TGFβ-induced SMAD4-dependent Apoptosis Proceeded by EMT in CRC

This article is featured in Highlights of This Issue, p. 1183

Colorectal cancer is one of the most common malignancy and the second leading cause of death by cancer in Western populations (1). However, there have been remarkable changes in the rate of colorectal cancer incidence and mortality over the past few decades in other ethnic population (2). In Saudi Arabia, colorectal cancer presents the number one cancer affecting Saudi males and the third most common among females (3). Few studies also have shown that colorectal cancer affecting Saudi population tends to be more aggressive and present with advanced stage (4, 5). Colorectal cancer has a complex pathogenesis involving multiple genetic and epigenetic alterations caused by somatic mutations that lead to uncontrolled tumor cell proliferation, invasiveness, and distant metastatic potential (6, 7).

Epithelial-to-mesenchymal transition (EMT) is a bidirectional process, by which epithelial cells lose their cell polarity and cell–cell adhesion, and gain a motile mesenchymal cell phenotype that is believed to cause metastasis (8). Furthermore, EMT has been shown to be attributed to chemotherapy resistance and enabling cancer cell survival (9). One of the well-known inducer of EMT is TGFβ (10, 11). TGFβ/Smad4 signaling pathway controls the signal transduction from cell membrane to nucleus and is responsible for many important cellular processes such as tumor initiation, progression, apoptosis, and migration (12).

Although TGFβ is one of the potent inducer of EMT, it has paradoxical effect as tumor growth suppressor by inducing cell-cycle arrest and apoptosis in early stages of tumor formation (13). As the main mediator of canonical TGFβ signaling pathway, Smad4 plays pivotal role in the switch of TGFβ function on tumorigenesis (14). Smad4 is a key mediator of the TGFβ signaling and act as tumor suppressor in colorectal cancer. Inactivation of Smad4 at the gene or protein level is associated with progression of several types of tumors including colorectal cancer (15–17).

Loss of Smad4 protein expression is found in 20%–40% of colorectal cancer cases and strongly correlated with poor prognosis of patients with colorectal cancer (17–21). Studies also showed that Smad4 has been mutated in about 15% of colorectal cancer (22). However, data of Smad4 alteration in Middle Eastern colorectal cancer is limited (23, 24). Therefore, we sought to explore the role of Smad4 inactivation in a large cohort of patients with colorectal cancer from Saudi Arabia and its prognostic value in this ethnic group. Furthermore, a recent study has shed light on the dichotomous effect of TGFβ/Smad4 signaling in pancreatic ductal adenocarcinoma (PDA; ref. 25) and its role in EMT and apoptosis. They revealed interesting results that TGFβ induces PDA cells to undergo Smad4-dependent EMT followed by apoptosis.

In this study, we identified high incidence of Smad4 alterations (mutation, deletion, and low expression) in colorectal cancer in Saudi Arabia, a country with a particularly high incidence of colorectal cancer developing at younger age, suggesting a potential role in colorectal cancer pathogenesis. Interestingly, low expression of Smad4 was an independent poor prognostic marker. Furthermore, functional analysis of relevant colorectal cancer cell lines support previously reported role of TGFβ/Smad4 on inducing EMT and expression of mesenchymal markers Snail, Twist, and Zeb and highlighted the role of Smad4 deficiency in mediating drug resistance (26–28). Similar to the observed role of TGFβ/Smad4 in PDA, this study also revealed that TGFβ/Smad4 signal inhibits the transcription of KLF5. Taken together, our study demonstrates that TGFβ induces Smad4-dependent EMT followed by apoptosis in colorectal cancer cells. This study further highlights the potential therapeutic role of KLF5 inhibition in colorectal cancer treatment.

A total of 1,050 archival tissue samples from patients diagnosed with colorectal cancer at the King Faisal Specialist Hospital and Research Centre (Riyadh, Saudi Arabia) were collected from Department of Pathology. Detailed clinicopathologic data were noted from case records and summarized in Supplementary Table S1. Of these 1,050 cases, 426 cases were subjected to targeted capture sequencing. All samples were obtained from patients with approval from institutional review board of the hospital. For the study, waiver of consent was obtained for archived paraffin tissue blocks from Research Advisory Council (RAC) under project RAC#2170 025.

All samples were analyzed in a tissue microarray (TMA) format. TMA construction was performed as described earlier (29). Briefly, tissue cylinders with a diameter of 0.6 mm were punched from representative tumor regions of each donor tissue block and brought into recipient paraffin block using a modified semiautomatic robotic precision instrument (Beecher Instruments). Two cores of colorectal cancer were arrayed from each case. Standard protocol was followed for IHC staining, details of which are provided in Supplementary Materials and Methods.

Smad4 IHC expression was seen predominantly in the nuclear compartment and nuclear expression was quantified using the intensity score as described previously (17). Briefly, the intensity of nuclear staining in the tumor cells were scored from 0 to 4, with score 0 indicating absent Smad4 staining and score 4 indicating highest intensity of staining. Cases showing ≥3+ intensity score were considered as normal for Smad4 expression. Score 0–2 were considered as low expression of Smad4.

All human colorectal cancer cell lines; CX-1, CACO-2, HCT116, COLO-320, HCT15, DLD1, HT29, SW480, and LOVO were purchased from ATCC and cultured in RPMI1640 media containing 10% FBS, 100 U/mL penicillin, 100 U/mL streptomycin at 37°C in a humidified atmosphere containing 5% CO₂. Authentication of all cell lines were performed in-house using short tandem repeats PCR. The results were similar to the published data. All treatment experiments were performed in reduced FBS (2%) condition. Apoptosis analysis was performed using annexin V/propidium iodide (PI) dual staining and measured by flow cytometry as described previously (30). Information of antibodies used for Western blotting are listed in Supplementary Materials and Methods.

Plasmid DNA encoding human KLF5 and short hairpin RNA (shRNA) targeting human Snail1, Zeb1, and KLF5 were purchased from OriGene. The overexpression of KLF5 and knockdown of Snail1, Zeb1, and KLF5 in colorectal cancer cells were performed using Lipofectamine2000 (Invitrogen) according to the manufacturer's protocol. Briefly, colorectal cancer cells were seeded in 6-well culture plates; when approximately 50% confluent, cells were transfected with 4 μg plasmid. Stable overexpression clones resistant to G418 and stable knockdown clones resistant to puromycin were isolated; overexpression of KLF5 and knockdown of Snail1, Zeb1, and KLF5 protein production were confirmed by immunoblotting.

Immunofluorescence assay was performed as described previously (31). Briefly, colorectal cancer cells were grown on glass coverslips in 6-well plates and fixed with ice-cold methanol. After permeabilization with 0.2% Triton X-100 and blocking with 5% horse serum in PBS solution, cells were stained with antibodies to Smad4 (1:100), or E-cadherin (1:200) for 1 hour at 37°C. After washing, cells were incubated with fluorescent-conjugated secondary antibodies, and mounted using DAPI. The cells were visualized using Olympus BX63 fluorescence microscope.

Cells (500/well) were plated on Corning 24-well ultralow attachment plates (Sigma-Aldrich) grown in serum-free DMEM-F12 (ATCC) supplemented with B27 (Thermo Fisher Scientific), 20 ng/mL EGF (Sigma-Aldrich), 0.4% BSA (Sigma-Aldrich), and 4 μg/mL insulin (Sigma-Aldrich). Fresh medium was supplemented every 2 days. The spheroids were counted and photographed at day 14. For secondary spheroid formation, the primary spheroids were dissociated into single cells, and cultured on 24-well ultralow attachment plates using spheroid culture medium for another 10 days.

Contingency table analysis and χ² tests were used to study the relationship between clinicopathologic variables and Smad4 mutation/expression. Survival curves were generated using the Kaplan–Meier method, with significance evaluated using the Mantel–Cox log-rank test. The limit of significance for all analyses was defined as P < 0.05; two-sided tests were used in the calculations. The JMP11.0 (SAS Institute, Inc.) software package was used for data analyses.

For all functional studies, data are expressed as the mean ± SD of triplicate samples and was repeated for at least two separate experiments. Statistical analysis was performed using one-way ANOVA, and P < 0.05 was considered statistically significant.

The characteristics of the 1,050 patients with colorectal cancer are summarized in Supplementary Table S1. The median age at the time of surgery was 56 years (inter quartile range, 47.0–68.0 years) with a male:female ratio of 1.1:1. Most of the tumors were located in the left colon (79.7%). A total of 77.9% of patients had a moderately differentiated tumor and 51% were either stage III or stage IV.

We first performed high depth capture sequencing of 426 colorectal cancer cases, and found 63 mutations in 13.6% of cases. Sanger sequencing confirmed 54 somatic mutations present in 12.2% (52/426) colorectal cancer cases (Supplementary Fig. S1A). Out of these, 68.5%, 18.5%, 11%, and 2% were missense, stop-gained, frame-shift, and in-frame mutations, respectively (Supplementary Table S2). The Smad4 mutation was significantly associated with low Smad4 protein expression examined by IHC (P = 0.0032). However, no association was found between Smad4 mutation and 5-year overall survival (Supplementary Table S3).

The target capture sequencing data of 426 tumors have been deposited to the European Nucleotide Archive with accession number PRJEB32337.

We also performed the FISH assay to examine the Smad4 deletion in the primary tumor tissues of colorectal cancer samples (Supplementary Fig. S1B). In our series of 1,050 colorectal cancer cases, FISH analysis was interpretable in 991 cases. TMA for 59 cases were nonrepresentative due to lack of tumor cells. Of the cases analyzed, only Smad4 gene deletion was detected in 4.2% (44) cases of colorectal cancer. Interestingly, chromosome 18q monosomy (only one Smad4 signal and one CEN18q signal) was detected in 559 (56.4%) cases (Supplementary Table S4A). Altogether, Smad4 deletion was identified in 603 (60.8%) of our cohort, and was significantly associated with distant metastasis (P = 0.0247) and low Smad4 protein expression (P < 0.0001). Smad4 deletion was also found to be significantly associated with microsatellite stable (MSS) tumor (P < 0.0001), which is in agreement with previous reports (refs. 32, 33; Supplementary Table S4B).

Of the 1,050 colorectal cancer cases investigated by IHC, low expression of Smad4 was noted in 67.1% (680/1,013) of cases by IHC (Supplementary Fig. S1C). Thirty-seven cases were noninterpretable due to loss of tissue cores or absent tumor cells in core. The staining pattern ranged from absent staining to intense 4+ nuclear staining. Low expression of Smad4 was found to be significantly associated with poor prognostic markers such as larger tumors (T3, P = 0.0008), lymph node involvement (P = 0.0006), tumor–node–metastasis (TNM) stage IV tumors (P = 0.0012), Dukes' stage C (P = 0.0006), and MSS tumors (P < 0.0001; Table 1).

Of particular importance was the association of low Smad4 protein expression with poor 5-year overall survival on univariate analysis using Kaplan–Meier curve (71.6% vs. 78.9%; χ² = 7.33; P = 0.0068; Supplementary Fig. S1D). On multivariate analysis using Cox proportional hazards regression model after adjusting for possible confounders of survival, low expression of Smad4 was indeed found to be an independent poor prognostic marker (HR = 1.77; 95% confidence interval, 1.26–2.55; P = 0.0009).

Smad4 alterations (mutation or low expression or deletion by FISH) were noted in 82.2% (342/416) of our cases. A significant association was noted between Smad4 alteration and lymph node involvement (P = 0.0439), as well as with MSS tumors (P < 0.0001; Supplementary Table S5).

To explore the role of Smad4 in TGFβ-induced apoptosis in colorectal cancer cells, we analyzed the expression of Smad4 in a panel of nine colorectal cancer cell lines by immunoblotting (Fig. 1A). On the basis of Smad4 expression, we selected two colorectal cancer cells lines, HCT116 and DLD1 that had overexpression of Smad4, as well as two other cell lines, HT29 and SW480 that had negligible Smad4 expression for further experimentation. Smad4 abundance in these cells was also verified by immunofluorescence analysis (Fig. 1B). We investigated the effect of TGFβ on inhibiting cell viability and inducing apoptosis in these cells. There was inhibition of cell viability and induction of apoptosis in both colorectal cancer cell lines that had overexpression of Smad4; however, TGFβ has no effect on cell viability and apoptosis in Smad4-low–expressing cells (Fig. 1C–F).

To verify the role of Smad4 in TGFβ-induced cell death, we knocked out Smad4 in Smad4-proficient cells (HCT116 and DLD1; Supplementary Fig. S2A) followed by treatment with TGFβ for 48 hours and analyzed cell proliferation and apoptosis. Silencing of Smad4 significantly suppressed TGFβ-induced inhibition of cell proliferation as confirmed by clonogenic assay (Supplementary Fig. S2B and S2C). As shown in Supplementary Fig. S2D, Smad4-knockout cells became resistant to TGFβ treatment as compared with Smad4-proficient control. Furthermore, we stably overexpressed Smad4 in Smad4-deficient cells (HT29 and SW480; Supplementary Fig. S2E). Forced expression of Smad4 significantly decreased cell proliferation after treatment with TGFβ (Supplementary Fig. S2F and S2G). Overexpression of Smad4 in HT29 and SW480 cells became sensitive to TGFβ treatment as compared with cells transfected with empty vector (Supplementary Fig. S2H).

TGFβ has been known to induce both EMT and apoptosis in vitro (34, 35). To understand these events in detail, we treated Smad4-proficient and Smad4-deficient colorectal cancer cells with TGFβ at different time points. As shown in Fig. 2A, Smad4-proficient colorectal cancer cells lost cellular polarity and started forming elongated spindle-like protrusions after 12 hours of TGFβ treatment. From 24 to 48 hours of TGFβ treatment, most spindle-like Smad4-proficient cells began to shrink and started dying. In contrast, Smad4-deficient cells retained its original morphology with negligible cell death throughout this period (Fig. 2A). Immunofluorescence analysis showed a downregulation of E-cadherin expression after TGFβ treatment in Smad4-proficient cells, whereas the Smad4-deficient cells retained E-cadherin expression (Fig. 2B).

As shown in Fig. 2C, there was a time course increase in apoptosis following treatment with TGFβ after 24 hours in Smad4-proficient cells confirming that this cell death was due to apoptosis. However, TGFβ had no effect on Smad4-deficient cells (Fig. 2C). We also investigated the expression of E-cadherin and apoptotic protein markers after TGFβ treatment at different time points (24 and 48 hours) on these cells by immunoblotting. As shown in Fig. 2D, treatment of TGFβ successfully downregulated E-cadherin expression and induced the cleavage of caspase-3 in Smad4-proficient colorectal cancer cells, whereas the expression of E-cadherin and caspase-3 remained unchanged in Smad4-deficient cells (Fig. 2D). Together, these findings suggest that TGFβ induces Smad4-dependent EMT followed by apoptosis in colorectal cancer cells.

Transcription factors, Snail1 and Zeb1, play a key role in initiating TGFβ-induced EMT process (36, 37). As shown in Fig. 3A and B, TGFβ treatment prominently induced the expression of Snail1 and Zeb1 after 48 hours in Smad4-proficient colorectal cancer cells but not in Smad4-deficient cells. Silencing of Snail1 and Zeb1 in Smad4-proficient cells markedly inhibited the TGFβ-induced E-cadherin downregulation (Fig. 3C). Besides, knockdown of Snail1 and Zeb1 suppressed the induction of EMT-like morphologic changes by TGFβ in Smad4-proficient cells (Fig. 3D). Next, we sought to determine the role of Snail1 and Zeb1 in TGFβ-induced apoptosis. As shown in Fig. 3E and F, silencing of Snail1 and Zeb1 significantly inhibited the TGFβ-induced apoptosis in Smad4-proficient cells, showing the essential role of these transcriptional factors on TGFβ-induced apoptosis.

Sox4 expression has shown to be elevated in a wide variety of human cancers (38–40). To understand the role of Sox4 in TGFβ-driven apoptosis, we silenced Sox4 in Smad4-proficient cells using two different shRNAs and treated with TGFβ for 48 hours and analyzed apoptosis. As shown in Fig. 4A, knockdown of Sox4 significantly suppressed TGFβ-induced apoptosis in Smad4-proficient colorectal cancer cells. There was no further significant increase in apoptosis even after addition of 5-fluorouracil (5-FU) in Sox4-depleted cells (Fig. 4A). This data clearly demonstrate that Sox4 is essential for TGFβ-driven apoptosis in Smad4-proficient cells. Next, we tested the role of Sox4 in TGFβ-induced EMT. Knockdown of Sox4 had no effect on EMT morphology (Fig. 4B) or E-cadherin expression (Fig. 4C). There was an increase in Sox4 expression in scramble control cells after TGFβ treatment (Fig. 4C). Moreover, silencing of Snail1 and Zeb1 had no effect on Sox4 expression (Fig. 4D). These results show that Sox4 has no role in TGFβ-induced EMT. To test the role of Sox4 in spheroid growth, we generated spheroids from Smad4-proficient cells and stemness of the spheroids was confirmed using stem cell markers (Fig. 4E). Next, we determined the expression of Sox4 in adherent and spheroid cells with and without TGFβ treatment. Cells grown as spheroids showed higher Sox4 expression than cells grown in adherent conditions (Fig. 4F). Moreover, there was an increase in Sox4 expression in both cells after TGFβ treatment (Fig. 4F). To verify the role of Sox4 in spheroid growth, we silenced Sox4 in Smad4-proficient cells and grown in spheroid medium. As shown in Fig. 4G and H, silencing of Sox4 significantly decreased the spheroid growth. This data were further verified using immunoblotting, where knockdown of Sox4 markedly decreased the stem cell marker proteins expression in spheroids (Fig. 4I). Above results indicate that Sox4 acts as a tumor initiator in Smad4-proficient cells. Taken together, these data suggest that during TGFβ-induced EMT, Sox4 switches its function from protumorigenic to proapoptotic.

KLF5 has been shown to promote cancer cell proliferation in several cancers including colorectal cancer (41, 42). To study the role of KLF5 in TGFβ-induced apoptosis, we treated Smad4-proficient and Smad4-deficient cells with TGFβ for 48 hours and analyzed the expression of KLF5 by immunoblotting. As shown in Fig. 5A, there was a prominent inhibition of KLF5 expression after TGFβ treatment in Smad4-proficient cells, but the KLF5 expression remained unchanged in Smad4-deficient cells. The basal expression of KLF5 was relatively higher in Smad4-deficient cells than Smad4-proficient cells (Fig. 5A). Next, we silenced KLF5 in Smad4-deficient cells using two different shRNAs and treated with TGFβ and 5-FU either alone or in combination for 48 hours and analyzed apoptosis. As shown in Fig. 5B, silencing of KLF5 markedly increased TGFβ-induced apoptosis in Smad4-deficient colorectal cancer cells. Furthermore, treatment of 5-FU significantly sensitized the KLF5-depleted cells (Fig. 5B). This data clearly demonstrate that inhibition of KLF5 is essential for TGFβ-driven apoptosis. This finding was in concordance with recent reports in pancreatic cancer cells (25). David and colleagues reported that TGFβ-induced EMT inhibits KLF5 in Smad4-proficient cells and induce apoptosis (25). To test the role of KLF5 in spheroid growth, we generated spheroids from Smad4-deficient cells and stemness of the spheroids was confirmed using stem cell markers (Fig. 5C). To test the role of KLF5 in spheroid growth, we silenced KLF5 in Smad4-deficient cells and were grown in spheroid medium. Silencing of KLF5 significantly decreased the spheroid growth (Fig. 5D and E) and stemness properties (Fig. 5F). Above results, clearly indicate that KLF5 acts as an oncogene in colorectal cancer cells regardless of Smad4 expression and inhibition of KLF5 is essential for TGFβ-induced apoptosis.

To test the role KLF5 in TGFβ-induced EMT, we stably overexpressed KLF5 in Smad4-proficient cells, treated with TGFβ, and analyzed the E-cadherin expression. As shown in Fig. 5G, overexpression of KLF5 inhibited TGFβ-induced downregulation of E-cadherin in Smad4-proficient cells. This data evidently demonstrate the association between KLF5 and EMT. Next, we tested the effect of KLF5 overexpression in TGFβ-induced apoptosis. Forced expression of KLF5 significantly inhibited TGFβ-induced apoptosis in Smad4-proficient cells (Fig. 5H). Above results, clearly exhibit the oncogenic function of KLF5. Finally, we silenced both Sox4 and KLF5 in Smad4-deficient cells to confirm the link between these proteins in TGFβ-induced apoptosis. As shown in Fig. 5I, depletion of KLF5 promotes apoptosis in Smad4-deficient cells after TGFβ treatment, whereas there was a significant inhibition of apoptosis on Sox4 depletion. All these results indicate that Sox4 promote apoptosis upon KLF5 depletion.

In this study, we used a large cohort of clinical samples to address the clinical significance of Smad4 alterations in Saudi colorectal cancer. We evaluated the Smad4 protein expression in more than 1,000 Saudi patients with colorectal cancer, as well as genomic mutation and gene deletion. The frequency of Smad4 alteration was similar to what has been reported in colorectal cancer from other ethnicities (17, 23, 43–45). Consistent with previous reports, loss of Smad4 expression exhibited strong correlation to poor overall survival in patients with colorectal cancer (46–48). Importantly, we show that low levels of Smad4 protein can identify a subset of patients with Dukes' C colorectal cancer who have a high probability of recurrence following potentially curative surgery and 5-FU–based adjuvant chemotherapy.

The role of TGFβ signaling in colorectal cancer progression and prognosis is very complex and has previously shown bilateral dependence based on Smad4 status (49). Smad4 is an important mediator of TGFβ signaling pathway and function as a tumor suppressor gene in colorectal cancer (50). Our study shows that Smad4 expression was reduced significantly with advancing tumor stage (Table 1). Our in vitro results using colorectal cancer cell lines identified a mechanistic basis for duality of TGFβ in colorectal cancer. We show that under conditions of TGFβ stimulation, presence of the common transactivator protein Smad4 in colorectal cancer cells undergo apoptosis, whereas in its absence cancer cells promote tumor growth. Interestingly, in agreement with recent report (25), TGFβ-induced Smad4-dependent apoptosis is preceded by EMT. Moreover, TGFβ-induced activation of Snail1 and Zeb1 were required for EMT-related apoptosis. TGFβ-induced EMT is generally considered as a protumorigenic event. However, our study demonstrates that TGFβ exerts tumor suppression in colorectal cancer cells by inducing EMT signal, which inhibits the transcription of KLF5 that in turn switches Sox4 from tumor promoter to suppressor.

Our study demonstrated the essential role of Sox4 on inducing TGFβ/Smad4-mediated apoptosis. However, Sox4 showed no direct effect on inducing EMT. Sox4 act as a tumor suppressor in those cells that undergo a TGFβ/Smad4-induced EMT, showing that EMT switches Sox4 function from protumorigenic to proapoptotic in colorectal cancer cells. Interestingly, knockdown of KLF5 was sufficient to revert protumorigenic function of TGFβ and induce apoptosis in Smad4-deficient cells. Furthermore, we show that TGFβ-mediated EMT downregulates KLF5 expression in Smad4-proficient cells that convert Sox4 from antiapoptotic to proapoptotic. These data indicate that KLF5 may be a vital determinant of Sox4 function in colorectal cancer. Moreover, forced expression of KLF5 in Smad4-proficient colorectal cancer cells prominently inhibited TGFβ-induced EMT and apoptosis, further confirming the oncogenic function of KLF5.

EMT has been previously shown to contribute to chemoresistance in various epithelial cancers (51, 52). Our data show that Smad4-proficient colorectal cancer cells were sensitive to 5-FU chemotherapy even under TGFβ stimulation, whereas Smad4-deficient cells were resistant to 5-FU chemotherapy. However, targeted inhibition of KLF5 using shRNA in Smad4-deficient colorectal cancer cells was able to revert the chemoresistance under TGF stimulation. This is important from translational stand point as it might suggest the possibility of therapeutic value of inhibiting KLF5 in patients with Smad4-deficient colorectal cancer. However, additional studies are needed to further validate these findings and move KLF5 inhibitor to preclinical testing.

In conclusion, our results show that Smad4 alteration is frequent in colorectal cancer from Middle Eastern ethnicity and can identify patients at advanced stage (Dukes' C) and significantly poor outcome. Our functional studies in colorectal cancer cells reinforce the role of Smad4 in TGFβ-induced EMT signaling cascade. Furthermore, our results revealed that TGFβ/Smad4 signal inhibits the transcription of KLF5 that in turn switches Sox4 from tumor promoter to suppressor. Therefore, we conclude that TGFβ induces Smad4-dependent lethal EMT in colorectal cancer cells.

No potential conflicts of interest were disclosed.

Conception and design: P. Pratheeshkumar, K.S. Al-Kuraya

Development of methodology: A.K. Siraj, K.S. Al-Kuraya

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): A.K. Siraj, P. Pratheeshkumar, S.P. Divya, S.K. Parvathareddy, N. Al-Sanea, L.H. Ashari, S. Al-Homoud, K.S. Al-Kuraya

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): P. Pratheeshkumar, S.P. Divya, S.K. Parvathareddy, R. Bu, T. Masoodi, Y. Kong

Writing, review, and/or revision of the manuscript: A.K. Siraj, P. Pratheeshkumar, K.S. Al-Kuraya

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): A.K. Siraj

Study supervision: A.K. Siraj, K.S. Al-Kuraya

Other (performed flow cytometry analysis): S. Thangavel

Other (provision of sample material and data): N. Al-Sanea, F. Al-Dayel

Other (clinical data provided from Colorectal Unit): L.H. Ashari, A. Abduljabbar

We thank Dr. Maqbool Ahmed, Roxanne Melosantos, Rafia Begum, Wael Haqawi, Zeeshan Qadri, Dionne Rae Rala, Ingrid Francesca Victoria, and Maha Al Rasheed, for technical assistance.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.