{"paper_id":"095b7e65-0c10-4345-a69f-dc2e8fb8ef64","body_text":"Identification and Multi-omic Analysis of Essential Coding and Long Non-conding Genes in Colorectal Cancer | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Identification and Multi-omic Analysis of Essential Coding and Long Non-conding Genes in Colorectal Cancer Yanguo Li, Tianci Han, Chengjiang Fan, Hao Rong, Huifang Wang, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4034323/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Essential genes are indispensable for the survival of cancer cell. CRISPR/Cas9-based pooled genetic screens have distinguished the pan-cancer essential genes and their functions in distinct cellular processes. Nevertheless, the landscape of essential genes at the single cell levels and the effect on the tumor microenvironment (TME) via cell-to-cell communication network (CCN) remains limited. Here, we identified 396 essential protein-coding genes (ESPs) by integration of 8 genome-wide CRISPR loss-of-function screen datasets of colorectal cancer (CRC) cell lines and a full-length single cells transcriptome data of CRC tissues. Then, 29 essential long non-coding genes (ESLs) were predicted using Hypergeometric Test (HT) and Personalized PageRank (PPR) algorithms based on ESPs and co-expressed network constructed from single cell transcriptome data. CRISPR/Cas9 knockout experiment verified the effect of several ESPs and ESLs on the survival of CRC cell line. Furthermore, multiple omics features of ESPs and ESLs were illustrated by examining their expression patterns and transcription factor (TF) regulatory network at the single cell level, as well as DNA mutation and DNA methylation events at bulk level. Finally, through integrating multiple intracellular regulatory networks with CCN, we elucidated that CD47 and MIF are regulated by multiple CRC essential genes and the anti-cancer drugs sunitinib can interfere the expression of them potentially. Our findings provide a comprehensive asset of CRC ESPs and ESLs, sheding light on the mining of potential therapy targets for CRC. Essential protein-coding genes essential long non-coding genes single-cell RNA sequencing CRISPR/Cas9 tumor microenvironment multi-omics Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Colorectal cancer is one of the most frequently diagnosed cancers worldwide. There were estimated 1.88 million new cases and 0.92 million deaths of CRC in 2020 1 . Although novel therapies and drugs have approximately doubled the average survival time over the past decade, patients with advanced CRC often succumb to the disease within three years 2 . Gene is defined as essential if its complete loss of function will lead to impaired cell development, gene activation, and cellular reprogramming 3-5 . The pooled CRISPR loss-of-function screen is a widely used method for identifying essential protein-coding genes involved in cancer-related biological processes including drug resistance 6 , immune suppression 7 , and immunotherapy response 8 . A number of studies have already obtained multiple sets of pan-cancer core ESPs and non-ESPs using human cancer cell lines, including Hart’s ESPs and Hart’s non-ESPs 9,10 , Achilles’ ESPs 11 , and Zhang’s ESPs 12 . Although these studies have highlighted the functions of ESPs and their regulatory networks in cancer tissues, how their impacts on the TME are still unknown and required further exploration and establishment. Besides, the genotypes and phenotypes of various human cancer cell lines and tissues were heterogeneous. It was reported that ESPs in breast, pancreatic, and ovarian cancer cell lines were partially overlapped, indicating certain ESPs may function in a tumor-specific fashion 13 . Nowadays, few studies were focused on specific ESPs in CRC. During the past several years, single-cell RNA sequencing (scRNA-seq) has become an important technique for resolving critical cancer genes and illustrating the composition of TME. Numerous studies have been conducted for a comprehensive investigation of the cellular and intercellular interactions in CRC, demonstrating the dynamic changes in tumor immunity and providing guidance for clinical treatment strategies 14,15 . Recently, the technology that combined scRNA-seq with CRISPR - activation 5 , CRISPR - knockout screen 16 , and CRISPR - interference 17 have been emerged to identify genetic regulators and lineage traces. For example, Vishnubalaji et.al integrated the scRNA-seq with CRISPR functional screen data from Achilles project 11 and identified potential therapeutic targets for different subtypes of Breast cancer 18 . These studies suggest that combined analysis of scRNA-seq and CRISPR functional screen data can more accurately identify essential genes in CRC and will broaden our understanding of their role in the high-dimensional TME. In this study, we first integrated genome-wide CRISPR/Cas9 loss-of-function screen data and scRNA-seq data to identify ESPs in CRC. Then CRC ESLs were predicted using HT and PPR algorithms based on gene co-expressed network constructed from scRNA-seq data of tumor epithelial cells. To further illustrate the characteristics of ESPs and ESLs, we examined their single cell expression patterns, pathway enrichment, DNA mutation, DNA methylation events, and transcription factor (TF) regulatory network. Finally, we integrated co-expressed network, TF regulatory network, and CCN to provide valuable insights into the impact of ESPs and ESLs on the remodeling of TME. The potential drugs that can interfere key factors of target signal process were also predicted. Collectively, our study presents novel insights into the essential roles of CRC ESPs and ELSs in tumor cell development, TF regulation, and TME remodeling, thus establishing a foundation for identification of novel therapeutic targets. Results Identification of CRC ESPs based on integration of CRISPR screen and scRNA-seq datasets Eight genome-wide CRISPR/Cas9 loss-of-function screen datasets of CRC were collected, including a total of 49 CRC cell line types and 301 samples with replicates ( Fig. 1a, Supplementary Table 1 ). To evaluate the essential degree of each gene, the sgRNAs depletion score for each gene in each sample was calculated by CERES algorithm and defined as CERES score 11 . A lower CERES score indicates a higher level of gene essentiality as all the screen datasets are negative selection in this study. Considering multiple datasets were involved, we integrated CERES scores for each gene across various datasets by utilizing Robust Rank Aggregation (RRA) algorithm 19 . The optimal cutoff to obtain ESPs was determined by calculating the degree of overlap between RRA results and other pan-cancer ESP sets previously identified, including Achilles’ ESPs, Hart’s ESPs, and Zhang’s ESPs. As a result, 1521 candidates of CRC ESPs were obtained ( Fig. 1b) . We further collected full-length scRNA-seq data of CRC (GSE81861) to analyze the detection fraction of candidate ESPs in tumor epithelial cells 20 . We found 396 ESPs were expressed in more than 50% of tumor epithelial cells and were thus considered as final CRC ESPs ( Supplementary Table 2 ). Among them, 74 ESPs were found to be overexpressed in tumor epithelial cells ( Fig. 1c, Supplementary Table 2 ). These ESPs were principally linked to cancer-related functions, such as MYC TARGETS V1/V2, positive regulation of signal transduction by p53 class mediator, and response to interferon-gamma ( Fig. 1d ). Notably, we found ESPs exhibit higher expression than other genes in both tumor cells and normal cells ( Supplementary Fig. a ), which corroborated the findings from a previous study 9 . Then we compared the CERES scores of CRC ESPs with those of pan-cancer ESPs and Hart’s non-ESPs to assess the quality of our results. As anticipated, Hart’s non-ESPs showed the highest CERES scores and indicated the lowest necessity, while the CRC ESPs were lower, implying their essential roles in CRC ( Fig. 1e ). Additionally, we observed larger overlap between CRC ESPs and other pan-cancer ESP sets than that of randomly selected genes, further reflecting the high quality of CRC ESPs ( Fig.1f ). Among 396 ESPs, 155 ESPs co-exist in all ESP sets, indicating the homogeneity of different tumors ( Fig. 1g ). Notably, we found three novel CRC ESPs, CTNNB1 , NDUFB9 , and SCNM1 that were not found in any other ESP sets ( Fig. 1g ). To determine whether these novel CRC ESPs are functionally important in the survival of CRC cancer cell lines, we detected the effect of CTNNB1 and NDUFB9 knockout on HCT116 cell line by CRISPR/Cas9. Remarkably, HCT116 cell line lost viability while CTNNB1 and NDUFB9 were knockout, respectively ( Fig. 1h ). CTNNB1 was reported as a driver gene in CRC according to the Pan-Cancer Analysis of Whole Genomes (PCAWG) study 21 . NDUFB9 , an important gene involved in mitochondrial function, has been identified as a prognostic biomarker for endometrial cancer 22 . However, the role of SCNM1 in CRC is needed to further investigate. Prediction of CRC ESLs LncRNAs play an important regulatory role in various cellular processes of CRC, including cell apoptosis, proliferation, and epithelial-mesenchymal transition 23 . To elucidate a comprehensive landscape of essential genes in CRC, we employed HT and PPR methods, which we proposed in our previous study, to predict essential lncRNAs 12 . These methods required the construction of a two-color network comprising of both protein-coding genes and lncRNAs, where lncRNAs essentiality were predicted based on their closeness to ESPs 12 . In order to look for the best two-color network, we built four different co-expression networks, consisting of an unweighted co-expression network and a topological overlap weighted co-expression network based on the RNA sequencing (RNA-seq) profiles of colorectal adenocarcinoma from The Cancer Genome Atlas (TCGA, referred to as unWCN-TCGA and WCN-TCGA, respectively), as well as an unweighted co-expression network and a topological overlap weighted co-expression network based on expression profiles of tumor epithelial cells from scRNA-seq dataset (referred to as unWCN-TEC and WCN-TEC, respectively) ( Fig. 2a, see method ). Then we calculated the adjusted p -value of each gene by HT to evaluate whether the directly related genes of a query gene are enriched in ESPs based on unWCN-TCGA and unWCN-TEC ( Fig. 2a) . We observed that the adjusted p -values of CRC ESPs, Achilles’ ESPs, Zhang’s ESPs, and Hart’s ESPs showed significantly smaller than those of Hart’s non-ESPs ( Fig. 2b ). Importantly, the adjusted p -values of ESPs based on unWCN-TEC were significantly lower than those based on unWCN-TCGA ( Fig. 2b ), suggesting the higher quality of co-expression network constructed from scRNA-seq datasets than that from TCGA RNA-seq datasets. Consequently, we identified 26 lncRNAs as potential candidates of CRC ESLs by using the cutoff of adjusted p -value lower than 0.01 based on unWCN-TEC. Next, we utilized CRC ESPs as original seeds to calculate the score of all genes by PPR method based on WCN-TCGA and WCN-TEC, respectively ( Fig. 2a ). The higher score represents the higher probability of a gene being essential. As expected, the PPR scores of CRC ESPs, Achilles’ ESPs, Zhang’s ESPs, and Hart’s ESPs were higher, and WCN-TEC also showed greater advantage than WCN-TCGA as higher PPR scores of ESPs in WCN-TEC were observed ( Fig. 2c ). We then selected the top 5% lncRNAs in PPR results based on WCN-TEC as CRC ESL candidates ( n =39). Results yielded 41 ESLs after combing the results of the HT and PPR methods, and 24 lncRNAs were found to be same ( Fig. 2a ). Finally, we retained the ESLs detected in more than 30% of CRC tumor epithelial cells (GSE81861) and defined them as CRC ESLs ( n =29, Supplementary Table 3 ). Expression analysis revealed that 11 CRC ESLs were significantly overexpressed in CRC tumor epithelial cells compared to normal ( Fig.2d, Supplementary Table 3 ). We selected 6 overexpressed ESLs, including H19 , MIF-AS1 , NORAD , SNHG17 , SNHG6 , and SNHG8 , for CRISPR/Cas9 knockout experiments. The results demonstrated that knockout of these ESLs in HCT116 cell line will result in the decreasing of survival cells ( Fig. 2e ), implying the dependent role of those ESLs in CRC. In our results, some ESLs were known as CRC associated lncRNAs according to previous reports, such as PVT1 , H19 , and GAS5 ( Fig. 2f, Supplementary Table 3 ). By analyzing the functions and clinical association of CRC ESLs in Lnc2Cancer 3.0 database 24 , we found that most CRC ESLs are involved in cell apoptosis, growth, survival, and metastasis ( Fig. 2f ). Among these ESLs, 11 lncRNAs were previously identified as part of Zhang’s pan-cancer ESLs ( n =97, Fig. 2f ).. To identify potential Gene Ontology (GO) biological process functions of CRC ESLs, we conducted enrichment analysis on the top 200 associated genes of each ESL based on WCN-TEC. The results revealed that a significant number of CRC ESLs are involved in ribosome/ncRNA/rRNA/DNA processing, regulation of translation, telomere maintenance, and signal transduction by p53 class mediator ( Supplementary Table 4 ), providing additional evidence for the crucial roles of CRC ESLs in fundamental metabolic processes of tumor cells. Expression, DNA mutation, and DNA methylation patterns of CRC essential genes To examine the expression characteristics of essential genes in different cell types of CRC, we utilized CRC ESPs and ESLs as gene sets to calculate enrichment scores in each cell type based on the 10X genomics scRNA-seq dataset of CRC (GSE132465) 15 . As anticipated, both CRC ESPs and ESLs gene set enrichment scores were higher in epithelial cells of tumor samples compared with normal samples ( Fig. 3a ). Interestingly, the gene set scores of CRC ESPs were particularly elevated in tumor epithelial cells compared to other pan-cancer ESP sets ( Fig. 3a ), indicating the specificity of CRC ESPs for CRC tumor epithelial cells. We also observed the expressions of ESPs and ESLs in other cell lineages, such as T cells and B cells ( Fig. 3a ), were lower than tumor epithelial cells, implying that target essential genes may have side effect on cell lineages in TME. Notably, 4 ESPs ( KRT8 , MYC , PRELID3B , and SCD ) were specifically expressed in tumor epithelial cells ( Fig. 3b, see methods ). Three of them, KRT8 , MYC , and SCD are well-known biomarkers or therapeutic targets for cancer 25-27 , while few research has performed on PRELID3B in cancer. CRISPR/Cas9 knockout experiments confirmed that function loss of these genes affects the survival of HCT116 cell line ( Fig. 3c ). Mutation of essential genes has been shown to disrupt normal cell development and promote cancer, such as MYC . We conducted mutation analysis using whole genome sequencing (WGS) datasets of CRC in PCAWG and whole exome sequencing (WES) datasets of CRC in TCGA to explore the genomic alteration of CRC ESPs and ESLs. In general, the mutation rates of ESPs and ESLs in WGS dataset ranged from 0% to 25% and 0% to 33%, respectively ( Fig. 3d-e ), while the mutation rates of ESPs were relatively low in WES dataset, ranging from 0% to 8% ( Supplementary Fig. b ). Among the mutated genes in WGS study, DYNC1H1 (25%), CHD4 (23%), CNOT1 (21%), and CTNNB1 (19%) exhibited the highest mutation frequency ( Fig. 3d ). Similarly, these genes also showed relatively high frequencies and ranked within the top 5 mutated genes in the TCGA CRC dataset ( Supplementary Fig. b ). Importantly, most CRC ESPs mutated in a co-occurrence manner in both the WGS and WES datasets ( Supplementary Fig. c-d ), suggesting that the mutation of ESPs typically activates synergistic oncogenic pathways. DNA methylation also plays a crucial role in influencing gene expression. Through analysis of differentially methylated positions (DMPs) in ESPs and ESLs in TCGA-COAD, we identified a total of 71 DMPs with 45 unique genes (adjusted p -value < 0.05, absolute log 2 FC > 0.2, Supplementary Fig. e ). Of these DMPs, 26 showed DNA hypermethylation (log 2 FC > 0.2), while 45 exhibited DNA hypomethylation (log 2 FC < -0.2). Subsequently, we calculated the correlation coefficients between the expression and DNA methylation levels of each gene. The results revealed three ESPs ( MYC , SCD , EIF6 ) and one ESLs ( PVT1 ) had negative correlation coefficients ( Supplementary Table 5 ). We observed hypomethylation of two probes cg08526705 and cg00163372 in the gene body may activate MYC expression in CRC ( Supplementary Fig. f-g ). These two probes were found to be related with chemotherapy drug resistance previously 28 . In PVT1 , three probes cg23898497, cg00780520, and cg10202727 were found to be associated with PVT1 expression ( Fig. 3f ), which were also observed in TCGA-READ ( Supplementary Table 5 ). PVT1 is an adjacent gene of MYC on chromosome 8q24.21 and has been shown to regulate MYC expression and contribute to cancer development 29 . Our findings indicated that the hypomethylation of MYC and PVT1 lead to the upregulation of their expression, and thus promote the development of CRC. TF regulatory network of essential genes in CRC tumor epithelial cells Based on 10X genomics scRNA-seq dataset of CRC epithelial cells (GSE81861), we generated a TF regulatory network of CRC by using SCENIC algorithm 30 . A total of 30,743 TF-target relationships associated with CRC ESPs and ESLs were established, including 177 TFs and 9,837 genes ( Supplementary Table 6 ). These relationships were further categorized into high- or low-confidence annotations 30 ( see methods ). Interestingly, 9 TFs from ESPs related to 80.63% ( n =24,788) of identified TF-target relationships. Among them, 3 TFs had a highest number of high-confidence relationships compared with others essential TFs, including BCLAF1 , MYC , and YY1 (Fig. 4a) . Area under the curve (AUC) scores from SCENIC were used to quantify the activity of TF targets (regulon). Differential AUC scores of regulons revealed the significant activation of 79 regulons in tumor epithelial cells, including BCLAF1 regulon , YY1 regulon , and MYC regulon ( Fig. 4b-c, Supplementary Table 7 ). Furthermore, three novel CRC ESPs ( CTNNB1 , NDUFB9 and SCNM1 ) that mentioned above were also regulated by BCLAF1 , YY1 , and MYC ( Fig. 4d ), suggesting that these TFs may act as oncogenes in CRC. Previously, knockdown of the L isoform of BCLAF1 in mouse tumor model inhibited the tumor growth, confirming the carcinogenic characteristics of BCLAF1 31 . Above all, our findings demonstrated the intricate regulatory networks between TFs and CRC ESPs, and the TFs arise from CRC ESPs may play significant roles in tumorigenesis. In the regulatory network, 424 relationships were linked between TFs ( n =108) and ESLs ( n =28, Fig. 4e , Supplementary Table 6 ). Notably, the lncRNAs such as FGD5 - AS1 , GAS5 , and PVT1 were regulated by more than 20 TFs. Among these, FGD5 - AS1 , has been emerged as a crucial regulator in CRC and other types of cancer via promoting cell proliferation, drug resistance, and epithelial-mesenchymal transition 32,33 . In our study, 14 TFs were found to regulate the expression of FGD5 - AS1 ( Fig.4f ), some of them were known well associated with CRC, such as EGR1 34 and XBP1 35 . Collectively, our findings provided a reliably regulatory network connecting TFs and CRC essential genes based on scRNA-seq data. Crosstalk between CRC essential genes and TME The CCN between cancer cells and cells in TME mediated by ligand-receptor interactions, plays a crucial role in shaping tumor behavior and TME remodeling. In our study, only three CRC ESPs, including COPA , RPS19 , and TFRC , were known to act as ligands or receptors. The majority of CRC ESPs may participate in the CCN by regulating the expression of ligands or receptors in tumor epithelial cells ( Fig. 5a ). To verify such relationships, we first constructed a CCN associated with tumor epithelial cells using CellphoneDB 36 . The CCN consists of 169 connections targeting tumor epithelial cells and 141 connections source from tumor epithelial cells ( Supplementary Table 8 ). The CCN involves 161 ligands/receptors, including JAG1 ligand, NOTCH receptors, and 11 protein complex such as the integrin complex and VEGFR complex ( Supplementary Table 8 ). On the whole, CRC tumor epithelial cells interacted more frequently with endothelial cells, fibroblast cells and myeloid cells ( Fig. 5b ), which are known to play important roles in angiogenesis, tumor invasion, and immune response. We then integrated the WCN-TEC, TF regulatory network, and CCN, resulting in a total of 1734 relationships ( Supplementary Table 9 ), where 356 ESPs (including 9 TFs) and 28 ESLs affect the expression of 85 ligands/receptors of tumor epithelial cells, suggesting that CRC ESPs and TFs have a widespread influence on CCN ( Supplementary Fig. h ). We observed that six ligands/receptors of tumor epithelial cells, including GPI , CDH1 , MIF , RPS19 , CXADR , and CD47 , are associated with more than 100 essential genes ( Supplementary Table 9 ), indicating these ligands/receptors may act as the critical mediators between essential genes and TME. Among them, MIF secreted from tumor epithelial cells could interact with CD74 , TNFRSF10D , and TNFRSF14 ( Fig. 5c ), the relationships may responsible for the suppression of antitumor immune response 37 . Additionally, it has been reported that the CD47-SIRPα/γ interaction protects tumor cells from killing by suppressing both macrophage phagocytosis and antigen presentation of dendritic cells 38 . Consistent with these findings, we observed a significant activation of CD47-SIRPα/γ interactions between tumor epithelial cells and myeloid cells ( Fig. 5c ). Clinically, CRC patients from TCGA with high expression levels of CD47 and MIF exhibited a significantly lower overall survival rate compared to those patients with low expression levels of CD47 and MIF ( Fig. 5d ). Our results supported the notion that essential genes can facilitate tumor cells evasion from immune system by regulating the expression of of receptors/ligands such as CD47 and MIF in CRC. Next, we aimed to identify existing drugs that could potentially repress CRC essential genes and TFs associated with CD47 and MIF ( n =304, Fig. 5a ). We utilized an integrative web platform, iLINCS 39 , to analyze the expression pattern of 304 targets in pre-computed signatures of 10 anti-cancer agents in 5 CRC cell lines (GSE116439) 40 . We found that drugs such as sunitinib, dasatinib, topotecan, and lapatinib were sensitive to cancer cells, as they can effectively inhibit the activity of target genes at appropriate concentration and treatment time ( Fig. 5e, Supplementary Table 10 ). For instance, sunitinib acts as an inhibitor of CSF1R , VEGF receptor, c-kit, PDGF receptor, and RET , and has been widely used in cancer therapy 41 . Treatment with sunitinib resulted in the down-regulation of MYC , KRT8 , CDC37 , and other genes in the CRC cell line KM12 ( Fig. 5e, signature ID: PG_4150). In addition, the patients with high CD47 expression benefit from sunitinib monotherapy in clear cell renal cell carcinoma 42 , further supporting sunitinib may act as CD47 inhibitor in CRC. On the other hand, CRC cell lines treated with cisplatin or vorinostat exhibited over-expression of target genes such as MYC and TAF7 ( Fig. 5e ), indicating that CRC cells may develop resistance to these drugs. In conclusion, our results suggest that CRC essential genes may impact the TME through CCN and provides valuable insights into potential therapeutic strategies for CRC. Discussion Accumulated genetic and epigenetic factors are widely believed to play a significant role in pathogenesis of CRC. However, the precise molecular mechanism remains challenging to uncover. Over the past decade, large-scale CRISPR loss-of-function screen data has been extensively used to identify essential genes for pan-cancer. Notably, the Dependency Map (DepMap) portal and BioGRID ORCS have provided an index of CRISPR screens research focused on identifying and understanding the landscape of pan-cancer ESPs 43,44 . However, due to the specificity of tissue and the heterogeneity of tumors, pan-cancer ESPs may not be suitable for certain types of cancer 13 , especially the application of the targeted therapy. In this study, we employed genome-wide CRISPR screen, scRNA-seq data, co-expression networks, and CRISPR/Cas9 knockout experiments to screen and validate ESPs and ESLs associated with CRC. We found CRC ESPs showed partially overlap with well known pan-cancer ESPs, exhibit high expression levels in tumor epithelial cells of CRC, contribute to fundamental and critical metabolic pathways and play crucial role in the regulation of TF regulatory network and CCN. Collectively, our results provide a comprehensive landscape of essential genes and their characteristics in CRC, enabling the identification of potential therapy strategies. Essential genes were reported to highly express in tumor than normal samples by using RNA-seq 12 . We then streamlined the candidate essential genes from 1521 to 396 considering that the essential genes should be expressed in more than 50% of tumor epithelial cells. Consistent with previous perspective, we have found that 155 CRC ESPs overlap with different sets of pan-cancer ESPs, indicating the homogeneity of different tumors and general cancer related functions. Such similar result across various functional genome screen suggests that common essential genes are generally required for tumor cells survival. Moreover, novel CRC ESP CTNNB1 has been reported to drive CRC evolution 21 because CTNNB1 mutation only occurs in early-stage of CRC.. As expected, knockout of CTNNB1 also significantly inhibited the proliferation of CRC, indicating the role of CTNNB1 in CRC tumor cells survival. Additionally, we revealed that most pan-cancer ESPs and CRC ESPs were not only expressed in tumor epithelial cells, but also immune cells and stromal cells in CRC, the finding indicates the low efficiency, side effect, and even failure of therapy will be occurred if targeting pan-cancer ESPs 45 . Our results yield new insights into the biological characteristic of CRC ESPs. Genome-wide CRISPR/Cas9 screen approaches are not well suitable for identifying lncRNAs 46 . Instead, co-expression networks including both protein coding genes and lncRNAs enable prediction of essential genes based on network features 47 . We utilized full-length scRNA-seq data to construct specific co-expression networks from tumor epithellial cells, which exhibited low noise and low redundancy compared to traditional bulk transcriptome in our study. Taking advantage of this, we can accurately predict CRC ESLs based on network features by performing HT and PPR algorithm. Similar to ESPs, we also observed that the identified ESLs were highly expressed in tumor epithelial cells than normal. Results showed that part of ESLs also existed in Zhang’s pan-cancer ESLs, and many of them are known to be involved in tumor-promoting, tumor-suppressing functions, and fundamental metabolic processes of tumor cells. We further validated the effect of ESLs on CRC cell survival through CRISPR/Cas9 knockout experiments, exhibiting the decisive role of them. Furthermore, the activity of ESLs also affected by DNA mutations and methylations. For instance, GAS5 , acts as a tumor suppressor 48,49 , showed high mutation rate in CRC patients from PCAWG and high expression level in CRC tumor epithelial cells compared to normal epithelial cells. Three CpG sites of PVT1 that identified in our study were significantly hypomethylated in CRC and was negatively correlated with PVT1 expression. As a novel epigenetic enhancer of MYC 29,50 , the epigenetic CpG sites of PVT1 may impact the expression of MYC and MYC-related signaling pathways. Our findings thus provide potential therapeutic targets in CRC. Understanding the intricate regulatory network between tumor cells and TME cells will advance our knowledge of cancer and facilitating the discovery of therapeutic strategies. Essential genes, which participate in critical signaling and metabolic pathways, may also receive extracellular signals and transmit signals to the TME through ligands/receptors within CCN. To identify potential mediators linking essential genes and TME, we integrated various intracellular regulatory networks with CCN. In the integrated network, signal flow comprises two axes: signal outgoing and signal incoming. In the signal outgoing axis, essential gene is influenced by transcription factors (TFs), DNA mutations, and DNA methylation. Then, these essential genes can impact the ligand/receptor expression of the tumor cells and then transmit signals to the TME. The signal incoming axis involves TME signals being transmitted to receptors on the surface of tumor cells. These signals then influence the expressions of essential genes via biological network. By employing integrative network analysis, we identified six ligands/receptors potentially regulated by over 100 essential genes and TFs, including MIF and CD47 . Elevated expression levels of MIF and CD47 in CRC have been linked to poor prognosis, possibly due to their involvement in evading immune responses of cancer cells 51,52 . Additionally, we extracted essential genes and TFs associated with MIF and CD47 from the integrated biological network, and discovered several anti-cancer drugs that can modulate the expressions of MIF and CD47 by targeting its associated genes. For instance, sunitinib, a known CD47 inhibitor, has demonstrated the ability to regulate CD47 expression in clear cell renal cell carcinoma 42 . This approach represents a potentially efficient method to interfere essential expression and the ligands/receptors of CCN for the purpose of cancer therapy. Although comprehensive analysis was performed in this study, there are still some drawbacks. Firstly, owning to the low sequencing depth in 10X genomic scRNA-seq, full-length scRNA-seq data of CRC must be used for constructing a co-expression network. However, there were only 270 tumor cells, further validation in larger-scale data is necessary in future. In addition, although we have elucidated the patterns of expression, DNA mutation, DNA methylation, TF regulation, and TME remodeling for essential genes, the affirmation of how essential genes perform irreplaceable functions in protein levels are urgently required. Finally, further experiments are needed to verify the key role of CRC essential genes in tumorigenesis. Materials and Methods Genome-scale CRISPR data collection and processing Eight genome-scale CRISPR loss-of-function screen datasets of CRC were downloaded from articles and the DepMap database (https://depmap.org/portal/, v20Q1). Then CERES algorithm, a computational method that estimates gene-dependency levels from CRISPR-Cas9 essentiality screens with consideration of copy number-specific effect, was applied. After that, we obtained the sgRNA depletion score for each gene in every cell line. To integrate the gene list by CERES scores across all cell lines, we utilized the Robust Rank Aggregation algorithm (RRA) in R (“RobustRankAggreg” package), and adjusted p -value using “Bonferroni” method. Once we obtained the integrated ranking list, we explored various proportions ranging from 1% to 23% (adjusted p -value < 0.05) to determine the optimized threshold to identify candidate ESPs. The optimized threshold was determined when the maximum value of the overlapping ratios between genes obtained from different ranking proportions and the union pan-cancer essential gene set of Hart's ESGs (1246 genes), Achilles' ESGs (2149 genes), and Zhang's ESGs (799 genes). Single-cell RNA sequencing data collection and processing We obtained full-length scRNA-seq data (GSE81861) from Gene Expression Omnibus database (GEO), consisting of 272 CRC tumor epithelial cells and 160 normal epithelial cells. The gene expression matrices of tumor and normal epithelial cells were combined and converted into a Seurat object using the Seurat R package for downstream analysis 53 . We also obtained 10X Genomics scRNA-seq data (GSE132465) from GEO, consisting of 56,465 cells from 23 patients with primary CRC samples and 10 matched normal mucosae samples. We performed quality filtering to remove cells with less than 500 expressed genes, less than 500 unique molecular identifiers (UMIs), more than 20% UMIs derived from mitochondrial genes, and log 10 (expressed genes / UMIs) greater than 0.78. Then, we normalized and scaled the gene expression matrices using SCTransform. Subsequently, harmony algorithm was employed to integrate scRNA-seq data across different patients 54 . For cell type identification, we first applied a graph-based clustering approach by using FindNeighbors and FindClusters function. Then the resulting clusters were further annotated into seven major cell types: epithelial cells ( EPCAM ), myeloid cells ( CD14 ), T cells ( CD3D ), B cells ( CD79A ), fibroblast cells ( COL1A1 ), endothelial cells ( PECAM1 ), and mast cells ( TPSAB1 ). The differentially expressed essential genes in different group or cell lineages were calculated using the Seurat FindMarkers function with “MAST” algorithm. The genes with default parameters of log 2 FC > 0.5 and adjusted p -value < 0.05 were considered as overexpressed genes. Specifically expressed genes in each cell type were defined as those detected in certain cell type at a percentage three times higher than other cell types. To assess the gene set score fore each cell type, AddModuleScore function in Seurat was applied. Cell culture and CRISPR/Cas9 knockout HCT116 (human colon carcinoma cell line, Ubigene, YC-C004) were cultured in RPMI-1640 (Thermo Fisher Scientific) with 10% fetal bovine serum (FBS; VISTECH) and 1% penicillin-streptomycin (Thermo Fisher Scientific). HEK293T cells (ATCC CRL3216) were cultured in DMEM (VISTECH) with 10% FBS and 1% penicillin-streptomycin. All cell lines were maintained at 37°C in a humidified incubator with 5% CO2. To clone individual gene targeting single-guide RNAs (sgRNAs), the lentiviral vector (pSLQ1373) was digested with BlpⅠ and BstXⅠ, and gel-purified 55 . For each gene, two sgRNA were designed on the CHOPCHOP website 56 ( Supplementary Table 11 ). SgRNA fragments were synthesized by Tsingke (Tsingke Biotechnology Co., Ltd.) as forward and reverse primers (Table S1), which were then annealed, gel-purified, and ligated to the linearized pSLQ1373 vector by homologous recombination (ClonExpress Ultra One Step Cloning Kit, Vazyme). To produce lentivirus, HEK293T cells were transiently transfected with polyethylenimine (PEI MAX, Polysciences) and packaging plasmids psPAX2 and pMD2.0G. The ratio of target plasmid to psPAX2 and pMD2.0G is 5:4:1. The ratio of the total mass of the added plasmid (μg) to the PEI MAX (μl) is 1:3. Lentivirus was collected by filtering the supernatant through a 0.45-μm filter 48 hours after transfection. The lentiviruses were collected and stored at -80°C. HCT116 cells were treated with 15 μg/μl BSD for 5 days to select Cas9-expressing cells, after being infected with EFS-spCas9-BSD lentivirus for 48 hours. The purified cells were sorted into 96-well plates using a flow cytometry sorter CytoFLEX SRT (BECKMAN COULTER). After 12-14 days, the well-growing clonal cell population in the 96-well plate was picked out, expanded, and frozen. HCT116-Cas9 was infected with a lentivirus construct expressing individual sgRNA, or two sgRNAs simultaneously for lncRNA knock-out. After 72 hours of infection, the infected cells were treated with 2 μg/ml puromycin for 24 hours and then recovered in fresh culture medium without puromycin for 48 hours. Gene knock-out cells were counted and seeded to a 12-well plate at 0.1 million cells/well. After 72 hours of growth, cells were uniformly passaged at a 1:10 and grown for another 72 hours. Cells were then collected, washed, and resuspended in phosphate-buffered saline (PBS). Ten microliters of the cell suspension were mixed with 0.4% trypan blue solution (Solarbio) at a 1:1 ratio and counted on a hemocytometer (Hirschmann). Each sample was counted three times to obtain a mean cell number. Collection and processing of gene expression, DNA mutation, and DNA methylation data We collected and combined the expression profiles, DNA mutation data, and DNA methylation data of TCGA-COAD and TCGA-READ from the UCSC data portal. The PCAWG mutation data of CRC was collected from International Cancer Genome Consortium (ICGC). The mutation data was processed, analyzed, and visualized using Maftools package in R. The DNA methylation data were processed and visualized by using ChAMP package in R. Differentially methylated positions (DMPs) were identified by champ.DMP function in ChAMP with adjusted p -value < 0.05 and absolute log 2 FC > 0.2. The Pearson correlation coefficients between the expression and DNA methylation levels of each gene were calculated to highlight the association of them. Co-expression network construction Two-color network with both protein-coding genes and lncRNAs were constructed based on bulk RNA-seq from TCGA without normal samples and full-length scRNA-seq data of tumor epithelial cells from GSE81861, respectively. For the construction of an unweighted co-expression network, the gene Pearson correlation coefficients were calculated. Fisher’s asymptotic test from WGCNA packages 57 and FDR correction were used to screen reliable relationships. Eventually, only the pairs with adjusted p -value lower than 0.01 and absolute value of correlation coefficient greater than 0.3 were reserved and further converted to undirected network. We also constructed topological overlap weighted co-expression network by using TOMsimilarityFromExpr function from WGCNA with parameter as follow: soft-thresholding power of 6 , Pearson correlation coefficient, and TOMDenom specifying as mean. Finally, the pairs with weight greater than 0 were exported and further converted to undirected network. Hypergeometrics Test and Personalized PageRank As described previously, we performed HT algorithm ( Fig. 2a ) to evaluate the degree of overlap between neighboring genes of a query gene in an unweighted co-expression network and ESPs 12 . For each lncRNA, p -values were calculated and adjusted using FDR correction, and were used to evaluate the possibility as ESLs. Candidate ESLs were defined as lncRNAs with adjusted p -values lower than 0.01. PPR is a well-known method derived from random walk with restart and can compute ranking score for each gene using multiple seeds (https://github.com/jinhongjung/pyrwr) in network. In this study, ESPs were designated as the seeds to calculate the ranking score of each lncRNA in the topological overlap weighted co-expression network. Only the top 5% of lncRNAs in the PPR results were considered as potential ESL candidates. Inferring TF regulatory network of CRC scRNA-seq data We utilized the R implementation of the SCENIC pipeline to infer TF regulatory network between transcription factors and essential genes. Two gene-motif rankings databases were employed to determine the cis -acting element around the transcription start site (TSS), including 10 kb around the TSS and 500 bp upstream of the TSS. The gene-motif annotations from cisTarget databases can be categorized as either high-confidence or low-confidence. The high-confidence annotations are “direct annotation” and “inferred by orthology” in the annotation source of gene-motif, while low-confidence annotations are “inferred by motif similarity”. The regulon named with the sufix “-extended” indicate lower confidence annotation. To assess the differential AUC score of the regulon between tumor and normal epithelial cells, t -test was conducted. CCN network construction and integration We employed CellPhoneDB to construct the CCN network of CRC based on 10X Genomics scRNA data (GSE132465). For downstream analysis, we only considered the ligand-receptor pairs with p -value lower than 0.05. We then integrating multiple intracellular regulatory networks, including co-expressed network, TF regulatory network, and CCN of tumor epithelial cells. First, we filtered the top 100 co-expressed pairs of each ESP and the pairs with weight great than 0.001 in WCN-TEC. Second, we filtered relationships associated with ESPs and ESLs from TF regulatory network. Finally, we integrated those two intracellular regulatory networks with CCN, the source node of integrated network are ESPs, ESLs, or TFs, while the target node are ligands or receptors of tumor epithelial cells. Function enrichment and survival analysis GO, KEGG, and HALLMARK function analyses of protein-coding genes were carried out by Metascape 58 . To determine the functions of lncRNAs, we first checked the existing reports from Lnc2Cancer 3.0. Second, top 200 associated genes for each ESL based on weighted co-expression network of tumor epithelial cells were selected. Then, GO biological process enrichment analysis was performed using “ClusterProfiler” R package. Additionally, we conducted survival analysis using GEPIA 59 , an interactive web server to analyze RNA sequencing data from TGCA. Declarations CRediT author statement and competing interests Yanguo Li : Methodology, Formal analysis, Software, Data Curation, Visualization, Writing - Original draft preparation. Tianci Han : Investigation, Formal analysis, Validation. Chengjiang Fan : Investigation. Hao Rong : Validation, Software. Chen Yu : Investigation, Resources, Writing - Reviewing and Editing, Funding acquisition. Yang Xi and Huifang Wang : Resources, Writing - Reviewing and Editing, Funding acquisition. Qi Liao : Conceptualization, Project administration, Methodology, Supervision, Writing - Reviewing and Editing, Funding acquisition. The authors declare no competing interests. Acknowledgements This work was supported by the Zhejiang Provincial Natural Science Foundation of China (No. LY21C060002, No. LY21C060001); the National Natural Science Foundation of China (No. 31970630); the Ningbo Natural Science Foundation of China (No. 2021J124), the Fundamental Research Funds for the Provincial Universities of Zhejiang (No. SJLZ2021001); Guangdong Basic and Applied Basic Research Foundation (No. 2023A1515011075); Shenzhen Bay Laboratory (No. 21250071); The Keynote Research Project of Ningbo City (No. 2023Z171); Science and Technology Project of Ningbo(No. 2020F032). References Sung, H. et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 71 , 209-249, doi:10.3322/caac.21660 (2021). McQuade, R. M., Stojanovska, V., Bornstein, J. C. & Nurgali, K. Colorectal Cancer Chemotherapy: The Evolution of Treatment and New Approaches. Curr Med Chem 24 , 1537-1557, doi:10.2174/0929867324666170111152436 (2017). Wang, T. et al. Identification and characterization of essential genes in the human genome. Science (New York, N.Y.) 350 , 1096-1101, doi:10.1126/science.aac7041 (2015). Shalem, O. et al. Genome-scale CRISPR-Cas9 knockout screening in human cells. Science (New York, N.Y.) 343 , 84-87, doi:10.1126/science.1247005 (2014). Alda-Catalinas, C. et al. A Single-Cell Transcriptomics CRISPR-Activation Screen Identifies Epigenetic Regulators of the Zygotic Genome Activation Program. Cell Syst 11 , 25-41 e29, doi:10.1016/j.cels.2020.06.004 (2020). Lin, K. H. et al. Using antagonistic pleiotropy to design a chemotherapy-induced evolutionary trap to target drug resistance in cancer. Nat Genet 52 , 408-417, doi:10.1038/s41588-020-0590-9 (2020). Cortez, J. T. et al. CRISPR screen in regulatory T cells reveals modulators of Foxp3. Nature 582 , 416-420, doi:10.1038/s41586-020-2246-4 (2020). Manguso, R. T. et al. In vivo CRISPR screening identifies Ptpn2 as a cancer immunotherapy target. Nature 547 , 413-418, doi:10.1038/nature23270 (2017). Hart, T., Brown, K. R., Sircoulomb, F., Rottapel, R. & Moffat, J. Measuring error rates in genomic perturbation screens: gold standards for human functional genomics. Mol Syst Biol 10 , 733, doi:10.15252/msb.20145216 (2014). Hart, T. et al. Evaluation and Design of Genome-Wide CRISPR/SpCas9 Knockout Screens. G3 (Bethesda) 7 , 2719-2727, doi:10.1534/g3.117.041277 (2017). Meyers, R. M. et al. Computational correction of copy number effect improves specificity of CRISPR-Cas9 essentiality screens in cancer cells. Nat Genet 49 , 1779-1784, doi:10.1038/ng.3984 (2017). Zhang, Y. et al. Genome-wide identification of the essential protein-coding genes and long non-coding RNAs for human pan-cancer. Bioinformatics 35 , 4344-4349, doi:10.1093/bioinformatics/btz230 (2019). Marcotte, R. et al. Essential gene profiles in breast, pancreatic, and ovarian cancer cells. Cancer Discov 2 , 172-189, doi:10.1158/2159-8290.CD-11-0224 (2012). Zhang, L. et al. Single-Cell Analyses Inform Mechanisms of Myeloid-Targeted Therapies in Colon Cancer. Cell 181 , 442-459 e429, doi:10.1016/j.cell.2020.03.048 (2020). Lee, H. O. et al. Lineage-dependent gene expression programs influence the immune landscape of colorectal cancer. Nat Genet 52 , 594-603, doi:10.1038/s41588-020-0636-z (2020). Replogle, J. M. et al. Combinatorial single-cell CRISPR screens by direct guide RNA capture and targeted sequencing. Nat Biotechnol 38 , 954-961, doi:10.1038/s41587-020-0470-y (2020). Genga, R. M. J. et al. Single-Cell RNA-Sequencing-Based CRISPRi Screening Resolves Molecular Drivers of Early Human Endoderm Development. Cell Rep 27 , 708-718 e710, doi:10.1016/j.celrep.2019.03.076 (2019). Vishnubalaji, R. & Alajez, N. M. Single-Cell Transcriptome Analysis Revealed Heterogeneity and Identified Novel Therapeutic Targets for Breast Cancer Subtypes. Cells 12 , doi:10.3390/cells12081182 (2023). Kolde, R., Laur, S., Adler, P. & Vilo, J. Robust rank aggregation for gene list integration and meta-analysis. Bioinformatics 28 , 573-580, doi:10.1093/bioinformatics/btr709 (2012). Li, H. et al. Reference component analysis of single-cell transcriptomes elucidates cellular heterogeneity in human colorectal tumors. Nat Genet 49 , 708-718, doi:10.1038/ng.3818 (2017). Gerstung, M. et al. The evolutionary history of 2,658 cancers. Nature 578 , 122-128, doi:10.1038/s41586-019-1907-7 (2020). Wang, J., Meng, F. & Mao, F. Single cell sequencing analysis and transcriptome analysis constructed the liquid-liquid phase separation(LLPS)-related prognostic model for endometrial cancer. Front Oncol 12 , 1005472, doi:10.3389/fonc.2022.1005472 (2022). Chen, S. & Shen, X. Long noncoding RNAs: functions and mechanisms in colon cancer. Mol Cancer 19 , 167, doi:10.1186/s12943-020-01287-2 (2020). Gao, Y. et al. Lnc2Cancer 3.0: an updated resource for experimentally supported lncRNA/circRNA cancer associations and web tools based on RNA-seq and scRNA-seq data. Nucleic Acids Res 49 , D1251-D1258, doi:10.1093/nar/gkaa1006 (2021). Scott, M. K. D. et al. A multi-scale integrated analysis identifies KRT8 as a pan-cancer early biomarker. Pac Symp Biocomput 26 , 297-308 (2021). Trop-Steinberg, S. & Azar, Y. Is Myc an Important Biomarker? Myc Expression in Immune Disorders and Cancer. Am J Med Sci 355 , 67-75, doi:10.1016/j.amjms.2017.06.007 (2018). Vriens, K. et al. Evidence for an alternative fatty acid desaturation pathway increasing cancer plasticity. Nature 566 , 403-406, doi:10.1038/s41586-019-0904-1 (2019). Krushkal, J. et al. Epigenome-wide DNA methylation analysis of small cell lung cancer cell lines suggests potential chemotherapy targets. Clin Epigenetics 12 , 93, doi:10.1186/s13148-020-00876-8 (2020). Tseng, Y. Y. et al. PVT1 dependence in cancer with MYC copy-number increase. Nature 512 , 82-86, doi:10.1038/nature13311 (2014). Aibar, S. et al. SCENIC: single-cell regulatory network inference and clustering. Nat Methods 14 , 1083-1086, doi:10.1038/nmeth.4463 (2017). Zhou, X. et al. BCLAF1 and its splicing regulator SRSF10 regulate the tumorigenic potential of colon cancer cells. Nat Commun 5 , 4581, doi:10.1038/ncomms5581 (2014). Zhu, L. et al. Expression and crucial role of long non-coding RNA FGD5-AS1 in human cancers. Am J Transl Res 13 , 10964-10976 (2021). Li, D. et al. Long noncoding RNA FGD5-AS1 promotes colorectal cancer cell proliferation, migration, and invasion through upregulating CDCA7 via sponging miR-302e. In Vitro Cell Dev Biol Anim 55 , 577-585, doi:10.1007/s11626-019-00376-x (2019). Kim, S. H. et al. Kruppel-Like Factor 12 Promotes Colorectal Cancer Growth through Early Growth Response Protein 1. PLoS One 11 , e0159899, doi:10.1371/journal.pone.0159899 (2016). Jiang, X. et al. Thapsigargin promotes colorectal cancer cell migration through upregulation of lncRNA MALAT1. Oncol Rep 43 , 1245-1255, doi:10.3892/or.2020.7502 (2020). Efremova, M., Vento-Tormo, M., Teichmann, S. A. & Vento-Tormo, R. CellPhoneDB: inferring cell-cell communication from combined expression of multi-subunit ligand-receptor complexes. Nat Protoc 15 , 1484-1506, doi:10.1038/s41596-020-0292-x (2020). Figueiredo, C. R. et al. Blockade of MIF-CD74 Signalling on Macrophages and Dendritic Cells Restores the Antitumour Immune Response Against Metastatic Melanoma. Frontiers in immunology 9 , 1132, doi:10.3389/fimmu.2018.01132 (2018). Kaur, S., Isenberg, J. S. & Roberts, D. D. CD47 (Cluster of Differentiation 47). Atlas Genet Cytogenet Oncol Haematol 25 , 83-102 (2021). Pilarczyk, M. et al. Connecting omics signatures and revealing biological mechanisms with iLINCS. Nat Commun 13 , 4678, doi:10.1038/s41467-022-32205-3 (2022). Monks, A. et al. The NCI Transcriptional Pharmacodynamics Workbench: A Tool to Examine Dynamic Expression Profiling of Therapeutic Response in the NCI-60 Cell Line Panel. Cancer research 78 , 6807-6817, doi:10.1158/0008-5472.CAN-18-0989 (2018). Demetri, G. D. et al. Efficacy and safety of sunitinib in patients with advanced gastrointestinal stromal tumour after failure of imatinib: a randomised controlled trial. Lancet 368 , 1329-1338, doi:10.1016/S0140-6736(06)69446-4 (2006). Jiang, W. et al. Immune inactivation by CD47 expression predicts clinical outcomes and therapeutic responses in clear cell renal cell carcinoma patients. Urol Oncol 40 , 166 e115-166 e125, doi:10.1016/j.urolonc.2021.11.024 (2022). Dempster, J. M. et al. Extracting Biological Insights from the Project Achilles Genome-Scale CRISPR Screens in Cancer Cell Lines. 720243, doi:10.1101/720243 %J bioRxiv (2019). Oughtred, R. et al. The BioGRID database: A comprehensive biomedical resource of curated protein, genetic, and chemical interactions. Protein Sci 30 , 187-200, doi:10.1002/pro.3978 (2021). Chang, L., Ruiz, P., Ito, T. & Sellers, W. R. Targeting pan-essential genes in cancer: Challenges and opportunities. Cancer Cell 39 , 466-479, doi:10.1016/j.ccell.2020.12.008 (2021). Liu, S. J. et al. CRISPRi-based genome-scale identification of functional long noncoding RNA loci in human cells. Science (New York, N.Y.) 355 , doi:10.1126/science.aah7111 (2017). Li, X., Li, W., Zeng, M., Zheng, R. & Li, M. Network-based methods for predicting essential genes or proteins: a survey. Brief Bioinform 21 , 566-583, doi:10.1093/bib/bbz017 (2020). Liu, L. et al. lncRNA GAS5 Inhibits Cell Migration and Invasion and Promotes Autophagy by Targeting miR-222-3p via the GAS5/PTEN-Signaling Pathway in CRC. Mol Ther Nucleic Acids 17 , 644-656, doi:10.1016/j.omtn.2019.06.009 (2019). Ni, W. et al. Long noncoding RNA GAS5 inhibits progression of colorectal cancer by interacting with and triggering YAP phosphorylation and degradation and is negatively regulated by the m(6)A reader YTHDF3. Mol Cancer 18 , 143, doi:10.1186/s12943-019-1079-y (2019). Shigeyasu, K. et al. The PVT1 lncRNA is a novel epigenetic enhancer of MYC, and a promising risk-stratification biomarker in colorectal cancer. Mol Cancer 19 , 155, doi:10.1186/s12943-020-01277-4 (2020). Logtenberg, M. E. W., Scheeren, F. A. & Schumacher, T. N. The CD47-SIRPalpha Immune Checkpoint. Immunity 52 , 742-752, doi:10.1016/j.immuni.2020.04.011 (2020). Otvos, B. et al. Cancer Stem Cell-Secreted Macrophage Migration Inhibitory Factor Stimulates Myeloid Derived Suppressor Cell Function and Facilitates Glioblastoma Immune Evasion. Stem Cells 34 , 2026-2039, doi:10.1002/stem.2393 (2016). Stuart, T. et al. Comprehensive Integration of Single-Cell Data. Cell 177 , 1888-1902 e1821, doi:10.1016/j.cell.2019.05.031 (2019). Korsunsky, I. et al. Fast, sensitive and accurate integration of single-cell data with Harmony. Nat Methods 16 , 1289-1296, doi:10.1038/s41592-019-0619-0 (2019). Liu, Y. et al. CRISPR Activation Screens Systematically Identify Factors that Drive Neuronal Fate and Reprogramming. Cell Stem Cell 23 , 758-771 e758, doi:10.1016/j.stem.2018.09.003 (2018). Labun, K. et al. CHOPCHOP v3: expanding the CRISPR web toolbox beyond genome editing. Nucleic Acids Res 47 , W171-W174, doi:10.1093/nar/gkz365 (2019). Langfelder, P. & Horvath, S. WGCNA: an R package for weighted correlation network analysis. BMC bioinformatics 9 , 559, doi:10.1186/1471-2105-9-559 (2008). Zhou, Y. et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat Commun 10 , 1523, doi:10.1038/s41467-019-09234-6 (2019). Tang, Z. et al. GEPIA: a web server for cancer and normal gene expression profiling and interactive analyses. Nucleic Acids Res 45 , W98-W102, doi:10.1093/nar/gkx247 (2017). Additional Declarations No competing interests reported. Supplementary Files SupplementaryFigures.pdf SupplementaryTables.xlsx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {\"props\":{\"pageProps\":{\"initialData\":{\"identity\":\"rs-4034323\",\"acceptedTermsAndConditions\":true,\"allowDirectSubmit\":true,\"archivedVersions\":[],\"articleType\":\"Research Article\",\"associatedPublications\":[],\"authors\":[{\"id\":277513869,\"identity\":\"ebeedef0-508c-4f02-a881-853cf8eee597\",\"order_by\":0,\"name\":\"Yanguo Li\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Ningbo University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Yanguo\",\"middleName\":\"\",\"lastName\":\"Li\",\"suffix\":\"\"},{\"id\":277513870,\"identity\":\"6fc2e123-3b4a-4e14-9ba4-7c8ea23b8bd3\",\"order_by\":1,\"name\":\"Tianci Han\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Shenzhen Bay Laboratory\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Tianci\",\"middleName\":\"\",\"lastName\":\"Han\",\"suffix\":\"\"},{\"id\":277513871,\"identity\":\"5be45c0e-e80f-40f1-ae89-d38c91af138f\",\"order_by\":2,\"name\":\"Chengjiang Fan\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Ningbo University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Chengjiang\",\"middleName\":\"\",\"lastName\":\"Fan\",\"suffix\":\"\"},{\"id\":277513872,\"identity\":\"31d31949-7217-4ec2-9908-b192feda18fa\",\"order_by\":3,\"name\":\"Hao Rong\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Ningbo University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Hao\",\"middleName\":\"\",\"lastName\":\"Rong\",\"suffix\":\"\"},{\"id\":277513873,\"identity\":\"d65524f9-3354-4a6c-8fd6-26a7f3fa74ff\",\"order_by\":4,\"name\":\"Huifang Wang\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"The First Affiliated Hospital of Ningbo University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Huifang\",\"middleName\":\"\",\"lastName\":\"Wang\",\"suffix\":\"\"},{\"id\":277513874,\"identity\":\"32a79971-f075-428c-9e18-c00b67d013e2\",\"order_by\":5,\"name\":\"Chen Yu\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Shenzhen Bay Laboratory\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Chen\",\"middleName\":\"\",\"lastName\":\"Yu\",\"suffix\":\"\"},{\"id\":277513875,\"identity\":\"b6890935-ad6c-41cd-86a5-d4fb0d63c951\",\"order_by\":6,\"name\":\"Yang Xi\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Ningbo University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Yang\",\"middleName\":\"\",\"lastName\":\"Xi\",\"suffix\":\"\"},{\"id\":277513876,\"identity\":\"1883f866-8aa8-4df7-894c-90cc28f7b231\",\"order_by\":7,\"name\":\"Qi Liao\",\"email\":\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAu0lEQVRIiWNgGAWjYBACPmYQaWDDYACieYjRwgbRkkaKFgh1mBQt7MwPH7wpOJ9nLpHA+OBtG4O8OWGHsRkbzjG4XWw5I4HZcG4bg+HOBoJaGMykeQxuJ264kcAmzdvGkGBwgKAW9m9ALedAWth/E6mFB2TLAbAtzMRqKQb6JTlxZ8/DZsk55yQMNxDSws9/fOODN3/sErezJx/88KbMRp6gLWAAiQ7GBiAhQYx6BiJjcBSMglEwCkYuAAAupTblkeqxAQAAAABJRU5ErkJggg==\",\"orcid\":\"\",\"institution\":\"Ningbo University\",\"correspondingAuthor\":true,\"prefix\":\"\",\"firstName\":\"Qi\",\"middleName\":\"\",\"lastName\":\"Liao\",\"suffix\":\"\"}],\"badges\":[],\"createdAt\":\"2024-03-08 01:26:13\",\"currentVersionCode\":1,\"declarations\":\"\",\"doi\":\"10.21203/rs.3.rs-4034323/v1\",\"doiUrl\":\"https://doi.org/10.21203/rs.3.rs-4034323/v1\",\"draftVersion\":[],\"editorialEvents\":[],\"editorialNote\":\"\",\"failedWorkflow\":false,\"files\":[{\"id\":52535899,\"identity\":\"ea5888b5-8a09-415a-8893-f9a52463676c\",\"added_by\":\"auto\",\"created_at\":\"2024-03-12 16:32:20\",\"extension\":\"png\",\"order_by\":1,\"title\":\"Figure 1\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":466742,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eIdentification of CRC ESPs using CRISPR loss-of-function screen and scRNA-seq datasets. \\u003cstrong\\u003ea.\\u003c/strong\\u003eThe pie chart displays the CRC cell line types in the CRISPR loss-of-function screen samples. \\u003cstrong\\u003eb.\\u003c/strong\\u003e Overlapping ratios between previously identified pan-cancer ESP sets and our results with different ranking proportion cutoffs. \\u003cstrong\\u003ec.\\u003c/strong\\u003e Average expression levels of CRC ESPs in tumor and normal epithelial cells. \\u003cstrong\\u003ed. \\u003c/strong\\u003eFunctional enrichment analysis of 74 overexpressed CRC ESPs. \\u003cstrong\\u003ee.\\u003c/strong\\u003e Comparison of CERES scores among different ESP sets. \\u003cstrong\\u003ef.\\u003c/strong\\u003e Overlap between CRC ESPs or randomly selected genes and three pan-cancer ESP sets. \\u003cstrong\\u003eg.\\u003c/strong\\u003e The venn diagram illustrates the intersection of different ESP sets. \\u003cstrong\\u003eh.\\u003c/strong\\u003e The effect of \\u003cem\\u003eCTNNB1\\u003c/em\\u003eand \\u003cem\\u003eNDUFB9\\u003c/em\\u003e knockout by CRISPR/Cas9 on the survival of HCT116 cell lines.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"1.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4034323/v1/e210ecbad78cd24f6350b9bc.png\"},{\"id\":52535759,\"identity\":\"31198f9a-9b8e-40a0-a511-fe7744d32573\",\"added_by\":\"auto\",\"created_at\":\"2024-03-12 16:32:10\",\"extension\":\"png\",\"order_by\":2,\"title\":\"Figure 2\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":373823,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003ePrediction of CRC ESLs using co-expression networks. \\u003cstrong\\u003ea.\\u003c/strong\\u003e The workflow of predicting CRC ESLs. The venn diagram displays the intersection of predicted CRC ESLs from the two methods. \\u003cstrong\\u003eb.\\u003c/strong\\u003e The distributions of adjusted \\u003cem\\u003ep\\u003c/em\\u003e-value for different ESP sets by HT method based on unWCN-TCGA and unWCN-TEC. \\u003cstrong\\u003ec.\\u003c/strong\\u003eThe distributions of PPR scores for different ESP sets by PPR method based on WCN-TCGA and WCN-TEC. \\u003cstrong\\u003ed.\\u003c/strong\\u003e The expression levels of CRC ESLs in CRC tumor and normal epithelial cells. Genes marked in red indicate significantly overexpression in tumor cells. \\u003cstrong\\u003ee.\\u003c/strong\\u003e The effect of ESLs knockout by CRISPR/Cas9 on survival of HCT116 cell line. \\u003cstrong\\u003ef. \\u003c/strong\\u003eThe functional annotation of CRC ESLs.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"2.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4034323/v1/5bbebf13aa93c558ff095955.png\"},{\"id\":52535793,\"identity\":\"dc8b8e7f-750a-4ee5-8e70-3ed9de166f62\",\"added_by\":\"auto\",\"created_at\":\"2024-03-12 16:32:11\",\"extension\":\"png\",\"order_by\":3,\"title\":\"Figure 3\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":439896,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eExpression, DNA mutation, and DNA methylation patterns of CRC essential genes. \\u003cstrong\\u003ea.\\u003c/strong\\u003eThe violin plot shows the gene sets enrichment scores of CRC ESPs, ESLs, other pan-cancer ESP sets, and Hart’s non-ESPs in different cell lineages between tumor and normal samples. The yellow line represents the mean score of CRC ESPs in tumor epithelial cells. \\u003cstrong\\u003eb.\\u003c/strong\\u003e Four CRC ESPs were specifically expressed in tumor epithelial cells compared with other cell lineages. \\u003cstrong\\u003ec.\\u003c/strong\\u003e The effect of four CRC ESPs knockout by CRISPR/Cas9 on survival of HCT116 cell lines. \\u003cstrong\\u003ed-e.\\u003c/strong\\u003e The top mutated CRC ESPs (d) and ESLs (e) based on WGS datasets of PCAWG. \\u003cstrong\\u003ef.\\u003c/strong\\u003e The correlation between \\u003cem\\u003ePVT1\\u003c/em\\u003e expression and DNA methylation levels of three probes.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"3.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4034323/v1/833015ad6695441e6e005c10.png\"},{\"id\":52535750,\"identity\":\"9ebd882e-de1b-40b0-9366-1a3ddf3a1544\",\"added_by\":\"auto\",\"created_at\":\"2024-03-12 16:32:05\",\"extension\":\"png\",\"order_by\":4,\"title\":\"Figure 4\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":557546,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eRegulatory network between CRC essential genes and TFs. \\u003cstrong\\u003ea. \\u003c/strong\\u003eThe number of target genes regulated by 9 TFs from ESPs. \\u003cstrong\\u003eb.\\u003c/strong\\u003e The tSNE plot displays the distribution of tumor and normal epithelial cells, along with the AUC scores of three regulons.\\u003cstrong\\u003ec.\\u003c/strong\\u003e The differential AUC scores of regulons between tumor and normal epithelial cells. \\u003cstrong\\u003ed. \\u003c/strong\\u003eThe network visualization of relationships between TFs and three novel CRC ESPs. \\u003cstrong\\u003ee.\\u003c/strong\\u003e The number of TFs regulating ESLs. \\u003cstrong\\u003ef.\\u003c/strong\\u003e The network visualization of TF-ESL relationships with high confidence. The width of the line represents the weight of associations.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"4.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4034323/v1/dfbaf3e1181098e137f48f89.png\"},{\"id\":52535902,\"identity\":\"eec96a0c-151b-4899-bcfa-97e2734abd65\",\"added_by\":\"auto\",\"created_at\":\"2024-03-12 16:32:21\",\"extension\":\"png\",\"order_by\":5,\"title\":\"Figure 5\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":473616,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eCRC essential genes involved in TME remodeling through the regulation of ligands/receptors expression. \\u003cstrong\\u003ea.\\u003c/strong\\u003e The relationships between CRC essential genes and TME. The targets of interest can be used to identify candidate drugs through iLINCS analysis. \\u003cstrong\\u003eb.\\u003c/strong\\u003e The CCN associated with tumor epithelial cells. The arrow indicates the direction of signal transduction, and the line width represents the strength of interactions. \\u003cstrong\\u003ec.\\u003c/strong\\u003eAn overview of selected ligands or receptors interactions between tumor epithelial cells and cells in TME. The values of means indicate the average expression levels of the interacting pairs. \\u003cstrong\\u003ed.\\u003c/strong\\u003e Overall survival rates of CRC patients with high or low expression levels of \\u003cem\\u003eCD47\\u003c/em\\u003e and \\u003cem\\u003eMIF\\u003c/em\\u003e. \\u003cstrong\\u003ee.\\u003c/strong\\u003eThe expression pattern of target genes in 5 CRC cell lines after treatment with different drugs.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"5.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4034323/v1/a9ea57d75faa03dc575bfed0.png\"},{\"id\":52589097,\"identity\":\"f64c1336-4a7b-4241-9092-a2ab289273a6\",\"added_by\":\"auto\",\"created_at\":\"2024-03-13 09:49:26\",\"extension\":\"pdf\",\"order_by\":0,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"manuscript-pdf\",\"size\":3270158,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"manuscript.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4034323/v1/c96fa0da-2c39-44de-852c-cc2581a2fbee.pdf\"},{\"id\":52535895,\"identity\":\"16b5d790-b20a-4a9a-afe6-58e8e98cb510\",\"added_by\":\"auto\",\"created_at\":\"2024-03-12 16:32:19\",\"extension\":\"pdf\",\"order_by\":1,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":1786287,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"SupplementaryFigures.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4034323/v1/f1b7049b78b3380fc388bace.pdf\"},{\"id\":52535760,\"identity\":\"337d4575-4215-4460-b90d-52cb4249fc04\",\"added_by\":\"auto\",\"created_at\":\"2024-03-12 16:32:10\",\"extension\":\"xlsx\",\"order_by\":2,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":2659045,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"SupplementaryTables.xlsx\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4034323/v1/9f1841351e12d01781b096bd.xlsx\"}],\"financialInterests\":\"No competing interests reported.\",\"formattedTitle\":\"Identification and Multi-omic Analysis of Essential Coding and Long Non-conding Genes in Colorectal Cancer\",\"fulltext\":[{\"header\":\"Introduction\",\"content\":\"\\u003cp\\u003eColorectal cancer is one of the most frequently diagnosed cancers worldwide. There were estimated 1.88 million new cases and 0.92 million deaths of CRC in 2020\\u0026nbsp;\\u003csup\\u003e1\\u003c/sup\\u003e. Although novel therapies and drugs have approximately doubled the average survival time over the past decade, patients with advanced CRC often succumb to the disease within three years\\u0026nbsp;\\u003csup\\u003e2\\u003c/sup\\u003e.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eGene is defined as essential if its complete loss of function will lead to impaired cell development, gene activation, and cellular reprogramming\\u0026nbsp;\\u003csup\\u003e3-5\\u003c/sup\\u003e. The pooled CRISPR loss-of-function screen is a widely used method for identifying essential protein-coding genes involved in cancer-related biological processes including drug resistance\\u0026nbsp;\\u003csup\\u003e6\\u003c/sup\\u003e, immune suppression\\u0026nbsp;\\u003csup\\u003e7\\u003c/sup\\u003e, and immunotherapy response\\u0026nbsp;\\u003csup\\u003e8\\u003c/sup\\u003e. A number of studies have already obtained multiple sets of pan-cancer core ESPs and non-ESPs using human cancer cell lines, including Hart\\u0026rsquo;s ESPs and Hart\\u0026rsquo;s non-ESPs\\u0026nbsp;\\u003csup\\u003e9,10\\u003c/sup\\u003e, Achilles\\u0026rsquo; ESPs\\u0026nbsp;\\u003csup\\u003e11\\u003c/sup\\u003e, and Zhang\\u0026rsquo;s ESPs\\u0026nbsp;\\u003csup\\u003e12\\u003c/sup\\u003e. Although these studies have highlighted the functions of ESPs and their regulatory networks in cancer tissues, how their impacts on the TME are still unknown and required further exploration and establishment. Besides, the genotypes and phenotypes of various human cancer cell lines and tissues were heterogeneous. It was reported that ESPs in breast, pancreatic, and ovarian cancer cell lines were partially overlapped, indicating certain ESPs may function in a tumor-specific fashion\\u0026nbsp;\\u003csup\\u003e13\\u003c/sup\\u003e. Nowadays, few studies were focused on specific ESPs in CRC.\\u003c/p\\u003e\\n\\u003cp\\u003eDuring the past several years, single-cell RNA sequencing (scRNA-seq) has become an important technique for resolving critical cancer genes and illustrating the composition of TME. Numerous studies have been conducted for a comprehensive investigation of the cellular and intercellular interactions in CRC, demonstrating the dynamic changes in tumor immunity and providing guidance for clinical treatment strategies\\u0026nbsp;\\u003csup\\u003e14,15\\u003c/sup\\u003e. Recently, the technology that combined scRNA-seq with CRISPR - activation\\u0026nbsp;\\u003csup\\u003e5\\u003c/sup\\u003e, CRISPR - knockout screen\\u0026nbsp;\\u003csup\\u003e16\\u003c/sup\\u003e, and CRISPR - interference\\u0026nbsp;\\u003csup\\u003e17\\u003c/sup\\u003e have been emerged to identify genetic regulators and lineage traces. For example, Vishnubalaji et.al integrated the scRNA-seq with CRISPR functional screen data from Achilles project\\u0026nbsp;\\u003csup\\u003e11\\u003c/sup\\u003e and identified potential therapeutic targets for different subtypes of Breast cancer\\u0026nbsp;\\u003csup\\u003e18\\u003c/sup\\u003e. These studies suggest that combined analysis of scRNA-seq and CRISPR functional screen data can more accurately identify essential genes in CRC and will broaden our understanding of their role in the high-dimensional TME.\\u003c/p\\u003e\\n\\u003cp\\u003eIn this study, we first integrated genome-wide CRISPR/Cas9 loss-of-function screen data and scRNA-seq data to identify ESPs in CRC. Then CRC ESLs were predicted using HT and PPR algorithms based on gene co-expressed network constructed from scRNA-seq data of tumor epithelial cells. To further illustrate the characteristics of ESPs and ESLs, we examined their single cell expression patterns, pathway enrichment, DNA mutation, DNA methylation events, and transcription factor (TF) regulatory network. Finally, we integrated co-expressed network, TF regulatory network, and CCN to provide valuable insights into the impact of ESPs and ESLs on the remodeling of TME. The potential drugs that can interfere key factors of target signal process were also predicted. Collectively, our study presents novel insights into the essential roles of CRC ESPs and ELSs in tumor cell development, TF regulation, and TME remodeling, thus establishing a foundation for identification of novel therapeutic targets.\\u003c/p\\u003e\"},{\"header\":\"Results\",\"content\":\"\\u003cp\\u003e\\u003cstrong\\u003eIdentification of CRC ESPs based on integration of CRISPR screen and scRNA-seq datasets\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eEight genome-wide CRISPR/Cas9 loss-of-function screen datasets of CRC were collected, including a total of 49 CRC cell line types and 301 samples with replicates (\\u003cstrong\\u003eFig. 1a,\\u0026nbsp;Supplementary Table 1\\u003c/strong\\u003e). To evaluate the essential degree of each gene, the sgRNAs depletion score for each gene in each sample was calculated by CERES algorithm and defined as CERES score\\u0026nbsp;\\u003csup\\u003e11\\u003c/sup\\u003e. A lower CERES score indicates a higher level of gene essentiality as all the screen datasets are negative selection in this study. Considering multiple datasets were involved, we integrated CERES scores for each gene across various datasets by utilizing Robust Rank Aggregation (RRA) algorithm\\u003csup\\u003e19\\u003c/sup\\u003e. The optimal cutoff to obtain ESPs was determined by calculating the degree of overlap between RRA results and other pan-cancer ESP sets previously identified, including Achilles\\u0026rsquo; ESPs, Hart\\u0026rsquo;s ESPs, and Zhang\\u0026rsquo;s ESPs. As a result, 1521 candidates of CRC ESPs were obtained (\\u003cstrong\\u003eFig. 1b)\\u003c/strong\\u003e. We further collected full-length scRNA-seq data of CRC (GSE81861) to analyze the detection fraction of candidate ESPs in tumor epithelial cells\\u003csup\\u003e20\\u003c/sup\\u003e. We found 396 ESPs were expressed in more than 50% of tumor epithelial cells and were thus considered as final CRC ESPs (\\u003cstrong\\u003eSupplementary Table 2\\u003c/strong\\u003e). Among them, 74 ESPs were found to be overexpressed in tumor epithelial cells (\\u003cstrong\\u003eFig. 1c, Supplementary Table 2\\u003c/strong\\u003e). These ESPs were principally linked to cancer-related functions, such as MYC TARGETS V1/V2, positive regulation of signal transduction by p53 class mediator, and response to interferon-gamma (\\u003cstrong\\u003eFig. 1d\\u003c/strong\\u003e). Notably, we found ESPs exhibit higher expression than other genes in both tumor cells and normal cells (\\u003cstrong\\u003eSupplementary Fig. a\\u003c/strong\\u003e), which corroborated the findings from a previous study\\u0026nbsp;\\u003csup\\u003e9\\u003c/sup\\u003e.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eThen we compared the CERES scores of CRC ESPs with those of pan-cancer ESPs and Hart\\u0026rsquo;s non-ESPs\\u0026nbsp;to assess the quality of our results. As anticipated, Hart\\u0026rsquo;s non-ESPs showed the highest CERES scores and indicated the lowest necessity, while the CRC ESPs were lower, implying their essential roles in CRC (\\u003cstrong\\u003eFig. 1e\\u003c/strong\\u003e). Additionally, we observed larger overlap between CRC ESPs and other pan-cancer ESP sets than that of randomly selected genes, further reflecting the high quality of CRC ESPs (\\u003cstrong\\u003eFig.1f\\u003c/strong\\u003e). Among 396 ESPs, 155 ESPs co-exist in all ESP sets, indicating the homogeneity of different tumors (\\u003cstrong\\u003eFig. 1g\\u003c/strong\\u003e). Notably, we found three novel CRC ESPs, \\u003cem\\u003eCTNNB1\\u003c/em\\u003e, \\u003cem\\u003eNDUFB9\\u003c/em\\u003e, and \\u003cem\\u003eSCNM1\\u0026nbsp;\\u003c/em\\u003ethat were not found in any other ESP sets (\\u003cstrong\\u003eFig. 1g\\u003c/strong\\u003e). To determine whether these novel CRC ESPs are functionally important in the survival of CRC cancer cell lines, we detected the effect of \\u003cem\\u003eCTNNB1\\u003c/em\\u003e and \\u003cem\\u003eNDUFB9\\u003c/em\\u003e knockout on HCT116 cell line by CRISPR/Cas9. Remarkably, HCT116 cell line lost viability while \\u003cem\\u003eCTNNB1\\u0026nbsp;\\u003c/em\\u003eand\\u003cem\\u003e\\u0026nbsp;NDUFB9\\u003c/em\\u003e were knockout, respectively (\\u003cstrong\\u003eFig. 1h\\u003c/strong\\u003e). \\u003cem\\u003eCTNNB1\\u003c/em\\u003e was reported as a driver gene in CRC according to the Pan-Cancer Analysis of Whole Genomes (PCAWG) study \\u003csup\\u003e21\\u003c/sup\\u003e. \\u003cem\\u003eNDUFB9\\u003c/em\\u003e, an important gene involved in mitochondrial function, has been identified as a prognostic biomarker for endometrial cancer \\u003csup\\u003e22\\u003c/sup\\u003e. However, the role of \\u003cem\\u003eSCNM1\\u0026nbsp;\\u003c/em\\u003ein CRC is needed to further investigate.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003ePrediction of CRC ESLs\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eLncRNAs play an important\\u0026nbsp;regulatory role in various cellular processes of CRC, including cell apoptosis, proliferation, and epithelial-mesenchymal transition\\u0026nbsp;\\u003csup\\u003e23\\u003c/sup\\u003e.\\u0026nbsp;To elucidate a comprehensive landscape of essential genes in CRC, we employed HT and PPR methods, which we proposed in our previous study, to predict essential lncRNAs\\u0026nbsp;\\u003csup\\u003e12\\u003c/sup\\u003e. These methods required the construction of a two-color network comprising of both protein-coding genes and lncRNAs, where lncRNAs\\u0026nbsp;\\u003cem\\u003eessentiality\\u003c/em\\u003e were predicted based on their closeness to\\u0026nbsp;ESPs\\u0026nbsp;\\u003csup\\u003e12\\u003c/sup\\u003e. In order to look for the best two-color network, we built four different co-expression networks, consisting of an unweighted co-expression network and a topological overlap weighted co-expression network based on the\\u0026nbsp;RNA sequencing (RNA-seq)\\u0026nbsp;profiles of colorectal adenocarcinoma from The Cancer Genome Atlas (TCGA, referred to as unWCN-TCGA and WCN-TCGA, respectively), as well as an unweighted co-expression network and a topological overlap weighted co-expression network based on expression profiles of tumor\\u0026nbsp;epithelial cells\\u0026nbsp;from scRNA-seq dataset (referred to as unWCN-TEC and WCN-TEC, respectively) (\\u003cstrong\\u003eFig. 2a, see method\\u003c/strong\\u003e). Then we calculated the adjusted \\u003cem\\u003ep\\u003c/em\\u003e-value of each gene by HT to evaluate whether the directly related genes of a query gene are enriched in ESPs based on unWCN-TCGA and unWCN-TEC (\\u003cstrong\\u003eFig. 2a)\\u003c/strong\\u003e. We observed that the adjusted \\u003cem\\u003ep\\u003c/em\\u003e-values of CRC ESPs, Achilles\\u0026rsquo;\\u0026nbsp;ESPs, Zhang\\u0026rsquo;s ESPs, and Hart\\u0026rsquo;s ESPs showed significantly smaller than those of\\u0026nbsp;Hart\\u0026rsquo;s non-ESPs (\\u003cstrong\\u003eFig. 2b\\u003c/strong\\u003e). Importantly, the\\u0026nbsp;adjusted\\u003cem\\u003e\\u0026nbsp;p\\u003c/em\\u003e-values of ESPs based on unWCN-TEC were significantly lower than those based on unWCN-TCGA\\u0026nbsp;(\\u003cstrong\\u003eFig. 2b\\u003c/strong\\u003e), suggesting the higher quality of co-expression network constructed from scRNA-seq datasets than that from TCGA RNA-seq datasets. Consequently, we identified 26 lncRNAs as potential candidates of CRC ESLs by using the cutoff of\\u0026nbsp;adjusted \\u003cem\\u003ep\\u003c/em\\u003e-value lower than 0.01 based on\\u0026nbsp;unWCN-TEC.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eNext, we utilized CRC ESPs as original seeds to calculate the score of all genes by PPR method based on WCN-TCGA\\u0026nbsp;and\\u0026nbsp;WCN-TEC,\\u0026nbsp;respectively\\u0026nbsp;(\\u003cstrong\\u003eFig. 2a\\u003c/strong\\u003e). The higher score represents the higher probability of a gene being essential. As expected, the PPR scores of\\u0026nbsp;CRC\\u0026nbsp;ESPs, Achilles\\u0026rsquo;\\u0026nbsp;ESPs, Zhang\\u0026rsquo;s ESPs, and Hart\\u0026rsquo;s ESPs were higher, and WCN-TEC also showed greater advantage than WCN-TCGA as higher PPR scores of ESPs in WCN-TEC were observed (\\u003cstrong\\u003eFig. 2c\\u003c/strong\\u003e). We then selected the top 5% lncRNAs in PPR results based on WCN-TEC as CRC ESL candidates (\\u003cem\\u003en\\u003c/em\\u003e=39). Results yielded 41 ESLs after combing the results of the HT and PPR methods, and 24 lncRNAs were found to be same (\\u003cstrong\\u003eFig. 2a\\u003c/strong\\u003e). Finally, we retained the ESLs detected in more than 30% of CRC tumor epithelial cells (GSE81861) and defined them as CRC ESLs (\\u003cem\\u003en\\u003c/em\\u003e=29,\\u0026nbsp;\\u003cstrong\\u003eSupplementary Table 3\\u003c/strong\\u003e).\\u0026nbsp;Expression analysis revealed that 11 CRC ESLs were significantly overexpressed in CRC tumor epithelial cells compared to normal (\\u003cstrong\\u003eFig.2d, Supplementary Table 3\\u003c/strong\\u003e). We selected 6 overexpressed ESLs, including \\u003cem\\u003eH19\\u003c/em\\u003e, \\u003cem\\u003eMIF-AS1\\u003c/em\\u003e, \\u003cem\\u003eNORAD\\u003c/em\\u003e, \\u003cem\\u003eSNHG17\\u003c/em\\u003e, \\u003cem\\u003eSNHG6\\u003c/em\\u003e, and \\u003cem\\u003eSNHG8\\u003c/em\\u003e, for CRISPR/Cas9 knockout experiments. The results demonstrated that knockout of these ESLs in HCT116 cell line will result in the decreasing of survival cells (\\u003cstrong\\u003eFig. 2e\\u003c/strong\\u003e), implying the dependent role of those ESLs in CRC.\\u003c/p\\u003e\\n\\u003cp\\u003eIn our results, some ESLs were known as CRC associated lncRNAs according to previous reports, such as \\u003cem\\u003ePVT1\\u003c/em\\u003e, \\u003cem\\u003eH19\\u003c/em\\u003e, and \\u003cem\\u003eGAS5\\u003c/em\\u003e (\\u003cstrong\\u003eFig. 2f, Supplementary Table 3\\u003c/strong\\u003e). By analyzing the functions and clinical association of CRC ESLs in Lnc2Cancer 3.0 database\\u003csup\\u003e24\\u003c/sup\\u003e, we found that most CRC ESLs are involved in cell apoptosis, growth, survival, and metastasis (\\u003cstrong\\u003eFig. 2f\\u003c/strong\\u003e). Among these ESLs, 11 lncRNAs were previously identified as part of Zhang\\u0026rsquo;s pan-cancer ESLs (\\u003cem\\u003en\\u003c/em\\u003e=97, \\u003cstrong\\u003eFig. 2f\\u003c/strong\\u003e)..\\u0026nbsp;To identify potential Gene Ontology (GO) biological process functions of CRC ESLs, we conducted enrichment analysis on the top 200 associated genes of each ESL based on WCN-TEC. The results revealed that a significant number of CRC ESLs are involved in ribosome/ncRNA/rRNA/DNA processing, regulation of translation, telomere maintenance, and signal transduction by p53 class mediator (\\u003cstrong\\u003eSupplementary Table 4\\u003c/strong\\u003e), providing additional evidence for the crucial roles of CRC ESLs in fundamental metabolic processes of tumor cells.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eExpression, DNA mutation, and DNA methylation patterns of CRC essential genes\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eTo examine the expression characteristics of essential genes in different cell types of CRC, we utilized CRC ESPs and ESLs as gene sets to calculate enrichment scores in each cell type based on the 10X genomics scRNA-seq dataset of CRC (GSE132465)\\u0026nbsp;\\u003csup\\u003e15\\u003c/sup\\u003e. As anticipated, both CRC ESPs and ESLs gene set enrichment scores were higher in epithelial cells of tumor samples compared with normal samples (\\u003cstrong\\u003eFig. 3a\\u003c/strong\\u003e). Interestingly, the gene set scores of CRC ESPs were particularly elevated in tumor epithelial cells compared to other pan-cancer ESP sets (\\u003cstrong\\u003eFig. 3a\\u003c/strong\\u003e), indicating the specificity of CRC ESPs for CRC tumor epithelial cells. We also observed the expressions of ESPs and ESLs in other cell lineages, such as T cells and B cells (\\u003cstrong\\u003eFig. 3a\\u003c/strong\\u003e), were lower than tumor epithelial cells, implying that target essential genes may have side effect on cell lineages in TME. Notably, 4 ESPs (\\u003cem\\u003eKRT8\\u003c/em\\u003e, \\u003cem\\u003eMYC\\u003c/em\\u003e, \\u003cem\\u003ePRELID3B\\u003c/em\\u003e, and \\u003cem\\u003eSCD\\u003c/em\\u003e) were specifically expressed in tumor epithelial cells (\\u003cstrong\\u003eFig. 3b, see methods\\u003c/strong\\u003e). Three of them, \\u003cem\\u003eKRT8\\u003c/em\\u003e, \\u003cem\\u003eMYC\\u003c/em\\u003e, and\\u003cem\\u003e\\u0026nbsp;SCD\\u003c/em\\u003e are well-known biomarkers or therapeutic targets for cancer\\u003csup\\u003e25-27\\u003c/sup\\u003e, while few research has performed on \\u003cem\\u003ePRELID3B\\u003c/em\\u003e in cancer. CRISPR/Cas9 knockout experiments confirmed that function loss of these genes affects the survival of HCT116 cell line (\\u003cstrong\\u003eFig. 3c\\u003c/strong\\u003e).\\u003c/p\\u003e\\n\\u003cp\\u003eMutation of essential genes has been shown to disrupt normal cell development and promote cancer, such as \\u003cem\\u003eMYC\\u003c/em\\u003e. We conducted mutation analysis using whole genome sequencing (WGS) datasets of CRC in PCAWG and whole exome sequencing (WES) datasets of CRC in TCGA to explore the genomic alteration of CRC ESPs and ESLs. In general, the mutation rates of ESPs and ESLs in WGS dataset ranged from 0% to 25% and 0% to 33%, respectively (\\u003cstrong\\u003eFig. 3d-e\\u003c/strong\\u003e), while the mutation rates of ESPs were relatively low in WES dataset, ranging from 0% to 8% (\\u003cstrong\\u003eSupplementary Fig. b\\u003c/strong\\u003e). Among the mutated genes in WGS study, \\u003cem\\u003eDYNC1H1\\u0026nbsp;\\u003c/em\\u003e(25%), \\u003cem\\u003eCHD4\\u0026nbsp;\\u003c/em\\u003e(23%), \\u003cem\\u003eCNOT1\\u0026nbsp;\\u003c/em\\u003e(21%), and \\u003cem\\u003eCTNNB1\\u003c/em\\u003e (19%) exhibited the highest mutation frequency (\\u003cstrong\\u003eFig. 3d\\u003c/strong\\u003e). Similarly, these genes also showed relatively high frequencies and ranked within the top 5 mutated genes in the TCGA CRC dataset (\\u003cstrong\\u003eSupplementary Fig. b\\u003c/strong\\u003e). Importantly, most CRC ESPs mutated in a co-occurrence manner in both the WGS and WES datasets (\\u003cstrong\\u003eSupplementary Fig. c-d\\u003c/strong\\u003e), suggesting that the mutation of ESPs typically activates synergistic oncogenic pathways.\\u003c/p\\u003e\\n\\u003cp\\u003eDNA methylation also plays a crucial role in influencing gene expression. Through analysis of differentially methylated positions (DMPs) in ESPs and ESLs in TCGA-COAD, we identified a total of 71 DMPs with 45 unique genes (adjusted \\u003cem\\u003ep\\u003c/em\\u003e-value \\u0026lt; 0.05, absolute log\\u003csub\\u003e2\\u003c/sub\\u003eFC \\u0026gt; 0.2, \\u003cstrong\\u003eSupplementary Fig. e\\u003c/strong\\u003e). Of these DMPs, 26 showed DNA hypermethylation (log\\u003csub\\u003e2\\u003c/sub\\u003eFC \\u0026gt; 0.2), while 45 exhibited DNA hypomethylation (log\\u003csub\\u003e2\\u003c/sub\\u003eFC \\u0026lt; -0.2). Subsequently, we calculated the correlation coefficients between the expression and DNA methylation levels of each gene. The results revealed three ESPs (\\u003cem\\u003eMYC\\u003c/em\\u003e, \\u003cem\\u003eSCD\\u003c/em\\u003e, \\u003cem\\u003eEIF6\\u003c/em\\u003e) and one ESLs (\\u003cem\\u003ePVT1\\u003c/em\\u003e) had negative correlation coefficients (\\u003cstrong\\u003eSupplementary Table 5\\u003c/strong\\u003e). We observed hypomethylation of two probes cg08526705 and cg00163372 in the gene body may activate \\u003cem\\u003eMYC\\u003c/em\\u003e expression in CRC (\\u003cstrong\\u003eSupplementary Fig. f-g\\u003c/strong\\u003e). These two probes were found to be related with chemotherapy drug resistance previously\\u003csup\\u003e28\\u003c/sup\\u003e. In \\u003cem\\u003ePVT1\\u003c/em\\u003e, three probes cg23898497, cg00780520, and cg10202727 were found to be associated with \\u003cem\\u003ePVT1\\u003c/em\\u003e expression (\\u003cstrong\\u003eFig. 3f\\u003c/strong\\u003e), which were also observed in TCGA-READ (\\u003cstrong\\u003eSupplementary Table 5\\u003c/strong\\u003e). \\u003cem\\u003ePVT1\\u003c/em\\u003e is an adjacent gene of \\u003cem\\u003eMYC\\u003c/em\\u003e on chromosome 8q24.21 and has been shown to regulate \\u003cem\\u003eMYC\\u003c/em\\u003e expression and contribute to cancer development \\u003csup\\u003e29\\u003c/sup\\u003e. Our findings indicated that the hypomethylation of \\u003cem\\u003eMYC\\u0026nbsp;\\u003c/em\\u003eand \\u003cem\\u003ePVT1\\u003c/em\\u003e lead to the upregulation of their expression, and thus promote the development of CRC.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eTF regulatory network of essential genes in CRC tumor epithelial cells\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eBased on 10X genomics scRNA-seq dataset of CRC epithelial cells (GSE81861), we generated a TF regulatory network of CRC by using SCENIC algorithm\\u003csup\\u003e30\\u003c/sup\\u003e. A total of 30,743 TF-target relationships associated with CRC ESPs and ESLs were established, including 177 TFs and 9,837 genes (\\u003cstrong\\u003eSupplementary Table 6\\u003c/strong\\u003e). These relationships were further categorized into high- or low-confidence annotations\\u0026nbsp;\\u003csup\\u003e30\\u003c/sup\\u003e(\\u003cstrong\\u003esee methods\\u003c/strong\\u003e). Interestingly, 9 TFs from ESPs related to 80.63% (\\u003cem\\u003en\\u003c/em\\u003e=24,788) of identified TF-target relationships. Among them, 3 TFs had a highest number of high-confidence relationships compared with others essential TFs, including \\u003cem\\u003eBCLAF1\\u003c/em\\u003e, \\u003cem\\u003eMYC\\u003c/em\\u003e, and \\u003cem\\u003eYY1\\u0026nbsp;\\u003c/em\\u003e\\u003cstrong\\u003e(Fig. 4a)\\u003c/strong\\u003e. Area under the curve (AUC) scores from SCENIC were used to quantify the activity of TF targets (regulon). Differential AUC scores of regulons revealed the significant activation of 79 regulons in tumor epithelial cells, including\\u003cem\\u003e\\u0026nbsp;BCLAF1\\u003c/em\\u003e regulon\\u003cem\\u003e, YY1\\u0026nbsp;\\u003c/em\\u003eregulon\\u003cem\\u003e,\\u003c/em\\u003e and \\u003cem\\u003eMYC\\u003c/em\\u003e regulon (\\u003cstrong\\u003eFig. 4b-c, Supplementary Table 7\\u003c/strong\\u003e). Furthermore, three novel CRC ESPs (\\u003cem\\u003eCTNNB1\\u003c/em\\u003e, \\u003cem\\u003eNDUFB9\\u003c/em\\u003e and \\u003cem\\u003eSCNM1\\u003c/em\\u003e) that mentioned above were also regulated by \\u003cem\\u003eBCLAF1\\u003c/em\\u003e,\\u003cem\\u003e\\u0026nbsp;YY1\\u003c/em\\u003e, and \\u003cem\\u003eMYC\\u003c/em\\u003e (\\u003cstrong\\u003eFig. 4d\\u003c/strong\\u003e), suggesting that these TFs may act as oncogenes in CRC. Previously, knockdown of the L isoform of \\u003cem\\u003eBCLAF1\\u003c/em\\u003e in mouse tumor model inhibited the tumor growth, confirming the carcinogenic characteristics of \\u003cem\\u003eBCLAF1\\u003c/em\\u003e\\u003csup\\u003e31\\u003c/sup\\u003e. Above all, our findings demonstrated the intricate regulatory networks between TFs and CRC ESPs, and the TFs arise from CRC ESPs may play significant roles in tumorigenesis.\\u003c/p\\u003e\\n\\u003cp\\u003eIn the regulatory network, 424 relationships were linked between TFs (\\u003cem\\u003en\\u003c/em\\u003e=108) and ESLs (\\u003cem\\u003en\\u003c/em\\u003e=28, \\u003cstrong\\u003eFig. 4e\\u003c/strong\\u003e, \\u003cstrong\\u003eSupplementary Table 6\\u003c/strong\\u003e). Notably, the lncRNAs such as \\u003cem\\u003eFGD5 - AS1\\u003c/em\\u003e, \\u003cem\\u003eGAS5\\u003c/em\\u003e, and \\u003cem\\u003ePVT1\\u003c/em\\u003e were regulated by more than 20 TFs. Among these, \\u003cem\\u003eFGD5 - AS1\\u003c/em\\u003e, has been emerged as a crucial regulator in CRC and other types of cancer via promoting cell proliferation, drug resistance, and epithelial-mesenchymal transition\\u003csup\\u003e32,33\\u003c/sup\\u003e. In our study, 14 TFs were found to regulate the expression of \\u003cem\\u003eFGD5 - AS1\\u003c/em\\u003e (\\u003cstrong\\u003eFig.4f\\u003c/strong\\u003e), some of them were known well associated with CRC, such as \\u003cem\\u003eEGR1\\u003c/em\\u003e\\u003cem\\u003e\\u003csup\\u003e34\\u003c/sup\\u003e\\u003c/em\\u003e and \\u003cem\\u003eXBP1\\u003c/em\\u003e\\u003cem\\u003e\\u003csup\\u003e35\\u003c/sup\\u003e\\u003c/em\\u003e. Collectively, our findings provided a reliably regulatory network connecting TFs and CRC essential genes based on scRNA-seq data.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eCrosstalk between CRC essential genes and TME\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eThe CCN between cancer cells and cells in TME mediated by ligand-receptor interactions, plays a crucial role in shaping tumor behavior and TME remodeling. In our study, only three CRC ESPs, including \\u003cem\\u003eCOPA\\u003c/em\\u003e, \\u003cem\\u003eRPS19\\u003c/em\\u003e, and \\u003cem\\u003eTFRC\\u003c/em\\u003e, were known to act as ligands or receptors. The majority of CRC ESPs may participate in the CCN by regulating the expression of ligands or receptors in tumor epithelial cells (\\u003cstrong\\u003eFig. 5a\\u003c/strong\\u003e). To verify such relationships, we first constructed a CCN associated with tumor epithelial cells using CellphoneDB\\u003csup\\u003e36\\u003c/sup\\u003e. The CCN consists of 169 connections targeting tumor epithelial cells and 141 connections source from tumor epithelial cells (\\u003cstrong\\u003eSupplementary Table 8\\u003c/strong\\u003e). The CCN involves 161 ligands/receptors, including JAG1 ligand, NOTCH receptors, and 11 protein complex such as the integrin complex and VEGFR complex (\\u003cstrong\\u003eSupplementary Table 8\\u003c/strong\\u003e). On the whole, CRC tumor epithelial cells interacted more frequently with endothelial cells, fibroblast cells and myeloid cells (\\u003cstrong\\u003eFig. 5b\\u003c/strong\\u003e), which are known to play important roles in angiogenesis, tumor invasion, and immune response.\\u003c/p\\u003e\\n\\u003cp\\u003eWe then integrated the WCN-TEC, TF regulatory network, and CCN, resulting in a total of 1734 relationships (\\u003cstrong\\u003eSupplementary Table 9\\u003c/strong\\u003e), where 356 ESPs (including 9 TFs) and 28 ESLs affect the expression of 85 ligands/receptors of tumor epithelial cells, suggesting that CRC ESPs and TFs have a widespread influence on CCN (\\u003cstrong\\u003eSupplementary Fig. h\\u003c/strong\\u003e). We observed that six ligands/receptors of tumor epithelial cells, including \\u003cem\\u003eGPI\\u003c/em\\u003e, \\u003cem\\u003eCDH1\\u003c/em\\u003e, \\u003cem\\u003eMIF\\u003c/em\\u003e, \\u003cem\\u003eRPS19\\u003c/em\\u003e, \\u003cem\\u003eCXADR\\u003c/em\\u003e, and \\u003cem\\u003eCD47\\u003c/em\\u003e, are associated with more than 100 essential genes (\\u003cstrong\\u003eSupplementary Table 9\\u003c/strong\\u003e), indicating these ligands/receptors may act as the critical mediators between essential genes and TME. Among them, \\u003cem\\u003eMIF\\u003c/em\\u003e secreted from tumor epithelial cells could interact with \\u003cem\\u003eCD74\\u003c/em\\u003e, \\u003cem\\u003eTNFRSF10D\\u003c/em\\u003e, and \\u003cem\\u003eTNFRSF14\\u003c/em\\u003e (\\u003cstrong\\u003eFig. 5c\\u003c/strong\\u003e), the relationships may responsible for the suppression of antitumor immune response\\u0026nbsp;\\u003csup\\u003e37\\u003c/sup\\u003e. Additionally, it has been reported that the CD47-SIRP\\u0026alpha;/\\u0026gamma;\\u0026nbsp;interaction protects tumor cells from killing by suppressing both macrophage phagocytosis and antigen presentation of dendritic cells\\u003csup\\u003e38\\u003c/sup\\u003e. Consistent with these findings, we observed a significant activation of CD47-SIRP\\u0026alpha;/\\u0026gamma;\\u0026nbsp;interactions between tumor epithelial cells and myeloid cells (\\u003cstrong\\u003eFig. 5c\\u003c/strong\\u003e). Clinically, CRC patients from TCGA with high expression levels of \\u003cem\\u003eCD47\\u003c/em\\u003e and \\u003cem\\u003eMIF\\u003c/em\\u003e exhibited a significantly lower overall survival rate compared to those patients with low expression levels of \\u003cem\\u003eCD47\\u003c/em\\u003e and \\u003cem\\u003eMIF\\u003c/em\\u003e (\\u003cstrong\\u003eFig. 5d\\u003c/strong\\u003e). Our results supported the notion that essential genes can facilitate tumor cells evasion from immune system by regulating the expression of of receptors/ligands such as \\u003cem\\u003eCD47\\u003c/em\\u003e and \\u003cem\\u003eMIF\\u0026nbsp;\\u003c/em\\u003ein\\u003cem\\u003e\\u0026nbsp;\\u003c/em\\u003eCRC.\\u003c/p\\u003e\\n\\u003cp\\u003eNext, we aimed to identify existing drugs that could potentially repress CRC essential genes and TFs associated with \\u003cem\\u003eCD47\\u003c/em\\u003e and \\u003cem\\u003eMIF\\u0026nbsp;\\u003c/em\\u003e(\\u003cem\\u003en\\u003c/em\\u003e=304, \\u003cstrong\\u003eFig. 5a\\u003c/strong\\u003e). We utilized an integrative web platform, iLINCS\\u0026nbsp;\\u003csup\\u003e39\\u003c/sup\\u003e, to analyze the expression pattern of 304 targets in pre-computed signatures of 10 anti-cancer agents in 5 CRC cell lines (GSE116439)\\u003csup\\u003e40\\u003c/sup\\u003e. We found that drugs such as sunitinib, dasatinib, topotecan, and lapatinib were sensitive to cancer cells, as they can effectively inhibit the activity of target genes at appropriate concentration and treatment time (\\u003cstrong\\u003eFig. 5e, Supplementary Table 10\\u003c/strong\\u003e). For instance, sunitinib acts as an inhibitor of \\u003cem\\u003eCSF1R\\u003c/em\\u003e, \\u003cem\\u003eVEGF\\u0026nbsp;\\u003c/em\\u003ereceptor, c-kit, \\u003cem\\u003ePDGF\\u0026nbsp;\\u003c/em\\u003ereceptor, and \\u003cem\\u003eRET\\u003c/em\\u003e, and has been widely used in cancer therapy\\u003csup\\u003e41\\u003c/sup\\u003e. Treatment with sunitinib resulted in the down-regulation of \\u003cem\\u003eMYC\\u003c/em\\u003e, \\u003cem\\u003eKRT8\\u003c/em\\u003e, \\u003cem\\u003eCDC37\\u003c/em\\u003e, and other genes in the CRC cell line KM12 (\\u003cstrong\\u003eFig. 5e,\\u0026nbsp;\\u003c/strong\\u003esignature ID: PG_4150). In addition, the patients with high \\u003cem\\u003eCD47\\u003c/em\\u003e expression benefit from sunitinib monotherapy in clear cell renal cell carcinoma\\u003csup\\u003e42\\u003c/sup\\u003e, further supporting sunitinib may act as CD47 inhibitor in CRC. On the other hand, CRC cell lines treated with cisplatin or vorinostat exhibited over-expression of target genes such as \\u003cem\\u003eMYC\\u003c/em\\u003e and \\u003cem\\u003eTAF7\\u003c/em\\u003e (\\u003cstrong\\u003eFig. 5e\\u003c/strong\\u003e), indicating that CRC cells may develop resistance to these drugs. In conclusion, our results suggest that CRC essential genes may impact the TME through CCN and provides valuable insights into potential therapeutic strategies for CRC.\\u003c/p\\u003e\"},{\"header\":\"Discussion\",\"content\":\"\\u003cp\\u003eAccumulated genetic and epigenetic factors are widely believed to play a significant role in pathogenesis of CRC. However, the precise molecular mechanism remains challenging to uncover. Over the past decade, large-scale CRISPR loss-of-function screen data has been extensively used to identify essential genes for pan-cancer. Notably, the Dependency Map (DepMap) portal and BioGRID ORCS have provided an index of CRISPR screens research focused on identifying and understanding the landscape of pan-cancer ESPs\\u0026nbsp;\\u003csup\\u003e43,44\\u003c/sup\\u003e. However, due to the specificity of tissue and the heterogeneity of tumors, pan-cancer ESPs may not be suitable for certain types of cancer\\u0026nbsp;\\u003csup\\u003e13\\u003c/sup\\u003e, especially the application of the targeted therapy. In this study, we employed genome-wide CRISPR screen, scRNA-seq data, co-expression networks, and CRISPR/Cas9 knockout experiments to screen and validate ESPs and ESLs associated with CRC. We found CRC ESPs showed partially overlap with well known pan-cancer ESPs, exhibit high expression levels in tumor epithelial cells of CRC, contribute to fundamental and critical metabolic pathways and play crucial role in the regulation of TF regulatory network and CCN. Collectively, our results provide a comprehensive landscape of essential genes and their characteristics in CRC, enabling the identification of potential therapy strategies.\\u003c/p\\u003e\\n\\u003cp\\u003eEssential genes were reported to highly express in tumor than normal samples by using RNA-seq\\u003csup\\u003e12\\u003c/sup\\u003e. We then streamlined the candidate essential genes from 1521 to 396 considering that the essential genes should be expressed in more than 50% of tumor epithelial cells. Consistent with previous perspective, we have found that 155 CRC ESPs overlap with different sets of pan-cancer ESPs, indicating the homogeneity of different tumors and general cancer related functions. Such similar result across various functional genome screen suggests that common essential genes are generally required for tumor cells survival. Moreover, novel CRC ESP \\u003cem\\u003eCTNNB1\\u003c/em\\u003e has been reported to drive CRC evolution\\u0026nbsp;\\u003csup\\u003e21\\u003c/sup\\u003e because \\u003cem\\u003eCTNNB1\\u003c/em\\u003e mutation only occurs in early-stage of CRC.. As expected, knockout of \\u003cem\\u003eCTNNB1\\u0026nbsp;\\u003c/em\\u003ealso significantly inhibited the proliferation of CRC, indicating the role of \\u003cem\\u003eCTNNB1\\u003c/em\\u003e in CRC tumor cells survival. Additionally, we revealed that most pan-cancer ESPs and CRC ESPs were not only expressed in tumor epithelial cells, but also immune cells and stromal cells in CRC, the finding indicates the low efficiency, side effect, and even failure of therapy will be occurred if targeting pan-cancer ESPs\\u003csup\\u003e45\\u003c/sup\\u003e. Our results yield new insights into the biological characteristic of CRC ESPs.\\u003c/p\\u003e\\n\\u003cp\\u003eGenome-wide CRISPR/Cas9 screen approaches are not well suitable for identifying lncRNAs\\u003csup\\u003e46\\u003c/sup\\u003e. Instead, co-expression networks including both protein coding genes and lncRNAs enable prediction of essential genes based on network features\\u003csup\\u003e47\\u003c/sup\\u003e. We utilized full-length scRNA-seq data to construct specific co-expression networks from tumor epithellial cells, which exhibited low noise and low redundancy compared to traditional bulk transcriptome in our study. Taking advantage of this, we can accurately predict CRC ESLs based on network features by performing HT and PPR algorithm. Similar to ESPs, we also observed that the identified ESLs were highly expressed in tumor epithelial cells than normal. Results showed that part of ESLs also existed in Zhang\\u0026rsquo;s pan-cancer ESLs, and many of them are known to be involved in tumor-promoting, tumor-suppressing functions, and fundamental metabolic processes of tumor cells. We further validated the effect of ESLs on CRC cell survival through CRISPR/Cas9 knockout experiments, exhibiting the decisive role of them. Furthermore, the activity of ESLs also affected by DNA mutations and methylations. For instance, \\u003cem\\u003eGAS5\\u003c/em\\u003e, acts as\\u003cem\\u003e\\u0026nbsp;\\u003c/em\\u003ea tumor suppressor\\u003csup\\u003e48,49\\u003c/sup\\u003e, showed high mutation rate in CRC patients from PCAWG and high expression level in CRC tumor epithelial cells compared to normal epithelial cells. Three CpG sites of \\u003cem\\u003ePVT1\\u003c/em\\u003e that identified in our study were significantly hypomethylated in CRC and was negatively correlated with \\u003cem\\u003ePVT1\\u0026nbsp;\\u003c/em\\u003eexpression. As a novel epigenetic enhancer of \\u003cem\\u003eMYC\\u0026nbsp;\\u003c/em\\u003e\\u003csup\\u003e29,50\\u003c/sup\\u003e, the epigenetic CpG sites of \\u003cem\\u003ePVT1\\u003c/em\\u003e may impact the expression of \\u003cem\\u003eMYC\\u003c/em\\u003e and MYC-related signaling pathways. Our findings thus provide potential therapeutic targets in CRC.\\u003c/p\\u003e\\n\\u003cp\\u003eUnderstanding the intricate regulatory network between tumor cells and TME cells will advance our knowledge of cancer and facilitating the discovery of therapeutic strategies. Essential genes, which participate in critical signaling and metabolic pathways, may also receive extracellular signals and transmit signals to the TME through ligands/receptors within CCN. To identify potential mediators linking essential genes and TME, we integrated various intracellular regulatory networks with CCN. In the integrated network, signal flow comprises two axes: signal outgoing and signal incoming. In the signal outgoing axis, essential gene is influenced by transcription factors (TFs), DNA mutations, and DNA methylation. Then, these essential genes can impact the ligand/receptor expression of the tumor cells and then transmit signals to the TME. The signal incoming axis involves TME signals being transmitted to receptors on the surface of tumor cells. These signals then influence the expressions of essential genes via biological network. By employing integrative network analysis, we identified six ligands/receptors potentially regulated by over 100 essential genes and TFs, including \\u003cem\\u003eMIF\\u003c/em\\u003e and \\u003cem\\u003eCD47\\u003c/em\\u003e. Elevated expression levels of \\u003cem\\u003eMIF\\u003c/em\\u003e and \\u003cem\\u003eCD47\\u003c/em\\u003e in CRC have been linked to poor prognosis, possibly due to their involvement in evading immune responses of cancer cells\\u003csup\\u003e51,52\\u003c/sup\\u003e. Additionally, we extracted essential genes and TFs associated with \\u003cem\\u003eMIF\\u003c/em\\u003e and \\u003cem\\u003eCD47\\u003c/em\\u003e from the integrated biological network, and discovered several anti-cancer drugs that can modulate the expressions of \\u003cem\\u003eMIF\\u003c/em\\u003e and \\u003cem\\u003eCD47\\u003c/em\\u003e by targeting its associated genes. For instance, sunitinib, a known \\u003cem\\u003eCD47\\u003c/em\\u003e inhibitor, has demonstrated the ability to regulate \\u003cem\\u003eCD47\\u003c/em\\u003e expression in clear cell renal cell carcinoma\\u003csup\\u003e42\\u003c/sup\\u003e. This approach represents a potentially efficient method to interfere essential expression and the ligands/receptors of CCN for the purpose of cancer therapy.\\u003c/p\\u003e\\n\\u003cp\\u003eAlthough comprehensive analysis was performed in this study, there are still some drawbacks. Firstly, owning to the low sequencing depth in 10X genomic scRNA-seq, full-length scRNA-seq data of CRC must be used for constructing a co-expression network. However, there were only 270 tumor cells, further validation in larger-scale data is necessary in future. In addition, although we have elucidated the patterns of expression, DNA mutation, DNA methylation, TF regulation, and TME remodeling for essential genes, the affirmation of how essential genes perform irreplaceable functions in protein levels are urgently required. Finally, further experiments are needed to verify the key role of CRC essential genes in tumorigenesis.\\u003c/p\\u003e\"},{\"header\":\"Materials and Methods\",\"content\":\"\\u003cp\\u003e\\u003cstrong\\u003eGenome-scale CRISPR data collection and processing\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eEight genome-scale CRISPR\\u0026nbsp;loss-of-function\\u0026nbsp;screen datasets of CRC were downloaded from articles and the DepMap database (https://depmap.org/portal/, v20Q1).\\u0026nbsp;Then CERES algorithm, a computational method that estimates gene-dependency levels from CRISPR-Cas9 essentiality screens with consideration of copy number-specific effect, was applied. After that, we obtained the sgRNA depletion score for each gene in every cell line. To integrate the gene list by CERES scores across all cell lines, we utilized the Robust Rank Aggregation algorithm (RRA) in R (“RobustRankAggreg” package), and adjusted \\u003cem\\u003ep\\u003c/em\\u003e-value using “Bonferroni” method. Once we obtained the integrated ranking list, we explored various proportions ranging from 1% to 23% (adjusted \\u003cem\\u003ep\\u003c/em\\u003e-value \\u0026lt; 0.05) to determine the optimized threshold to identify candidate ESPs. The optimized threshold was determined when the maximum value of the overlapping ratios between genes obtained from different ranking proportions and the union pan-cancer essential gene set of Hart's ESGs (1246 genes), Achilles' ESGs (2149 genes), and Zhang's ESGs (799 genes).\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eSingle-cell RNA sequencing data\\u0026nbsp;\\u003c/strong\\u003e\\u003cstrong\\u003ecollection\\u0026nbsp;\\u003c/strong\\u003e\\u003cstrong\\u003eand processing\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eWe obtained\\u0026nbsp;full-length\\u0026nbsp;scRNA-seq data (GSE81861)\\u0026nbsp;from Gene Expression Omnibus\\u0026nbsp;database (GEO), consisting of 272 CRC tumor epithelial cells and 160 normal epithelial cells. The gene expression matrices of tumor and normal epithelial cells were combined and converted into a Seurat object using the Seurat R package for downstream analysis\\u0026nbsp;\\u003csup\\u003e53\\u003c/sup\\u003e.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eWe also obtained 10X Genomics scRNA-seq data (GSE132465) from GEO, consisting of 56,465 cells from 23 patients with primary CRC samples and 10 matched normal mucosae samples. We performed quality filtering to remove cells with less than 500 expressed genes, less than 500 unique molecular identifiers (UMIs), more than 20% UMIs derived from mitochondrial genes, and log\\u003csub\\u003e10\\u003c/sub\\u003e(expressed genes / UMIs) greater than 0.78. Then, we normalized and scaled the gene expression matrices using\\u0026nbsp;SCTransform. Subsequently, harmony algorithm was employed to integrate scRNA-seq data across different patients\\u0026nbsp;\\u003csup\\u003e54\\u003c/sup\\u003e. For cell type identification, we first applied a graph-based clustering approach by using FindNeighbors and FindClusters function. Then the resulting clusters were further annotated into seven major cell types: epithelial cells (\\u003cem\\u003eEPCAM\\u003c/em\\u003e), myeloid cells (\\u003cem\\u003eCD14\\u003c/em\\u003e), T cells (\\u003cem\\u003eCD3D\\u003c/em\\u003e), B cells (\\u003cem\\u003eCD79A\\u003c/em\\u003e), fibroblast cells (\\u003cem\\u003eCOL1A1\\u003c/em\\u003e), endothelial cells (\\u003cem\\u003ePECAM1\\u003c/em\\u003e), and mast cells (\\u003cem\\u003eTPSAB1\\u003c/em\\u003e).\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eThe differentially expressed essential genes in different group or cell lineages were calculated using the Seurat FindMarkers function with “MAST” algorithm. The genes with default parameters of log\\u003csub\\u003e2\\u003c/sub\\u003eFC \\u0026gt; 0.5 and adjusted \\u003cem\\u003ep\\u003c/em\\u003e-value \\u0026lt; 0.05 were considered as\\u0026nbsp;overexpressed genes. Specifically expressed genes in each cell type were defined as those detected in certain cell type at a percentage three times higher than other cell types. To assess the gene set score fore each cell type,\\u0026nbsp;AddModuleScore function in Seurat was applied.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eCell culture and CRISPR/Cas9 knockout\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eHCT116 (human colon carcinoma cell line, Ubigene, YC-C004) were cultured in RPMI-1640 (Thermo Fisher Scientific) with 10% fetal bovine serum (FBS; VISTECH) and 1% penicillin-streptomycin (Thermo Fisher Scientific). HEK293T cells (ATCC CRL3216) were cultured in DMEM (VISTECH) with 10% FBS and 1% penicillin-streptomycin. All cell lines were maintained at 37°C in a humidified incubator with 5% CO2.\\u003c/p\\u003e\\n\\u003cp\\u003eTo clone individual gene targeting single-guide RNAs (sgRNAs), the lentiviral vector (pSLQ1373) was digested with BlpⅠ and BstXⅠ, and gel-purified\\u003csup\\u003e55\\u003c/sup\\u003e. For each gene, two sgRNA were designed on the CHOPCHOP website\\u0026nbsp;\\u003csup\\u003e56\\u003c/sup\\u003e (\\u003cstrong\\u003eSupplementary Table 11\\u003c/strong\\u003e). SgRNA fragments were synthesized by Tsingke (Tsingke Biotechnology Co., Ltd.) as forward and reverse primers (Table S1), which were then annealed, gel-purified, and ligated to the linearized pSLQ1373 vector by homologous recombination (ClonExpress Ultra One Step Cloning Kit, Vazyme).\\u003c/p\\u003e\\n\\u003cp\\u003eTo produce lentivirus, HEK293T cells were transiently transfected with polyethylenimine (PEI MAX, Polysciences) and packaging plasmids psPAX2 and pMD2.0G. The ratio of target plasmid to psPAX2 and pMD2.0G is 5:4:1. The ratio of the total mass of the added plasmid (μg) to the PEI MAX (μl) is 1:3. Lentivirus was collected by filtering the supernatant through a 0.45-μm filter 48 hours after transfection. The lentiviruses were collected and stored at -80°C.\\u003c/p\\u003e\\n\\u003cp\\u003eHCT116 cells were treated with 15 μg/μl BSD for 5 days to select Cas9-expressing cells, after being infected with EFS-spCas9-BSD lentivirus for 48 hours. The purified cells were sorted into 96-well plates using a flow cytometry sorter CytoFLEX SRT (BECKMAN COULTER). After 12-14 days, the well-growing clonal cell population in the 96-well plate was picked out, expanded, and frozen.\\u003c/p\\u003e\\n\\u003cp\\u003eHCT116-Cas9 was infected with a lentivirus construct expressing individual sgRNA, or two sgRNAs simultaneously for lncRNA knock-out. After 72 hours of infection, the infected cells were treated with 2 μg/ml puromycin for 24 hours and then recovered in fresh culture medium without puromycin for 48 hours.\\u003c/p\\u003e\\n\\u003cp\\u003eGene knock-out cells were counted and seeded to a 12-well plate at 0.1 million cells/well. After 72 hours of growth, cells were uniformly passaged at a 1:10 and grown for another 72 hours. Cells were then collected, washed, and resuspended in phosphate-buffered saline (PBS). Ten microliters of the cell suspension were mixed with 0.4% trypan blue solution (Solarbio) at a 1:1 ratio and counted on a hemocytometer (Hirschmann). Each sample was counted three times to obtain a mean cell number.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eCollection\\u0026nbsp;\\u003c/strong\\u003e\\u003cstrong\\u003eand processing of gene\\u0026nbsp;\\u003c/strong\\u003e\\u003cstrong\\u003eexpression, DNA mutation, and DNA methylation data\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eWe collected and combined the expression profiles, DNA mutation data, and DNA methylation data of TCGA-COAD and TCGA-READ from the UCSC data portal. The PCAWG mutation data of CRC was collected from International Cancer Genome Consortium (ICGC). The mutation data was processed, analyzed, and visualized using Maftools package in R. The DNA methylation data were processed and visualized by using ChAMP package in R. Differentially methylated positions (DMPs) were identified by champ.DMP function in ChAMP with adjusted \\u003cem\\u003ep\\u003c/em\\u003e-value \\u0026lt; 0.05 and absolute log\\u003csub\\u003e2\\u003c/sub\\u003eFC \\u0026gt; 0.2. The Pearson correlation coefficients between the expression and DNA methylation levels of each gene were calculated to highlight the association of them.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eCo-expression network construction\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eTwo-color network with both protein-coding genes and lncRNAs were constructed based on bulk RNA-seq from TCGA without normal samples and\\u0026nbsp;full-length\\u0026nbsp;scRNA-seq data of tumor epithelial cells from\\u0026nbsp;GSE81861,\\u0026nbsp;respectively. For the construction of an unweighted co-expression network, the gene Pearson correlation coefficients were calculated. Fisher’s asymptotic test from WGCNA packages\\u003csup\\u003e57\\u003c/sup\\u003e and FDR correction were used to screen reliable relationships. Eventually, only the pairs with adjusted \\u003cem\\u003ep\\u003c/em\\u003e-value lower than 0.01 and absolute value of correlation coefficient greater than 0.3 were reserved and further converted to undirected network. We also constructed topological overlap weighted co-expression network by using TOMsimilarityFromExpr function from WGCNA with parameter as follow: soft-thresholding power of 6 , Pearson correlation coefficient, and TOMDenom specifying as mean. Finally, the pairs with weight greater than 0 were exported and further converted to undirected network.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eHypergeometrics Test and Personalized PageRank\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eAs described\\u0026nbsp;previously, we performed\\u0026nbsp;HT algorithm (\\u003cstrong\\u003eFig. 2a\\u003c/strong\\u003e) to evaluate the degree of overlap between neighboring genes of a query gene in an unweighted co-expression network and ESPs\\u0026nbsp;\\u003csup\\u003e12\\u003c/sup\\u003e. For each lncRNA,\\u0026nbsp;\\u003cem\\u003ep\\u003c/em\\u003e-values were calculated and adjusted using FDR correction, and were used to evaluate the possibility as ESLs. Candidate ESLs were defined as lncRNAs with adjusted \\u003cem\\u003ep\\u003c/em\\u003e-values lower than 0.01.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003ePPR is a well-known method derived from random walk with restart and can compute ranking score for each gene using multiple seeds (https://github.com/jinhongjung/pyrwr) in\\u0026nbsp;network. In this study, ESPs were designated as the seeds to calculate the ranking score of each lncRNA in the\\u0026nbsp;topological overlap\\u0026nbsp;weighted co-expression network. Only the top 5% of lncRNAs in the PPR results were considered as potential ESL candidates.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eInferring TF regulatory network of CRC scRNA-seq data\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eWe utilized the R implementation of the SCENIC pipeline to infer TF regulatory network between transcription factors and essential genes. Two gene-motif rankings databases were employed to determine the \\u003cem\\u003ecis\\u003c/em\\u003e-acting element around the transcription start site (TSS), including 10 kb around the TSS and 500 bp upstream of the TSS. The gene-motif annotations from cisTarget databases can be categorized as either high-confidence or low-confidence. The high-confidence annotations are “direct annotation” and “inferred by orthology” in\\u0026nbsp;the\\u0026nbsp;annotation source of gene-motif, while low-confidence annotations are “inferred by motif similarity”. The regulon named with the sufix “-extended” indicate lower confidence annotation. To assess the differential AUC score of the regulon between tumor and normal epithelial cells,\\u003cem\\u003e\\u0026nbsp;t\\u003c/em\\u003e-test was conducted.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eCCN network construction and integration\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eWe employed CellPhoneDB to construct the CCN network of CRC based on\\u0026nbsp;10X Genomics scRNA data (GSE132465). For downstream analysis, we only considered the ligand-receptor pairs with \\u003cem\\u003ep\\u003c/em\\u003e-value lower than 0.05. We then\\u0026nbsp;integrating multiple intracellular regulatory networks, including\\u0026nbsp;co-expressed network,\\u0026nbsp;TF regulatory network, and CCN of tumor epithelial cells.\\u0026nbsp;First, we filtered the top 100 co-expressed pairs of each ESP and\\u0026nbsp;the pairs with weight great than 0.001 in\\u0026nbsp;WCN-TEC. Second, we filtered\\u0026nbsp;relationships associated with ESPs and ESLs from TF regulatory network.\\u0026nbsp;Finally, we integrated those two intracellular regulatory networks with CCN, the source node of integrated network are ESPs, ESLs, or TFs, while the target node are ligands or receptors of tumor epithelial cells.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eFunction enrichment and survival analysis\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eGO, KEGG, and HALLMARK function analyses of protein-coding genes were carried out by Metascape\\u0026nbsp;\\u003csup\\u003e58\\u003c/sup\\u003e. To determine\\u0026nbsp;the functions of lncRNAs, we first checked the existing reports from Lnc2Cancer 3.0. Second, top 200 associated genes for each ESL based on weighted co-expression network of tumor epithelial cells were selected. Then, GO biological process enrichment analysis was performed using “ClusterProfiler” R package. Additionally, we conducted survival analysis using GEPIA\\u0026nbsp;\\u003csup\\u003e59\\u003c/sup\\u003e, an interactive web server to analyze RNA sequencing data from TGCA.\\u003c/p\\u003e\"},{\"header\":\"Declarations\",\"content\":\"\\u003cp\\u003e\\u003cstrong\\u003eCRediT author statement and competing interests\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eYanguo Li\\u003c/strong\\u003e: Methodology, Formal analysis, Software, Data Curation, Visualization, Writing - Original draft preparation. \\u003cstrong\\u003eTianci Han\\u003c/strong\\u003e: Investigation, Formal analysis, Validation. \\u003cstrong\\u003eChengjiang Fan\\u003c/strong\\u003e: Investigation. \\u003cstrong\\u003eHao Rong\\u003c/strong\\u003e: Validation, Software. \\u003cstrong\\u003eChen Yu\\u003c/strong\\u003e: Investigation, Resources, Writing - Reviewing and Editing, Funding acquisition. \\u003cstrong\\u003eYang Xi and Huifang Wang\\u003c/strong\\u003e: Resources, Writing - Reviewing and Editing, Funding acquisition. \\u003cstrong\\u003eQi Liao\\u003c/strong\\u003e: Conceptualization, Project administration, Methodology, Supervision, Writing - Reviewing and Editing, Funding acquisition.\\u003c/p\\u003e\\n\\u003cp\\u003eThe authors declare no competing interests.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eAcknowledgements\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eThis work was supported by the Zhejiang Provincial Natural Science Foundation of China (No. LY21C060002, No. LY21C060001); the National Natural Science Foundation of China (No. 31970630); the Ningbo Natural Science Foundation of China (No. 2021J124), the Fundamental Research Funds for the Provincial Universities of Zhejiang (No. SJLZ2021001); Guangdong Basic and Applied Basic Research Foundation (No. 2023A1515011075); Shenzhen Bay Laboratory (No. 21250071); The Keynote Research Project of Ningbo City (No. 2023Z171); Science and Technology Project of Ningbo(No. 2020F032).\\u003c/p\\u003e\"},{\"header\":\"References\",\"content\":\"\\u003col\\u003e\\n\\u003cli\\u003eSung, H.\\u003cem\\u003e et al.\\u003c/em\\u003e Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. \\u003cem\\u003eCA Cancer J Clin\\u003c/em\\u003e \\u003cstrong\\u003e71\\u003c/strong\\u003e, 209-249, doi:10.3322/caac.21660 (2021).\\u003c/li\\u003e\\n\\u003cli\\u003eMcQuade, R. M., Stojanovska, V., Bornstein, J. C. \\u0026amp; Nurgali, K. Colorectal Cancer Chemotherapy: The Evolution of Treatment and New Approaches. \\u003cem\\u003eCurr Med Chem\\u003c/em\\u003e \\u003cstrong\\u003e24\\u003c/strong\\u003e, 1537-1557, doi:10.2174/0929867324666170111152436 (2017).\\u003c/li\\u003e\\n\\u003cli\\u003eWang, T.\\u003cem\\u003e et al.\\u003c/em\\u003e Identification and characterization of essential genes in the human genome. \\u003cem\\u003eScience (New York, N.Y.)\\u003c/em\\u003e \\u003cstrong\\u003e350\\u003c/strong\\u003e, 1096-1101, doi:10.1126/science.aac7041 (2015).\\u003c/li\\u003e\\n\\u003cli\\u003eShalem, O.\\u003cem\\u003e et al.\\u003c/em\\u003e Genome-scale CRISPR-Cas9 knockout screening in human cells. \\u003cem\\u003eScience (New York, N.Y.)\\u003c/em\\u003e \\u003cstrong\\u003e343\\u003c/strong\\u003e, 84-87, doi:10.1126/science.1247005 (2014).\\u003c/li\\u003e\\n\\u003cli\\u003eAlda-Catalinas, C.\\u003cem\\u003e et al.\\u003c/em\\u003e A Single-Cell Transcriptomics CRISPR-Activation Screen Identifies Epigenetic Regulators of the Zygotic Genome Activation Program. \\u003cem\\u003eCell Syst\\u003c/em\\u003e \\u003cstrong\\u003e11\\u003c/strong\\u003e, 25-41 e29, doi:10.1016/j.cels.2020.06.004 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eLin, K. H.\\u003cem\\u003e et al.\\u003c/em\\u003e Using antagonistic pleiotropy to design a chemotherapy-induced evolutionary trap to target drug resistance in cancer. \\u003cem\\u003eNat Genet\\u003c/em\\u003e \\u003cstrong\\u003e52\\u003c/strong\\u003e, 408-417, doi:10.1038/s41588-020-0590-9 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eCortez, J. T.\\u003cem\\u003e et al.\\u003c/em\\u003e CRISPR screen in regulatory T cells reveals modulators of Foxp3. \\u003cem\\u003eNature\\u003c/em\\u003e \\u003cstrong\\u003e582\\u003c/strong\\u003e, 416-420, doi:10.1038/s41586-020-2246-4 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eManguso, R. T.\\u003cem\\u003e et al.\\u003c/em\\u003e In vivo CRISPR screening identifies Ptpn2 as a cancer immunotherapy target. \\u003cem\\u003eNature\\u003c/em\\u003e \\u003cstrong\\u003e547\\u003c/strong\\u003e, 413-418, doi:10.1038/nature23270 (2017).\\u003c/li\\u003e\\n\\u003cli\\u003eHart, T., Brown, K. R., Sircoulomb, F., Rottapel, R. \\u0026amp; Moffat, J. Measuring error rates in genomic perturbation screens: gold standards for human functional genomics. \\u003cem\\u003eMol Syst Biol\\u003c/em\\u003e \\u003cstrong\\u003e10\\u003c/strong\\u003e, 733, doi:10.15252/msb.20145216 (2014).\\u003c/li\\u003e\\n\\u003cli\\u003eHart, T.\\u003cem\\u003e et al.\\u003c/em\\u003e Evaluation and Design of Genome-Wide CRISPR/SpCas9 Knockout Screens. \\u003cem\\u003eG3 (Bethesda)\\u003c/em\\u003e \\u003cstrong\\u003e7\\u003c/strong\\u003e, 2719-2727, doi:10.1534/g3.117.041277 (2017).\\u003c/li\\u003e\\n\\u003cli\\u003eMeyers, R. M.\\u003cem\\u003e et al.\\u003c/em\\u003e Computational correction of copy number effect improves specificity of CRISPR-Cas9 essentiality screens in cancer cells. \\u003cem\\u003eNat Genet\\u003c/em\\u003e \\u003cstrong\\u003e49\\u003c/strong\\u003e, 1779-1784, doi:10.1038/ng.3984 (2017).\\u003c/li\\u003e\\n\\u003cli\\u003eZhang, Y.\\u003cem\\u003e et al.\\u003c/em\\u003e Genome-wide identification of the essential protein-coding genes and long non-coding RNAs for human pan-cancer. \\u003cem\\u003eBioinformatics\\u003c/em\\u003e \\u003cstrong\\u003e35\\u003c/strong\\u003e, 4344-4349, doi:10.1093/bioinformatics/btz230 (2019).\\u003c/li\\u003e\\n\\u003cli\\u003eMarcotte, R.\\u003cem\\u003e et al.\\u003c/em\\u003e Essential gene profiles in breast, pancreatic, and ovarian cancer cells. \\u003cem\\u003eCancer Discov\\u003c/em\\u003e \\u003cstrong\\u003e2\\u003c/strong\\u003e, 172-189, doi:10.1158/2159-8290.CD-11-0224 (2012).\\u003c/li\\u003e\\n\\u003cli\\u003eZhang, L.\\u003cem\\u003e et al.\\u003c/em\\u003e Single-Cell Analyses Inform Mechanisms of Myeloid-Targeted Therapies in Colon Cancer. \\u003cem\\u003eCell\\u003c/em\\u003e \\u003cstrong\\u003e181\\u003c/strong\\u003e, 442-459 e429, doi:10.1016/j.cell.2020.03.048 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eLee, H. O.\\u003cem\\u003e et al.\\u003c/em\\u003e Lineage-dependent gene expression programs influence the immune landscape of colorectal cancer. \\u003cem\\u003eNat Genet\\u003c/em\\u003e \\u003cstrong\\u003e52\\u003c/strong\\u003e, 594-603, doi:10.1038/s41588-020-0636-z (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eReplogle, J. M.\\u003cem\\u003e et al.\\u003c/em\\u003e Combinatorial single-cell CRISPR screens by direct guide RNA capture and targeted sequencing. \\u003cem\\u003eNat Biotechnol\\u003c/em\\u003e \\u003cstrong\\u003e38\\u003c/strong\\u003e, 954-961, doi:10.1038/s41587-020-0470-y (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eGenga, R. M. J.\\u003cem\\u003e et al.\\u003c/em\\u003e Single-Cell RNA-Sequencing-Based CRISPRi Screening Resolves Molecular Drivers of Early Human Endoderm Development. \\u003cem\\u003eCell Rep\\u003c/em\\u003e \\u003cstrong\\u003e27\\u003c/strong\\u003e, 708-718 e710, doi:10.1016/j.celrep.2019.03.076 (2019).\\u003c/li\\u003e\\n\\u003cli\\u003eVishnubalaji, R. \\u0026amp; Alajez, N. M. Single-Cell Transcriptome Analysis Revealed Heterogeneity and Identified Novel Therapeutic Targets for Breast Cancer Subtypes. \\u003cem\\u003eCells\\u003c/em\\u003e \\u003cstrong\\u003e12\\u003c/strong\\u003e, doi:10.3390/cells12081182 (2023).\\u003c/li\\u003e\\n\\u003cli\\u003eKolde, R., Laur, S., Adler, P. \\u0026amp; Vilo, J. Robust rank aggregation for gene list integration and meta-analysis. \\u003cem\\u003eBioinformatics\\u003c/em\\u003e \\u003cstrong\\u003e28\\u003c/strong\\u003e, 573-580, doi:10.1093/bioinformatics/btr709 (2012).\\u003c/li\\u003e\\n\\u003cli\\u003eLi, H.\\u003cem\\u003e et al.\\u003c/em\\u003e Reference component analysis of single-cell transcriptomes elucidates cellular heterogeneity in human colorectal tumors. \\u003cem\\u003eNat Genet\\u003c/em\\u003e \\u003cstrong\\u003e49\\u003c/strong\\u003e, 708-718, doi:10.1038/ng.3818 (2017).\\u003c/li\\u003e\\n\\u003cli\\u003eGerstung, M.\\u003cem\\u003e et al.\\u003c/em\\u003e The evolutionary history of 2,658 cancers. \\u003cem\\u003eNature\\u003c/em\\u003e \\u003cstrong\\u003e578\\u003c/strong\\u003e, 122-128, doi:10.1038/s41586-019-1907-7 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eWang, J., Meng, F. \\u0026amp; Mao, F. Single cell sequencing analysis and transcriptome analysis constructed the liquid-liquid phase separation(LLPS)-related prognostic model for endometrial cancer. \\u003cem\\u003eFront Oncol\\u003c/em\\u003e \\u003cstrong\\u003e12\\u003c/strong\\u003e, 1005472, doi:10.3389/fonc.2022.1005472 (2022).\\u003c/li\\u003e\\n\\u003cli\\u003eChen, S. \\u0026amp; Shen, X. Long noncoding RNAs: functions and mechanisms in colon cancer. \\u003cem\\u003eMol Cancer\\u003c/em\\u003e \\u003cstrong\\u003e19\\u003c/strong\\u003e, 167, doi:10.1186/s12943-020-01287-2 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eGao, Y.\\u003cem\\u003e et al.\\u003c/em\\u003e Lnc2Cancer 3.0: an updated resource for experimentally supported lncRNA/circRNA cancer associations and web tools based on RNA-seq and scRNA-seq data. \\u003cem\\u003eNucleic Acids Res\\u003c/em\\u003e \\u003cstrong\\u003e49\\u003c/strong\\u003e, D1251-D1258, doi:10.1093/nar/gkaa1006 (2021).\\u003c/li\\u003e\\n\\u003cli\\u003eScott, M. K. D.\\u003cem\\u003e et al.\\u003c/em\\u003e A multi-scale integrated analysis identifies KRT8 as a pan-cancer early biomarker. \\u003cem\\u003ePac Symp Biocomput\\u003c/em\\u003e \\u003cstrong\\u003e26\\u003c/strong\\u003e, 297-308 (2021).\\u003c/li\\u003e\\n\\u003cli\\u003eTrop-Steinberg, S. \\u0026amp; Azar, Y. Is Myc an Important Biomarker? Myc Expression in Immune Disorders and Cancer. \\u003cem\\u003eAm J Med Sci\\u003c/em\\u003e \\u003cstrong\\u003e355\\u003c/strong\\u003e, 67-75, doi:10.1016/j.amjms.2017.06.007 (2018).\\u003c/li\\u003e\\n\\u003cli\\u003eVriens, K.\\u003cem\\u003e et al.\\u003c/em\\u003e Evidence for an alternative fatty acid desaturation pathway increasing cancer plasticity. \\u003cem\\u003eNature\\u003c/em\\u003e \\u003cstrong\\u003e566\\u003c/strong\\u003e, 403-406, doi:10.1038/s41586-019-0904-1 (2019).\\u003c/li\\u003e\\n\\u003cli\\u003eKrushkal, J.\\u003cem\\u003e et al.\\u003c/em\\u003e Epigenome-wide DNA methylation analysis of small cell lung cancer cell lines suggests potential chemotherapy targets. \\u003cem\\u003eClin Epigenetics\\u003c/em\\u003e \\u003cstrong\\u003e12\\u003c/strong\\u003e, 93, doi:10.1186/s13148-020-00876-8 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eTseng, Y. Y.\\u003cem\\u003e et al.\\u003c/em\\u003e PVT1 dependence in cancer with MYC copy-number increase. \\u003cem\\u003eNature\\u003c/em\\u003e \\u003cstrong\\u003e512\\u003c/strong\\u003e, 82-86, doi:10.1038/nature13311 (2014).\\u003c/li\\u003e\\n\\u003cli\\u003eAibar, S.\\u003cem\\u003e et al.\\u003c/em\\u003e SCENIC: single-cell regulatory network inference and clustering. \\u003cem\\u003eNat Methods\\u003c/em\\u003e \\u003cstrong\\u003e14\\u003c/strong\\u003e, 1083-1086, doi:10.1038/nmeth.4463 (2017).\\u003c/li\\u003e\\n\\u003cli\\u003eZhou, X.\\u003cem\\u003e et al.\\u003c/em\\u003e BCLAF1 and its splicing regulator SRSF10 regulate the tumorigenic potential of colon cancer cells. \\u003cem\\u003eNat Commun\\u003c/em\\u003e \\u003cstrong\\u003e5\\u003c/strong\\u003e, 4581, doi:10.1038/ncomms5581 (2014).\\u003c/li\\u003e\\n\\u003cli\\u003eZhu, L.\\u003cem\\u003e et al.\\u003c/em\\u003e Expression and crucial role of long non-coding RNA FGD5-AS1 in human cancers. \\u003cem\\u003eAm J Transl Res\\u003c/em\\u003e \\u003cstrong\\u003e13\\u003c/strong\\u003e, 10964-10976 (2021).\\u003c/li\\u003e\\n\\u003cli\\u003eLi, D.\\u003cem\\u003e et al.\\u003c/em\\u003e Long noncoding RNA FGD5-AS1 promotes colorectal cancer cell proliferation, migration, and invasion through upregulating CDCA7 via sponging miR-302e. \\u003cem\\u003eIn Vitro Cell Dev Biol Anim\\u003c/em\\u003e \\u003cstrong\\u003e55\\u003c/strong\\u003e, 577-585, doi:10.1007/s11626-019-00376-x (2019).\\u003c/li\\u003e\\n\\u003cli\\u003eKim, S. H.\\u003cem\\u003e et al.\\u003c/em\\u003e Kruppel-Like Factor 12 Promotes Colorectal Cancer Growth through Early Growth Response Protein 1. \\u003cem\\u003ePLoS One\\u003c/em\\u003e \\u003cstrong\\u003e11\\u003c/strong\\u003e, e0159899, doi:10.1371/journal.pone.0159899 (2016).\\u003c/li\\u003e\\n\\u003cli\\u003eJiang, X.\\u003cem\\u003e et al.\\u003c/em\\u003e Thapsigargin promotes colorectal cancer cell migration through upregulation of lncRNA MALAT1. \\u003cem\\u003eOncol Rep\\u003c/em\\u003e \\u003cstrong\\u003e43\\u003c/strong\\u003e, 1245-1255, doi:10.3892/or.2020.7502 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eEfremova, M., Vento-Tormo, M., Teichmann, S. A. \\u0026amp; Vento-Tormo, R. CellPhoneDB: inferring cell-cell communication from combined expression of multi-subunit ligand-receptor complexes. \\u003cem\\u003eNat Protoc\\u003c/em\\u003e \\u003cstrong\\u003e15\\u003c/strong\\u003e, 1484-1506, doi:10.1038/s41596-020-0292-x (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eFigueiredo, C. R.\\u003cem\\u003e et al.\\u003c/em\\u003e Blockade of MIF-CD74 Signalling on Macrophages and Dendritic Cells Restores the Antitumour Immune Response Against Metastatic Melanoma. \\u003cem\\u003eFrontiers in immunology\\u003c/em\\u003e \\u003cstrong\\u003e9\\u003c/strong\\u003e, 1132, doi:10.3389/fimmu.2018.01132 (2018).\\u003c/li\\u003e\\n\\u003cli\\u003eKaur, S., Isenberg, J. S. \\u0026amp; Roberts, D. D. CD47 (Cluster of Differentiation 47). \\u003cem\\u003eAtlas Genet Cytogenet Oncol Haematol\\u003c/em\\u003e \\u003cstrong\\u003e25\\u003c/strong\\u003e, 83-102 (2021).\\u003c/li\\u003e\\n\\u003cli\\u003ePilarczyk, M.\\u003cem\\u003e et al.\\u003c/em\\u003e Connecting omics signatures and revealing biological mechanisms with iLINCS. \\u003cem\\u003eNat Commun\\u003c/em\\u003e \\u003cstrong\\u003e13\\u003c/strong\\u003e, 4678, doi:10.1038/s41467-022-32205-3 (2022).\\u003c/li\\u003e\\n\\u003cli\\u003eMonks, A.\\u003cem\\u003e et al.\\u003c/em\\u003e The NCI Transcriptional Pharmacodynamics Workbench: A Tool to Examine Dynamic Expression Profiling of Therapeutic Response in the NCI-60 Cell Line Panel. \\u003cem\\u003eCancer research\\u003c/em\\u003e \\u003cstrong\\u003e78\\u003c/strong\\u003e, 6807-6817, doi:10.1158/0008-5472.CAN-18-0989 (2018).\\u003c/li\\u003e\\n\\u003cli\\u003eDemetri, G. D.\\u003cem\\u003e et al.\\u003c/em\\u003e Efficacy and safety of sunitinib in patients with advanced gastrointestinal stromal tumour after failure of imatinib: a randomised controlled trial. \\u003cem\\u003eLancet\\u003c/em\\u003e \\u003cstrong\\u003e368\\u003c/strong\\u003e, 1329-1338, doi:10.1016/S0140-6736(06)69446-4 (2006).\\u003c/li\\u003e\\n\\u003cli\\u003eJiang, W.\\u003cem\\u003e et al.\\u003c/em\\u003e Immune inactivation by CD47 expression predicts clinical outcomes and therapeutic responses in clear cell renal cell carcinoma patients. \\u003cem\\u003eUrol Oncol\\u003c/em\\u003e \\u003cstrong\\u003e40\\u003c/strong\\u003e, 166 e115-166 e125, doi:10.1016/j.urolonc.2021.11.024 (2022).\\u003c/li\\u003e\\n\\u003cli\\u003eDempster, J. M.\\u003cem\\u003e et al.\\u003c/em\\u003e Extracting Biological Insights from the Project Achilles Genome-Scale CRISPR Screens in Cancer Cell Lines. 720243, doi:10.1101/720243 %J bioRxiv (2019).\\u003c/li\\u003e\\n\\u003cli\\u003eOughtred, R.\\u003cem\\u003e et al.\\u003c/em\\u003e The BioGRID database: A comprehensive biomedical resource of curated protein, genetic, and chemical interactions. \\u003cem\\u003eProtein Sci\\u003c/em\\u003e \\u003cstrong\\u003e30\\u003c/strong\\u003e, 187-200, doi:10.1002/pro.3978 (2021).\\u003c/li\\u003e\\n\\u003cli\\u003eChang, L., Ruiz, P., Ito, T. \\u0026amp; Sellers, W. R. Targeting pan-essential genes in cancer: Challenges and opportunities. \\u003cem\\u003eCancer Cell\\u003c/em\\u003e \\u003cstrong\\u003e39\\u003c/strong\\u003e, 466-479, doi:10.1016/j.ccell.2020.12.008 (2021).\\u003c/li\\u003e\\n\\u003cli\\u003eLiu, S. J.\\u003cem\\u003e et al.\\u003c/em\\u003e CRISPRi-based genome-scale identification of functional long noncoding RNA loci in human cells. \\u003cem\\u003eScience (New York, N.Y.)\\u003c/em\\u003e \\u003cstrong\\u003e355\\u003c/strong\\u003e, doi:10.1126/science.aah7111 (2017).\\u003c/li\\u003e\\n\\u003cli\\u003eLi, X., Li, W., Zeng, M., Zheng, R. \\u0026amp; Li, M. Network-based methods for predicting essential genes or proteins: a survey. \\u003cem\\u003eBrief Bioinform\\u003c/em\\u003e \\u003cstrong\\u003e21\\u003c/strong\\u003e, 566-583, doi:10.1093/bib/bbz017 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eLiu, L.\\u003cem\\u003e et al.\\u003c/em\\u003e lncRNA GAS5 Inhibits Cell Migration and Invasion and Promotes Autophagy by Targeting miR-222-3p via the GAS5/PTEN-Signaling Pathway in CRC. \\u003cem\\u003eMol Ther Nucleic Acids\\u003c/em\\u003e \\u003cstrong\\u003e17\\u003c/strong\\u003e, 644-656, doi:10.1016/j.omtn.2019.06.009 (2019).\\u003c/li\\u003e\\n\\u003cli\\u003eNi, W.\\u003cem\\u003e et al.\\u003c/em\\u003e Long noncoding RNA GAS5 inhibits progression of colorectal cancer by interacting with and triggering YAP phosphorylation and degradation and is negatively regulated by the m(6)A reader YTHDF3. \\u003cem\\u003eMol Cancer\\u003c/em\\u003e \\u003cstrong\\u003e18\\u003c/strong\\u003e, 143, doi:10.1186/s12943-019-1079-y (2019).\\u003c/li\\u003e\\n\\u003cli\\u003eShigeyasu, K.\\u003cem\\u003e et al.\\u003c/em\\u003e The PVT1 lncRNA is a novel epigenetic enhancer of MYC, and a promising risk-stratification biomarker in colorectal cancer. \\u003cem\\u003eMol Cancer\\u003c/em\\u003e \\u003cstrong\\u003e19\\u003c/strong\\u003e, 155, doi:10.1186/s12943-020-01277-4 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eLogtenberg, M. E. W., Scheeren, F. A. \\u0026amp; Schumacher, T. N. The CD47-SIRPalpha Immune Checkpoint. \\u003cem\\u003eImmunity\\u003c/em\\u003e \\u003cstrong\\u003e52\\u003c/strong\\u003e, 742-752, doi:10.1016/j.immuni.2020.04.011 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eOtvos, B.\\u003cem\\u003e et al.\\u003c/em\\u003e Cancer Stem Cell-Secreted Macrophage Migration Inhibitory Factor Stimulates Myeloid Derived Suppressor Cell Function and Facilitates Glioblastoma Immune Evasion. \\u003cem\\u003eStem Cells\\u003c/em\\u003e \\u003cstrong\\u003e34\\u003c/strong\\u003e, 2026-2039, doi:10.1002/stem.2393 (2016).\\u003c/li\\u003e\\n\\u003cli\\u003eStuart, T.\\u003cem\\u003e et al.\\u003c/em\\u003e Comprehensive Integration of Single-Cell Data. \\u003cem\\u003eCell\\u003c/em\\u003e \\u003cstrong\\u003e177\\u003c/strong\\u003e, 1888-1902 e1821, doi:10.1016/j.cell.2019.05.031 (2019).\\u003c/li\\u003e\\n\\u003cli\\u003eKorsunsky, I.\\u003cem\\u003e et al.\\u003c/em\\u003e Fast, sensitive and accurate integration of single-cell data with Harmony. \\u003cem\\u003eNat Methods\\u003c/em\\u003e \\u003cstrong\\u003e16\\u003c/strong\\u003e, 1289-1296, doi:10.1038/s41592-019-0619-0 (2019).\\u003c/li\\u003e\\n\\u003cli\\u003eLiu, Y.\\u003cem\\u003e et al.\\u003c/em\\u003e CRISPR Activation Screens Systematically Identify Factors that Drive Neuronal Fate and Reprogramming. \\u003cem\\u003eCell Stem Cell\\u003c/em\\u003e \\u003cstrong\\u003e23\\u003c/strong\\u003e, 758-771 e758, doi:10.1016/j.stem.2018.09.003 (2018).\\u003c/li\\u003e\\n\\u003cli\\u003eLabun, K.\\u003cem\\u003e et al.\\u003c/em\\u003e CHOPCHOP v3: expanding the CRISPR web toolbox beyond genome editing. \\u003cem\\u003eNucleic Acids Res\\u003c/em\\u003e \\u003cstrong\\u003e47\\u003c/strong\\u003e, W171-W174, doi:10.1093/nar/gkz365 (2019).\\u003c/li\\u003e\\n\\u003cli\\u003eLangfelder, P. \\u0026amp; Horvath, S. WGCNA: an R package for weighted correlation network analysis. \\u003cem\\u003eBMC bioinformatics\\u003c/em\\u003e \\u003cstrong\\u003e9\\u003c/strong\\u003e, 559, doi:10.1186/1471-2105-9-559 (2008).\\u003c/li\\u003e\\n\\u003cli\\u003eZhou, Y.\\u003cem\\u003e et al.\\u003c/em\\u003e Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. \\u003cem\\u003eNat Commun\\u003c/em\\u003e \\u003cstrong\\u003e10\\u003c/strong\\u003e, 1523, doi:10.1038/s41467-019-09234-6 (2019).\\u003c/li\\u003e\\n\\u003cli\\u003eTang, Z.\\u003cem\\u003e et al.\\u003c/em\\u003e GEPIA: a web server for cancer and normal gene expression profiling and interactive analyses. \\u003cem\\u003eNucleic Acids Res\\u003c/em\\u003e \\u003cstrong\\u003e45\\u003c/strong\\u003e, W98-W102, doi:10.1093/nar/gkx247 (2017).\\u003c/li\\u003e\\n\\u003c/ol\\u003e\"}],\"fulltextSource\":\"\",\"fullText\":\"\",\"funders\":[],\"hasAdminPriorityOnWorkflow\":false,\"hasManuscriptDocX\":true,\"hasOptedInToPreprint\":true,\"hasPassedJournalQc\":\"\",\"hasAnyPriority\":false,\"hideJournal\":true,\"highlight\":\"\",\"institution\":\"\",\"isAcceptedByJournal\":false,\"isAuthorSuppliedPdf\":false,\"isDeskRejected\":\"\",\"isHiddenFromSearch\":false,\"isInQc\":false,\"isInWorkflow\":false,\"isPdf\":false,\"isPdfUpToDate\":true,\"isWithdrawnOrRetracted\":false,\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true},\"keywords\":\"Essential protein-coding genes, essential long non-coding genes, single-cell RNA sequencing, CRISPR/Cas9, tumor microenvironment, multi-omics\",\"lastPublishedDoi\":\"10.21203/rs.3.rs-4034323/v1\",\"lastPublishedDoiUrl\":\"https://doi.org/10.21203/rs.3.rs-4034323/v1\",\"license\":{\"name\":\"CC BY 4.0\",\"url\":\"https://creativecommons.org/licenses/by/4.0/\"},\"manuscriptAbstract\":\"Essential genes are indispensable for the survival of cancer cell. CRISPR/Cas9-based pooled genetic screens have distinguished the pan-cancer essential genes and their functions in distinct cellular processes. Nevertheless, the landscape of essential genes at the single cell levels and the effect on the tumor microenvironment (TME) via cell-to-cell communication network (CCN) remains limited. Here, we identified 396 essential protein-coding genes (ESPs) by integration of 8 genome-wide CRISPR loss-of-function screen datasets of colorectal cancer (CRC) cell lines and a full-length single cells transcriptome data of CRC tissues. Then, 29 essential long non-coding genes (ESLs) were predicted using Hypergeometric Test (HT) and Personalized PageRank (PPR) algorithms based on ESPs and co-expressed network constructed from single cell transcriptome data. CRISPR/Cas9 knockout experiment verified the effect of several ESPs and ESLs on the survival of CRC cell line. Furthermore, multiple omics features of ESPs and ESLs were illustrated by examining their expression patterns and transcription factor (TF) regulatory network at the single cell level, as well as DNA mutation and DNA methylation events at bulk level. Finally, through integrating multiple intracellular regulatory networks with CCN, we elucidated that CD47 and MIF are regulated by multiple CRC essential genes and the anti-cancer drugs sunitinib can interfere the expression of them potentially. Our findings provide a comprehensive asset of CRC ESPs and ESLs, sheding light on the mining of potential therapy targets for CRC.\",\"manuscriptTitle\":\"Identification and Multi-omic Analysis of Essential Coding and Long Non-conding Genes in Colorectal Cancer\",\"msid\":\"\",\"msnumber\":\"\",\"nonDraftVersions\":[{\"code\":1,\"date\":\"2024-03-12 16:23:42\",\"doi\":\"10.21203/rs.3.rs-4034323/v1\",\"editorialEvents\":[{\"type\":\"communityComments\",\"content\":0}],\"status\":\"published\",\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true}}],\"origin\":\"\",\"ownerIdentity\":\"8db03f43-346e-40f9-a179-92cee4333905\",\"owner\":[],\"postedDate\":\"March 12th, 2024\",\"published\":true,\"recentEditorialEvents\":[],\"rejectedJournal\":[],\"revision\":\"\",\"amendment\":\"\",\"status\":\"posted\",\"subjectAreas\":[],\"tags\":[],\"updatedAt\":\"2024-03-13T09:48:41+00:00\",\"versionOfRecord\":[],\"versionCreatedAt\":\"2024-03-12 16:23:42\",\"video\":\"\",\"vorDoi\":\"\",\"vorDoiUrl\":\"\",\"workflowStages\":[]},\"version\":\"v1\",\"identity\":\"rs-4034323\",\"journalConfig\":\"researchsquare\"},\"__N_SSP\":true},\"page\":\"/article/[identity]/[[...version]]\",\"query\":{\"redirect\":\"/article/rs-4034323\",\"identity\":\"rs-4034323\",\"version\":[\"v1\"]},\"buildId\":\"qtupq5eGEP_6zYnWcrvyt\",\"isFallback\":false,\"isExperimentalCompile\":false,\"dynamicIds\":[84888],\"gssp\":true,\"scriptLoader\":[]}","source_license":"CC-BY-4.0","license_restricted":false}