Evaluation of colorectal cancer in ulcerative colitis reveals key immune factors during its malignant transformation

doi:10.21203/rs.3.rs-3339612/v1

Download PDF

Research Article

Evaluation of colorectal cancer in ulcerative colitis reveals key immune factors during its malignant transformation

https://doi.org/10.21203/rs.3.rs-3339612/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

BACKGROUND & AIMS:

Ulcerative colitis (UC) is linked to an increased risk of colitis-associated colorectal cancer (CAC), which accounts for approximately 15% of UC-related deaths. Despite this significant impact on patients, the mechanism behind how UC promotes cancer development remains unknown. The present study aims to investigate alterations in the immune microenvironment during the malignant transformation of UC, shedding light on the underlying mechanisms of UC carcinogenesis.

METHODS

We collected single-cell transcriptome samples of 41 healthy samples, 45 UC samples, and 148 colorectal cancer(CRC) samples from public databases. Using the UC-CRC signature, we were able to screen for CAC-like samples. Based on those datasets, several bioinformatics analyses were performed on 228,538 immune cells to evaluate the immune microenvironment from UC to CAC.

RESULTS

Using predefined UC-CRC signature, we screened 14 CAC-like samples and revealed an immune remodeling process from healthy tissue to UC and CAC-like samples, particularly involving the VEGFA_Macro cells and Treg cells. VEGFA_Macro cells was significantly enriched in UC and CAC-like samples, showed a phenotype alteration during disease progression, and expressed more inflammation-related genes and signal pathways. Additionally, the proportion of Treg cells gradually increased with disease progression, potentially promoting an immunosuppressive microenvironment. Comparative analysis of the immune microenvironment between CAC-like and sporadic CRC(sCRC) samples revealed higher levels of myeloid cells but reduced CD8 + T cells in CAC-like samples. Finally, we simplified the UC-CRC signature for ease of clinical use in screening CAC-like samples.

CONCLUSIONS

Our results may help improve the understanding dynamic change of immune microenvironment from UC to CAC and provide clues for further exploration of strategies to prevent carcinogenesis of UC.

Ulcerative colitis (UC) is a chronic, idiopathic, relapsing inflammatory disease of the gastrointestinal (GI) tract that can lead to one of its most severe complications: colitis-associated colorectal cancer (CAC)[1]. Patients with UC have an approximately 2–3 fold increased risk of developing colorectal cancer(CRC), which contributes to about 15% of UC-related mortality[2]. The risk of developing cancer in UC patients gradually increases with the length of the disease. According to a meta-analysis by Eaden et al.[3], the incidence of CAC in UC patients was found to be 2%, 8%, and 18% after 10, 20, and 30 years of diagnosis, respectively. Accumulating evidence suggests that CAC may emerge through a distinct pathway of tumorigenesis compared to sporadic CRC (sCRC)[4]. Patients with CAC experience earlier morbidity and a poorer prognosis[5]. Thus, the effective differentiation of CAC from sCRC may provide valuable insights for the treatment of clinical patients.

CAC is infiltrated by various types of immune cells, much like other solid malignancies. CAC develops in the context of chronic inflammation and pro-inflammatory immune cells, and their products may contribute to the development and progression of cancer[6]. However, CAC also occurs in the context of endogenous anti-inflammatory or inhibitory mechanism activation[7]. While the tumor microenvironment (TME) plays a vital role in the carcinogenesis of UC, studies have primarily focused on the bulk profiling of advanced-stage tumors, while largely ignoring premalignant lesions. Consequently, the role of the immune microenvironment in the transformation of malignant UC into CAC remains unclear. Therefore, to comprehend the progression of changes in the TME, it is essential to gain a deeper understanding of the molecular drivers behind this transformation during the transition from normal to UC and ultimately to CAC.

Recent advances in single-cell transcriptome sequencing (scRNA-seq) have revolutionized our ability to analyze the transcriptional state of thousands of individual cells in a given tissue, which facilitates an unbiased exploration of the tumor microenvironment (TME). Collecting both precancerous and early-stage cancer samples, scRNA-seq can fulfill an analytical method that tracks TME dynamics transitioning between pre-cancerous and cancerous states. For instance, Greenleaf investigated poly adenoma and CRC and found that advanced polyps contain increasing numbers of Treg cells (regulatory T cells)[8, 9]. As no study has been done on directly collected samples that have experienced the full transition process from UC to CAC, the TME’s detailed changes over the extended period between UC and early-stage CAC remain unexplored.

In this study, we collected single-cell transcriptome samples from healthy, UC, and CRC patients in public datasets. We defined a UC-CRC signature based on previous research and identified 14 CAC-like single-cell samples out of the 148 colorectal cancer samples (Fig. 1a). We comprehensively analyzed a single-cell atlas of 228,538 immune cells from 14 CAC-like samples, 45 UC samples, and 41 healthy intestinal samples. We found that immune remodeling from healthy tissue to UC and early-stage CAC-like, particularly the VEGFA_Macro cells and Treg cells subtypes. Those result sequencialy validated in three extra cohorts. Further analysis show higher levels of myeloid cells but reduced CD8 + T cells in CAC-like samples comparing to those in sCRC samples. In addition, we simplified the UC-CRC signature for clinical use in screening CAC-like samples. Collectively, our results identify the rich tumor environment changes between UC and CAC-like and provide valuable clues leads for future studies to prevent UC’s carcinogenesis.

Allocating CAC-like samples from single-cell RNA-sequencing datasets of colorectal cancer

As different gene expression patterns have been reported between CAC and sCRC[10, 11], we hypothesize that CAC can be efficiently distinguished from sCRC using a list of differentially expressed genes. For this purpose, we employed a recently published list of DEGs obtained from analyzing 17 CAC and 47 sCRC transcriptome samples[12], which we refer to as the UC-CRC signature. To find the biological meaning of the signature, we then conducted pathway enrichment analysis based on the gene set[13, 14]. We identified the staphylococcus aures infection and the cell adhesion molecules (CAMs) pathways as the most significant signaling pathways (Extended Data Figs. 3a). Bacterial infection is one of the potential causes of UC[15], and CAMs are small molecular peptides or glycoproteins that mediate the cell-to-cell and/or cell-to-extracellular-matrix interactions that are involved in various pathophysiological processes such as immune response, inflammation, and tumor metastasis[16, 17]. To determine whether the UC-CRC signature could serve as a screening tool for identifying colorectal cancer samples with CAC properties, we assessed the signature in 9 datasets with a total of 2,277 samples [18–24](Supplementary Table 2). Using the GSVA algorithm[25], we calculated the UC-CRC score for each sample. By selecting the top 5%, 8%, and 10% of samples with the highest UC-CRC scores, we found that over half of the samples in all 9 datasets were categorized as CMS4 subtype (Figs. 1b, Extended Data Figs. 1a-1d, Extended Data Figs. 2a-2d). CAC are predominantly of CMS4 subtype[26], which is recognized as the “inflamed” subtype of colorectal tumors. Furthermore, our analysis of the TCGA-COAD and GSE39582 datasets revealed that patients with higher UC-CRC scores had a shorter overall survival (OS) than those with lower scores (Figs. 1c, Extended Data Figs. 3b-3f), although only the differences in the top 5% scores reached statistical significance. Overall, the distinct gene expression patterns observed between CAC and sCRC suggest that the UC-CRC signature may serve as an interpretable and workable tool to distinguish CAC from sCRC.

Next, we were going to assign samples into CAC based on single cell dataset. We assembled 148 single-cell samples of colorectal cancer from the GEO database obtained from 7 cohorts[27–33]. After quality control and conversion into pseudo-bulk data, UC-CRC score were calculated for each sample (Supplementary Table 3). To determine the optimal cut-off value for identifying CAC-like samples, we utilized a finite mixture model (FMM). FMM is a statistical model that assumes the presence of unobserved groups or “latent classes” within a population. Each latent class can be modeled with its own regression function, which may have a linear or generalized linear response. We used the FMM-recommended optimal number of classifications to classify the 14 samples with the highest GSVA scores into one category (Fig. 1d, Supplementary Table 4). To validate this classification with an alternative algorithm, K-means clustering was used to show that the 14 samples with the highest GSVA scores could be assigned to the same category (Extended Data Figs. 3g, Supplementary Table 5). Therefore, we believe that these 14 samples exhibit similar CAC characteristics and, as such, we refer to them as “CAC-like single-cell samples.”

Single-cell immune atlas of healthy, UC and CAC-like samples

Subsequent to obtaining the CAC-like samples, we undertook a sequential investigation regarding the role of the immune microenvironment in the malignant transformation of UC to CAC. To accomplish this, we integrated 41 healthy samples and 45 UC samples that stemmed from 6 studies[34–39] with the 14 CAC-like samples (Figs. 1a). After filtering out low-quality cells and doublets, we obtained a total of 289,751 high-quality single cells for further analysis. We employed dimensionality reduction and unsupervised cell clustering to identify distinct cell populations based on gene expression patterns, followed by removing batch effects among multiple samples. As a result, almost all subsets were represented in all specimens (Extended Data Figs. 4a), suggesting a satisfying quality for the next analyses. Based on the expression of known markers (Extended Data Figs. 4b-4c), we found that the atlas was composed primarily of four cell types, including immune cells, endothelial cells, epithelial cells and fibroblast cells(Supplementary Table 6). We found samples of all three disease states are primarily composed of immune cells (Extended Data Figs. 4d). Subsequently, we subset the immune type for further analysis (Supplementary Table 6). After marker-based annotations (Extended Data Figs. 5a, Supplementary Table 7), five main cell clusters were identified in both healthy, UC, and CAC-like samples (Extended Data Figs. 5b), namely: mast cells, myeloid cells, B cells, plasma cells, and T_NK cells. With disease progression, there was an overall increase in myeloid and a decrease in plasma-cell populations compared to healthy tissues (Extended Data Figs. 5c), pointing towards a redirected immune response, and implicating the immune response in the transformation of UC to CAC.

VEGFA_Macro cells contributes the promotion process from UC carcinoma to CAC

Among the five identified cell types, myeloid cells exhibited a remarkable increase in number with the progression of the disease. We divided myeloid cells into 11 subpopulations (Figs. 2a, Supplementary Table 8) and designated six subpopulations as macrophages and four subpopulations as dendritic cells (DCs) based on high or low CD14 expression levels (Figs. 2b). One subpopulation of plasmacytoid DCs (pDCs) and three subpopulations of conventional DCs (CLEC9A_cDC1, CD1C_cDC2, and CCR7_cDC3) were identified, in keeping with the findings of single-cell analyses in recent studies[40, 41](Figs. 2b). As the disease progressed, CLEC9A_cDC1 and CD1C_cDC2 populations gradually decreased, while the CCR7_cDC3 population gradually increased (Figs. 2c). The mobilization of DCs to lymphoid organs relies on chemokine stimulation and the activation of chemokine receptor CCR7 - a key regulator that governs DC populations in both inflammatory and steady-state conditions[42–44]. The rise in CCR7_cDC3 cells could reflect immune response activation. A recent study on Crohn’s disease noted that CCR7_cDC3 was linked to anti-TNF-resistant therapy[45]. Of the six macrophage subpopulations identified (Figs. 2a-2b), VEGFA_Macro cells exhibited the most pronounced change in proportion, being predominantly enriched in UC and CAC-like samples (Figs. 2c).

To further characterize the function of VEGFA_Macro cells in CAC progression process, we combined all VEGFA_Macro cells and constructed a single-cell pseudotime trajectory by utilizing the monocle3 R package. We assigned the healthy samples as the root state manually (Figs. 2d) and noticed that these samples generally gathered close to the root of the pseudotime trajectory. However, UC and CAC-like samples tended to be dispersed uniformly throughout the middle and the end portions of the trajectory, which may represent UC and CAC-like VEGFA_Macro cells had similar characteristics. To explore which genes drive the differentiation of VEGFA_Maro from healthy state to disease state. We perfomed differentially expressed analysis by the pseudotime trajectory and calculated their co-expression modules (Extended Data Figs. 6a). We subsequently selected gene modules that increased gradually as the disease progressed and assessed them for enrichment (Figs. 2e). Those result suggested that inflammation-related pathways, such as the NF-kappa B signaling pathway, TNF signaling pathway, and IL-17 signaling pathway were significantly changed in VEGFA_Macro cells during the disease progression. With the progression of the disease, these pathways increased (Extended Data Figs. 6b), indicating the crucial role of inflammatory signaling pathways in promoting the transformation of VEGFA_Macro cells from a healthy phenotype to an inflammation and cancer phenotype. The most significant pathway is the NF-kappa B signaling pathway, which is a family of transcription factors that regulate the transcription of pro-inflammatory cytokines, chemokines, and other inflammatory mediators, thereby promoting and sustaining the inflammatory response[46]. As a core regulator of inflammatory responses, NF-κB has been implicated in the development of UC[47]. Additionally, it has been found that NF-κB contributes to CAC by primarily activating the transcription of pro-inflammatory cytokines and promoting tumor growth and metastasis through increased expression of angiogenesis-related genes[48, 49]. In comparison to the healthy samples, we identified another pathway - the HIF-1 signaling pathway - that was significantly enriched in UC and CAC-like samples (Figs. 4e and Extended Data Figs. 5b). To note, the key gene of the HIF-1 signaling pathway, HIF1A (hypoxia-inducible transcription factor 1), gradually increased in expression (Figs. 2f). The HIF-1 signaling pathway plays a role in hypoxic disorders, tumorigenesis and development, as well as inflammation. HIF-1 can activate the transcription of genes critical for cancer, including signaling pathways that regulate cell proliferation, angiogenesis, and cell death[50, 51]. HIF-1a levels rise in hypoxic environments and up-regulate various angiogenic factors, especially VEGF (vascular endothelial growth factor)[52], to induce angiogenesis. The genes VEGFA and EREG (epidermal growth factor) related to angiogenesis are specifically expressed in VEGFA_Macro cells, and their expression gradually increases with disease progression (Figs. 2f). To gain a deeper understanding of VEGFA_Macro cells’ role in the UC and CAC-like tumor microenvironment (TME), we examined interactions between VEGFA_Macro cells and different immune cell populations, as well as the strength of cell interactions in healthy, UC, and CAC-like samples. Based on the receptor-ligand database, we inferred a putative cellular interaction network (Extended Data Figs. 6d). VEGFA_Macro cells subpopulations had the most robust interaction with other immune cell types among the CAC-like samples. These findings suggested that VEGFA_Macro cells populations may play critical roles in UC progression to CAC.

To establish a stronger connection between VEGFA_Macro cells and the risk of developing UC carcinogenesis, we conducted additional analysis on three cohorts: GSE109142, GSE128682, and GSE150961[53–55]. The GSE109142 cohort contains rectal mucosal biopsies collected from new-onset pediatric ulcerative colitis patients before treatment. It also includes the pediatric ulcerative colitis activity index (PUCAI), which evaluates the severity of ulcerative colitis in children. Meanwhile, GSE128682 comprises RNA-seq data from mucosal biopsies of UC patients in remission, those in activation, and healthy individuals. Lastly, the GSE150961 cohort, a multicenter study, explores the response to standardized colitis therapy and includes rectal RNAseq biopsies from 21 patients who received colectomy and 310 who did not. We used the DEG of the VEGFA_Macro cells cluster to derive the gene set variation analysis (GSVA), allowing us to quantify VEGFA_Macro cells in each transcriptome sample. Our results showed that the VEGFA_Macro cells signature positively correlates with the severity of UC in all three cohorts (Figs. 2f-2i). This finding emphasizes the significant role that VEGFA_Macro cells plays in promoting the transformation of UC into CAC.

T_NK cells remodeling from healthy, UC to CAC, especially Treg cells

In addition to myeloid cells, we investigated B and plasma cells. We clustered a total of five types of B cells and two types of plasma cells (Extended Data Figs. 7a-7b, Supplementary Table 9). We observed a significant reduction in the IGHA1_IGHA2_Plasma cell compartment in UC and CAC-like tissue (Extended Data Figs. 7c). Moreover, we identified 15 subpopulations of T and NK cells (Figs. 3a). Based on the expression of CD4 and CD8A, we divided the 14 T cell subsets into 4 CD8 + T cells and 10 CD4 + T cells (Figs. 3b, Supplementary Table 10). Among the clusters, we observed the CD8_STMN1_GZMB_Tex cluster consisting almost exclusively of cells from CAC-like samples (Figs. 3c). Our findings suggest that the immune microenvironment of CAC-like tissue is in a state of exhaustion. We also found a gradual increase in CD4_RORA_Th17 cells in all three conditions. This subset of cells expresses multiple inflammatory cytokines, including IL-17A and IL-22. While the role of CD4_RORA_Th17 in CAC-like pathogenesis is not yet clear, studies in mouse models have shown that IL-17A ablation reduces CAC-like tumorigenesis[56]. Notably, Treg cells showed the largest percentage change in disease progression, mainly enriched in UC and CAC-like samples.

Enrichment of Treg cells is associated with the risk of carcinogenesis of UC

Treg cells is a type of cell that suppresses aberrant immune responses against self-antigens. However, it can also suppress anti-tumor immune responses, leading to poor prognosis when infiltrated into tumor tissues[57, 58]. Chronic inflammation recruits Treg cells[59], which increases the proportion of Treg cells in UC and CAC-like tissues compared to healthy tissues. Several molecules, such as CTLA4 (cytotoxic T-lymphocyte-associated protein 4) and TIGIT (T cell immune receptor with Ig and ITIM domains), are involved in Treg cells-mediated suppression mechanisms[60–62]. Our data shows that the expression levels of TIGIT and CTLA4 in Treg cells significantly increase with disease progression (Figs. 3d-3e). Furthermore, we found that the communication intensity between Treg cells and other immune cells significantly increases as the disease progresses (Figs. 3e), indicating an increased immunosuppressive effect of Treg cells on other cells. In addition, we verified that the higher the risk of malignancy in UC, the more enriched Treg cells exist (Figs. 3f-3h). To summarize, chronic inflammation in UC may create an immunosuppressive environment with excessive recruitment of Treg cells, which prevents adequate immunosurveillance and promotes the evolution from UC to CAC.

CAC-like and sCRC had different immune microenvironments.

To confirm that the findings are specific to CAC, we compared the immune microenvironment of CAC-like and sCRC samples based on their original classification. We first screened the top 50% pseudo-bulk CRC samples with a lower UC-CRC score and integrated them with CAC-like samples (Extended Data Figs. 8a). We classified the immune cells into five major cell types, including B cells, T_NK cells, myeloid cells, plasma cells, and mast cells (Extended Data Figs. 8d-8e). For a more comprehensive analysis, we designated six subsets as macrophages and four myeloid subsets as dendritic cells (DCs) (Extended Data Figs. 9a-9b, Supplementary Table 11), six T subsets as CD8 + T cells, and seven T subsets as CD4 + T cells (Extended Data Figs. 9c-9d, Supplementary Table 12). Lastly, we identified six B subsets and two plasma subsets (Extended Data Figs. 9e-9f, Supplementary Table 13) to further refine our analysis.

We used the Roe score to compare changes in immune cells between CAC-like and sCRC samples (Figs. 4a). The myeloid system in CAC-like samples tends to expand, including VEGFA_Marco, while T cells in CAC-like samples show an overall downward trend, especially CD8 + T cells. CD8 + T cells are the main effector cells with anti-tumor function, and their decreased proportion may lead to a decline in the immune system’s anti-tumor function. This may explain the poorer prognosis of CAC compared to sCRC. Given our research results, we hypothesized that VEGFA_Marco might be functionally distinct across different disease statuses. To pinpoint the difference imprinted by the unique tumor microenvironment, we performed differential gene expression analysis of VEGFA_Marco in CAC-like and sCRC samples. Pathway enrichment analyses showed strong enrichment of inflammation-related pathways in VEGFA_Marco from CAC-like samples, such as the IL-17 signaling pathway and NF-kappa B signaling pathway (Figs. 4b). This highlights the fundamental role of inflammation in determining macrophage function. Notably, we found that VEGFA_Marco in sCRC samples were enriched with oxidative phosphorylation and the chemical carcinogenesis-reactive oxygen species signaling pathway. Additionally, some key genes mentioned earlier, such as VEGFA, EREG, and HIF1A, had significantly higher expression levels in CAC-like samples than in sCRC samples (Figs. 4c). We also found that CTLA4 and TIGIT, which are essential for the function of Treg cells, had much higher expression levels in CAC-like samples compared to sCRC samples (Figs. 4d). These data confirm that different disease statuses may mediate the phenotypic and functional shift of immune cells in response to distinct microenvironmental cues in such disease.

Simplified the UC-CRC signature with VSOLassoBag algorithm.

The genes involved in UC-CRC signature discussed in this paper mainly come from a reference document and comprise up to 870 genes. Our analysis shows that these genes can screen for samples with CAC characteristics. However, in clinical applications, the detection of so many genes face complex data processing and cost issues. Therefore, to reduce the number of genes to make sample differentiation clinically accessible and actionable, we applied the VSOLassobag[63] method to shrink the genelist that also keep the classification ability. The VSOLassoBag algorithm integrates bagging ensemble learning and LASSO algorithm, which can efficiently and stably screen high-confidence biomarkers from limited genomic data. We used the bagging method(Figs. 5a) of VOSLassobag to simplify all genes in pseudo-bulk CRC samples and obtained 899 genes(Supplementary Table 14). After intersecting with UC-CRC signature, we obtained 45 genes(Supplementary Table 15) that can be used to distinguish between CAC and sCRC(Figs. 5b). Some of these genes are closely related to UC and CAC, such as VCAM1,C3. Vascular Cell Adhesion Molecule-1 (VCAM-1) is an endothelial cell adhesion molecule that mediates leukocyte binding[64] and plays a central role in the progression of ulcerative colitis (UC) and the development CAC[65]. The Complement component C3 level is significantly elevated in patients with UC[66] and have a promoting effect on CAC[67]. Depletion of complement in the (AOM/DSS) mouse model has been shown to suppress tumor growth in (AOM/DSS) mice[68]. Importantly, we found that these 45 genes can effectively distinguish CAC-like samples, and their performance is almost equivalent to that of the 870 genes(Supplementary Table 16). The signature of these 45 genes in pseudo-bulk data can achieve effects similar to those of the UC-CRC signature. Therefore, we believe that these 45 genes can serve as a gene panel to replace the above classification and have clinical usability.

Patients with UC have an increased incidence of CAC. Previous studies have demonstrated the significant role of the immune microenvironment in the carcinogenesis of UC. CAC typically develops against a background of chronic inflammation and pro-inflammatory immune cells whose products contribute to cancer development and progression. Additionally, the effects of the immunosuppressive microenvironment in CAC development and progression have gained more attention. However, there is no research using single-cell data to systematically compare changes in the immune microenvironment throughout the process of UC carcinogenesis.

This study, for the first time, to explored changes in the immune microenvironment during malignant transformation from UC to CAC at the single-cell level. As UC approaches malignancy, the myeloid lineage expands significantly, especially VEGFA_Macro. Different disease states show that VEGFA_Macro cells has different phenotypes, and as the disease progresses, it significantly up-regulates inflammation-related cellular pathways that suggests it increased immune activity. In addition, the more malignant the disease, the more VEGFA_Macro cells will express genes that promote the growth of cancer cells, such as VEGFA and EREG. Although there are many reports on the role of macrophages that express VEGFA in colorectal cancer, which can promote tumor angiogenesis and cell proliferation[69, 70], no study previously connected VEGFA_Macro cells with the transformation of UC into CAC. Furthermore, as a main part of immune suppress cell type, Treg cells gradually increase, promoting the occurrence of malignant tumor. Treg cells is a type of immunosuppressive cell, whose increase can produce an immunosuppressive microenvironment, thus promoting the occurrence of malignant tumors. Previous literature reports show that Treg cells were strongly increased in number, and transient ablation of regulatory T cells improves antitumor immunity in a mouse model of CAC[71, 72]. Additionally, by comparing the differences in the immune microenvironment between CAC-like and sCRC. We found that CAC myeloid system expands significantly while CD8 + T cells decrease significantly. The decrease of CD8 + T cells may lead to a significant weakening of anti-tumor effect, which may be one of the reasons why CAC has higher mortality rates than sCRC.

Interventions in disease progression in precancerous lesions are expected to provide easy-to-implement methods for preventing cancer deaths. However, most previous work has focused on advanced tumors and not precancerous lesions. To explore changes in the immune microenvironment in UC malignancy, we have established a strategy for screening CAC-like samples. This strategy may provide a favorable basis for patient management in the clinic. Based on this signature, we screened CAC-like samples from hundreds of samples and make feasible comparisons. Feature analysis showed that the selected CAC-like samples are consistent with CAC characteristics, such as more CMS4 typing and worse survival rates. In clinical practice, particularly in retrospective studies, there are no direct methods for distinguishing CAC samples. Therefore, in this study, we were only able to indirectly determine CAC samples by using the UC-CRC score. Additionally, we simplified the signature into a panel of only 45 genes for further application of the strategy. We believe this method could serve as a useful tool for distinguishing CAC samples from patients with colorectal cancer.

Collectively, this work has revealed that the immune microenvironment of UC undergoes significant changes during its progression to CAC. Specifically, VEGFA_Macro cells and Treg cells appear to play critical roles in the transformation of UC into CAC. The inhibition of both functions may have the potential to prevent UC from developing into CAC.

Sample and dataset collections

We downloaded the single-cell RNA-seq gene expression data from GEO (Gene Expression Omnibus, https://www.ncbi.nlm.nih.gov/geo/) with accession number GSE125527, GSE150115, GSE12527, GSE150516, GSE162335, GSE132465, GSE144735, GSE188711, GSE178318, GSE164522, GSE146771, GSE178341, GSE200997, Single Cell Portal(https://singlecell.broadinstitute.org/single_cell) with accession number SCP259, and Gut Cell Survey(https://www.gutcellatlas.org/). Colorectal cancer bulk transcriptome data and CMS typing information are from CRCSC(Colorectal Cancer Subtyping Consortium, https://www.synapse.org/#!Synapse:syn2634742). Bulk RNA-seq data of UC were downloaded from the GEO with accession number GSE109142, GSE128682, and GSE150961.

Bioinformatics analysis

Gene set variation analysis(GSVA). Firstly, we carried out quality control on the single cell data of 148 cases of colorectal cancer, eliminated 8 samples of poor quality, and converted the remaining samples into pseudo-bulk data. Then we performed “gsva” to derive the enrichment score of UC-CRC signature in pseudo-bulk data using a R package called GSVA(v.1.38.2).

Finite mixture models(FMM). We assume that some of the colorectal cancer samples are CAC and the other is sCRC, then the UC-CRC score calculated by GSVA will have different distributions. In order to get the best cut off value, we used finite mixture model(FMM) by using R package Flexmix (v.2.3–18)[73]. FMM is a statistical model that assumes the presence of unobserved groups, called latent classes, within an overall population. Each latent class can be fit with its own regression model, which may have a linear or generalized linear response function. We choose the best model according to Bayesian information criterion(BIC).

BIC = kln(n)-2ln(L)

Where k is the number of model parameters, n is the number of samples, and L is the likelihood function. Model evaluation criteria: the lower the BIC, the better. According to the optimal number of classifications recommended by the FMM, we classify the 14 samples with the highest GSVA scores into one category

K-means clustering algorithm. We use k-means clustering algorithm(KCA) to verify the cut off obtained by FMM. KCA is an iterative clustering analysis algorithm. Its step is to divide the data into K groups and randomly select K objects as the initial clustering centers, then calculate the distance between each object and each seed clustering center and assign each object to the nearest clustering center. Cluster centers and the objects assigned to them represent a cluster. For each sample assigned, the cluster center of the cluster will be recalculated according to the existing objects in the cluster. This process will be repeated until a termination condition is met. The termination condition can be that no (or minimum number of) objects are reassigned to different clusters, no (or minimum number of) clustering centers change again, and the sum of square errors is locally minimum. Under the recommended optimal number of classifications, we also classify the 14 samples with the highest GSVA scores into one category

ScRNA-seq: Quality control and doublet removal. As for quality control of scRNA-seq data, we set the cut off min.cells as 10 in the CreateSeuratObject function in Seurat (v.4.1.0)[74]. The data were filtered to remove cells with fewer than 200 unique genes per cell or greater than 2500 genes per cell, and remove cells with mitochondrial genes greater than 5%. To remove the potential doublets, we perform doublet remove by using DoubletFinder (v.2.0.3)[75].

The data were then processed with Seurat’s standard pipeline. First, NormalizeData was run using the method LogNormalize and scale.factor of 10,000. We scaled data with top 2000 most variable genes by using FindVariableFeatures function in R package Seurat (v.4.1.0). We used variable genes for principal component analysis (PCA), used FindNeighbors in Seurat to get nearest neighbors for graph clustering based on PCs, and used FindCluster in Seurat to obtain cell subtypes, and visualized cells with the uniform manifold approximation and projection (UMAP) algorithm. To eliminate the batch effect, we performed harmony algorithm in Harmony (v.0.1.0)[76] to remove batch correction before clustering analysis, and applied FindNeighbors and FindCluster in Seurat to obtain cell subtypes. The resolution was selected variably according to the number of cells. The major cell types were annotated by comparing the canonical marker genes and the differentially expressed genes (DEGs) for each cluster. We defined 33 immune cell types in CRC based on the gene signatures of each cell type and known lineage markers.

Differential-expression analysis. We used the “FindAllMarkers” function in Seurat to identify genes that are differentially expressed between clusters with the following parameters: min.pct = 0.25, logfc.threshold = 0.25. The non-parametric Wilcoxon rank-sum test was used to obtain p-values for comparisons, and the adjusted p-values, based on Bonferroni correction, for all genes in the dataset. We used heatmap to visualize DEGs based on gene expression after the log-transformed and scaling.

Cell pseudotime trajectory analysis. Monocle3(v.1.0.0)[77] was used to construct the pseudotime trajectory. In detail, the “cell_data_set” was built from the Seurat object of all the tumor cells using the data slot of the integrated assay. Dimension reduction was done by UMAP, “learn_graph” and “order_cells” functions were used to establish the trajectory. Each single cell was projected on the tree and formed the trajectory by DDRThree. The “root_state” of this trajectory tree was manually appointed in cells from healthy samples.

Using "graph_test" function to find differential genes of pseudotime trajectories(we set 1e-10 as q_value cut-off value) and calculate the co-expression gene module of the differential genes(Extended Data Figs. 4a).

Pathway enrichment analysis and Gene set variation analysis(GSVA). We select the co-expression gene modules that gradually increase with the state of the disease. We next performed the pathway enrichment analysis of those differential expressed genes by using cluterProfiler (v.3.18.1) (enricher function, KEGG gene sets or cancer hallmark genesets from msigdb). For GSVA analysis, e performed “gsva” to derive the enrichment score of these differentially expressed genes in the UC bulk transcriptome data using a R package GSVA(v.1.38.2).

Interaction map in each patient sample using CellPhoneDB. To analyze receptor-ligand pairing in healthy, UC, CAC-like samples, we used CellphoneDB(v.3.0.2)[78], a cell-cell interaction network tool for scRNA-seq datasets. The interactions between distinct cell subtypes via putative ligand-receptor pairs were visualized by a heatmap.

Simplified the UC-CRC signature VOSLassobag. To streamline the UC-CRC signature, we used the VOSLassobag to shrink genes, and further selecting the intersection of genes with a frequency greater than 1 and the UC-CRC signature.

Statistical Analysis.

Statistical analysis was performed in R v.4.0.3. Comparisons of numerical variables according to disease state were carried out using the Wilcox test and the ANOVA test. Survival analysis was performed by R package. Survival (V3.3-1) and survminer (V0.4.9) were adopted to generate the Kaplan-Meier plot and the P-value was calculated through log-rank test. Hazard ratio (HR) was calculated by Cox proportional hazards model and 95% CI was reported.

FUNDING

National Key Research And Development Plan (2022YFC3401000), National Natural Science Foundation of China (82022037), Guangdong Basic and Applied Basic Research Foundation (2021B1515230009)，Key Research and Development Plan of Guangdong Province(2020B0101030006）.

Code availability

The code for this article has been deposited with GitHub(https://github.com/jiaojiao1223/jiaojiao1223/tree/main/CAC).

Kaplan GG: The global burden of IBD: from 2015 to 2025. Nature reviews Gastroenterology & hepatology 2015, 12(12):720-727.
Leong RW, Koo JH: Colorectal cancer in inflammatory bowel disease. Journal of gastroenterology and hepatology 2009, 24(4):503-505.
Eaden JA, Abrams KR, Mayberry JF: The risk of colorectal cancer in ulcerative colitis: a meta-analysis. Gut 2001, 48(4):526-535.
Adams SV, Ahnen DJ, Baron JA, Campbell PT, Gallinger S, Grady WM, LeMarchand L, Lindor NM, Potter JD, Newcomb PA: Survival after inflammatory bowel disease-associated colorectal cancer in the Colon Cancer Family Registry. World journal of gastroenterology 2013, 19(21):3241-3248.
Dienstmann R, Vermeulen L, Guinney J, Kopetz S, Tejpar S, Tabernero J: Consensus molecular subtypes and the evolution of precision medicine in colorectal cancer. Nature reviews Cancer 2017, 17(4):268.
Waldner MJ, Neurath MF: Mechanisms of Immune Signaling in Colitis-Associated Cancer. Cellular and molecular gastroenterology and hepatology 2015, 1(1):6-16.
Terzic J, Grivennikov S, Karin E, Karin M: Inflammation and colon cancer. Gastroenterology 2010, 138(6):2101-2114 e2105.
Becker WR, Nevins SA, Chen DC, Chiu R, Horning AM, Guha TK, Laquindanum R, Mills M, Chaib H, Ladabaum U et al: Single-cell analyses define a continuum of cell state and composition changes in the malignant transformation of polyps to colorectal cancer. Nature genetics 2022, 54(7):985-995.
Chen B, Scurrah CR, McKinley ET, Simmons AJ, Ramirez-Solano MA, Zhu X, Markham NO, Heiser CN, Vega PN, Rolong A et al: Differential pre-malignant programs and microenvironment chart distinct paths to malignancy in human colorectal polyps. Cell 2021, 184(26):6262-6280 e6226.
Reynolds IS, O'Toole A, Deasy J, McNamara DA, Burke JP: A meta-analysis of the clinicopathological characteristics and survival outcomes of inflammatory bowel disease associated colorectal cancer. International journal of colorectal disease 2017, 32(4):443-451.
Ullman TA, Itzkowitz SH: Intestinal inflammation and cancer. Gastroenterology 2011, 140(6):1807-1816.
Rajamaki K, Taira A, Katainen R, Valimaki N, Kuosmanen A, Plaketti RM, Seppala TT, Ahtiainen M, Wirta EV, Vartiainen E et al: Genetic and Epigenetic Characteristics of Inflammatory Bowel Disease-Associated Colorectal Cancer. Gastroenterology 2021, 161(2):592-607.
Liberzon A, Birger C, Thorvaldsdottir H, Ghandi M, Mesirov JP, Tamayo P: The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell systems 2015, 1(6):417-425.
Yu G, Wang LG, Han Y, He QY: clusterProfiler: an R package for comparing biological themes among gene clusters. Omics : a journal of integrative biology 2012, 16(5):284-287.
Kobayashi T, Siegmund B, Le Berre C, Wei SC, Ferrante M, Shen B, Bernstein CN, Danese S, Peyrin-Biroulet L, Hibi T: Ulcerative colitis. Nature reviews Disease primers 2020, 6(1):74.
Beauchemin N, Arabzadeh A: Carcinoembryonic antigen-related cell adhesion molecules (CEACAMs) in cancer progression and metastasis. Cancer metastasis reviews 2013, 32(3-4):643-671.
Kong DH, Kim YK, Kim MR, Jang JH, Lee S: Emerging Roles of Vascular Cell Adhesion Molecule-1 (VCAM-1) in Immunological Disorders and Cancer. International journal of molecular sciences 2018, 19(4).
Cancer Genome Atlas N: Comprehensive molecular characterization of human colon and rectal cancer. Nature 2012, 487(7407):330-337.
Jorissen RN, Gibbs P, Christie M, Prakash S, Lipton L, Desai J, Kerr D, Aaltonen LA, Arango D, Kruhoffer M et al: Metastasis-Associated Gene Expression Changes Predict Poor Outcomes in Patients with Dukes Stage B and C Colorectal Cancer. Clinical cancer research : an official journal of the American Association for Cancer Research 2009, 15(24):7642-7651.
Smith JJ, Deane NG, Wu F, Merchant NB, Zhang B, Jiang A, Lu P, Johnson JC, Schmidt C, Bailey CE et al: Experimentally derived metastasis gene expression profile predicts recurrence and death in patients with colon cancer. Gastroenterology 2010, 138(3):958-968.
Skrzypczak M, Goryca K, Rubel T, Paziewska A, Mikula M, Jarosz D, Pachlewski J, Oledzki J, Ostrowski J: Modeling oncogenic signaling in colon tumors by multidirectional analyses of microarray data directed for maximization of analytical reliability. PloS one 2010, 5(10).
de Sousa EMF, Colak S, Buikhuisen J, Koster J, Cameron K, de Jong JH, Tuynman JB, Prasetyanti PR, Fessler E, van den Bergh SP et al: Methylation of cancer-stem-cell-associated Wnt target genes predicts poor prognosis in colorectal cancer patients. Cell stem cell 2011, 9(5):476-485.
Laibe S, Lagarde A, Ferrari A, Monges G, Birnbaum D, Olschwang S, Project COL: A seven-gene signature aggregates a subgroup of stage II colon cancers with stage III. Omics : a journal of integrative biology 2012, 16(10):560-565.
Marisa L, de Reynies A, Duval A, Selves J, Gaub MP, Vescovo L, Etienne-Grimaldi MC, Schiappa R, Guenot D, Ayadi M et al: Gene expression classification of colon cancer into molecular subtypes: characterization, validation, and prognostic value. PLoS medicine 2013, 10(5):e1001453.
Hanzelmann S, Castelo R, Guinney J: GSVA: gene set variation analysis for microarray and RNA-seq data. BMC bioinformatics 2013, 14:7.
Shah SC, Itzkowitz SH: Colorectal Cancer in Inflammatory Bowel Disease: Mechanisms and Management. Gastroenterology 2022, 162(3):715-730 e713.
Lee HO, Hong Y, Etlioglu HE, Cho YB, Pomella V, Van den Bosch B, Vanhecke J, Verbandt S, Hong H, Min JW et al: Lineage-dependent gene expression programs influence the immune landscape of colorectal cancer. Nature genetics 2020, 52(6):594-603.
Zhang L, Li Z, Skrzypczynska KM, Fang Q, Zhang W, O'Brien SA, He Y, Wang L, Zhang Q, Kim A et al: Single-Cell Analyses Inform Mechanisms of Myeloid-Targeted Therapies in Colon Cancer. Cell 2020, 181(2):442-459 e429.
Liu Y, Zhang Q, Xing B, Luo N, Gao R, Yu K, Hu X, Bu Z, Peng J, Ren X et al: Immune phenotypic linkage between colorectal cancer and liver metastasis. Cancer cell 2022, 40(4):424-437 e425.
Che LH, Liu JW, Huo JP, Luo R, Xu RM, He C, Li YQ, Zhou AJ, Huang P, Chen YY et al: A single-cell atlas of liver metastases of colorectal cancer reveals reprogramming of the tumor microenvironment in response to preoperative chemotherapy. Cell discovery 2021, 7(1):80.
Pelka K, Hofree M, Chen JH, Sarkizova S, Pirl JD, Jorgji V, Bejnood A, Dionne D, Ge WH, Xu KH et al: Spatially organized multicellular immune hubs in human colorectal cancer. Cell 2021, 184(18):4734-4752 e4720.
Guo W, Zhang C, Wang X, Dou D, Chen D, Li J: Resolving the difference between left-sided and right-sided colorectal cancer by single-cell sequencing. JCI insight 2022, 7(1).
Khaliq AM, Erdogan C, Kurt Z, Turgut SS, Grunvald MW, Rand T, Khare S, Borgia JA, Hayden DM, Pappas SG et al: Refining colorectal cancer classification and clinical stratification through a single-cell atlas. Genome biology 2022, 23(1):113.
Mitsialis V, Wall S, Liu P, Ordovas-Montanes J, Parmet T, Vukovic M, Spencer D, Field M, McCourt C, Toothaker J et al: Single-Cell Analyses of Colon and Blood Reveal Distinct Immune Cell Signatures of Ulcerative Colitis and Crohn's Disease. Gastroenterology 2020, 159(2):591-608 e510.
Boland BS, He Z, Tsai MS, Olvera JG, Omilusik KD, Duong HG, Kim ES, Limary AE, Jin W, Milner JJ et al: Heterogeneity and clonal relationships of adaptive immune cells in ulcerative colitis revealed by single-cell analyses. Science immunology 2020, 5(50).
Gettler K, Levantovsky R, Moscati A, Giri M, Wu Y, Hsu NY, Chuang LS, Sazonovs A, Venkateswaran S, Korie U et al: Common and Rare Variant Prediction and Penetrance of IBD in a Large, Multi-ethnic, Health System-based Biobank Cohort. Gastroenterology 2021, 160(5):1546-1557.
Devlin JC, Axelrad J, Hine AM, Chang S, Sarkar S, Lin JD, Ruggles KV, Hudesman D, Cadwell K, Loke P: Single-Cell Transcriptional Survey of Ileal-Anal Pouch Immune Cells From Ulcerative Colitis Patients. Gastroenterology 2021, 160(5):1679-1693.
Smillie CS, Biton M, Ordovas-Montanes J, Sullivan KM, Burgin G, Graham DB, Herbst RH, Rogel N, Slyper M, Waldman J et al: Intra- and Inter-cellular Rewiring of the Human Colon during Ulcerative Colitis. Cell 2019, 178(3):714-730 e722.
Elmentaite R, Kumasaka N, Roberts K, Fleming A, Dann E, King HW, Kleshchevnikov V, Dabrowska M, Pritchard S, Bolt L et al: Cells of the human intestinal tract mapped across space and time. Nature 2021, 597(7875):250-255.
Sun Y, Wu L, Zhong Y, Zhou K, Hou Y, Wang Z, Zhang Z, Xie J, Wang C, Chen D et al: Single-cell landscape of the ecosystem in early-relapse hepatocellular carcinoma. Cell 2021, 184(2):404-421 e416.
Zhang Q, He Y, Luo N, Patel SJ, Han Y, Gao R, Modak M, Carotta S, Haslinger C, Kind D et al: Landscape and Dynamics of Single Immune Cells in Hepatocellular Carcinoma. Cell 2019, 179(4):829-845 e820.
Brandum EP, Jorgensen AS, Rosenkilde MM, Hjorto GM: Dendritic Cells and CCR7 Expression: An Important Factor for Autoimmune Diseases, Chronic Inflammation, and Cancer. International journal of molecular sciences 2021, 22(15).
Forster R, Schubel A, Breitfeld D, Kremmer E, Renner-Muller I, Wolf E, Lipp M: CCR7 coordinates the primary immune response by establishing functional microenvironments in secondary lymphoid organs. Cell 1999, 99(1):23-33.
Ohl L, Mohaupt M, Czeloth N, Hintzen G, Kiafard Z, Zwirner J, Blankenstein T, Henning G, Forster R: CCR7 governs skin dendritic cell migration under inflammatory and steady-state conditions. Immunity 2004, 21(2):279-288.
Martin JC, Chang C, Boschetti G, Ungaro R, Giri M, Grout JA, Gettler K, Chuang LS, Nayar S, Greenstein AJ et al: Single-Cell Analysis of Crohn's Disease Lesions Identifies a Pathogenic Cellular Module Associated with Resistance to Anti-TNF Therapy. Cell 2019, 178(6):1493-1508 e1420.
Yu H, Lin L, Zhang Z, Zhang H, Hu H: Targeting NF-kappaB pathway for the therapy of diseases: mechanism and clinical study. Signal transduction and targeted therapy 2020, 5(1):209.
Atreya I, Atreya R, Neurath MF: NF-kappaB in inflammatory bowel disease. Journal of internal medicine 2008, 263(6):591-596.
Schottelius AJ, Baldwin AS, Jr.: A role for transcription factor NF-kappa B in intestinal inflammation. International journal of colorectal disease 1999, 14(1):18-28.
Frigerio S, Lartey DA, D'Haens GR, Grootjans J: The Role of the Immune System in IBD-Associated Colorectal Cancer: From Pro to Anti-Tumorigenic Mechanisms. International journal of molecular sciences 2021, 22(23).
Semenza GL: Targeting HIF-1 for cancer therapy. Nature reviews Cancer 2003, 3(10):721-732.
Han J, Jackson D, Holm J, Turner K, Ashcraft P, Wang X, Cook B, Arning E, Genta RM, Venuprasad K et al: Elevated d-2-hydroxyglutarate during colitis drives progression to colorectal cancer. Proceedings of the National Academy of Sciences of the United States of America 2018, 115(5):1057-1062.
Harris AL: Hypoxia--a key regulatory factor in tumour growth. Nature reviews Cancer 2002, 2(1):38-47.
Majumder S, Shivaji UN, Kasturi R, Sigamani A, Ghosh S, Iacucci M: Inflammatory bowel disease-related colorectal cancer: Past, present and future perspectives. World journal of gastrointestinal oncology 2022, 14(3):547-567.
Haberman Y, Karns R, Dexheimer PJ, Schirmer M, Somekh J, Jurickova I, Braun T, Novak E, Bauman L, Collins MH et al: Ulcerative colitis mucosal transcriptomes reveal mitochondriopathy and personalized mechanisms underlying disease severity and treatment response. Nature communications 2019, 10(1):38.
Fenton CG, Taman H, Florholmen J, Sorbye SW, Paulssen RH: Transcriptional Signatures That Define Ulcerative Colitis in Remission. Inflammatory bowel diseases 2021, 27(1):94-105.
Hyun YS, Han DS, Lee AR, Eun CS, Youn J, Kim HY: Role of IL-17A in the development of colitis-associated cancer. Carcinogenesis 2012, 33(4):931-936.
Tanaka A, Sakaguchi S: Regulatory T cells in cancer immunotherapy. Cell research 2017, 27(1):109-118.
Wing JB, Tanaka A, Sakaguchi S: Human FOXP3(+) Regulatory T Cell Heterogeneity and Function in Autoimmunity and Cancer. Immunity 2019, 50(2):302-316.
Belkaid Y, Oldenhove G: Tuning microenvironments: induction of regulatory T cells by dendritic cells. Immunity 2008, 29(3):362-371.
Joller N, Lozano E, Burkett PR, Patel B, Xiao S, Zhu C, Xia J, Tan TG, Sefik E, Yajnik V et al: Treg cells expressing the coinhibitory molecule TIGIT selectively inhibit proinflammatory Th1 and Th17 cell responses. Immunity 2014, 40(4):569-581.
Dai E, Zhu Z, Wahed S, Qu Z, Storkus WJ, Guo ZS: Epigenetic modulation of antitumor immunity for improved cancer immunotherapy. Molecular cancer 2021, 20(1):171.
Togashi Y, Shitara K, Nishikawa H: Regulatory T cells in cancer immunosuppression - implications for anticancer therapy. Nature reviews Clinical oncology 2019, 16(6):356-371.
Liang J, Wang C, Zhang D, Xie Y, Zeng Y, Li T, Zuo Z, Ren J, Zhao Q: VSOLassoBag: a variable-selection oriented LASSO bagging algorithm for biomarker discovery in omic-based translational research. Journal of genetics and genomics = Yi chuan xue bao 2023, 50(3):151-162.
Cook-Mills JM: VCAM-1 signals during lymphocyte migration: role of reactive oxygen species. Molecular immunology 2002, 39(9):499-508.
Gulubova MV, Manolova IM, Vlaykova TI, Prodanova M, Jovchev JP: Adhesion molecules in chronic ulcerative colitis. International journal of colorectal disease 2007, 22(6):581-589.
Sugihara T, Kobori A, Imaeda H, Tsujikawa T, Amagase K, Takeuchi K, Fujiyama Y, Andoh A: The increased mucosal mRNA expressions of complement C3 and interleukin-17 in inflammatory bowel disease. Clinical and experimental immunology 2010, 160(3):386-393.
Markov AV, Savin IA, Zenkova MA, Sen'kova AV: Identification of Novel Core Genes Involved in Malignant Transformation of Inflamed Colon Tissue Using a Computational Biology Approach and Verification in Murine Models. International journal of molecular sciences 2023, 24(5).
Ning C, Li YY, Wang Y, Han GC, Wang RX, Xiao H, Li XY, Hou CM, Ma YF, Sheng DS et al: Complement activation promotes colitis-associated carcinogenesis through activating intestinal IL-1beta/IL-17A axis. Mucosal immunology 2015, 8(6):1275-1284.
Wang H, Tian T, Zhang J: Tumor-Associated Macrophages (TAMs) in Colorectal Cancer (CRC): From Mechanism to Therapy and Prognosis. International journal of molecular sciences 2021, 22(16).
Chen Y, Zheng X, Wu C: The Role of the Tumor Microenvironment and Treatment Strategies in Colorectal Cancer. Frontiers in immunology 2021, 12:792691.
Pastille E, Bardini K, Fleissner D, Adamczyk A, Frede A, Wadwa M, von Smolinski D, Kasper S, Sparwasser T, Gruber AD et al: Transient ablation of regulatory T cells improves antitumor immunity in colitis-associated colon cancer. Cancer research 2014, 74(16):4258-4269.
Olguin JE, Medina-Andrade I, Molina E, Vazquez A, Pacheco-Fernandez T, Saavedra R, Perez-Plasencia C, Chirino YI, Vaca-Paniagua F, Arias-Romero LE et al: Early and Partial Reduction in CD4(+)Foxp3(+) Regulatory T Cells during Colitis-Associated Colon Cancer Induces CD4(+) and CD8(+) T Cell Activation Inhibiting Tumorigenesis. Journal of Cancer 2018, 9(2):239-249.
Scharl T, Gru B, Leisch F: Mixtures of regression models for time course gene expression data: evaluation of initialization and random effects. Bioinformatics 2010, 26(3):370-377.
Butler A, Hoffman P, Smibert P, Papalexi E, Satija R: Integrating single-cell transcriptomic data across different conditions, technologies, and species. Nature biotechnology 2018, 36(5):411-420.
McGinnis CS, Murrow LM, Gartner ZJ: DoubletFinder: Doublet Detection in Single-Cell RNA Sequencing Data Using Artificial Nearest Neighbors. Cell systems 2019, 8(4):329-337 e324.
Korsunsky I, Millard N, Fan J, Slowikowski K, Zhang F, Wei K, Baglaenko Y, Brenner M, Loh PR, Raychaudhuri S: Fast, sensitive and accurate integration of single-cell data with Harmony. Nature methods 2019, 16(12):1289-1296.
Cao J, Spielmann M, Qiu X, Huang X, Ibrahim DM, Hill AJ, Zhang F, Mundlos S, Christiansen L, Steemers FJ et al: The single-cell transcriptional landscape of mammalian organogenesis. Nature 2019, 566(7745):496-502.
Efremova M, Vento-Tormo M, Teichmann SA, Vento-Tormo R: CellPhoneDB: inferring cell-cell communication from combined expression of multi-subunit ligand-receptor complexes. Nature protocols 2020, 15(4):1484-1506.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Evaluation of colorectal cancer in ulcerative colitis reveals key immune factors during its malignant transformation

Status:

Version 1

Abstract

BACKGROUND & AIMS:

METHODS

RESULTS

CONCLUSIONS

Figures

INTRODUCTION

RESULTS

Allocating CAC-like samples from single-cell RNA-sequencing datasets of colorectal cancer

Single-cell immune atlas of healthy, UC and CAC-like samples

VEGFA_Macro cells contributes the promotion process from UC carcinoma to CAC

T_NK cells remodeling from healthy, UC to CAC, especially Treg cells

Enrichment of Treg cells is associated with the risk of carcinogenesis of UC

DISCUSSION

METHODS

Sample and dataset collections

Bioinformatics analysis

Statistical Analysis.

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1