Upregulation of p16^INK4A in Peripheral White Blood Cells as a Novel Screening Marker for Colorectal Carcinoma

Abstract

Objective:

Screening of colorectal cancer (CRC) is important for the early detection. CRC is relating to aging and immuno-senescence. One such senescent marker is p16^INK4A expression in immune cells. The objective of the study is to investigate the protein expression of p16^INK4A in peripheral white blood cells as a screening marker for colorectal cancer.

Methods:

A case-control studies were conducted. Cases were patients with colorectal cancer and controls were matched with cases based on age and sex. Peripheral blood was collected from patients and controls and the protein p16^INK4A was measured with immunofluorescent techniques. The p16^INK4A levels from cases and controls were evaluated using ROC analysis to be used as a screening marker in CRC patients. Mean fluorescent intensity of p16^INK4A of cases and controls were analyzed in CD45+, CD3+ or CD14+ cells. The p16^INK4A levels of cases were also correlated with clinical data.

Result:

Statistically significant increased expression of p16^INK4A levels were found in cases compared to controls. p16^INK4A in peripheral immune cells had 78% sensitivity and 71% specificity which can possibly be used as a diagnosis tool for colorectal cancer. P16^INK4A-positive cell percentage and mean florescent intensity were significantly higher in CD45+ cells, CD3 positive cells and CD14 positive cells. No significant correlation was observed with the clinical data and p16^INK4A level of CRC patients.

Conclusion:

The significant increase of p16 ^INK4A expression level in peripheral immune cells represents potential for use as a CRC screening marker.

Introduction

Colorectal cancer (CRC) is the third most common cancer and second highest leading cause of cancer deaths worldwide (Inadomi and Jung, 2020), accounting for about 10% of cancer incidence and 9.4% of cancer deaths around the world in 2020 (Xi and Xu, 2021). Disease incidence can vary with sex, age, geographical regions and lifestyles differences, where higher incidence has been reported in males than in females and is higher in more-developed regions than in less-developed regions (Rawla et al., 2019). The majority of CRC patients have sporadic diseases with no family history of CRC, while about 5% to 10% of CRC cases are hereditary (Rawla et al., 2019).

CRC is more common in aging population. Most sporadic CRC patients are >50 years of age, with about 75% of rectal cancer patients and 80% of colon cancer patients diagnosed in their 60s and above (Kuipers et al., 2015). Aging relates to the colonic epithelial proliferation (Roncucci et al., 1988; Holt et al., 2009), epigenetic alterations and genomic instability (Mutirangura, 2019). The other important thing is the collective changes of the immune system, termed as immuno-senescence, can also be seen with age (Xu et al., 2020). Although immune responses may inhibit cancer growth, they can also stimulate cancer growth and metastasis, as a consequence of immuno-senescence (Pawelec, 2017) and tumor-associated inflammation (Grizzi et al., 2013). One of the aging markers is p16^INK4A.

p16^INK4A is a cell cycle regulator, functioning as a cyclin-dependent kinase inhibitor of the INK4 family (Laphanuwat and Jirawatnotai, 2019) and tumor suppressor protein (Buj and Aird, 2019). It in its canonical pathway (RB pathway) can act as an aging marker (LaPak and Burd, 2014). Increased expression of p16^INK4A is observed in pathophysiological conditions such as tumorigenesis and senescence (Liu et al., 2009).

Therefore, it is worth to elucidate p16^INK4A in CRC patients’ immune cells as a marker. This research aimed to measure the protein expression of p16^INK4A in peripheral white blood cell of CRC patients for application as a diagnostic tool. Then, we investigated the CD3+ cells represented T lymphocytes and CD14+ cells represented monocytes to find whether or not they manifest the p16^INK4A protein.

Materials and Methods

Study design and target population

Cross-sectional case-control studies were done at the King Chulalongkorn Memorial Hospital. All samples were collected from May 2021 to December 2021. Normal samples were collected from patients without a family history of cancer, autoimmune diseases and showed negative CRC screening results from colonoscopy. This group served as healthy controls for this study. Sample size was calculated from the pilot study using the formula N = 2(Zα/2 + Zβ)² * Ϭ² / (X̅₁-X̅₂)² with α significance level at 0.05 and power β of 90%, which required at least 53 participants in each group. Colorectal cancer staging was assessed by the American Joint Committee on Cancer TNM system by a pathologist. Venous whole blood (2 mL) with anticoagulant EDTA was collected from participant based on WHO guidelines.

p16 ^INK4A Assay

Difference in p16^INK4A clones can affect the outcomes of the experiments (Sawicka et al., 2013; Shain et al., 2018). Therefore, we validate the p16^INK4A antibodies with cell lines: Hela (American Type Culture Collection: ATCC^® CCL-2™), HCT116 (American Type Culture Collection: ATCC^® CCL-247™), and keratinocytes. Antibodies to p16^INK4A (cell signaling) were rabbit monoclonal antibodies. D3W8G (92803) was raised against synthetic peptide residues surrounding Ala34 of human p16^INK4A protein. D7CM1(80772) was produced from the synthetic peptide residues surrounding Ala143 of human p16^INK4A protein. D3W8G and D7CM1 antibodies showed positive staining stain in Hela cells (Figure 1A and 1B). D7CM1 antibodies had strong cytoplasmic staining in keratinocytes (Figure 1C). D3W8G antibodies showed nuclear staining in keratinocytes (Figure 1D). Hct116 cells were used as negative control for p16^INK4A staining (Figure 1E and 1F) (Myöhänen et al., 1998). Cytoplasmic p16^INK4A could have various functions other than its canonical function of cell cycle regulation and senescence (Buj and Aird, 2019). Therefore, we focused on the D3W8G antibodies.

Figure 1.

Immunofluorescence Staining of p16^INK4A Antibodies in Cell Lines, (a) Hela cells with p16^INK4A D7CM1 antibodies showing strong cytoplasmic staining, (b) Hela cells with p16 ^INK4A D3W8G antibodies, (c) Rat epidermal keratinocytes with p16^INK4A D7CM1 antibodies showing cytoplasmic staining, (d) Rat epidermal keratinocytes with p16^INK4A D3W8G antibodies showing nuclear staining, (e) HCT116 with p16 ^INK4A D7CM1 antibodies as negative control, (f) HCT116 with p16 ^INK4A D3W8G antibodies as negative control

Buffy coat layer preparation

Whole blood was centrifuged at 1000g for 12 min at room temperature. The buffy coat layer was collected, and the red blood cells (RBCs) were lysed with 20 volumes of RBC lysis solution (Abcam, USA). The lysis reaction was terminated by adding an equal volume of phosphate buffered saline (PBS) and centrifuged for 5 min at 400g. Cells were washed with PBS 3 times and white blood cells (WBCs) were counted with a counting chamber. WBCs were then fixed with 4% paraformaldehyde for 15 min (1 million cells/mL fixative), and washed 3 times with deionized water. WBCs were placed in 96-well plates (approximately 40,000 cells/ well) and stored at room temperature before immuno-fluorescence staining.

Immunofluorescence staining

Sample wells containing WBCs were washed with PBS 3 times. One hundred microliter of 0.5% Triton X-100 was added into each well and incubated for 10 min at room temperature. Wells were then washed with PBS 5 times. Then, 100μl of 5% (w/v) BSA was added into each well for 30 min. One hundred μL of diluted primary antibodies (anti-p16^INK4A and anti-CD45) in 1% (w/v) BSA, were added into each well and incubated overnight (16 h) at 4°C. Dilutions of anti-p16^INK4A primary rabbit antibody (Cell Signaling) and anti-CD45 primary mouse antibody (Abcam) were 1:1,000. Anti-CD3 primary mouse antibody (Cell signaling) and anti-CD14 primary mouse antibody (Cell signaling) were diluted 1:500. Wells were washed with 100 μL of 0.1% Tween PBS 3 times. One hundred μL of 1:1,000 diluted secondary fluorescent antibodies (anti-rabbit FITC and anti-mouse Cy3) (cell signaling) in 1% (w/v) BSA, were added into each well and incubated for 2 h at room temperature in the dark. Wells were washed with PBS 3 times. One hundred μL of DAPI (final concentration of 1μg/ml) (for nuclear staining) was added into each well for 10 min and washed with PBS for 3 times.

Image analysis

Three-color images were captured as follows: blue (DAPI 358 - 461 nm); green (FITC 500 - 520 nm) and red (Cy3 560 - 570 nm) with a motorized fluorescence microscope IX83 (Olympus Co., Ltd., USA). All images were acquired using a defined experimental protocol throughout the study. Briefly, 20 fields of 20X objective (5 columns and 5 rows) were taken with the fixed exposure time for DAPI (7s), FITC (200ms) and Cy3 (300ms). The fluorescence intensity of cell was calculated using CellSens imaging software after setting the range of intensity values (70 to 150) and perimeter values (18 to 55 nm). Mean fluorescence intensity (MFI) was then acquired from the average fluorescence intensity of cells from the whole image. In double immunofluorescence staining, the positive region was identified by using CancerScreen.exe program (Puttipanyalears et al., 2021). P16^INK4A-positive/ CD-positive cells were identified by the signals of red and green in the same spot with the criteria of intensity value (60 to 150) and perimeter value (18-100 nm).

Ethical approval

This study was conducted under the regulations of the Institutional Review Board, Faculty of Medicine, Chulalongkorn University, COA number: 1580/2021. This study was conducted in accordance with the Declaration of Helsinki. All study participants provided informed consent.

Statistical analysis

Normal distribution values of the study groups were assessed by Shapiro-wilk test. To detect the significant differences of measurements between CRC and healthy control participants, independent sample t-tests was used. For comparison between more than two groups, ANOVA was calculated. Data are expressed as mean±SD, 95% CI and p-value. Receiver operating characteristic (ROC) curve and Youden index were used to evaluate the sensitivity and specificity of the marker. GraphPad prism-8 (GraphPad Software Inc., USA) was used for statistical analysis and graphical illustrations.

Results

p16 ^INK4A protein level in white blood cells of cases and controls

p16^INK4A levels were measured in the peripheral white blood cells of CRC patients (Figure 2A) and healthy controls. Since p16^INK4A levels can depend on donor age, the healthy control was age- and sex-matched to the CRC cohort. Comparison of cases with controls revealed a statistically significant increase in MFI per cell between CRC patients compared to controls (Figure 2B) (95% CI = 2.8 to 11.9, p = 0.003).

Figure 2.

(a) Immunofluorescence staining of WBCs showing DAPI staining (i), p16^INK4A staining (ii), CD45 staining (iii) and merged image (iv), the bar marker on the images represents 20μm, (b) Scatter plot of p16^INK4A MFI of healthy controls and p16^INK4A MFI of CRC group (age and sex adjusted cohort study), (c) Scatter plot of p16^INK4A MFI of healthy controls and p16^INK4A MFI of CRC group. The bar on the scatter plot represents mean value

Then, we increased the number of samples both in CRC and control, not adjusted by age and sex. The total number of CRC patients in this study was 72 with mean age of 64.54±11.3 years, while that of healthy controls was 79 with a mean age of 63.91±8.8 years. Cases and controls demographics are shown in Table 1. The minimum age of study participants was 31 years and maximum age was 89 years. The values in samples were normally distributed according to Shapiro-Wilk test. Therefore, we calculated the difference of the two groups with the independent sample t-test. Among the total number of CRC 72 and control 79, there was also a statistically significant increase MFI of p16^INK4A in WBC of CRC patients compared to controls (Figure 2C) (95% CI = 4.5 to 8.3, p<0.001).

Table 1.

Cases and Controls Demographics and Clinical Criteria of CRC Patients

Parameters	N	%	Elderly population for each category (Age 65 and above) N(%)"
Age (range) of Normal Controls
<50	3	3.8	0
50-64	39	49.4	0
65 and above	37	46.8	37 (100)
Gender of Normal Controls
Male	31	39.2	16 (51.6)
Female	48	60.8	21 (43.8)
Age (range) of CRCs
<50	5	6.9	0
50-64	27	37.5	0
65 and above	40	55.6	40 (100)
Gender of CRCs
Male	45	62.5	24 (53.3)
Female	27	37.5	16 (59.3)
Stages of CRCs
I	6	8.6	3 (50.0)
II	17	24.3	13 (76.5)
III	26	37.1	12 (46.2)
IV	21	30	10 (47.6)
Tumor size of CRCs
T1	4	5.7	2 (50)
T2	8	11.4	5 (62.5)
T3	42	60	22 (52.4)
T4	16	22.9	9 (56.3)
Lymph node involvement of CRCs
N0	29	41.4	19 (65.5)
N1	29	41.4	15 (51.7)
N2	12	17.1	4 (33.3)
Metastasis status of CRCs
Metastasis	21	30	11 (52.4)
Non metastasis	49	70	27 (55.1)
Tumor histological types of CRCs
Well differentiated adenocarcinoma	14	20	8 (57.1)
Moderately differentiated adenocarcinoma	47	67.1	24 (51.1)
Poorly differentiated adenocarcinoma	1	1.4	1 (100)
Epithelioid cell carcinoma	1	1.4	0
Mucinous adenocarcinoma	5	7.1	4 (80)
Signet ring cell carcinoma	2	2.9	1 (50)
Left or right-side tumor of CRCs
Left side	62	88.6	32 (51.6)
Right side	8	11.4	6 (75.0)
Tumor location of CRCs
Cecum	2	2.9	2 (100.0)
Ascending colon	4	5.7	3 (75.0)
Hepatic flexure	1	1.4	0
Transverse colon	1	1.4	1 (100.0)
Descending colon	5	7.1	5 (100.0)
Sigmoid colon	12	17.1	7 (58.3)
Recto-sigmoid colon	2	2.9	1 (50.0)
Rectum	43	61.4	19 (44.2)

Evaluation of p16 ^INK4A as a marker in CRC patients

p16^INK4A levels in peripheral immune cells were observed to be used as a marker in CRC patients, as the p16^INK4A level of peripheral white blood cells increased in CRC patients compared to controls. ROC curve was calculated using MFI of p16^INK4A in CD45+ cells of patients and controls. Area under the curve was 78% (p <0.001) with the standard error value of 0.04. There was 78% sensitivity and 71% specificity with Youden index of 0.48, cut-off value of 83.26 MFI (Figure 3). The 2x2 table with the numbers of true positive (TP), false negative (FN), false positive (FP) and true negative (TN) was shown in Table 2. From 2x2 table, sensitivity (TP/TP+FN) is 78%, specificity (TN/FP+TN) is 71%, positive predictive value (TP/TP+FP) is 71%, negative predictive value (TN/FN+TN) is 78% and accuracy (TP+TN / TP+FP+FN+TN) is 74%.

Figure 3.

ROC Analysis of p16^INK4A in White Blood Cells of CRC Patients and Healthy Controls. The Youden index point at the vertical solid line and horizontal solid line showing 78% sensitivity and 71% specificity

Table 2.

2x2 Table of Cases and Controls

	CRCs	Controls
Test positive	TP (56)	FP (23)
Test negative	FN (16)	TN (56)

p16 ^INK4A expression according to clinical data of CRC patients

p16^INK4A protein levels in CRC samples were analyzed based on clinical parameters such as the stages of CRC, tumor size, lymph node involvement, metastasis status, patients age and histological types of tumors (Figure 4 A-F). Of the 72 patients from CRC group, clinical data of 2 patients were not available, resulting their exclusion from the analysis. CRC samples were categorized based on the age as <50-year, 50-64 year, and 65 and above year. Half of patients (55.6%) were in the population of 65 and above. There is no significant difference of p16^INK4A level among the age groups (F (2, 69) = 0.4, p = 0.7). Regarding to gender, 62.5% of patients in the study were male and showed higher MFI of p16^INK4A, but no statistically significant difference (t=1.3, p=0.2). When the p16^INK4A levels were observed according to the stages, size and lymph node involvement of tumors, their levels were similar, showing the statistical values of stage (F (3, 66) = 0.1, p = 0.9), tumor size (F (3, 66) = 0.3, p = 0.8) and lymph node involvement (F (3, 66) = 0.2, p = 0.8). Thirty percent of patients had metastasis, and their p16^INK4A MFI were not statistically different from that of non-metastasis patients (t = 0.5, p = 0.6). Clinical data of CRC patients were shown in Table 1.

Figure 4.

Correlation of p16^INK4A Mean Fluorescence Intensity of CRC Patients with Clinical Parameters. a, Data from CRC patients were grouped according to age; b, gender; c, Clinical stages of CRC; d, tumor sizes; e, LN involvement; f, metastasis or non-metastasis

Expression of p16 ^INK4A in peripheral blood CD45+ cells, CD3+ cells and CD14+ cells of cases and controls

We used double immunofluorescence staining to investigate lymphocytes and monocytes positive for p16^INK4A. Antibodies of p16^INK4A and CD3 were used to detect p16^INK4A-positive lymphocytes (Figure 5A). For monocytes population, antibodies of p16^INK4A and CD14 were used (Figure 5B). Mean fluorescence intensity of p16^INK4A in CD3+ cells and CD14+ cells were calculated in CRC patients and healthy controls. A significant increase in levels of p16^INK4A were observed in CD3+ cells (95% CI = 0.2 to 9.1, p = 0.04) and CD14+ cells (95% CI = 0.2 to 6.3, p = 0.04) of CRC patients. Expression of p16^INK4A was higher in CD3+ cells than in CD14+ cells (95% CI = 0.4 to 6.6, p = 0.03) (Figure 5C).

Figure 5.

(a) Immunofluorescence staining of CD3+ peripheral immune cells showing DAPI staining (i), p16^INK4A staining (ii), CD3 staining (iii) and merged image (iv), (b) Immunofluorescence staining of CD14+ peripheral immune cells showing DAPI staining (i), p16^INK4A staining (ii), CD3 staining (iii) and merged image (iv), (c) Scatter plot of p16^INK4A MFI of healthy controls and p16^INK4A MFI of CRC group in CD3+ and CD14+ cells. The bar on the scatter plot represents mean value

The number of p16^INK4A-positive white blood cells increased in CRC patients, revealing a significant difference in CD45+ subset (95% CI = 3.2 to 24.3, p = 0.01), in CD3+ subset (95% CI = 0.06 to 17.1, p = 0.05) and in CD14+ subset (95% CI = 1.0 to 13.2, p = 0.02) (Figure 6A). When the number of p16^INK4A positive cells and MFI of each cell subset were combined in a graph, there were increasing trend of higher value in p16^INK4A and percent positive cells in CRC groups (Figure 6B, 6C and 6D).

Figure 6.

(a) Comparison of p16INK4A positive cells in white blood cells, CD3+ cells and CD14+ cells between CRC group and healthy control group, (b) Correlation of p16^INK4A-positive cells percentage and p16^INK4A mean fluorescent intensity of CRC patients and healthy controls in white blood cells, (c) Correlation of p16^INK4A-positive cells percentage and p16^INK4A mean fluorescent intensity of CRC patients and healthy controls in CD3+ cells, (d) Correlation of p16^INK4A-positive cells percentage and p16^INK4A mean fluorescent intensity of CRC patients and healthy controls in CD14+ cells

Discussion

With immunofluorescence technique, we can measure the intensity of the positive cells and number of positive cells. The MFI of p16^INK4A in WBC of CRC patients had a significantly higher value than that of healthy control group in this study. P16^INK4A in peripheral immune cells represents 78% sensitivity and 71% specificity for application as a potential colorectal cancer screening marker. Proteins, DNA, RNA and metabolites from tissue, blood, stool and urine have been used in CRC for screening, diagnosis and monitoring, but have varying degree of success to be used as an effective biomarker (Loktionov, 2020). Colonoscopy has been used as the gold standard for the diagnosis of CRC (Hazewinkel and Dekker, 2011). Blood-based protein biomarkers for detection of CRC have various sensitivity and specificity, while using protein panel increase the detection rate of CRC (Loktionov, 2020).

The increased expressions of p16^INK4A have been notified in the peripheral blood cells of testicular cancer survivors (Bourlon et al., 2020) and breast cancer survivors (Sanoff et al., 2014). Presence of p16^INK4A in the immune cells can represent immuno-senescence (Liu et al., 2009). Immuno-senescence is initiated earlier in men than in women, likely due to hormonal differences between males and females, as estrogen enhances immune responses, while progesterone and androgens favor immune suppressive actions (Ostan et al., 2016). Therefore, we investigated age and sex-adjusted CRC patients and controls, the MFI in CRC group is significantly higher, meaning that immuno-senescence can contribute to CRC patients regardless of age and sex. Immuno-senescence can result from oxidative stress, cellular and DNA damage, chronic inflammation and cytotoxic therapy (De Padova et al., 2021). According to Giunco et al., (2019) immuno-senescence in CRC patients can lead to negative consequences like disease relapse, progression and death.

CRC group showed increased p16^INK4A levels compared to healthy controls (mean age 63.91±8.8 yrs). However, stages of CRC were not correlated to the level of p16^INK4A which is similar to the finding of Milde-Langosch et al., (2001) that reported p16^INK4A expression was not correlated with the clinical stages of breast cancer. MFI value of p16^INK4A were higher in some categories and was found to be relatively high in the categories of late stage, male patients and old age, but there were no statistically significant differences. The immune landscape in elderly population and metastasis patients differ from other patients in both innate and adaptive immune system (Weng, 2006; Fulop et al., 2017; Blomberg et al., 2018). The various factors can influence the level of p16^INK4A in immune cells, as p16^INK4A has an impact on immune surveillance (Sznurkowski et al., 2017; Leon et al., 2021). Limitations of this study include the small sample size (n) of some categories and due to the nature of this cross-sectional study, we could not prove that p16^INK4A level correlates to the prognosis of the patients.

Protein expression of p16^INK4A was higher in T cell subsets of CRC patients. T cells are important for tumor immunity and immunotherapy (Woolaver et al., 2021). Senescent T cells relate to progression of cancers (Vicente et al., 2016). Senescent T cells in CRC patients also relate to patient negative outcomes such as disease relapse, disease progression and death (Giunco et al., 2019). p16^INK4A expression in peripheral blood T cells is associated with chronologic age, molecular age, and IL-6 production (Liu et al., 2009; Burd et al., 2020). IL-6 is important in human frailty (Soysal et al., 2016) and relates to cellular senescence (Kojima et al., 2013). p16^INK4A protein expression increase in T cells of peripheral blood and bone marrow of acute lymphoblastic leukemia and its expression correlates to senescent features (Chebel et al., 2007).

Monocytes also expressed p16 ^INK4A protein in CRC patients in this study. Monocytes are a heterogenous group of cells. A typical portion (about 85%) of monocytes are classical monocytes (CD14high CD16-), the rest (5 to 10 % each) are intermediate monocytes (CD14+ CD16+) and non-classical monocytes (CD14dim CD16high) (Coillard and Segura, 2019; Kapellos et al., 2019). Monocyte recruitment to the inflammation sites and cancer tissue is an important process and it can occur within hours of onset of inflammation (Coillard and Segura, 2019). Moreover, monocytes can be transformed into monocyte-derived macrophages (Mo-mac) and monocyte-derived dendritic cells (mo-DC) (Coillard and Segura, 2019). On the other side, p16^INK4A can stimulate the maturation of mo-DC through MAPK pathway (Sunthamala et al., 2020). Mo-DC can be found especially in the intestines (Watchmaker et al., 2014). In addition, p16^INK4A can involve in macrophage polarization (Cudejko et al., 2011; Kuo et al., 2011; Hall et al., 2017).

In conclusion, the well-known aging marker p16^INK4A in peripheral blood of CRC patients has been observed in this study. The statistically significant increase of p16^INK4A level in WBCs revealed a potential diagnostic tool for CRC. However, the protein expressions of p16^INK4A were divergent with CRC patients’ clinical data. Functions of p16^INK4A in CRC patient immune cells are interesting topics that warrant further studies.

Abbreviation

ATCC: American Type Culture Collection

CRC: Colorectal cancer

MFI: Mean fluorescence intensity

PBS: Phosphate buffer saline

RBC: Red blood cell

ROC: Receiver operating characteristic

WBC: White blood cell

Author Contribution Statement

The experiments were conducted and designed by CP and AM. The clinical samples were collected by KAT, PA and CP. KAT and CP performed and optimized the immunofluorescent staining. The results were analyzed and interpreted by KAT, CP and AM. KAT wrote the manuscript. CP, SWE and AM reviewed and edited the manuscript. All authors read and approved the final manuscript.