) had homozygous mutant AA genotype. In control population, 90 subjects (45.69%) had GG genotype, 101 (51.27%) had AG and 6 (3.04%) had AA genotypes. The presence of AG and AA alleles were assoiated with colorectal cancer risk (OR = 1.80, CI = 1.18‒2.74, P = 0.01 for AG and OR = 5.00, CI = 1.90‒3.18, P = 0.001 for AA). The association between AKT1 rs1130233 polymorphism and colorectal cancer was also assessed using dominent and recessive models. For dominant model (GG vs AG+AA), homozygous wild type (GG) was present in 60 patients (30.0%) and 90 controls (45.69%) while heterozygous and homozygous mutatant alleles (AG+AA) were collectively present in 140 patients (70.00%) and 107 controls (54.31%). An increased risk for colorectal cancer (OR = 1.96, CI = 1.30‒2.96, P = 0.001) was observed for dominant model. For recessive model, (AA vs GG+AG), homozygous polymorphism was observed in 20 patients (10.00%) and 6 controls (3.04%) while homozygous wild type and heterozygous polymorphism (GG+AG) was collectively observed in 180 patients (90.00%) and 191 controls (96.96%). An increased risk for colorectal cancer was also observed for recessive model (OR = 0.28, CI = 0.11‒0.72, P = 0.01). Similarly, G and A allele frequencies were 0.60 and 0.40 respectively for cases and 0.7132 and 0.2868 respecively for control and hence follows Hardy Weinberg equilibrium. Overall genotype frequency of GG, AG and AA for both cases and control was 150 (37.78%), 221 (55.67%) and 26 (6.55%) respectively indicating that heterozygous AG genotype is more prevalent than GG and AA.

Figure 1

The colorectal cancer patients were sub grouped into colon and rectum cancer pateints and their association AKT1 rs1130233 polymorphism was determined (Table 4). Among overall 200 colorectal cancer pateints, 102 (51%) were colon cancer patients while 98 (49%) were rectum cancer patients. Among colon cancer patients, GG, AG and AA genoptes frequency were 29.41, 59.80 and 10.79% respectively. The AG and AA genotypes were assoiated with higher risk for development of colon cancer (OR = 1.81, CI = 1.08‒3.05; P = 0.02 for AG and OR = 5.50, CI = 1.87‒16.15; P = 0.001 for AA). An increased risk for colon cancer was observed for dominant (OR = 2.02, CI = 1.21‒3.36; P = 0.006) and recessive (OR = 3.85, CI = 1.38‒10.73; P = 0.01) models.

The association between AKT1 rs1130233 polymorphism and rectum cancer was also evalutaed (Table 5). Among 98 rectum cancer patients, GG, AG and AA genotypes frequencies were 30.61, 60.20, and 9.19% respectively. Both AG (OR = 1.75, CI = 1.04‒2.96; P = 0.04) and AA (OR = 4.50, CI = 1.48‒13.69; P = 0.008) were assoiated with higher risk for development of recctum cancer. An increased risk for rectum cancer (OR = 1.91, CI = 1.14‒3.18; P = 0.01) was observed for dominant and recessive models (OR = 3.22, CI = 1.11-9.32; P = 0.03).

On the basis of tumor location, the colorectal cancer patients were separted as colon and rectum cancer pateints and the association of AKT1 polymorphism was assessed. Among overall 200 colorectal cancer pateints, 102 pateints (51%) had colon cancer while 98 patients (49%) were rectum cancer patients. Among 102 colon cancer patients, 32 pateints (31.37%) had GG, 67 patients (65.69%) had AG and 3 patients (2.94%) possessed AA genotypes. Among 98 rectum cancer patients, 36 pateints (36.74%) had GG genotype, 55 patients (56.12%) had AG genotype while remaining 7 patients (7.14%) had AA genotype. The AG and AA polymorphism had equaleffect on the prevalce of both colon and rectum cancer as both of them showed non-significant difference (P > 0.05) for AG and AA transition.

To find out whether rs1130233 G to A transition can have effect on AKT expression, lymphocytes were isolated from various subjects of different genotypes (GG = 21, AG = 25 and AA = 06) and their AKT and pAKT proteins levels were determined using immunoblotting. The data indicated that G to A transition decreased AKT expression in both healthy in various individuals independent of their age, sex and health status (Figure 2A). The densitometry analysis revealed GG genotype carriers had significantly (P < 0.05) higher level of AKT followed by heterozygous AG carriers while the AKT expression was lowest in AA genotypes (Figure 2B). pAKT represents the active kinase, therefore, the phosphorylation status of AKT was also determined in GG, AG and AA carriers. pAKT level also showed a decreased intensity in GG>AG>AA order (Figure 2C). The data shows that substitution of G by A have a significant impact on AKT expression and activation and hence could have an effect on colorectal cancer development in different ways.

Figure 2

To find out the association of rs1130233 with genome integrity, DNA damage was assessed by comet assay. Because cancer patients have multiple genes mutation leading to DNA damages, therefore comet assay was performed only in control individuals carrying GG (n = 13), AG (n = 15) or AA (n = 4) alleles. Because age and life style can have an impact on DNA damage, therefore comet assay was performed in individuals of similar age groups, non-smoking subjects and subjects with similar dietary habits. Moreover, the frequency of AA genotype carriers in control individuals was very less, therefore, the combined total comet score was calculated for AG and AA genotype carriers (AG+AA) (Figure 3). The total comet score for individuals carrying AA genotype was 190 ± 30.5, AG genotype (110 ± 20.54) and GG 63 ± 15.70. The total comet score of AA genotypes was significantly (P < 0.05) greater than AG and GG genotypes. Also, AG carriers had significantly higher (P < 0.05) comet score than GG genotype indicating a greater DNA damage. The data thus indicates that GG allele of AKT1 contributes to genome stability.

Figure 3

Previously we have reported that AKT downregulates BTG2 expression in various types of cells.20 Moreover, BTG2 has been shown to be involved in DNA damage repair, we therefore determined the expression of BTG2 in GG (n = 20), AG (n = 24) and AA (n = 6) carriers of AKT1 rs1130233 single nucleotide polymorphism. BTG2 expression was determined in lymphocytes of control individuals of similar age groups and life style. The BTG2 expression profile and its densitometric analysis is given in Figure 4. The data shows that an inverse association between AKT activation and BTG2 expression in genotype dependent manner. BTG2 expression was significantly (P < 0.05) higher in AA and AG carriers compared to GG individuals indicating that various genotypes of AKT differentially regulate Btg2 gene expression and hence will impart distinct effect of various cellular processes.

Figure 4