Unveiling the protein landscape for early detection of colorectal precancerous lesions

Abstract

Colorectal cancer (CRC) has emerged as the second most prevalent cause of cancer-related mortality globally. Early identification of precancerous lesions prone to malignant transformation is pivotal in CRC prevention. Proteins, as microscopic reflections of cellular functional states, offer insights into pathological alterations within precancerous lesions through changes in their expression and function. Our review summarizes the protein research on colorectal adenomas under different sample conditions, including traditional adenomas, serrated lesions, LST, FAP and IBD. It highlights the changes in the expression patterns of key proteins and their potential mechanisms underlying the transition from precancerous to cancerous states. Additionally, it summarizes the research on post-translational modifications of characteristic protein families and associated signaling pathways, while discussing current techniques for studying protein expression and function in colorectal cancer, such as proteomics and artificial intelligence. However, current research limitations, such as small sample sizes, limited sample types, and insufficient in-depth mechanistic analysis, hinder comprehensive understanding. Future research should expand study cohorts, diversify sample types, and leverage machine learning and multi-omics approaches to develop predictive models. By doing so, a more comprehensive understanding of protein profiles during the progression from colorectal precancerous to cancerous lesions can be obtained, facilitating early CRC diagnosis and the development of targeted therapeutic interventions.

Introduction

Colorectal cancer (CRC) remains a significant global health concern, ranking as the third most common malignant tumor worldwide [1]. It is characterized by a high incidence and mortality rate, underscoring the urgent need for effective prevention and treatment strategies. Precancerous lesions of the colon and rectum, such as traditional adenomas, serrated lesions, familial adenomatous polyposis (FAP), and dysplasia in inflammatory bowel disease (IBD), have been established as precursors to CRC. These lesions carry a heightened risk of malignant transformation and are classified by pathological type, according to the 2019 World Health Organization (WHO) criteria [2].

Colorectal precancerous lesions are precursors to colorectal malignant tumors, exhibiting a risk of malignant transformation that increases with age. This progression is widely thought to be a gradual and long-term process. Traditional colorectal adenomas, the most studied precancerous lesions, are implicated in approximately 85% of CRCs, following the normal-adenoma-carcinoma (N-A-C) sequence [3]. During this progression, numerous proteins and signaling pathways undergo alterations. Besides, the prevalence of laterally spreading tumors in the colorectum is rising, with research suggesting that these lesions also involve alterations in various proteins, making them a potential category of colorectal precancerous lesions [4].

Proteins are the essential building blocks of life, and their expression patterns play a crucial role in various biological processes. Recent advancements in protein detection technologies have revealed significant changes in protein abundance within precancerous lesions of the colon. As these lesions progress to CRC, abnormal protein abundance may change by influencing gene mutations, CpG island methylation, or impaired mismatch repair, promoting malignant transformation [5, 6]. These findings highlight the potential of protein biomarkers to aid in the early diagnosis and targeted treatment of precancerous colorectal lesions, paving the way for further investigations into the underlying mechanisms of CRC carcinogenesis. This review comprehensively summarizes protein abundance alterations and their functional implications in colorectal precancerous lesions over the past two decades. We synthesize the types, expression changes, and potential oncogenic mechanisms of analogous proteins. Additionally, we summaries proteomics research techniques and address the limitations of current research and outline promising avenues for future development. Our goal is to provide a robust foundation for in-depth investigations into colorectal precancerous lesions.

Traditional colorectal adenoma

Adenomas represent the most common precancerous lesion of the colon and rectum. They are classified based on their pathological characteristics into three main categories: tubular adenoma, villous adenoma, and tubulovillous adenoma. Among these lesions, villous adenoma has the highest malignant transformation rate, while tubular adenoma has the lowest [7]. Analyzing protein abundance differences in these samples can help elucidate the progression of lesions from the N-A-C.

Tissue-specific protein pattern

The N-A evolutionary process

The formation of colorectal adenoma mostly starts from the biallelic mutation of APC in the basal stem cells of normal intestinal crypts, which leads to the loss of APC protein function, and then destroys the β-catenin degradation complex, activates the Wnt/β-catenin signaling pathway, drives cell proliferation out of control and inhibits differentiation [8, 9], eventually forming adenomatous polyps, a precancerous lesion with malignant potential.

The pathological progression from normal mucosa to adenoma involves alterations in the expression of several proteins. Using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry and immunohistochemical detection, researchers have identified that S100 calcium-binding protein A11 (S100A11), S100P, Annexin A3 (ANXA3), and eukaryotic translation initiation factor 5A1 are upregulated in adenomas compared to normal mucosa. Conversely, the expression of S100A9, ligand galectin-1, and liver fatty acid binding protein is downregulated [10]. Among these proteins, S100A11, S100P, and S100A9 are members of the S100 calcium-binding protein family and play roles in CRC progression. Upregulation of S100A11 expression has been shown to promote epithelial-mesenchymal transition (EMT) and facilitate cancer cell migration [11]. Similarly, overexpression of S100P has been demonstrated to activate the Wnt/β-catenin and ERK signaling pathways, abnormally regulate the cell cycle, and promote cancer cell invasion and metastasis [12]. In contrast to the downregulation of S100A9 expression observed in adenomas, S100A9 is upregulated in inflammatory cells at the invasive front of CRC tissue. This upregulation has been demonstrated to regulate the PI3K/AKT and MAPK/ERK signaling pathways, thereby promoting the malignant transformation of cells [13]. This phenomenon can be attributed to the lack of inflammatory response characteristic of adenomas, distinguishing them from malignant tumors in CRC [10]. Future studies investigating the overexpression of S100A9 in adenomas may provide insights into its potential role as an initiating signal for the progression toward malignant tumors. Notably, researchers have identified through Olink analysis that WNT1-inducible signaling pathway protein 1, endothelial-cell-specific molecule-1, tissue factor pathway inhibitor 2, and hepatocyte growth factor are upregulated in colorectal cancer progression and serve as potential biomarkers for colorectal cancer [14].

The A-C evolutionary process

The progression from colorectal adenoma to CRC is referred to as the A-C evolutionary process. This process involves alterations in the expression levels of various proteins and aberrant signaling pathways. Among these, Cyclin D1 and certain membrane-associated proteins can regulate the cell cycle to influence the proliferation and apoptosis of cancer cells; binding proteins interact with various molecules to collectively promote the occurrence of colorectal cancer; while the Mucin (MUC) family can promote the transformation of colorectal adenomas into cancer through multiple pathways.

The Cyclin protein family is a group of proteins that regulate the cell cycle. Elevated Cyclin expression can trigger uncontrolled cell cycles and contribute to cancer development. The Wnt/β-catenin signaling pathway is known to promote the progression of CRC. β-catenin is a key signaling molecule in this pathway, and the expression of Cyclin D1, a downstream molecule, is regulated by it. Elevated expression of them accelerates cell cycle progression and promotes cancer cell proliferation [15, 16].

Binding proteins play crucial roles in regulating various biological functions by interacting with other molecules. A group of colorectal adenoma samples subjected to long-term clinical follow-up were analyzed, revealing differences in the abundance of three proteins [17]. Among these, complement component 1q subcomponent binding protein (C1QBP) and premature ovarian failure protein 1B were found to be overexpressed in adenoma samples compared to normal mucosa and cancer tissues [18]. C1QBP is believed to promote cell proliferation and migration while also inhibiting apoptosis [19]. Conversely, integrin-alpha 1 (ITGA1) expression was reduced in adenomas. Previous studies have shown that c-Myc can regulate ITGA1 expression in CRC, influencing tumor invasion and migration through the Ras/ERK pathway [20]. Serpin family H member 1 (SERPINH1), also known as heat shock protein 47 (HSP47), is a collagen-binding protein located on the cell surface. SERPINH1 is a pivotal protein in the progression of adenomas to carcinomas and can enhance the malignant transformation of cells by activating the Wnt/β-catenin signaling pathway. Furthermore, Procollagen-lysine,2-oxoglutarate 5-dioxygenases 3 was identified as a distinctive protein associated with the progression of tubulovillous adenoma to colorectal cancer [21]. U3 small nucleolar RNA-associated protein 18 homolog (UTP18) is an RNA-binding protein. Proteomic analysis and immunofluorescence have shown that UTP18 is progressively upregulated in the progression of adenomas to carcinomas and can abnormally regulate the cell cycle, activate Hippo signaling pathways related to cell proliferation, promote abnormal cell proliferation and metastasis [22]. UTP18 may be a potential diagnostic biomarker for colorectal adenomas and cancer.

Membrane-related proteins play essential roles in cell proliferation, differentiation, structural formation, and signal transduction. Toll-like receptor 4, a transmembrane receptor, is highly expressed in adenomas and can activate the NF-κB and MAPK signaling pathways, leading to enhanced cell proliferation, anti-apoptosis, and immune evasion [23]^,[24]. Furthermore, high-mobility group box 2 was found to be upregulated in adenomas. This protein is primarily involved in abnormal cell cycle regulation and affects tumor cell apoptosis [25]. Therefore, alterations in the expression levels of these proteins may serve as reliable biomarkers for assessing cancer risk and diagnosing early-stage colorectal cancer. Members of the MUC family exhibit differential expression patterns in adenoma and cancer tissues. MUC2, the primary component of the intestinal mucus layer, is downregulated during adenoma progression to cancer and is associated with cell proliferation, apoptosis, and migration. MUC4, a transmembrane mucin, is also downregulated during carcinogenesis and may be associated with tumorigenesis, progression, and EMT [26]. MUC1, another transmembrane mucin, exhibits a similar mechanism but is upregulated and considered an important CRC biomarker [27]. The secreted glycoprotein MUC5AC is highly expressed in adenomas and is associated with tumor location, differentiation, infiltration depth, and lymph node metastasis [28]. MUC17, on the other hand, is downregulated during adenoma carcinogenesis and can effectively differentiate between hyperplastic polyps and adenomas, aiding in the diagnosis of precancerous lesions [29]. Overall, the expression changes of the mucin family are closely linked to the evolution of adenomas to carcinomas.

Recent research on precancerous lesions has focused on comparing various biological materials, including normal colorectal mucosa, adenoma/polyp/carcinoma adjacent tissue, and cancer tissue. This approach has provided a more comprehensive understanding of protein abundance changes throughout the progression from precancerous lesions to CRC. Runt-related transcription factor 3 (RUNX3) is a tumor suppressor protein consistently downregulated in colon polyps and CRC tissues, its expression can activate the TGF-β signaling pathway and influence Smad homologue 4 transcription, facilitating cancer cell proliferation and invasion [30]. Immunohistochemical analysis showed that Secreted protein acidic and rich in cysteine (SPARC) expression was decreased in adenoma and CRC tissues compared to normal colorectal tissue, while matrix metalloproteinase (MMP)−2 expression was increased [31]. Reduced SPARC expression may facilitate increased MMP-2 expression, enabling cancer cells to invade surrounding tissues and accelerate metastasis [32]. So, increasing SPARC expression may be a promising strategy for inhibiting tumorigenesis and development.

The above studies overlap in their assertion that protein abundance changes occur during the progression from normal mucosa to carcinoma, with many proteins consistently expressed in both adenoma and cancer tissues, as illustrated in Fig. 1. Proteins that support cell proliferation and survival are upregulated, while those that inhibit tumor response are downregulated. These shared expression patterns may be pivotal in the progression of adenomas to cancer. Molecules that interact with proteins also play a significant role in cancer progression. For instance, calcium-binding proteins of the S100 and ANXA families bind to calcium ions, inducing expression changes and suggesting the importance of calcium ions in CRC development [13]. Further research into the functions and interactions of these proteins may enhance our understanding of the early stages of CRC and aid in the differentiation of early malignant lesions.

Fig. 1.

Expression and function of characteristic proteins in colorectal adenomas. Full name of protein: C1QBP: Complement component 1q subcomponent binding protein, PLOD3: Procollagen-lysine,2-oxoglutarate 5-dioxygenases 3, UTP18: U3 small nucleolar RNA-associated protein 18 homolog, S100: S100 calcium-binding protein, MUC: Mucin, RUNX: Runt-related transcription factor, ITGA1: Integrin-alpha 1, PRSS8: Serine protease 8, FCGBP: Fc fragment of IgG-binding protein, TFR: Transferrin receptor, C4A: Complement C4-A. FGB: fibrinogen beta chain. FGA: fibrinogen alpha chain. (Created with BioRender.com)

Blood-based protein analysis

Certain proteins involved in tumorigenesis and development are released into the bloodstream, providing real-time information about disease progression. Blood plasma contains numerous vital proteins, including glycoproteins, complement, fibrinogen, transferrin, and others, essential for bodily functions. A pivotal study revealed that fibrinogen beta chain (FGB) and complement C4-A (C4A) were down-regulated in plasma and tissue samples from patients with advanced adenomas compared to normal [33]. Indeed, FGB is a coagulation factor, and its downregulation can lead to abnormal coagulation in adenoma patients. Besides, C4A mediates inflammatory processes and facilitates tumor cell growth. Tang et al.‘s study identified significant expression changes in transferrin receptor 1 (TFR1), S-adenosylhomocysteine hydrolase, and immuno-globulin heavy variable region 3–7 in adenomatous polyps [34]. Elevated levels of TFR1 activate the IL-6/IL-11-Stat3 signaling pathway, promoting colon epithelial cell proliferation, exacerbating intestinal mucosal damage, and contributing to the development of CRC. Longitudinal and in-depth proteomic analysis pointed out that S100A8 and S100A9 were overexpressed in plasma samples of CRC patients and healthy people, which could be used as potential biomarkers to distinguish CRC and healthy people [35]. It is worth noting that the Tandem Mass Tag (TMT) based liquid chromatography-tandem mass spectrometry (LC-MS/MS) technology was explored and verified by data-independent acquisition (DIA)-mass spectrometry (MS)-based diagnostic method. Researchers confirmed that high levels of total fibronectin 1 (FN1), haptoglobin, S100A9 and fibrinogen alpha chain (FGA) in extracellular vesicles were significantly associated with colorectal cancer progression, and FGA was the most important candidate biomarker [36].

Blood samples are widely used for protein detection and offer the advantages of real-time assessment, minimal invasiveness, and reusability. They are valuable tools for research into the molecular characteristics of diseases, biomarkers, and molecular typing. In recent years, body fluid biopsy technology mostly combines protein glycosylation with LC-MS and artificial intelligence to predict the occurrence and development of CRC. Although this is an important exploration of a more cost-effective colorectal cancer screening method, the individual differences in protein glycosylation cannot be ignored. In addition, the development of colorectal cancer risk prediction model combining plasma proteomics, polygenic score and traditional non genetic risk factors can not only effectively identify high-risk people, but also significantly improve the accuracy of risk prediction [37, 38].

Figure 1 summarizes the changes in protein abundance observed in blood samples in the aforementioned studies, which can promote or inhibit the malignant transformation of cells. These findings have positive implications for exploring the potential of blood proteins as biomarkers in CRC development and developing less invasive screening and diagnostic methods for precancerous lesions.

Stool-based proteomic analysis

In addition to the two aforementioned sample types, non-invasive fecal sample collection is valuable for studying precancerous lesions. Fecal samples are typically analyzed using proteomic techniques and fecal immunochemical tests (FIT). Soares and colleagues demonstrated that proteins in fecal samples from individuals with precancerous lesions are involved in specific and non-specific immune responses [39]. A protein group comprising haptoglobin, lysosome-associated membrane protein 1, spectrin repeat containing nuclear envelope protein 2, and ANXA6 has been identified as a potential marker for detecting high-risk adenomas (with sensitivity and specificity of 53% and 95%). Another protein group, including haptoglobin, leucine-rich alpha-2 glyco-protein-1(LRG), retinol-binding protein 4, and fibronectin 1, can be used to predict the risk of high-risk adenomas and CRC (with sensitivity of 66% and 62%, specificity of 95%), suggesting their potential as fecal biomarkers for early detection [40]. Since FIT has low sensitivity for colon adenomas, using a biomarker panel may improve sensitivity for high-risk adenomas, increasing the detection rate of malignant lesions and potentially enhancing the current FIT-based screening method.

The Fc fragment of IgG-binding protein (FCGBP) shares a high degree of structural similarity with the MUC2 protein [41] and may be involved in colorectal epithelial development, inflammatory response, and Wnt signaling pathway regulation [42], potentially contributing to the development of colorectal cancer. Bosch’s research team [43] observed that the expression of FCGBP and MUC2 was decreased in fecal samples from patients with adenomas and colorectal cancer during the progression of adenomas to cancer, as determined by liquid chromatography-tandem mass spectrometry. These proteins may function as protective factors, reducing the risk of colorectal cancer. The study also identified reduced expression of serine protease 8 (PRSS8). Downregulation of PRSS8 expression has been shown to impede the inhibitory effects on the EMT, Wnt/β-catenin signaling pathways, facilitating tumor cell proliferation, invasion, and migration [44].

In addition, the pathogenesis of colorectal cancer is related to changes in the gut microbiota. Due to the extremely high complexity of the obtained microbial samples, in recent years, by improving the analysis of mass spectrometry data, metaproteomics can provide functional information of intestinal flora that is of great value for pathogenesis research and has become a powerful method for characterizing intestinal microbial proteins. At present, some studies have used quantitative metaproteomics methodology to discover that the protein abundances of 341 kinds of microbiota are significantly different between CRC patients and healthy people and are related to oxidative stress responses, DNA replication, recombination, and repair [45]. Thus, it can be known that in the future, metaproteomics can better screen for biomarkers existing in such complex samples and be used for the diagnosis and treatment of precancerous lesions.

Figure 1 summarizes the expression changes and mechanisms of proteins identified in fecal samples through protein studies. These proteins contribute to predicting the risk of carcinogenesis in precancerous lesions. However, current research is limited by factors such as small sample sizes, low specificity, and unclear mechanisms. Additional basic experiments and clinical studies are necessary to validate the reliability of these potential biomarkers.

A substantial body of research is currently focused on examining the unique protein profiles associated with colorectal adenomas. Figure 2 summarizes the signaling pathways involved in the characteristic proteins identified in tissue, blood, and fecal samples of adenoma patients. These pathways include Wnt/β-catenin, PI3K/AKT, TGF-β, and NF-κB, among others. This research provides a deeper understanding of adenoma pathogenesis and carcinogenesis, facilitates the development of effective treatment strategies, and establishes a scientific foundation for CRC prevention and intervention.

Fig. 2.

Some characteristic proteins and their associated signaling pathways in colorectal adenomas. Full name of protein: PRSS8: Serine protease 8, FCGBP: Fc fragment of IgG-binding protein, S100: S100 calcium-binding protein, PLOD3: Procollagen-lysine,2-oxoglutarate 5-dioxygenases 3, ITGA1: Integrin-alpha 1, UTP18: U3 small nucleolar RNA-associated protein 18 homolog, RUNX: Runt-related transcription factor, TFR1: Transferrin receptor 1 (Created with BioRender.com)

Serrated lesions

Colorectal serrated lesions are classified into four categories: hyperplastic polyps (HP), sessile serrated lesions (SSL), traditional serrated adenomas (TSA), and sessile adenomas of unspecified classification. Of these, SSL and TSA are associated with an increased risk of malignancy. Most studies indicate that changes in protein expression in serrated lesions are primarily associated with CpG island methylation and mismatch repair, and alterations in proteins such as MUC5AC, ANXA10, and HSP70 also play a critical role.

CpG island methylation is a primary mechanism underlying CRC in serrated lesions. Research has shown that methylation of the Runx3 gene promoters may induce downregulation or deletion of RUNX3 protein abundance [46]. In a study of serrated lesions, Fang Yuan and colleagues observed a significant increase in the methylation rate of the Wifi-1 gene and abnormal positive expression of β-catenin protein. These findings suggest that methylation of the Wifi-1 gene promoter region may alter Wifi-1 protein abundance and activate the Wnt/β-catenin signaling pathway, contributing to carcinogenesis [5]. The downregulation of these proteins plays a crucial role in the serrated pathway and may serve as initiating events in the progression of serrated lesions.

Mismatch repair (MMR) is a critical mechanism for repairing DNA damage. MMR prevents cancer development by correcting base mismatches that occur during cellular replication, transcription, and other processes. Cytokeratin 20 expression was significantly increased in tissue samples from non-tender serrated lesions compared to conventional serrated adenomas. At this stage of the study, cytokeratin 20 expression was found to be an independent poor prognostic factor for MMR-positive CRC, potentially associated with lymph node metastasis and tumor infiltration [6].

The MUC family demonstrates not only changes in expression in colorectal adenomas but also exhibits similar patterns of expression alterations in serrated lesions. Researchers have shown that MUC5AC expression is upregulated in HP, SSL, and TSA, MUC2 expression is downregulated during the malignant transformation of TSA [47], while trefoil factor 1 protein is also overexpressed in HP and SSL [48]. Notably, this study found a higher positivity rate of ANXA10 in SSL than TSA, it has been identified as a specific marker of serrated lesions. Its expression can suppress autophagy-mediated transferrin receptor degradation, inhibiting cell iron death and hindering cancer progression [49]. Overall, ANXA10 may represent a promising target for the diagnosis and treatment of precancerous lesions and CRC. Besides, most studies report that HSP70 are upregulated in colorectal serrated polyps and participate in cancer cell growth, invasion, metastasis, anti-apoptosis, angiogenesis, and drug resistance. Erythropoietin hepatoma (Eph) B2, which promotes cancer cell development, exhibits a decreasing trend from normal colon tissue to serrated polyps to CRC. Conversely, the expression of the Ephrin(Erythropoietin hepatocyte kinase receptor ligand)B2 ligand increases [50]. This suggests that EphB2 may inhibit CRC development, while high expression of receptor EphB4 and its specific ligand EphrinB2 may promote tumor growth and metastasis. However, the precise mechanisms of these factors remain to be fully elucidated.

The aforementioned research suggests that the primary mechanism of serrated lesion carcinogenesis involves CpG island methylation, and mismatch repair, which are implicated in cell proliferation, apoptosis, migration, and drug resistance (Fig. 3). Furthermore, the MUC and ANXA protein families exhibit similar expression patterns throughout the progression from colorectal adenomas and serrated lesions to CRC. These findings offer new insights into the overall progression of colorectal precancerous lesions. By combining similar proteins, it may be possible to develop a biomarker for detecting CRC, improving the specificity and accuracy of early screening for CRC lesions.

Fig. 3.

Expression and function of characteristic proteins in serrated lesions and laterally spreading tumors. Full name of protein: HSP: Heat shock protein, CK20: Cytokeratin 20, LGR: Leucine-rich repeat-containing G protein-coupled Receptor, CDK: Cyclin-Dependent Kinase, RNF: Ring finger protein. (Created with BioRender.com)

Laterally spreading tumors

LSTs are a unique type of colorectal tumor with a distinct morphology. Recently, they have been identified as potential precursor lesions of CRC. It has a high propensity for malignant transformation. When their diameter reaches 2 cm or more, the risk of cancer can be as high as 31%, and this risk increases with increasing lesion size [51].The classic Wnt/β-catenin signaling pathway has been confirmed to play a crucial role in the development of LST [52]. Therefore, alterations in proteins along this signaling pathway are most prominent in LST.

Compared to normal tissues, colorectal protruding adenomas, LST, and colorectal cancer tissues, the expression of leucine-rich repeat-containing G protein-coupled receptor 5 (LGR5), cyclin-dependent kinase 5 (CDK5), and β-catenin exhibited a gradual increase. LGR5 and CDK5 may regulate cell proliferation and EMT through the Wnt/β-catenin signaling pathway, contributing to LST growth and carcinogenesis. Previous studies have shown that LGR5 expression also gradually increases during the progression from normal mucosa to adenoma to carcinoma, potentially correlating with the degree of dysplasia [53]. Ring finger protein 6 (RNF6) expression is elevated in tumor tissues with higher degrees of intraepithelial neoplasia. As a transcription factor, it activates the JAK/STAT3 and Wnt/β-catenin signaling pathways [54]. A research group used isotope-labeled relative and absolute quantitative proteomic detection techniques to identify significant increases in Lipocalin-2 (LCN-2) and MMP-9 expression in tumor tissues and serum samples from patients with colorectal LST [55]. The LCN-2-MMP-9 complex has been shown to accelerate CRC progression and may serve as a valuable diagnostic and prognostic marker.

Despite the lack of a consensus on the development of LSTs in colorectal cancer precancerous lesions, research in this area is increasing due to their high risk of carcinogenesis and rapid progression. Current research on LST has primarily focused on the Wnt/β-catenin signaling pathway and the characteristic proteins that regulate cell proliferation, apoptosis, migration, and epithelial-mesenchymal transition (Fig. 3). A comprehensive analysis of the protein profile of lateral developing tumors can provide a deeper understanding of their progression toward colorectal cancer and contribute to advancements in scientific research and the clinical management of LSTs as precancerous lesions.

Others

Familial adenomatous polyposis is an autosomal dominant disorder caused by mutations in the adenomatous polyposis coli (APC) gene. Individuals with FAP have a nearly 100% lifetime risk of developing colorectal cancer. Early detection and diagnosis are crucial for the management of FAP patients and associated colorectal cancers, as timely intervention can significantly improve prognosis.

Experimental studies on mouse models revealed that the expression of β-catenin, HSP70, and P53 was significantly upregulated in malignant tissues. A strong correlation between P53 and HSP70 suggested a potential role of HSP70 in the malignant transformation of FAP through its interaction with P53 [56]. Moreover, recent research has indicated that P53 can contribute to malignant changes via the Wnt signaling pathway, highlighting the significance of increased β-catenin and P53 expression in the development and malignant progression of FAP intestinal polyps [57]. An iTRAQ analysis of FAP samples during the N-A-C process demonstrated upregulation of β-catenin and S100A10, while decorin (DCN) and septin-7 expression was downregulated. These findings suggest a close association between the expression of β-catenin, DCN, septin-7, and S100A10 in the development of human hereditary polyposis colorectal cancer [58].

IBD, encompassing ulcerative colitis and Crohn’s disease, is a chronic inflammatory condition of the intestinal tract associated with an increased risk of colorectal cancer compared to the general population. Immunohistochemical analysis has revealed elevated expression of hepatocyte nuclear factor 4α protein and β-catenin in the colitis-tumor sequence, suggesting their potential as biomarkers for predicting low-grade dysplasia [59]. In addition, by using metaproteomics to study Crohn’s disease and ulcerative colitis fecal samples, some researchers have found that the metaproteins gamma-glutamyl hydrolase, resistin 50 olfactomedin-4 and trefoil factor 252 as well as the metaproteins of the C4A, and plasma protease C1 inhibitor may be correlated with these diseases [60]. The investigators identified the upregulation of six inflammation-related proteins by comparison of plasma proteins from pre-morbid ulcerative colitis and healthy controls using the Olink Target 96-plex panels analyses identified multiple potential key regulators. Included among these proteins, five proteins were found to be effective in differentiating between patients with ulcerative colitis and healthy patients, with potential disease diagnostic value [61].

Given the strong correlation between the onset of FAP and genetic mutations, genetic testing is a reliable method for accurately screening for the association between IBD and CRC. This topic has emerged as a distinct area of research and is therefore not discussed in detail in this article.

Post-translational modifications of characteristic protein families and related signaling pathways

Post-translational modification (PTM) such as ubiquitination, acetylation, and glycosylation are closely related to tumor occurrence, progression, and metastasis. In the process of colorectal cancer development, the ubiquitination levels of S100A11 are increased, which suppresses cellular senescence, promotes tumor cell proliferation and invasion, and is significantly associated with poor prognosis in CRC patients [62]. Furthermore, hypomethylation of MUC5AC accurately identifies serrated lesions with BRAF mutations, CIMP, or microsatellite instability (MSI), suggesting that MUC5AC hypomethylation aids in assessing the malignant risk of precancerous lesions [63]. Additionally, aberrant glycosylation alters the structure and expression levels of MUC2, significantly enhancing cell proliferation, migration, and anti-apoptotic capacity, thereby inducing the occurrence and progression of CRC [64].

A comprehensive analysis of colorectal precancerous lesions identified several protein families, including S100, MUC, ANX, and HSP, as key players. The S100 family of calcium-binding proteins is involved in a wide range of intracellular Ca²⁺-dependent functions, including cell proliferation, differentiation, apoptosis, energy metabolism, cell signaling, and Ca²⁺ homeostasis regulation [65]. Compared to normal mucosal tissues, S100A11 and S100P were upregulated, while S100A9 and S100A10 were downregulated in FAP tissues during the progression to adenomas, suggesting their crucial roles in precancerous lesion development. Abnormal MUC expression can activate signaling pathways such as NF-κB, MAPK, Wnt, and JAK-STAT, leading to precancerous lesion development and further MUC expression. During the progression of colorectal precancerous lesions to cancer, MUC1 and MUC5AC are upregulated, while MUC2 and MUC4 are downregulated, primarily affecting pathways such as apoptosis, cell cycle, DNA damage, and EMT [26]. The ANX family, as key tumor regulatory molecules, plays a critical role in carcinogenesis. Upregulation of ANXA3 and ANXA10 can activate cell membrane and cytoskeleton formation, as well as related signaling pathways, promoting cancer cell invasion and metastasis. In cancer, HSP family members are typically highly expressed to enable tumor cells to evade apoptosis [66]. Our study also revealed upregulation of HSP47 and HSP70 during malignant transformation, promoting cancer cell development, progression, and metastasis.

Table 1.

Post-translational modifications of characteristic protein families and related signaling pathways

Name and expression changes		Check out location	Type	Molecular Mechanism	Post-translational modifications	Role in Cell Phenotypes or Behavior	Refs.
S100	S100A9↓	Human tissue	CRA	PI3K/AKT, MAPK/ERK	/	Cell proliferation	[13]
	S100A10↑	Human tissue/Mouse tissue	FAP	Wnt/β-catenin	/	Cell proliferation, Migration	[58]
	S100A11↑	Human tissue/Mouse tissue	CRA	TGF-β	Ubiquitination	Cell migration, EMT	[11]
	S100P↑	Human tissue	CRA	Wnt/β-catenin, ERK	/	Cell cycle, Cell migration	[12]
MUC	MUC1↑	Human tissue	CRA	JAK/STAT	Glycosylation	Cancer progression, EMT	[27]
	MUC2↓	Human tissue/ feces	CRA, SL	NF-κB	Sulfation、 Methylation	Cell proliferation Migration, Apoptosis	[28]
	MUC4↓	Human tissue/Mouse tissue	CRA	Wnt/β-catenin, TGF-β, PI3K/AKT	Glycosylation	Cell migration	[26]
	MUC5AC↑	Human tissue/ feces	CRA, SL	Wnt/β-catenin	Glycosylation、 Methylation	Cell proliferation, Apoptosis, Immune escape	[28, 47]
ANXA	ANXA3↑	Human tissue	CRA	HIF-1α	/	Cell proliferation, Migration, Drug resistance	[10]
	ANXA6	Human feces	CRA	MAPK, PI3K-Akt	/	Cell proliferation, Apoptosis	[40]
	ANXA10↑	Mouse tissue	SL	TGF-β	/	Cell proliferation, Migration, Iron death	[49]
HSP	HSP47↑	Human tissue	CRA	Wnt/β-catenin, TGF-β	/	Stress response, Proliferation, Apoptosis, Migration	[21]
	HSP70↑	Mouse tissue	SL, FAP	PI3K/AKT	/	Cell proliferation, Migration	[56]

Furthermore, certain proteins exhibited consistent differential expression patterns across different samples of the same lesion type and within the same sample of different lesion types (Table 2). MUC2 and MUC5AC, for instance, were detectable in both tissue and fecal samples from CRA patients and displayed identical expression changes in both types of samples. Similarly, these two proteins demonstrated the same expression changes in serrated lesions, with MUC2 downregulated and MUC5AC upregulated. RUNX3 was downregulated, while LGR5 was upregulated in CRA and LST, potentially promoting cancer development through activation of the Wnt/β-catenin signaling pathway. Up-regulation of HSP70 was primarily observed in tissue samples from SL and FAP, promoting tumor cell proliferation and metastasis by activating the PI3K/AKT signaling pathway. β-catenin, a pivotal molecule in tumorigenesis, was overexpressed in all colorectal precancerous lesions discussed in this study. It primarily drives the malignant transformation of precancerous lesions by activating the Wnt/β-catenin signaling pathway.

Table 2.

Expression and role of the same protein in different types of precancerous lesions

Name and expression changes	Check out location	Type	Molecular Mechanism	Role in Cell Phenotypes or Behavior	Refs.
MUC2↓	Tissue, Stool samples from patients with adenoma	CRA, SL	NF-κB	Cell proliferation Migration, Apoptosis	[28]
MUC5AC↑	Tissue, Stool samples from patients with adenoma	CRA, SL	Wnt/β-catenin	Cell proliferation, Apoptosis, Immune escape	[28, 47]
HSP70↑	Tissue	SL, FAP	PI3K/AKT	Cell proliferation, Migration	[56]
β-catenin↑	Tissue	CRA, SL, LST, FAP, IBD	Wnt/β-catenin	Cell proliferation, Migration	[15, 58, 59]
RUNX3↓	Tissue	CRA, LST	TGF-β, Wnt/β-catenin	Cell proliferation, Migration	[46]
LGR5↑	Tissue	CRA, LST	Wnt/β-catenin	EMT	[53]

The differential expression of these proteins’ sheds light on the pathogenesis of colorectal cancer, transitioning from precancerous to malignant states. This not only enhances our understanding of the etiology and progression of CRC but also provides potential biomarkers for disease diagnosis and treatment, driving advancements in colorectal disease research. To further validate the specific roles and underlying mechanisms of these proteins, additional fundamental experiments and clinical investigations are essential.

Investigating protein-related techniques in colorectal precancerous lesions

Proteomics

Currently, proteomics represents an advanced scientific methodology extensively utilized in the investigation of colorectal precancerous lesions, underscoring its substantial importance. Recent advancements in proteomics technology have significantly enhanced our understanding and identification of alterations in protein types and abundances throughout the malignant transformation of precancerous lesions. Accordingly, we have systematically reviewed the proteomics technologies and identified biomarkers pertinent to colorectal lesions (see Table 3), with the objective of offering more rigorous scientific approaches and fostering innovative perspectives for researchers in the field.

Table 3.

Proteomics for colorectal precancer detection

Technology	Details	Indicators in colorectal diseases	Refs.
liquid chromatography-tandem mass spectrometry (LC-MS/MS)	Triple Quadrupole: Targeted quantification (such as biomarker validation).	TFR1, SAHH, and HV307 may be considered as potential biomarkers for CRC screening.	[34, 67]
	Quadrupole Time of Flight: Non targeted screening, metabolite identification.
	Orbitrap: Deep proteomics and post-translational modification analysis.
	Ion Trap: Small molecule structure analysis and multi-level fragmentation research.
Metaproteomics	Proteins were digested into peptides and analyzed with an Orbitrap LC-MS/MS, while MetaProteomeAnalyzer software identified proteins and interpreted taxonomic and functional data.	The metaproteins gamma-glutamyl hydrolase, resistin 50 olfactomedin-4 and trefoil factor 252 as well as the metaproteins of the complement system C4-A, and plasma protease C1 inhibitor may be correlated with IBD.	[60]
Tandem Mass Tag (TMT)/Isobaric tags for relative and absolute quantification (iTRAQ)	Labelling with isotope tags, followed by enzymatic digestion of amino acid N-terminals and lysine residues, specific labelling with TMT/iTRAQ reagents, and finally tandem mass spectrometry analysis, allowing simultaneous comparison of the relative amounts of protein.	β-catenin, DCN, septin-7, and S100A10 are the differentially expressed proteins involved in the development of CRC. RRP12 and SERPINH1 may play an important role in the N-A and A-C processes. Potential clinical application of UTP18 in predicting colorectal adenoma recurrence.	[21, 22, 58]
Matrix-assisted laser desorption/ionization-time of flight mass spectrometry(MALDI-TOF/TOF MS)	In MALDI-TOF MS, a laser bombards the sample-matrix film, causing the matrix to absorb energy and ionize biomolecules via proton transfer. These ions are accelerated through a flight tube by an electric field, and their time of arrival at the detector determines their mass-to-charge ratio (m/z) and signal value, forming corresponding peaks.	eIF5A-1, S100P, ANXA3, S100A11 and FABPL may play a crucial role in early CRC development.	[10]
Olink Target 96-plex panels	Olink Target 96 technology uses a pair of antibodies, each linked to a specific DNA strand, to detect proteins. When both antibodies bind to a protein, their DNA strands complement and form a double-stranded DNA template via enzyme extension, allowing protein quantification through Proximity extension assay (PEA).	WISP-1, ESM-1, and TFPI-2 are closely associated with the occurrence of colorectal cancer.	[14]

In the study of precancerous lesions, LC-MS/MS is indispensable for its high precision in exploring deep mechanisms and targeted validation, but its throughput limits large-scale applications. The high throughput of iTRAQ/TMT greatly promotes biomarker screening, but its inherent “compression effect” severely threatens the accuracy of quantitative results, potentially leading to misinterpretation or omission of early key molecular events.

However, emerging proteomics technologies are fundamentally reshaping our understanding of tumor heterogeneity and early events in carcinogenesis. Single-cell proteomics overcomes the limitations of traditional bulk analysis, directly revealing hidden cell subpopulations, driver cells, and activated abnormal signaling pathways at the single-cell level in precancerous lesions, thereby identifying early malignant transformation drivers that traditional methods cannot capture. DIA-MS achieves high reproducibility and high-precision deep proteome quantification by systematically and unbiasedly collecting all peptide fragment information, significantly overcoming the “compression effect” and batch variation issues and enabling longitudinal tracking of dynamic evolution in precancerous lesions and multicenter cohort validations [17, 36]. Spatial proteomics preserves the spatial distribution information of proteins in situ, directly mapping molecular gradients, immune microenvironment boundaries, and heterogeneous regions’ protein network profiles within lesion tissues, revealing spatially specific oncogenic signaling niches. Based on these emerging proteomics technologies, future efforts can validate biomarker performance in independent cohorts, integrate multi-biomarkers using machine learning to build diagnostic models, and convert validated biomarkers into ELISA or liquid chip kits for non-invasive blood/stool screening, further reducing invasiveness and improving therapeutic precision.

Multi-omics strategy

Beyond proteins, current biomarker research for precancerous lesions encompasses studies at multiple levels, including genes, metabolites, and microorganisms. For example, genomic testing of tissues and blood from patients with traditional or sessile serrated adenomas has identified a subset of 20 genes that can differentiate colorectal adenomas from adenocarcinomas, providing insights into the somatic mutation status of precancerous lesions [68]. Additionally, alterations in microbial communities have been implicated in the malignant transformation of cells. A cohort study with long-term follow-up has linked colorectal precancerous lesions and colorectal cancer to the incidence of streptococcal endocarditis. This association may be attributed to the role of streptococcus in promoting colorectal cancer progression via the bloodstream. The gut microbiome and its metabolites have been shown to be associated with the occurrence and development of CRC. Microbiome-derived or related metabolites can enter the circulatory system and exert their effects [69]. By systematically integrating epigenomic, transcriptomic, proteomic, and metabolomic data to construct spatiotemporal dynamic network models, we can overcome the limitations of single-omics approaches and comprehensively capture cancer-driving events. For example, when mRNA and protein levels are inconsistent, integrating phosphoproteomics and lipid metabolomics can identify key functional nodes where pathways go awry. Additionally, machine learning can fuse methylation markers, protein combinations, and metabolite ratios to build cross-omics diagnostic models, significantly improving the detection rate of early lesions.These findings contribute to the identification of precancerous lesions and the selection of suitable candidates for targeted monitoring and preventive interventions, ultimately aiming to reduce the incidence of cancer.

Artificial intelligence

Artificial intelligence is revolutionizing the early detection of precancerous lesions in colorectal cancer by efficiently mining high-value biomarkers from complex, multidimensional data. Machine learning algorithms analyze endoscopic images, histopathological slides, multi-omics data, and clinical information to automatically identify subtle morphological abnormalities and molecular patterns that are difficult to capture with traditional methods, significantly improving sensitivity and specificity for detecting precancerous lesions [70, 71]. Moreover, AI integrates multi-source heterogeneous data to build predictive biomarker models—for instance, discovering novel pre-cancer-related protein clusters through unsupervised learning in proteomics data or screening non-invasive diagnostic markers in blood/stool samples using supervised learning [72–74]. These algorithms not only overcome the quantitative biases of traditional high-throughput technologies (e.g., TMT/iTRAQ) but also uncover nonlinear interactions between biomarkers, enabling the identification of multimodal biomarker combinations with higher clinical translation potential. This provides a critical driving force for developing minimally invasive and precise screening tools for precancerous lesions, ultimately reducing the incidence and mortality of colorectal cancer.

Conclusion and perspectives

Given the crucial role of precancerous lesions in the prevention and treatment of colorectal cancer, and the potential biomarker function of proteins, research in this area has garnered significant attention in recent years. This review summarizes protein profile-related research on colorectal cancer precancerous lesions over the past two decades. These studies primarily employ traditional protein detection techniques, such as Western blotting and immunohistochemistry, as well as high-throughput proteomics. We categorize the types of proteins identified, their expression changes, and potential carcinogenic mechanisms. Additionally, we provide a detailed summary of the proteomics currently used to study precancerous lesions, and we aim to provide a comprehensive reference for in-depth studies of precancerous lesions in colorectal cancer. The goal is to provide a comprehensive reference for in-depth research on colorectal cancer precancerous lesions.

Although the number of studies on colorectal precancerous lesions has steadily increased, several limitations persist. Our review highlights that most investigations into proteins in these lesions primarily focus on detection and expression changes, with limited mechanistic studies due to the scarcity of clinical samples. Additionally, the low utilization of samples is a significant constraint. While adenomas have been extensively studied, other precancerous lesions, such as serrated lesions and laterally developed tumors, have primarily relied on tissue samples, with limited exploration of other types. To enhance the evidence base for colorectal precancerous lesions, future research should include larger sample sizes, multicenter prospective cohort studies with diverse sample types, and long-term follow-up. Expanding the range of molecules investigated, such as extracellular vesicles and microRNA, can create a comprehensive biological sample library, broadening the scope of research. Additionally, integrating multi-omics data to develop biomarker clusters, combined with artificial intelligence techniques, can achieve technical standardization, enhancing the predictive power of models. Further validation through clinical trials can promote clinical application, thereby advancing disease diagnosis and improving patient outcomes. This knowledge will enable the development of more rapid, accurate, and efficient methods for predicting and diagnosing precancerous lesions, ultimately reducing the morbidity and mortality associated with CRC.

Author contributions

Yuanke Luo. Chong Xiao. and Xi Fu. Xueke Li. wrote the main manuscript text. Simin Luo and Yifang Jiang prepared figures and tables. Chuan Zheng and Fengming You substantively revised it. All authors reviewed the manuscript.

Funding

This research was funded by THE NATIONAL NATURE SCIENCE FOUNDATION OF CHINA, grant number:82205072.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable

Consent for publication

All authors have read and agreed to the published version of the manuscript.

Competing interests

The authors declare no competing interests.