The number of individuals diagnosed with colorectal cancer (CRC) has been on an alarming upward trajectory over the past decade. In some countries, this cancer represents one of the most frequently diagnosed types of neoplasia. Therefore, it is an important demand to study the pathology underlying this disease to gain insights into the mechanism of resistance to treatment. Resistance of tumors to chemotherapy and tumor aggressiveness have been associated with a minor population of neoplastic cells, which are considered to be responsible for tumor recurrence. These types of neoplastic cells are known as cancer stem cells, which have been previously reported to serve an important role in pathogenesis of this malignant disease. Slovakia has one of the highest incidence rates of CRC worldwide. In the present study, the aim was to classify the abundance of selected stem cell markers (CD133, CD166 and Lgr5) in CRC tumors using flow cytometry. In addition, the methylation status of selected genomic regions of CRC biomarkers (
Colorectal cancer (CRC) is a prevalent neoplastic disease (
Epigenetic changes, in particular DNA methylation, has been reported to serve a vital role in colorectal carcinogenesis (
CD166 [(activated leukocyte cell adhesion molecule (ALCAM)] is a transmembrane type-1 glycoprotein (
Leucine-rich repeat-containing G protein-coupled receptor 5 (Lgr5) belongs to the G protein-coupled receptors family and is located on the surfaces of intestinal stem cells (
O6-methyguanine-DNA methyltransferase (MGMT) is a DNA repair enzyme that removes mutagenic and cytotoxic adducts from O6-guanine molecules in the DNA (
Therefore, the aim of presented study was to determine the abundance of selected stem cell markers (CD133, CD166 and Lgr5) on the surfaces of CRC cancer cells isolated from surgically resected colorectal tissues from patients using flow cytometry. In addition, the methylation status in the promoter regions of selected genes (
Primary (PT) and metastatic tumor (MTS) samples were obtained during surgical resection from 30 patients with diagnosed CRC in collaboration with The Department of Pathological Anatomy and Clinic of Surgery and Transplant Center, Martin University Hospital, Comenius University in Bratislava, Jessenius Faculty of Medicine in Martin (Martin, Slovakia). Sample collection and processing protocols were reviewed and approved by The Ethics Committee of Jessenius Faculty of Medicine in Martin (approval no. EK 1856/2016). All patients signed the informed consent document prior to surgery. The patient's clinical protocols were reviewed for clinical data, diagnosis, sex and age (
Paraffin blocks were generated from the CRC tissues and fixed in 10% formalin. (room temperature, 24 h). Tissue was dehydrated by passing the tissue through series of alcohol [70% ethanol (1×), 96% ethanol (4×) and xylene (3×)] and finally placed in warm paraffin wax. Tissue sections (5 µm) rehydrated with decreasing strengths of alcohol and finally with water were then submitted for hematoxylin-eosin staining. The hematoxylin nuclei staining step (8 min, room temperature) was followed by washing with tap water and then distilled water. The histology sections were counterstained with 1% eosin alcoholic solution (1 min, room temperature) and washed again. The slides were then observed and images were recorded using a BX53 light microscope (Olympus Corporation). Representative images are shown in
After surgical resection, the tumors were transferred to the Department of Pathological Anatomy and dissected for histological evaluation. Small pieces (~0.5 cm) of tissue were stored in RNAlater buffer (cat. no. AM7020; Thermo Fisher Scientific, Inc.) at −80°C for future DNA analysis. Any residual tissues were placed into a 50-ml Falcon tube containing DMEM/F-12 + GlutaMAX™ medium (cat. no. 10565-018; Gibco; Thermo Fisher Scientific, Inc.) supplemented with Penicillin-Streptomycin (cat. no. 15070063; Gibco; Thermo Fisher Scientific, Inc.) and 10% FBS (cat. no. 10082-147, Gibco; Thermo Fisher Scientific, Inc.) and stored at 4°C until further analysis. Before analysis, tumor samples were sterilized with 30–50% ethanol for 10 sec (at room temperature) as described previously (
After the collection of samples, frozen aliquots were quickly defrosted, washed 2X with sterile PBS and counted. Cells were first blocked using a FACS buffer [PBS, pH 7.4; 0.001 M EDTA; and 5% mouse serum (cat. no. ab7486, Abcam)] for 30 min on ice. After blocking, ~100,000 primary cancer cells were again re-suspended in 100 µl FACS buffer, transferred into 5-ml FACS tubes and stained with fluorescently-labeled antibodies [PE-anti-human Lgr5 (GPR49) antibody, PE-anti-human CD166 antibody and APC-anti-human CD133/2 antibody, see
From the total of 30 CRC samples, three samples were excluded due to insufficient DNA concentration (
Bisulfite-converted DNA was amplified using the PyroMark PCR kit (Qiagen, Inc.) in accordance with manufacturer's protocol. For the analysis of the selected regions of
PCR products were visualized by electrophoretic analysis (1.75% agarose gel) under UV light. Analyses were conducted according to the manufacturer's protocol, which was described previously (
The data were explored and analyzed in R ver. 4.0.5 (
The discriminative ability of the studied markers for discriminating between PT and MTS was assessed using the RF machine learning algorithm as implemented in R library randomForestSRC (
Expression of CD133, CD166 and Lgr5 markers was detected with fluorescent antibodies by flow cytometry on the surface of the cells isolated from tumor samples in a cohort of patients diagnosed with CRC and the percentage of the cells positive for these markers was recorded. All 30 tissue samples were analyzed for the expression of CD133 and Lgr5 markers. However, only 24 samples were analyzed for the expression of CD166, due to the insufficient number of cells in the six clinical samples. Tumor cells showed the highest average rate of CD166 (63. 7% of positive cells), followed by Lgr5 (31%) and CD133 (26.5%). For Lgr5 and CD166 expression, the difference were not found to be significant between the MTS and PT subgroups. Kruskal-Wallis test of the median expression of CD133 at stages 1, 2, 3 and 4 [Stage 1 represents (T1-T2, N0 and M0), Stage 2 (T3-T4,N0,M0), Stage 3 (any T, N1-N2, M0) and Stage 4 (any T, any N, M1) on TNM scale] revealed significance (P=0.034;
To assess the ability of the studied markers for discriminating between PT and MTS, the imbalanced RF machine learning algorithm was trained before the importance measure was used to rank the predictors. Since TNM staging appeared to be the most important predictor, another RF was trained without TNM staging. The discriminative ability of the RF without TNM staging was found to be substantially lower (AUC=64%,
Samples treated with sodium bisulfite were used for pyrosequencing analysis. In the PCR reaction, cytosine was converted to uracil or thymine in the PCR product, whereas methylated cytosines remained unchanged. In this part of the study detection of methylated regions in the sequences of five genes,
Following the application of Kruskal-Wallis test on the median expression levels of
There was a significant association between Sex and the origin of the tumor (P=0.042;
The main aim of the present study was to measure the expression of three previously reported cancer stem cell markers CD133, CD166 and Lgr5 in addition to assessing the methylation status of specific regions of selected genes
In present study, flow cytometry analysis of selected surface markers on tumor tissues isolated by surgery was performed, to measure their levels of expression in CRC. In addition, the extent of DNA methylation of five selected genes that have been previously documented to serve a role in CRC carcinogenesis was measured. The percentage of cells, positive for CD133, a known stem cell marker for colon CSCs (
Differences in the expression of CD166, another marker that has been previously found to be associated with CSCs in CRC (
Subsequently, the methylation status of two biomarkers,
Cancer-specific promoter hypermethylation of
To conclude, flow cytometry analysis of three cancer stem cell markers was performed in a cohort of 30 patients with CRC. Positivity for each of the three individual marker was found to be associated with age, sex, PTs and MTS, TNM stage, grade, presence of invasions (lymphatic, venous and perineural invasion) and the sidedness of tumor. Analysis of DNA methylation in the specified genomic regions of the five CRC biomarkers was also performed, which revealed the highest level of methylation in the
Not applicable.
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
JS, SM and ZL designed the study. SM, JS and KJ performed the experiments. MG conducted the statistical analysis. MK, JM, MV, RK, MP, JJ, PM, LL and EG obtained and handled colorectal specimens and clinical data. JS, SM, MG, ZL and EM analyzed the data and were major contributors in writing the manuscript. MK and JM performed the histological examination of the samples and clinical data. ZL, JS and EH supervised the entire study and participated in study design and coordination. SJ and MS confirm the authenticity of all the raw data. All authors read and approved the final manuscript.
Sample collection and processing protocols were reviewed and approved by The Ethics Committee of Jessenius Faculty of Medicine in Martin (approval no. EK 1856/2016; Martin, Slovakia). All patients signed the informed consent document prior to surgery.
Not applicable.
The authors declare that they have no competing interests.
Pathological analysis. Primary colonic adenocarcinoma tissues with (A) cribriform, (B) papillary or (C) tubular/tubulocystic growth patterns, showing typical ‘dirty’ necrosis and peritumoral desmoplastic reactions. (D-F) Metastatic colorectal adenocarcinoma in the liver with ‘high-grade’ architectural morphology. (D) Glandular structures with abortive luminas. (E) Solid trabecular growth pattern and ‘high-grade’ histocytologic morphology. (F) Enlarged pleomorphic nuclei of (E) with high nucleus/cytoplasmic ratio and hyperchromasia. Scale bars, 100 µm.
CD133 expression in the cells among the different TNM stages. The boxplot was overlaid with swarmplot. *P<0.05 (Dunn's post hoc test).
Comparison of marker expression between primary tumor and metastatic tumor in addition to between the presence and absence of lymphatic invasion. Boxplots were overlaid using swarm plot. (A) CD133, (B) CD166 and (C) Lgr5 expression in MTS and PT samples. (D) CD133 expression between patients showing tumor propagation in lymphatic vessels and those negative for this. Lgr5, leucine-rich repeat-containing G-protein-coupled receptor 5.
Comparison of MGMT methylation among the four TNM stages. The boxplot was overlaid using swarmplot. P-value is from Kruskal-Wallis test. *P<0.05 (Dunn's post hoc test). MGMT, O6-methyguanine-DNA methyltransferase.
Comparison of CD133 methylation among the four TNM stages. The boxplot was overlaid with swarmplot. P-value is from Kruskal-Wallis test. *P<0.05 (Dunn's post hoc test). A gross outlier in stage 1 was removed from the plot.
Patient characteristics.
Parameter | N (%) |
---|---|
Sex | |
Male | 17 ( |
Female | 13 ( |
Age | |
MTS | 9 ( |
PT | 21 (70) |
Side of primary tumor | |
R | 11 ( |
L | 10 ( |
N/A (metastasis) | 9 ( |
TNM stage | |
1 | 7 ( |
2 | 4 ( |
3 | 9 ( |
4 | 10 ( |
Grade | |
Low | 22 (73) |
High | 8 ( |
Lymphatic invasion | |
No | 14 ( |
Yes | 16 ( |
Venous invasion | |
No | 20 (67) |
Yes | 10 ( |
Perineural invasion | |
No | 24 (80) |
Yes | 6 ( |
Mismatch repair | |
Proficient | 19 (63) |
Deficient | 5 ( |
N/A | 6 ( |
Data are presented as N (%).
Patient characteristics after the exclusion of three samples due to insufficient DNA concentration.
Parameter | N (%) |
---|---|
Sex | |
Male | 16 ( |
Female | 11 ( |
Age | |
MTS | 8 ( |
PT | 19 (70) |
Side of primary tumor | |
R | 10 ( |
L | 9 ( |
N/A (metastasis) | 8 ( |
TNM Stage | |
1 | 6 ( |
2 | 4 ( |
3 | 8 ( |
4 | 9 ( |
Grade | |
Low | 19 (70) |
High | 8 ( |
Lymphatic invasion | |
No | 11 ( |
Yes | 16 ( |
Venous invasion | |
No | 17 (63) |
Yes | 10 ( |
Perineural invasion | |
No | 22 (81) |
Yes | 5 ( |
Mismatch repair | |
Proficient | 16 ( |
Deficient | 5 ( |
N/A | 6 ( |
P-values from the Dunn's post hoc comparisons of CD133 expression,
TNM stage comparison | CD133 expression (flow cytometry) | ||
---|---|---|---|
1 vs. 2 | 0.96 | 0.06 | 0.02 |
1 vs. 3 | 0.21 | 0.05 | 0.83 |
1 vs. 4 | 0.01 | 0.01 | 0.06 |
2 vs. 3 | 0.27 | 0.86 | 0.03 |
2 vs. 4 | 0.03 | 0.82 | 0.44 |
3 vs. 4 | 0.16 | 0.61 | 0.07 |
MGMT, O6-methyguanine-DNA methyltransferase.
Comparisons of each marker among each clinicopathological parameter.
A, Grade | |||||||
---|---|---|---|---|---|---|---|
Venous invasion | Sex | ||||||
Category | CD133 | CD166 | Lgr5 | Positive | Negative | Male | Female |
Low (n=22) | 19 ( |
67 ( |
20 ( |
8 ( |
14 (64) | 13 ( |
9 ( |
High (n=8) | 26 ( |
79 (64,86) | 38 ( |
2 ( |
6 (75) | 4 ( |
4 ( |
P-value | 0.8 | 0.087 | 0.3 | 0.7 | 0.7 | ||
B, Lymphatic invasion | |||||||
Sex | |||||||
Category | CD133 | CD166 | Lgr5 | Venous invasion | Male | Female | |
Negative (n=14) | 14 ( |
70 ( |
15 ( |
2 ( |
12 (86) | 7 ( |
7 ( |
Positive (n=16) | 31 ( |
71 ( |
29 ( |
8 ( |
8 ( |
10 (62) | 6 ( |
P-value | 0.026 | 0.6 | 0.12 | 0.058 | 0.5 | ||
C, Venous invasion | |||||||
Sex | |||||||
Category | CD133 | CD166 | Lgr5 | Venous invasion | Male | Female | |
Negative (n=20) | 24 ( |
73 ( |
23 ( |
- | 8 ( |
12 ( |
|
Positive (n=10) | 25 ( |
63 ( |
24 ( |
- | 9 (90) | 1 ( |
|
P-value | >0.9 | 0.6 | 0.8 | - | 0.017 | ||
D, Perineural invasion | |||||||
Sex | |||||||
Category | CD133 | CD166 | Lgr5 | Venous invasion | Male | Female | |
Negative (n=24) | 25 ( |
68 ( |
20 ( |
7 ( |
17 (71) | 14 ( |
10 ( |
Positive (n=6) | 22 ( |
75 ( |
29 ( |
3 ( |
3 ( |
3 ( |
3 ( |
P-value | 0.5 | 0.4 | 0.4 | 0.4 | >0.9 | ||
E, Sidedness of tumor | |||||||
Sex | |||||||
Category | CD133 | CD166 | Lgr5 | Venous invasion | Male | Female | |
Right (n=11) | 21 ( |
78 (64,81) | 33 ( |
4 ( |
7 (64) | 8 (73) | 3 ( |
Left (n=10) | 16 ( |
52 ( |
14 ( |
1 ( |
9 (90) | 1 ( |
9 (90) |
P-value | 0.3 | 0.4 | 0.14 | 0.3 | 0.008 |
Data represents the N (%) or the median (interquartile range) within each clinicopathological category. Fisher's exact test or Wilcoxon-Mann-Whitney test (where normality was not tenable) with the one-side alternative were used for categorical and continuous variables, respectively. Lgr5, leucine-rich repeat-containing G-protein-coupled receptor 5.
Comparisons of each marker among each clinicopathological parameter.
A, Grade | |||||
---|---|---|---|---|---|
Category | ALCAM | CD133 | Lgr5 | MGMT | ADAMTS16 |
Low (n=22) | 3.00 (2.75,3.25) | 5.00 (4.50,5.75) | 6.00 (5.50,6.75) | 15 ( |
77 (63,82) |
High (n=8) | 3.12 (2.69,3.38) | 4.75 (4.44,5.25) | 6.38 (5.88,7.06) | 22 ( |
81 (77,86) |
P-value | 0.9 | 0.6 | 0.3 | 0.4 | 0.2 |
B, Lymphatic invasion | |||||
Category | ALCAM | CD133 | Lgr5 | MGMT | ADAMTS16 |
Negative (n=14) | 3.25 (3.00,3.38) | 5.50 (4.75,6.88) | 6.25 (5.88,6.75) | 13 ( |
78 (76,84) |
Positive (n=16) | 3.00 (2.75,3.25) | 4.75 (4.25,5.31) | 6.12 (5.50,7.00) | 15 ( |
77 (63,83) |
P-value | 0.085 | 0.063 | 0.6 | 0.6 | 0.5 |
C, Venous invasion | |||||
Category | ALCAM | CD133 | Lgr5 | MGMT | ADAMTS16 |
Negative (n=20) | 3.00 (3.00,3.25) | 5.25 (4.75,6.25) | 6.25 (5.50,7.00) | 23 ( |
77 (73,83) |
Positive (n=10) | 2.88 (2.75,3.25) | 4.62 (4.06,4.94) | 6.00 (4.94,6.25) | 11 ( |
80 (66,84) |
P-value | 0.6 | 0.049 | 0.2 | 0.1 | 0.7 |
D, Perineural invasion | |||||
Category | ALCAM | CD133 | Lgr5 | MGMT | ADAMTS16 |
Negative (n=24) | 3.00 (2.81,3.25) | 5.12 (4.75,6.00) | 6.25 (5.56,7.00) | 15 ( |
78 (68,83) |
Positive (n=6) | 2.75 (2.50,3.00) | 4.25 (4.00,4.50) | 5.50 (4.75,6.25) | 14 ( |
81 (73,83) |
P-value | 0.078 | 0.012 | 0.2 | 0.8 | >0.9 |
E, Sidedness of tumour | |||||
Category | ALCAM | CD133 | Lgr5 | MGMT | ADAMTS16 |
Left (n=11) | 3.12 (3.00,3.25) | 5.25 (4.75,6.62) | 6.00 (5.31,6.50) | 14 ( |
76 ( |
Right (n=10) | 3.00 (3.00,3.25) | 4.75 (4.50,5.50) | 6.25 (5.75,6.25) | 33 ( |
77 (73,79) |
P-value | >0.9 | 0.6 | 0.6 | 0.11 | 0.6 |
Data are presented as the median (interquartile range). Wilcoxon-Mann-Whitney test was used. ALCAM, activated leukocyte cell adhesion molecule; lgr5, leucine-rich repeat-containing G-protein-coupled receptor 5; MGMT, O6-methyguanine-DNA methyltransferase; ADAMTS, A disintegrin and metalloproteinase with thrombospondin motifs 16.
Assessment of association between origin of tumor and each clinicopathological parameter.
Parameter | Metastasis (n=9) | Primary tumor (n=21) | P-value | Right-side tumor (n=11) | Left-side tumor (n=10) | P-value |
---|---|---|---|---|---|---|
Sex | 0.042 | 0.008 | ||||
Male | 8 (89) | 9 ( |
8 (73) | 1 ( |
||
Female | 1 ( |
12 ( |
3 ( |
9 (90) | ||
Age | 68 ( |
73 (65,77) | 0.041 | 73 (65,76) | 74 (70,78) | 0.5 |
Stage | <0.001 | 0.4 | ||||
1 | 0 (0) | 7 ( |
2 ( |
5 ( |
||
2 | 0 (0) | 4 ( |
3 ( |
1 ( |
||
3 | 0 (0) | 9 ( |
5 ( |
4 ( |
||
4 | 9 (100) | 1 (4.8) | 1 (9.1) | 0 (0) | ||
Mismatch repair | >0.9 | 0.012 | ||||
Deficient | 0 (0) | 5 ( |
0 (0) | 5 ( |
||
Proficient | 3 (100) | 16 (76) | 11 (100) | 5 ( |
Data are presented as either N (%) or the median (interquartile range). Fisher's exact test was used for comparing N (%) variables whereas Wilcoxon-Mann-Whitney test was used for comparing median (interquartile range) variables.
Association between each parameter and TNM staging.
TNM | |||||
---|---|---|---|---|---|
Parameters | 1 | 2 | 3 | 4 | P-value |
Sex | 0.021 | ||||
Male | 2 ( |
3 (75) | 3 ( |
9 (90) | |
Female | 5 (71) | 1 ( |
6 (67) | 1 ( |
|
Age | 71 (64,76) | 73 (70,76) | 76 (71,79) | 66 ( |
0.068 |
Origin of tumour | <0.001 | ||||
Metastasis | 0 (0) | 0 (0) | 0 (0) | 9 (90) | |
Primary tumour | 7 (100) | 4 (100) | 9 (100) | 1 ( |
|
Lymphatic invasion | <0.001 | ||||
Positive | 0 (0) | 0 (0) | 8 (89) | 8 (80) | |
Negative | 7 (100) | 4 (100) | 1 ( |
2 ( |
|
Venous invasion | 0.048 | ||||
Positive | 0 (0) | 2 ( |
2 ( |
6 ( |
|
Negative | 7 (100) | 2 ( |
7 (78) | 4 ( |
Data are presented as N (%) or the median (interquartile range). Fisher's exact test or Kruskal-Wallis test (for continuous variables) was used for comparison.