Transcriptomic profiling disclosed the role of DNA methylation and histone modifications in tumor-infiltrating myeloid-derived suppressor cell subsets in colorectal cancer
Increased numbers of myeloid-derived suppressor cells (MDSCs) are positively correlated with poor prognosis and reduced survivals of cancer patients. They play central roles in tumor immune evasion and tumor metastasis. However, limited data are available on phenotypic/transcriptomic characteristics of the different MDSCs subsets in cancer. These cells include immature (I-MDSCs), monocytic (M-MDSCs), and polymorphonuclear/granulocytic (PMN-MDSCs).
Phenotypic characterization of myeloid subsets from 27 colorectal cancer (CRC) patients was assessed by flow cytometric analyses. RNA-sequencing of sorted I-MDSCs, PMN-MDSCs, and antigen-presenting cells (APCs) was also performed.
We found that the levels of I-MDSCs and PMN-MDSCs were increased in tumor tissues (TT), compared with normal tissues (NT) in colorectal cancer. Our functional annotation analyses showed that genes associated with histone deacetylase (HDAC) activation- and DNA methylation-mediated transcriptional silencing were upregulated, and histone acetyl transferase (HAT)-related genes were downregulated in tumor-infiltrating I-MDSCs. Moreover, pathways implicated in cell trafficking and immune suppression, including Wnt, interleukin-6 (IL-6), and mitogen-activated protein kinase (MAPK) signaling, were upregulated in I-MDSCs. Notably, PMN-MDSCs showed downregulation in genes related to DNA methylation and HDAC binding. Using an ex vivo model, we found that inhibition of HDAC activation or neutralization of IL-6 in CRC tumor tissues downregulates the expression of genes associated with immunosuppression and myeloid cell chemotaxis, confirming the importance of HDAC activation and IL-6 signaling pathway in MDSC function and chemotaxis.
This study provides novel insights into the epigenetic regulations and other molecular pathways in different myeloid cell subsets within the CRC tumor microenvironment (TME), giving opportunities to potential targets for therapeutic benefits.
KeywordsMyeloid-derived suppressor cells Colorectal cancer Transcriptomic profile DNA methylation Histone modifications
Myeloid-derived suppressor cells
Immature myeloid-derived suppressor cells
Mitogen-activated protein kinase
Monocytic myeloid-derived suppressor cells
Polymorphonuclear/granulocytic myeloid-derived suppressor cells
Colorectal cancer (CRC) is the third most common cancer worldwide associated with high morbidity rates every year . Inflammation is recognized as a key driver for CRC pathogenesis . In inflammation-related cancers, myelopoiesis is disrupted and leads to accumulation of a heterogeneous population of myeloid cells, halted at varying stages of maturation/differentiation and with a potent immunosuppressive activity, referred to as myeloid-derived suppressor cells (MDSCs) [3, 4]. MDSCs express myeloid markers but lack expression of MHC class II molecule, HLA-DR, and are mainly divided into three phenotypically distinct subpopulations; CD33+HLA-DR−/lowCD14−CD15− early-stage, or immature MDSCs (e-MDSC/I-MDSC), which consist of immature myeloid progenitors; CD33+HLA-DR−/lowCD14+CD15− monocytic MDSCs (M-MDSCs), which represent suppressive monocytes; and CD33+HLA-DR−/lowCD14−CD15+ polymorphonuclear or granulocytic MDSCs (PMN-MDSCs or G-MDSCs), which are phenotypically distinct from mature neutrophils and possess strong suppressive activity [3, 5, 6, 7].
MDSCs have been implicated in the pathogenesis of cancer [5, 8], where they act as potent suppressors of T cell-mediated responses against tumor cells within the tumor microenvironment (TME)  and lymphoid organs . MDSCs enhance the progression of various types of tumors by promoting immune suppression, cancer-associated fibroblasts (CAF) activation, angiogenesis, and tumor growth and metastasis  via the expression of co-inhibitory receptors, such as PD-L1, and the release of a vast array of molecules, such as arginase-1, inducible nitric oxide synthase (iNOS), nitric oxide (NO), and reactive oxygen/nitrogen species (ROS/RNS) [11, 12]. Additionally, the number of circulating MDSCs has been positively correlated with poor prognosis and low survival rates in cancer patients, including those with CRC [13, 14, 15]. However, limited studies have explored the molecular/functional characteristics of different MDSC populations. In addition, the underlying mechanisms behind gene transcriptional regulation and signal transduction in these populations remain to be elucidated. Further insights into these mechanisms could provide potential candidates to target MDSC populations in cancer.
Elevated levels of PMN-MDSCs and I-MDSCs in CRC tumor tissues
Characteristic features of study populations
Next, we selected six CRC patients to sort different myeloid cell subpopulations (I-MDSCs, PMN-MDSCs, and APCs) from TT and NT and analyze their transcriptional profiling using RNA-Seq. However, the library preparation and subsequent RNA-sequencing were succeeded from two patients only (#07 and #08). Gating strategy for sorting these subpopulations from the two patients is shown in Fig. 2e and f.
Genes associated with HDAC activation, DNA methylation, and IL-6 signaling pathway are upregulated in tumor-infiltrating I-MDSCs
Collectively, our data show that the transcriptional profile of tumor-infiltrating I-MDSCs could be altered by histone post-translational modifications induced by HDAC activation, and/or DNA methylation. In addition, we speculate that I-MDSCs promote tumorigenesis by activating signaling pathways, such as Wnt and IL-6 pathways, which in turn enhance MDSCs recruitment to the TME and support MDSCs suppressive function.
Genes associated with chemotaxis, cell migration, and anti-apoptosis are upregulated in tumor-infiltrating PMN-MDSCs
Transcriptional regulation in tumor-infiltrating I-MDSCs and PMN-MDSCs is mediated via distinct epigenetic mechanisms
We also investigated the differences in the transcriptional profile between tumor-infiltrating PMN-MDSCs and APCs. The hierarchal clustering of PMN-MDSCs and APCs is shown in Fig. 6c and Additional file 2: Table S1. We found 1803 genes that were distinct in PMN-MDSCs, compared with APCs. Out of 1803, 1098 genes were significantly upregulated and 705 were downregulated in tumor-infiltrating PMN-MDSCs, compared with APCs (with a fold of change > 2 and P value cutoff < 0.05). Interestingly, 15 genes related to CRC progression, 19 genes related to cell migration and 17 genes related to NF-kB/IL-1β-mediated recruitment of MDSCs were upregulated in tumor-infiltrating PMN-MDSCs (Fig. 6d and Additional file 3: Table S2). Moreover, 13 genes related to HDAC binding and 17 genes related to DNA methylation were significantly downregulated in tumor-infiltrating PMN-MDSCs (Fig. 6d). Together, these results suggest that upregulation of genes involved in cell migration and MDSCs recruitment in tumor-infiltrating PMN-MDSCs may occur as a result of HDAC inactivation and DNA demethylation.
Additionally, we compared the transcription profile between tumor-infiltrating PMN-MDSCs and I-MDSCs. The hierarchal clustering is shown in Fig. 6e and Additional file 2: Table S1. We found that 1395 genes were significantly distinct in tumor-infiltrating PMN-MDSCs. Out of 1395, 633 genes were significantly upregulated and 762 were downregulated in tumor-infiltrating PMN-MDSCs. Functional annotation analyses showed that 6 genes related to tumor progression and 6 genes related to migration were significantly upregulated, while DNA methylation-related genes were significantly downregulated in PMN-MDSCs, compared with I-MDSCs (Fig. 6f and Additional file 3: Table S2). These data are consistent with our findings above (PMN-MDSCs TT vs. NT, Fig. 5; I-MDSCs TT vs. NT, Fig. 3). In addition, we found that genes associated with transcriptional regulation (KAT2B, OAS1, UST, CD38, and CELF2) and signal transduction (CD40 and RPA4) were differentially expressed in PMN-MDSCs and I-MDSCs (Fig. 6g). We validated the expression of the genes above in tumor-infiltrating PMN-MDSCs, I-MDSCs, and APCs (the latter used as a control) by RT-PCR. We found that mRNA expression levels in PMN-MDSCs vs. I-MDSCs (Fig. 6h) were consistent with RNA-Seq data (Fig. 6g). Statistical significance was obtained by comparing the mRNA expression levels in PMN-MDSCs vs. I-MDSCs (Fig. 6h). Interestingly, our functional network analysis shows that HIF-1 pathway was active in tumor-infiltrating I-MDSCs (Additional file 1: Figure S2). These data are consistent with previous reports demonstrating that HIF-1α is essential for the infiltration and differentiation of myeloid cells within the TME [21, 22]. Additionally, we validated the functional clusters for patients #07 and #08 in four independent data sets (Additional file 1: Figure S3). Adhesion molecules, such as integrins, have been implicated in cancer progression . It was reported that recruitment of neutrophils (a granulocytic subset that display similar phenotypic/functional characteristics as PMN-MDSCs) to tumor sites is associated with increased expression of CD11a integrin on neutrophils (also known as LFA-1) . In this study, our data from RNA-Seq analysis showed that ITGAL gene encoding CD11a integrin is upregulated in tumor-infiltrating PMN-MDSCs, compared with I-MDSCs (Additional file 1: Figure S4A). We validated this finding by flow cytometric analysis and showed that CD11a is highly expressed on tumor-infiltrating PMN-MDSCs, compared with I-MDSCs (Additional file 1: Figure S4B).
Genes associated with immunosuppression and myeloid cell recruitment in CRC tumor tissue are downregulated upon HDAC or IL-6 inhibition
Tumor-infiltrating MDSCs are key players, which negatively modulate the anti-tumor immune response, and promote angiogenesis and tumor growth/metastasis [26, 27]. Elevated levels of circulating and tumor-infiltrating PMN-MDSCs and I-MDSCs have been detected in CRC patients [13, 28]. Additionally, the expansion of tumor-infiltrating PMN-MDSCs in CRC patients has been positively correlated with advanced disease stages and high histological grades . However, the phenotypic and molecular characteristics of these MDSCs subsets and their role in CRC progression remain elusive. Therefore, we investigated the transcriptional profiles and functional characteristics of I-MDSCs and PMN-MDSCs within the CRC microenvironment.
The phenotypic and functional characterization of MDSCs is widely known to be challenging due to the heterogeneous nature of these cells . In this study, we utilized myeloid cell markers to identify the different subsets of MDSCs, to calculate their levels in colorectal tissues as previously described . We found that the numbers of PMN-MDSCs and I-MDSCs are higher in colorectal tissues, compared to adjacent normal colorectal tissues. This falls in agreement with other reports demonstrating the expansion of different MDSCs subsets in the TME of several malignancies [29, 30].
In this study, transcriptional profiling analyses of I-MDSCs and PMN-MDSCs revealed novel mechanisms/signaling pathways, which may govern the MDSCs-mediated immune suppression and tumor progression in CRC. Our canonical and functional pathway analyses of I-MDSCs showed that Wnt, IL-6, and MAPK signaling pathways are significantly upregulated in TT, compared with NT, suggesting a role of these pathways in tumor-infiltrating I-MDSCs within the CRC microenvironment. The involvement of Wnt signaling in CRC pathogenesis and progression has been previously established [31, 32, 33]. Using a co-culture system with breast cancer cell lines, Pukrop et al. showed that Wnt signaling is crucial for macrophage-mediated tumor invasiveness and tumor cell migration via the production of matrix metalloproteinases (MMPs) . Wnt pathway has been shown to antagonize the recruitment and differentiation of MDSCs in tumor sites . However, our data suggest a potential role of Wnt signaling in one MDSCs subset (I-MDSCs) from CRC tissues. This discrepancy could be due to the heterogeneity of MDSCs and suggests that signaling pathways could impact subsets of MDSCs differentially. The importance of IL-6 signaling pathway in MDSCs recruitment and immune suppression has been previously demonstrated . Using breast cancer in vitro and in vivo models, Jiang et al. reported that exogenous IL-6 augments the accumulation of MDSCs and enhances T cell suppression by inhibiting SOCS3 pathway . In another study, it was demonstrated that IL-6 upregulates the expression of indoleamine-pyrrole 2,3-dioxygenase (IDO) enzyme via NF-κB activation, thereby enhancing the immunosuppressive function of MDSCs in breast cancer . MAPK signaling is another pathway, which we found to be significantly upregulated in tumor-infiltrating I-MDSCs from CRC patients. Studies have demonstrated that MAPK signaling pathway is essential for the activation and differentiation of myeloid cells [37, 38]. By combining these findings, we could anticipate that I-MDSCs in CRC microenvironment are activated and recruited to tumor sites where they differentiate into mature phenotypes and become immune suppressive under the influence of MAPK- and IL-6-mediated signaling pathways. Targeting these pathways in I-MDSCs could suppress their activity and reduce their accumulation within the TME, which in turn may improve the anti-tumor immune response and diminish tumor growth/metastasis.
Maintaining transcriptional homeostasis requires a dynamic equilibrium of HAT and HDAC activity, where the former favors gene transcription and the latter suppresses it . In pathological conditions, including cancer, this equilibrium can be dysregulated by certain factors, causing either the silencing of tumor suppressor genes or inducing the expression of genes promoting tumor growth/metastasis and immunosuppression [40, 41, 42, 43]. Based on functional annotation analyses, we found that 44% of HDAC- and 29% of DNA methylation-related genes along with 13% of transcriptional repressors were upregulated in tumor-infiltrating I-MDSCs. Notably, 90% of HAT-related genes were downregulated in tumor-infiltrating I-MDSCs. These data indicate that transcriptional regulation in tumor-infiltrating I-MDSCs could be controlled by DNA methylation and post-translational histone modifications, possibly, mediated by HDAC activation and the suppression of HAT activity. Such epigenetic changes can be responsible for downregulating genes involved in tumor suppression or T cell activation [44, 45]. Unlike I-MDSCs, we found that DNA methylation- and HDAC-related genes were downregulated in tumor-infiltrating PMN-MDSCs, suggesting that the regulation of gene expression in these MDSCs subsets is governed via differing epigenetic mechanisms. Therefore, we could propose a novel therapeutic approach by which specific populations of MDSCs can be targeted in CRC. Rather than endemic targeting of MDSCs, targeting specific populations of MDSCs by the use of appropriate epigenetic modifiers could be employed to design a well-targeted therapy for CRC patients.
Interestingly, we found that genes of JNK- and SNARE-mediated pathways were upregulated in tumor-infiltrating I-MDSCs, compared with APCs. These pathways could have a functional impact on I-MDSCs activity and CRC progression. Han et al. reported the upregulation of JNK pathway in tumor-infiltrating myeloid cells, from mouse liver cancer model, is crucial for promoting disease pathogenesis and progression . Based on animal studies, it was suggested that specific targeting of JNK in myeloid cells could offer therapeutic benefits for patients with hepatocellular carcinoma . In addition, JNK signaling pathway in myeloid cells has been associated with the production of Treg chemoattractant agents, such as CCL17 and CCL22 , indicating its potential role in MDSCs suppressive function. SNARE proteins have been associated with the regulation of cytokine secretion macrophages and dendritic cells, and inhibition of phagocytosis and antigen presentation , suggesting their potential function in I-MDSCs via similar means.
Our functional annotation and transcriptomic analyses showed that genes related to cell-cell signaling, cell proliferation, anti-apoptosis, and cellular migration were upregulated in tumor-infiltrating PMN-MDSCs. These data imply the potential of PMN-MDSCs function in CRC progression, which may involve the enhancement of tumor growth/metastasis and immune cell trafficking. Consistent with this, Ouzounova et al. reported that enrichment of lung TME with activated PMN-MDSCs was associated with increased expression of genes promoting tumor growth and metastasis . Within the CRC microenvironment, we found that NF-κB- and IL-1β production-related genes were upregulated in PMN-MDSCs, compared with APCs. It has been reported that tumor-derived IL-1β enhances MDSCs infiltration, accumulation, and immunosuppressive activity within the TME, thereby promoting tumor progression [38, 50]. Another study by Tu et al. showed that IL-1β overexpression induces gastric inflammation and cancer through the activation of NF-κB, resulting in the accumulation of MDSCs within tumor sites . In turn, NF-κB activation exerts an anti-inflammatory function on MDSCs, enhances MDSCs infiltration in tumor sites by regulating chemotaxis , and induces immunosuppression by upregulating the expression of IDO . Therefore, we could suggest that the suppressive activity of tumor-infiltrating PMN-MDSCs in CRC is positively regulated by IL-1β/NF-κB signaling pathway.
Hypoxia is a common feature of solid tumors, which arise from reduced levels of oxygen and nutrient supply as a result of tumor outgrowth . Hypoxia can mediate resistance to chemotherapy, promote tumor metastasis, and enhance MDSCs accumulation within TME . Kumar et al. reported that the upregulation of HIF-1α enhances the immunosuppressive activity of myeloid cells and promotes the rapid conversion of MDSCs to tumor-associated macrophages (TAMs)  via the upregulation of iNOS and arginase . In addition, HIF-1α can bind to the promotor of PD-L1, co-inhibitory receptor, and increase its expression on myeloid cells has been demonstrated . In our study, the upregulation of HIF-1 pathway was more profound in I-MDSCs (immature myeloid cells), compared with PMN-MDSCs (mature myeloid cells), suggesting the potential importance of this pathway in I-MDSCs function/biology. On these grounds, we could speculate that I-MDSCs differentiate into the mature phenotype, TAMs, by upregulating HIF-1 pathway. In renal carcinoma, the upregulation of HIF-1 and constitutive activation of HIF-1 pathway occur due to a loss of tumor suppressor gene/function, von Hippel-Lindau (VHL), either by mutation or hypermethylation . This subsequently results in the expression of angiogenic factors and genes promoting disease oncogenesis . This is consistent with the finding that I-MDSCs regulate gene transcription by DNA methylation, which might be associated with tumor suppressor gene silencing. Employing therapeutic strategies to target HIF-1 could be beneficial in reducing the suppressive activity of MDSCs and the differentiation of TAMs, which ultimately should improve anti-tumor responses in patients. Using an ex vivo approach, we confirmed the importance of HDAC activation and IL-6 signaling pathway in regulating the expression of genes associated with MDSC suppressive function, such as ARG1, and genes related to myeloid cell chemotaxis, such as CCR2 and ITGAL.
This study provides novel insights into the epigenetic mechanisms and signaling pathways regulating the transcriptional profile, and perhaps the function, of different myeloid cell subsets within the CRC TME. Furthermore, findings from this study indicate the importance of HDAC activation and IL-6 signaling pathway in regulating MDSC suppressive function and perhaps the recruitment of MDSCs to the CRC microenvironment.
Sample collection and storage
Tumor tissues (TT) and paired, adjacent normal tissues (NT) were obtained from 27 CRC patients who underwent surgery at Hamad Medical Corporation, Doha, Qatar. All patients included in the study were treatment-naïve prior to surgery and provided written informed consent prior to sample collection. Table 1 shows the clinical and pathological characteristics of all participating patients. All experiments were performed in accordance with relevant guidelines and regulations. This study was executed under ethical approvals from Hamad Medical Corporation, Doha, Qatar (protocol no. MRC-02-18-012) and Qatar Biomedical Research Institute, Doha, Qatar (protocol no. 2018-018). Tissue specimens were cut into small pieces and frozen in 1 ml of freezing medium (10% dimethylsulphoxide (DMSO; Sigma-Aldrich, Missouri, USA), 50% fetal calf serum (FCS; Hyclone, GE Healthcare Life Sciences, Utah, USA), and 40% RPMI-1640 medium (Life Technologies, New York, USA)), then stored in liquid nitrogen to be used in batches for subsequent analyses.
Cells were isolated from TT by mechanical disaggregation. Briefly, tissues frozen in freezing media were thawed and washed with phosphate-buffered saline (PBS) and then mechanically cut into small pieces (~ 2–4 mm) using a surgical scalpel. Further disaggregation was performed on gentleMACS dissociator (Miltenyi Biotech, Bergisch Gladbach, Germany) without using any enzymes. The cell suspension was then passed through a 100-μM cell strainer to remove debris and aggregates. The single cell suspension was washed with PBS and stained for flow cytometric analyses and FACS sorting.
Multi-parametric flow cytometry
Cells isolated from tissues were washed with PBS and re-suspended in 100 μl flow cytometry staining buffer (PBS with 1% FCS and 0.1% sodium azide). Fc receptors (FcR) were first blocked using FcR Blocker (Miltenyi Biotech, Bergisch Gladbach, Germany). 7-AAD viability dye (eBioscience, San Diego, USA) was used to gate live cells. Cells were then stained with cell surface antibodies against CD33-Fluorescein isothiocyanate (clone HIM3-4; BD Biosciences, Oxford, UK), CD14-phycoerythrin-Cy7 (clone M5E2; BD Biosciences), CD15-allophycocyanin (clone HI98; BioLegend, San Diego, USA), HLA DR-phycoerythrin (clone G46-6; BD Biosciences) or CD11a-phycoerythrin (clone G43-25B; BD Pharmingen, San Jose, USA) and incubated at 4 °C for 30 min. Cells were then washed twice with flow cytometry staining buffer and data were acquired by BD LSRFortessa X-20 flow cytometer (BD Biosciences).
For sorting, cells were re-suspended in Pre-Sort buffer (BD Biosciences). BD FACSAria III SORP cell sorter with BD FACSDiva software (BD Biosciences) was used. Applicable measures were taken to ensure minimal sorter-induced cell stress (SICS). Data analyses were performed on FlowJo V10 software (FlowJo, Ashland, USA).
Cells were isolated from TT by mechanical disaggregation, as described above. 0.5 × 106 cells per well were seeded onto a 24-well tissue culture-treated plate and cultured in complete RPMI-1640 medium (Life Technologies). Cells were untreated or pre-treated with 100 nM of HDAC inhibitor Vorinostat (SAHA, MK0683, Selleckchem, Texas, USA) or 2 μg/ml of anti-IL-6 monoclonal antibody (Sino Biological, Beijing, China) . After a 2-h incubation period, 100 ng/ml of Lipopolysaccharides (LPS from E. coli, Sigma-Aldrich) was added to all the wells. Cells were kept in a humidified incubator at 37 °C in 5% CO2 for 5 days. Every other day, HDAC inhibitor and anti-IL-6 mAb were added to the wells at the same concentration. Cells were harvested on day 5 post treatment for RNA extraction.
Pure 1000 CD33+HLA-DR−CD14−CD15− (I-MDSCs), CD33+HLA-DR−CD14−CD15+ (PMN-MDSCs) and CD33+HLA-DR+CD14+CD15− (APCs) cells were sorted from TT and NT. Cells were initially sorted from six CRC NT and TT samples. Library preparation and subsequent RNA-Seq were succeeded in subsets from two patients only (#07 and #08). This could be due to the lower percentage and lower expression level of MDSC subsets in some TT and NT samples. Similarly, CTLA-4+ and CTLA-4− Tregs from head and neck cancer patients were compared in very small number of samples . The cDNA libraries were generated using QIAseq FX Single Cell RNA Library Kit (Qiagen, Hilden, Germany) following the manufacturer’s instructions. Briefly, sorted cells were spun down and lysed immediately. The gDNA was then removed by using gDNA Wipeout buffer (Qiagen). The gDNA-removed lysates were used to generate double-stranded DNA, which were subsequently amplified using REPLI-g sc SensiPhi DNA polymerase. Small fractions of amplified products were cleaned using PureLink PCR Purification Kit (Thermo Fisher Scientific, Massachusetts, USA) and quality checked using Agilent High Sensitivity DNA Kit (Agilent Technologies, California, USA). The quality passed DNA (> 2000 bp) were quantified using Qubit dsDNA HS Assay Kit (Invitrogen, California, USA). Five hundred nanograms to 1 μg DNA was enzymatically fragmented and ligated using paired adaptors. DNA was further purified using Agencourt AMPure XP beads (Beckman Coulter, California, USA). The yield and size distribution of libraries (500–1000 bp) were determined using Qubit dsDNA HS Assay Kit (Invitrogen) and Agilent High Sensitivity DNA Kit (Invitrogen).
RNA-sequencing data analyses
Pair end reads were quality-trimmed and were aligned to the hg19 human reference genome in CLC Genomics Workbench-12 (Qiagen) using default settings [57, 58]. The abundance of the expression of transcripts was measured as the score of TPM (transcripts per million) mapped reads in CLC Genomics Workbench 12. Abundance data were subsequently subjected to differential gene expression using built-in statistical analyses recommended in CLC Genomics protocol with 2.0-fold change and P value cutoff < 0.05.
Functional annotation analyses using DAVID platform
Database for Annotation, Visualization, and Integrated Discovery (DAVID) platform was used to analyze the functional annotation, in addition to Kyoto Encyclopedia of genes and Genomes (KEGG)/BioCarta network [59, 60]. The “Functional Annotation,” “KEGG pathway,” and “BioCarta” in the online version of DAVID 6.8 was performed (https://david.ncifcrf.gov/) by matching with Homo sapiens gene list. The upregulated genes and downregulated genes were trimmed with both expression consistency of individual genes (> 2 FC) and P value cutoff < 0.05. The upregulated genes and downregulated genes were uploaded separately and focused on the Gene-Ontology, Functional-Categories, Gene-Annotations and Pathways. To visualize the network pathways, we used both KEGG and BioCarta. The related genes from functional analyses were depicted as heat plots.
RNA extraction and reverse transcription
Total RNA was extracted from the sorted tumor-infiltrating myeloid cell populations (APCs, I-MDSCs and PMN-MDSCs) isolated from six CRC patients (#05, 07, 08, 09, 44, and 53), using RNAqueous-Micro Total RNA isolation Kit (Thermo Fisher Scientific). Total RNA was also extracted from cultured cells isolated from tumor tissue of four CRC patients (#26, 56, 60, and 63) using Total RNA Purification Plus Micro Kit (Norgen, Ontario, Canada). RNA was then amplified using 5X MessageAmp II aRNA Amplification Kit (Thermo Fisher Scientific). RNA concentrations before and after amplification were determined by Qubit RNA HS and Broad Range Assay Kit, respectively (Invitrogen). One microgram of RNA was reverse transcribed into cDNA using QuantiTect Reverse Transcription Kit (Qiagen).
Quantitative real-time reverse transcriptase PCR (qRT-PCR)
qRT-PCR was performed using QuantStudio 6/7 Flex Real-time PCR system (Applied Biosystems, California, USA) for KAT2B, CD40, RPA4, UST, CELF2, CD38, OAS1, HDAC1, HDAC2, HDAC3, HDAC4, ARG1, IDO, ITGAL, CCR2, Wnt5a, Wnt5b, and β-actin with PowerUp SYBR Green Master Mix (Applied Biosystems). Quantification of relative gene expression was determined, using 2−ΔΔCT, and normalized to β-actin. Sequences for the primers are shown in Additional file 2: Table S1.
Statistical analyses were performed using GraphPad Prism 8 software (GraphPad Software, California, USA). Paired t tests were performed on samples that passed the Shapiro-Wilk normality test and Mann-Whitney tests were performed for samples that did not show normal distribution. A P value of > 0.05 was considered statistically non-significant. The P values are represented as follows: ***P < 0.001, **P < 0.01, and *P < 0.05. Data are presented as mean ± standard error of the mean (SEM).
We are grateful to patients for donating their samples. We also would like to thank the genomics core facility at Qatar Biomedical Research Institute for performing RNA-sequencing.
VN and RS performed experimental work, data analyses, and wrote the manuscript. ST and RT assisted in experimental work. AAA and MAK contributed to sample collection and analyzing the clinical data. KM helped in studying patients’ samples and provided the clinicopathological data. NMA performed bioinformatics and data analysis and reviewed the manuscript. MAN contributed to the study design, sample collection, and reviewing the manuscript. EE conceived the idea, designed the study, supervised the project, analyzed and interpreted data, and wrote and revised the manuscript. All authors were involved in the final approval of the manuscript.
This work was supported by a start-up grant [VR04] for Dr Eyad Elkord from Qatar Biomedical Research Institute, Qatar Foundation.
Ethics approval and consent to participate
This study was executed under ethical approvals from Hamad Medical Corporation, Doha, Qatar (protocol no. MRC-02-18-012), and Qatar Biomedical Research Institute, Doha, Qatar (protocol no. 2018-018).
Consent for publication
The authors declare that they have no competing interests.
- 6.Talmadge JE, Cole K, Britton H, Dafferner A, Warkentin P. Human myeloid derived suppressor cell (MDSC) subset phenotypes. The Journal of Immunology. 2017;198:211.212.Google Scholar
- 11.Toor SME, Elkord: Myeloid-derived suppressor cells. eLS John Wiley 2015. https://doi.org/10.1002/9780470015902.a0024245.
- 19.Jiang M, Chen J, Zhang W, Zhang R, Ye Y, Liu P, Yu W, Wei F, Ren X, Yu J. Interleukin-6 trans-signaling pathway promotes immunosuppressive myeloid-derived suppressor cells via suppression of suppressor of cytokine signaling 3 in breast cancer. Front Immunol. 2017;8:1840.PubMedPubMedCentralCrossRefGoogle Scholar
- 23.Reina M, Espel E. Role of LFA-1 and ICAM-1 in Cancer. Cancers (Basel). 2017;9.Google Scholar
- 38.Trikha P, Carson WE 3rd. Signaling pathways involved in MDSC regulation. Biochim Biophys Acta. 1846;2014:55–65.Google Scholar
- 47.Hefetz-Sela S, Stein I, Klieger Y, Porat R, Sade-Feldman M, Zreik F, Nagler A, Pappo O, Quagliata L, Dazert E, et al. Acquisition of an immunosuppressive protumorigenic macrophage phenotype depending on c-Jun phosphorylation. Proc Natl Acad Sci U S A. 2014;111:17582–7.PubMedPubMedCentralCrossRefGoogle Scholar
- 49.Ouzounova M, Lee E, Piranlioglu R, El Andaloussi A, Kolhe R, Demirci MF, Marasco D, Asm I, Chadli A, Hassan KA, et al. Monocytic and granulocytic myeloid derived suppressor cells differentially regulate spatiotemporal tumour plasticity during metastatic cascade. Nat Commun. 2017;8:14979.PubMedPubMedCentralCrossRefGoogle Scholar
- 51.Tu S, Bhagat G, Cui G, Takaishi S, Kurt-Jones EA, Rickman B, Betz KS, Penz-Oesterreicher M, Bjorkdahl O, Fox JG, Wang TC. Overexpression of interleukin-1beta induces gastric inflammation and cancer and mobilizes myeloid-derived suppressor cells in mice. Cancer Cell. 2008;14:408–19.PubMedPubMedCentralCrossRefGoogle Scholar
- 56.Matoba T, Imai M, Ohkura N, Kawakita D, Ijichi K, Toyama T, Morita A, Murakami S, Sakaguchi S, Yamazaki S. Regulatory T cells expressing abundant CTLA-4 on the cell surface with a proliferative gene profile are key features of human head and neck cancer. Int J Cancer. 2019;144:2811–22.PubMedCrossRefGoogle Scholar
- 60.Huang DW, Sherman BT, Tan Q, Kir J, Liu D, Bryant D, Guo Y, Stephens R, Baseler MW, Lane HC, Lempicki RA. DAVID bioinformatics resources: expanded annotation database and novel algorithms to better extract biology from large gene lists. Nucleic Acids Res. 2007;35:W169–75.PubMedPubMedCentralCrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.