The stepwise nature of sporadic CRC is still poorly understood, even though several mechanisms have been described to be involved in its initiation and progression. Epidemiological studies have found a relationship between CRC and chronic exposure to environmental risk factors (see above section) with strong pro-inflammatory potential. Moreover, increasing evidence suggests that intestinal microbiota and its products (e.g., butyrate and bacterial toxins) play a pivotal role in all CRC steps (initiation, progression, and metastasis) (19–21) ( Figure 1 ). CRC patients display a reduced bacterial diversity and richness compared to healthy individuals, reflecting a distinctive intestinal microbial dysbiosis (22). Dysbiosis causes alteration in gut mucosa integrity and permeability, due to alteration of intercellular tight junctions. This condition, enhancing the colocyte susceptibility to mutagenic/carcinogenic factors and pathogenic bacteria, also promotes the activation of Mucosal Associated Lymphatic Tissue. Moreover, Th2-derived cytokines induced by pathogens or autoantigens may result in myeloid cell recruitment (neutrophils and macrophages) and, consequently in Reactive Oxygen Species (ROS) production.

As known, inflammation and oxidative stress are tightly coupled; in fact, a chronic activation of inflammatory cells and production of pro-inflammatory mediators (e.g., cyclooxygenase 2, prostaglandin E2, tumor necrosis factor α, and transforming growth factor β) enhance ROS generation and dysregulate the activity of signal transduction pathways, including Transducer and Activator of Transcription 3, Nuclear Factor-kappa B (NF-kB), hypoxia-inducible factor-1α (HIF1α), NF-E2 related factor-2 (NRF2) (23–25). In addition, it has been reported that ROS overproduction may result in genetic/epigenetic changes, such as single-strand cleavage, point mutations, miscoding, abnormal amplification, oncogene activation, and immune suppression, leading to possible precancerous lesions (i.e., adenomatous polyps) (26, 27).

Mutations providing a selective growth advantage to the cells within their microenvironment can potentially drive cancer development. Generally, two mutated gene drivers can lead to a net cell gain and detectable benign lesions, but over three gene mutations promote the invasion through the basement membrane, thereby leading to malignancy. In CRC, the first mutations usually concern the APC gene resulting in a proliferative advantage of epithelial cells promoting benign lesions (small adenomas) ( Figure 2 ). APC mutated small adenomas have a slow growth rate, but further mutations of the KRAS gene can increase their proliferation. However, the mutational status may be worsen by sequential mutations of genes such as PIK3CA, SMAD4, and TP53 that promote the onset of malignant tumors capable of infiltrating surrounding tissue and metastasizing distant organs ( Figure 2 ) (28). In addition, both colorectal adenomas and CRC are linked to epigenetic alterations such as aberrant DNA methylation in key tumor-suppressor and oncogenes and dysregulation of miRNA expression (29, 30).

Well-known examples of the carcinogenic role of ROS in CRC are missense mutations at p53 suppressor gene, activation of canonical Wnt signaling pathway (Wnt/β-catenin), which is involved in cancer stem cell renewal process, and PI3K/Akt signaling pathway, which regulates cell proliferation (31, 32). Moreover, the most frequent ROS-dependent pre-mutagenic DNA lesion is represented by 8-oxoguanine (8-oxoG) (33).

Notably, oxidative stress, due to either ROS overproduction or a reduced activity of the enzymatic and non-enzymatic antioxidant systems, is often involved in development and progression of several cancers via activation of redox-responsive signaling pathways leading to uncontrolled cell growth and oxidation of lipids, carbohydrates, and proteins (i.e., cancer initiation, promotion, and progression stages). Accordingly, oxidative stress is one of the main cancer research topics given its involvement in both genetic and metabolic cell damage (34–36). Increasing evidence showed that oxidative stress control biogenesis of cancer-associated microRNA (miRNA) via targeting various transcription and epigenetic factors. Recently, CRC-associated miRNAs (e.g., miR-106b-5p, miR-335-5p, miR-193a-5p, miR-378a-3p and miR-423-5p) are becoming attractive biomarkers as they are expressed from the early stage of tumor development (37).

Moreover, either up-regulation or down-regulation of miRNAs also known as “onco-miRNAs” are involved in CRC progression and metastasis contributing to dysregulate several signaling pathways, including mitogen-activated protein kinases (e.g., miR-422a, miR-195), Wnt (e.g., miR-135a, miR-135b, miR-155, miR-17–5p, miR-224), transforming growth factor-β (e.g., miR-224, miR-20a-5p) and epithelial-to-mesenchymal transition (EMT) (e.g., miRNA-155, miR-34) (38–40).

Oncogene, oncosuppressor and metabolic gene mutations contribute to the profound metabolic alterations found in cancer cells, i.e., impaired respiration, increased fermentation and anabolism (41). In most cases, the metabolism of cancer cells favors aerobic glycolysis (the Warburg effect) rather than oxidative metabolism to fulfill their biosynthetic and bioenergetic demands of rapid and sustained proliferation. Mitochondrial Oxidative Phosphorylation System (OXPHOS) is not necessarily defective in tumorigenic cells, and it can take place proportionally to the oxygen supply. Indeed, it has been shown that cancer stem cells are able to revert glycolysis to TCA cycle to better satisfy their metabolic needs and overproduce ROS (42). The Warburg phenotype has been demonstrated to be driven by overexpression of oncogenes such as c-Myc and HIF-1α (43). The inhibition of pyruvate dehydrogenase activity and the increase of lactate dehydrogenase activity lead to the conversion of pyruvate to lactate following the mass action law (44). Moreover, lactate is also generated from catabolism of glutamine and it is considered a metabolite eliciting a broad spectrum of effects useful to sustain cancer progression and metastasis (45). Cancer cells are capable of adapting to metabolic-derived acidosis via monocarboxylate transporters (MCTs), which export lactate and favor intracellular alkalinization. Thus, the lactate exported by tumor cells can be imported by cells of tumor microenvironmental where it acts as important intracellular signaling for angiogenesis (46, 47).

Notably, cancer cells well adapt to ROS by triggering a powerful antioxidant response mainly driven by glutathione (GSH) and antioxidant enzymes, such as superoxide dismutase, catalase, peroxiredoxins, GSH peroxidases, and thioredoxins (48). Thus, the maintenance of the oxidative balance enables cancer cells to perform their biological functions such as proliferation, differentiation, and migration (49–51).

Chemotherapy agents are usually used after a surgical excision as the treatment of choice to eradicate Minimal Residual Disease (MRD) in high-risk stage II and stage III patients and to increase the overall survival rate in stage IV patients (62–65). However, the only use of chemotherapy as standard-of-care ( Table 1 ) can represent a limit due to the high systemic toxicity, unsatisfactory response rate, the onset of drug resistance, and the low tumor-specific selectivity. Therefore, massive investments have been earmarked to develop new approaches to improve patient outcomes. The identification of point mutations in specific oncogenes (KRAS, NRAS, and BRAF), amplification of human epidermal growth factor receptor 2 (HER2), the MSI status, the DNA mismatch repair status (deficiency or proficiency), has provided a framework for finding additional approaches, as well as new prognostic perspective (66). Up to today it is possible to hit the cancer more effectively, by administering the most suitable biological agents with the standard chemotherapy taking into account the genetic setting of patients (67, 68). In this regard, targeted therapies (i.e., antibodies and small molecules) and immunotherapies, which actively or passively target the patient’s immune system, are widely used in combination with FOLFOX or FOLFIRI as a first-/second-line setting or alone as a third-line setting to improve the overall survival (OS) and progression-free survival (PFS) of advanced/metastatic cancer patients (69, 70). Moreover, antitumor immunity exerted by vaccines, specialized dendritic cells or new generation of cytotoxic T cells are currently under investigation in clinical trials (71, 72) ( Table 2 ).

In patients with left-sided KRAS wild-type tumors, for instance, the administration of anti-EGFR (i.e., cetuximab or panitumumab) in combination with standard-of-care chemotherapy as a first-line setting shows improvement in both OS and progression-free survival (PFS) (73–75). Additionally, anti-EGFR can be used alone in chemo-refractory patients with advanced CRC (76).

Recently, in patients with BRAF-mutated mCRC, the use of anti-EGFR in combination with a selective inhibitor of BRAF kinase (encorafenib) and a reversible inhibitor of the kinase activity of mitogen-activated extracellular signal-regulated kinase 1 (MEK1) and MEK2 (binimetinib) has been proposed as the third line treatment to improve the prognosis (77, 78).

Among the first- and second line interventions for mCRC, also the VEGF inhibitors (i.e., Bevacizumab, Aflibercept) in combination with standard-of-care chemotherapy contribute to improve OS and PFS in patients (79, 80).

Another targeted therapy to treat advanced/metastatic CRC refractory to all standard treatments is represented by the diphenylurea-based multikinase inhibitor (i.e., Regorafenib). This anti-tumoral drug targeting multiple protein kinases regulating angiogenesis, proliferation, immunity, and metastases, increases OS of heavily pre-treated patients (79, 81–83).

KRAS-targeted drugs, such as sotorasib and adagrasib, are emerging for their anti-cancer activity in heavily pre-treated patients harboring the KRAS^G12C mutation. These drugs are small molecules that keep KRAS in its inactive state, allowing apoptosis. However, their use in CRC treatment is still under investigation (84).

The main limiting factor for cancer patients is the onset of multi-drug resistance (MDR), which makes cancer cells tolerant to anti-cancer drugs. In fact, the combined chemotherapy and the development of different administration schedules are not always sufficient to avoid these issues due to the biological complexity of the tumor (103).

Tumor MDR is a highly-complex phenomenon that encompasses a plethora of molecular mechanisms involving not only cancer cells, but also infiltrating cells (e.g., endothelial, hematopoietic, and stromal cells) and the resulting tumor microenvironment (104). The constant interactions between tumor cells and their surrounding stroma result in alterations of many different cellular processes. Moreover, the presence of a sub clonal variation among cancer cells allows greater adaptability of the tumor to therapy, promoting its evolution.

Multiple molecular mechanisms have been identified as contributing factors to MDR development. Among these, the interplay between pre-existing and drug-induced mechanisms, including defects in the apoptotic machinery, mitochondrial dysfunction, altered autophagy activity, aberrant cell signaling, reduction in drug concentration and genetic and epigenetic changes, plays a significant role (105–108).

Moreover, the major cause of primary therapy resistance is represented by unresponsive subpopulations, such as cancer stem cells (CSCs) that can increase by up to 30% following long-term drug treatment (109).

Stochastically CSCs are distributed within tumors, but preferably they reside in specific niches, characterized by hypoxia, low pH, and fewer nutrients, which in addition to conferring them stemness features, allow the generation of differentiated progenies (110, 111).

CSCs are frequently quiescent and poorly differentiated cell populations with a lower level of intracellular ROS that share with normal stem cells both properties (i.e., self-renewal, self-sufficiency, and differentiation), and stemness signaling pathways (e.g., Notch, Sonic hedgehog, WNT/β-Catenin, JAK/STAT, and NF-kB). Their origin is still debated but it has been suggested that, at the moment of tumor initiation, the acquisition of CSC phenotype from either transformed differentiated cells (stochastic model) or transformed tissue-resident stem cells (hierarchical model) is promoted by the overexpression of oncogenes and the inhibition of tumor suppressor genes (e.g., APC, TP53, TGFBR2, SMAD4, PTEN, and RAS). Instead, following chemotherapy or radiotherapy regimen, new CSCs derive from either non-CSC subpopulations or therapy-induced senescent tumor cells (112).

Standard chemotherapy is not a valid therapeutic option for CSCs because they can effectively counteract the chemotherapy-induced oxidative stress through their free radical scavenging systems, such as GSH, and overexpression of the anti-apoptotic protein B-cell lymphoma 2 (Bcl2) (113–115). Moreover, the enhancement of ATP-binding cassette (ABC) transporters and aldehyde dehydrogenase (ALDH) expression, the increased resistance to apoptosis, and the activation of DNA damage sensor and repair machinery contribute to give to CSCs a survival advantage against anti-cancer therapy (116). Additionally, CSCs can transiently and reversibly switch between epithelial and mesenchymal states and vice versa (i.e., epithelial-mesenchymal plasticity) via Wnt/b-catenin signaling (117, 118). Such versatility, consequently, results in metabolic reprogramming in cellular bioenergetics, where energy supply can alternatively depend on aerobic glycolysis or mitochondrial OXPHOS.

It has been shown that metformin potentially offers therapeutic advantages by inhibiting the mitochondrial respiration, forces CSCs to a metabolic shift from OXPHOS to glycolysis (119, 120). The temporary OXPHOS suppression renders CSCs more prone to apoptosis. However, tumor relapse under metformin treatment cannot be excluded since CSCs can acquire resistance mainly due to MYC overexpression, promoting a Warburg-like glycolytic phenotype (120). Also, progressive increase in glycolysis-derived lactate may promote the activation of proteases, leading to ECM degradation, and resistance to chemotherapy (121).

The release of IL-4 from colorectal CSCs promotes their survival and hampers the CD8+T cell-mediated antitumor immune response, while the presence of inflammatory cytokines, including IL-1, IL-4, IL-6, IL-8, IL-10, and TGF-β, fuels an inflammatory loop, via Stat3/NF-κB pathways, for stimulating the self-renewal of CSCs (118, 122). Moreover, the tumorigenic and self-renewal capacity of CSCs also depend on the hyperactivation of β-catenin, Notch, and Hedgehog signaling pathways (123, 124).

Although several CSC biomarkers have been identified for CRC, their preclinical application is still unavailable due to the intrinsic features of CSCs, i.e., phenotypic heterogeneity, and the influence of the TME or CSC behavior. In this regard, previous studies have focused the attention on both the CSC-related signature and immune cell infiltration as important prognostic factors. The correlation between infiltrating portion of immune cells, i.e., tumor immune microenvironment (TIME), and hallmark gene sets may represent a possible starting point for developing CSC-targeted therapeutic strategies (122, 125).

Increasing evidence suggests that the use of nanoscale nanoparticles (NPs) as drug delivery systems (DDS), including liposomes, nanoemulsions, hydrogels, multifunctional inorganic materials (e.g., carbon nanotubes, gold nanoparticles, quantum dots), and peptides, could provide a novel therapeutic approach useful in overcoming MDR and improving the pharmacokinetics and biodistribution of anticancer compounds, resulting in reduced side effects (126). These NPs refers to nanometer scale systems (10-1000 nm) capable of protecting encapsuled molecules from degradation and passively or actively delivering drugs, small molecules, proteins, peptides, DNA and RNA into specific targets. However, their bio-distribution and clearance in the body depend not only by the NP chemical, physical and biological properties (e.g., size, stability, surface charge, solubility, and so on) but also by factors, such as the administration route (e.g., intravenous, oral, pulmonary and dermal administration) and host environment (e.g., pre-existing inflammation) (127).

Recently, it has been reported that NPs can accumulate in tumor tissues by passive or active delivery. Passive delivery of drug-loaded nanoparticles (i.e., the Enhanced Permeability and Retention, EPR, effect) is mainly due to fenestrated and immature new tumor vessels (128–131) while, the active delivery is due to a ligand-binding mechanism (e.g., nanoparticles targeting EpCAM, the folate receptor, EGFR and CD44) (132–134) ( Figure 3 ). However, regardless of NP delivery, nanoparticle-protein complex, namely protein corona (PC), can permanently change the NP fate. The protein profile of the corona complex does not have a standard composition, because it varies not only among NPs of different chemical designs but even across the NPs of the same type. This latter is explained by the so-called Vroman effect, in which protein turnover depends not only by the high-affinity binding of proteins, but also on their exchange kinetics (135). In general, on the basis of the binding affinity between plasma proteins and NP surfaces a “hard corona” and a “soft corona” are distinguished, respectively (136). Moreover, proteins participating in the complex influence the cell recognition pathway by the reticuloendothelial system (RES) and promote biological processes against NPs, including aggregation, opsonization, and phagocytosis. Therefore, either second-generation NPs or PEGylation technique enhance the effect of cancer therapy by ensuring drug delivery within the tumor and evading phagocytosis (137). In this regard, emerging self-assembly pH-sensitive pegylated nano drug delivery systems, namely HA-mPEG-Cis NPs, are able to target CD44-CRC-positive cells and dissolve the hydrated PEG in the acidic tumor environment. These drug delivery systems improve drug circulation time and tumor targeting while reducing the side effects of the loaded drug (138) ( Figure 3 ).

Inorganic nanocarrier, of controlled size and shape, such as gold NPs (AuNPs), show a certain versatility of use, including chemical sensing, imaging, and drug delivery due to their favourable optical and physical properties coupled with a reasonable biocompatibility with regard to biological environment (139). Interestingly, AuNPs, including gold nanorods, nanocages, nanostars, nanocubes, and nanospheres, find application in photodynamic therapy (PDT) for their specific physical features (i.e., optical and Surface Plasmon Resonance properties, proton-capture cross-section) (140) ( Figure 3 ).

Lipid-based NPs are already FDA-approved for various therapeutic purposes, including cancer treatment (e.g., Doxil^®, DaunoXome^®, Myocet^®, DepoCyt^®, Marqibo^® and Onivyde^®), severe infections or immunocompromised conditions (e.g. AmBisome^®) and RNAi therapeutic (Onpattro^®) (127, 141) ( Figure 3 ).

Liposomes are small-size vesicles consisting of an outer lipid bilayer, synthetic or natural, and an aqueous core, widely used to encapsulate/entrap drugs or nucleic acids (i.e. gene therapy) (142). Currently, the manipulation of liposome lipid membrane components (e.g., neutral and/or negatively charged lipids plus cholesterol, sphingomyelin plus cholesterol, hydrogenated soy phosphatidylcholine plus cholesterol) as well as specific key parameters (e.g., size and shape) has improved their biological performance, in term of enhanced delivery efficiency, maximizing so-called nano-bio interactions (143).

Nanoemulsions (NEs) are another system to deliver hydrophobic drugs and hydrophilic or hydrophobic compounds through different routes of administration (e.g., aerosols, ingestion, and injections). NEs are made as single (i.e., oil-in-water [o/w], water-in-oil [w/o]) or dual (w/o/w, o/w/o) emulsions with biocompatible and FDA-approved biodegradable oils (143). Previous in vitro studies have shown that natural active compounds encapsulated within NEs, acting synergistically with chemotherapy, can improve the therapeutic value of treatment despite the use of a lower dosage of drug (144, 145). Also, the entrapment of active or cytotoxic drugs within nanoemulsions can be useful to sensitize CSCs to apoptosis (146) ( Figure 3 ).

Over the past decade, nanotechnology has been widely explored to develop cytotoxic drug carriers. Although further improvements are needed, different types of NPs are already considered reliable systems for drug delivery due to their ability in targeting the tumor before releasing the drug.