Cannabinoids and Endocannabinoid System Changes in Intestinal Inflammation and Colorectal Cancer

2.1. Introduction to the Cannabinoid System

To accomplish their action, cannabinoids mainly bind to seven-transmembrane G_i/o-coupled receptors (GPCRs), usually the inhibitory type [21,40,41]. When cannabinoids bind to CB1 or CB2 receptors, there is a decrease in levels of cyclic adenosine monophosphate (cAMP) and expression of adenylate cyclase [40,41]. CB1 and CB2 receptors share only 44% of the protein homology and 68% in the transmembrane domains with binding sites for cannabinoids [42].

In in situ receptor/G protein reconstruction techniques, the activation of CB1 receptors results in high-affinity interactions with both G_i and G_o, whereas CB2 activation results in high-affinity interaction with G_o [43]. For CB1 receptors, the synthetic compounds HU210, WIN 55,212-2, and endocannabinoid anandamide (AEA) have a maximum activation effect on the G_i cascade. The stimulation of G_o is achieved only by HU210, whereas AEA, Δ9-THC, WIN 55,212-2 only have a 60–75% effect on the G_o compared to HU210 (partial agonists) [43]. Similar data were obtained for CB2 receptors; in the ligand interaction experiments, CB2 stimulation is induced by HU210, which demonstrates maximal activation (a full agonist). Other compounds showed submaximal effects on CB2 receptors, with WIN 55,212-2 having 64%, AEA 42%, and Δ9-THC 44% (partial agonists) [43]. Such agonists induce different conformational changes in CB receptors, thereby allowing us to choose the CB targeted therapy that modifies G-protein signaling more selectively [43]. One of the pathway for cannabinoid action is comprised of GPCR kinase-3 and β-arrestin-2, which causes desensitization and relates to the development of tolerance that results in CB1 internalization [44]. Receptor desensitization is one of the molecular mechanisms that underlies the onset of cell tolerance to a particular stimuli [44] and can become one of the reasons for cannabinoid treatment resistance.

Ionic channels are placed in the membrane as multimeric complexes that form passage pathways for selected molecules triggered by mechanical or chemical signals. Cannabinoids can stimulate A-type potassium channels, which causes an increased efflux of potassium from the cell. Moreover, by binding to CB1 receptors, cannabinoids may inhibit N- and P/Q-types of voltage-dependent calcium channels and activate inwardly, rectifying potassium channels, which results in decreased calcium influx and increased potassium efflux from the cells [44,45,46,47,48]. The ionic channels that respond to cannabinoids are transient receptor ion channels (TRPs), such as TRPV1, TRPV2, TRPM8, and TRPA1 [45,46,47,48,49].

In addition to CB1 and CB2, there are multiple other receptors that may respond to cannabinoids [40]. The most studied receptors are GPR119, GPR55, peroxisome proliferating activated receptor α (PPARα), and PPARγ [35,49,50]. We will discuss some receptors and pumps listed in different sections throughout this review regarding their effects on CRC and intestinal inflammation.

2.2. Endocannabinoid System

The endocannabinoid system (ECS) includes CB receptors, cannabinoid enzymes, and endocannabinoids. Endocannabinoids are mostly represented by arachidonoyl ethanolamine, or anandamide (AEA), and 2-arachidonoylglycerol (2-AG), which are derivatives of membrane phospholipids [21,40] and are produced by multiple pathways. Thus, the inhibition of enzymes that take part in their synthesis will not always result in changes in endocannabinoid levels and may affect the amounts of other cell mediators [51,52]. Endocannabinoids are synthesized from cell membrane’s phospholipids on demand due to intracellular calcium elevation. Both endocannabinoids, AEA and 2-AG, signal through GPCRs, ion channels (TRPs), and nuclear receptors (PPARs) [53]. PPARs are the subfamily receptors that act with retinoic X receptors of the nuclear hormone receptor superfamily, regulating the expression of target genes by binding to peroxisome proliferator response elements in the genes [54].

The first endocannabinoid discovered was AEA [55]. AEA is a part of the larger family of N-acylethanolamines, and 2-AG belongs to a family of 2-acylglycerols. There are four main routes of AEA synthesis: N-acyl-phosphatidiyl-ethanolamine-hydrolyzing phosphatase D (NAPE-PLD) [56], NAPE-phospholipase C followed by phosphatase [57], dual hydrolysis of the acyl groups by the phospholipase D, and (Lyso)-N-acylphosphatidylethanolamine lipase (ABHD4) followed by hydrolysis by glycerophosphodiester phosphodiesterase 1 (GDE1) [58,59,60]. Alternatively, AEA can be produced by α/β lysosomal hydrolase 4. After AEA accomplishes its action, the cell reuptakes it for enzymatic degradation by fatty acid amid hydrolase (FAAH) [61], or monoacylglycerol lipase (MAGL) [62,63,64]. The third route of AEA degradation is COX-2, which results in the production of prostamides [65]. Finally, the fourth degradation process of AEA is via N-acylethanolamine-hydrolyzing acid amidase (NAAA) [66].

AEA is a partial agonist of CB1 receptors [51]. Furthermore, AEA binds to CB2 with low efficacy and can act as an antagonist of CB2. AEA is present in the extracellular space and is accumulated in cells via facilitated diffusion due to its transmembrane concentration gradient and does not require ATP and sodium [51]. In a low CB receptor expression, AEA can be an antagonist to high efficacy agents. However, CB2 receptors can be induced up to 100 fold by AEA, especially during tissue injuries or inflammation [67,68]. When CB1 receptors are expressed at a normal level, AEA can directly stimulate hydrolysis of PIP2, releasing inositol-1,4,5-phosphate (IP3), which eventually causes Ca²⁺ release from ER [69]. Activation of the PI3K pathway may indirectly affect levels of PIP2 and regulate calcium release within cells [70]. Except for CB1 and CB2 receptor interaction, AEA also binds to L-type calcium channels that can produce non-CB-mediated effects in vivo [71,72].

Another endocannabinoid, 2-AG, is produced from phospholipids in a two-step process. First, by the enzyme phospholipase C, 1,2-diacylglycerol is produced; next, diacylglycerol lipase (DAGL) converts it to 2-AG. The degradation of 2-AG is similar to AEA and it is held by MAGL, COX-2, lipoxygenase, or cytochrome P450 enzymes [73]. Additionally, 2-AG is an agonist of CB1 and CB2 receptors [68]. The experimental inactivation of endocannabinoid degradation pathways, such as blocking MAGL, may result in CB receptor desensitization in the tested tissues with subsequent elevation of 2-AG [74,75].

Knowledge of endocannabinoid’s synthesis and degradation process would help us understand the adaptational and pathophysiological changes of ECS in intestinal inflammation and malignancies, as well as develop newer and better treatment options for these diseases.

2.3. Normal Intestinal Cells and Their Functions

The normal large intestinal lumen consists of epithelial cells organized into anatomical structures called crypts of Lieberkühn [76]. Crypts are the source of intensively renewing and proliferating cells. The crypt base is composed of actively dividing stem cells, from which most intestinal cell lineages differentiate [4,76]. Colonocytes are the most common crypts cell type taking part in intestinal absorption. The other cells present in the crypts include mucin-producing Goblet cells and Tuft cells, which sense the intestinal content, enteroendocrine cells that produce hormones as a reaction to internal and environmental stimuli, and the microfold cells transporting luminal antigens [76,77].

The signaling pathway that plays a crucial role in the proliferation, maintenance, and differentiation of intestinal cells is Wingless-related integration site (WNT)/β-catenin. This pathway stimulates the division and differentiation of stem cells [4]. Moreover, it is important to emphasize that WNT, Notch, bone morphogenic protein (BMP), and transforming growth factor β (TGF-β) signaling are essential for the homeostasis of normal colon epithelial cells, and they are often dysregulated in intestinal malignancies [78]. Some studies have shown that aging is associated with increased genomic instability within intestinal crypts, even in the histologically healthy epithelium. Over time, deletions, translocations, duplications, and gene-conversions accumulate, forming a strong niche for CRC development [8,79].

3.1. Colitis and Changes in ECS

Intestinal biopsies of patients with different inflammatory diseases such as IBDs, diverticulitis, and celiac disease revealed higher expressions of cannabinoid receptors and endocannabinoids than in intact intestinal tissues. The experiments provided by D’Argelio et al. (2006) showed that there is a two-fold elevation of AEA in patients with untreated UC, which can correlate with the activity of the disease [133]. Transcriptome analysis performed by Grill et al. (2019) revealed that AEA, OEA, and 2-AG were elevated in patients with IBDs; however, only PEA and OEA were elevated in CRCs [20]. Additionally, in the CD study group, CB1 and GPR119 receptor transcription were significantly lower, emphasizing the protective role of cannabinoid receptors in intestinal inflammation [20]. In contrast, the CB2 expression in CD was increased, especially in the damaged intestinal crypts [91,134]. Other experiments showed higher levels of CB2, DAGLα, and MAGL expression in mild and moderate forms of colitis, although in the light type of intestinal inflammation, CB1, CB2, and DAGL-α were decreased and NAPE-PLD was elevated, especially in patients treated with aspirin or aspirin and corticosteroids. At the cellular level, MAGL and FAAH were particularly increased in the immune cells during acute pancolitis, but dropped after treatment with non-steroidal anti-inflammatory drugs (NSAIDs) [89]. Thus, studies showed that in case of light forms of colitis there were increased expressions of CB receptors and higher productions of endocannabinoids; however, in moderate types of colonic inflammation caused by IBDs, CB1/CB2 expression decreases, as do endocannabinoid levels.

Some studies pointed out that the crucial response of GIT to ECS activation depends on CB1 receptor expression levels. These receptors mediate protective reactions against colonic inflammation. In animal models representing colitis-associated tumors, pharmacological stimulation of CB1 and CB2 receptors alleviated signs of experimental bowel inflammation [133,135,136]. For instance, in the mustard oil colitis model, direct activations of CB1 and CB2 receptors by their agonists (arachidonoyl-chloro-ethanolamide and JWH-133) revealed their protective roles on intestinal mucosa [137]. The inhibition of cannabinoids degrading FAAH and MAGL enzymes was shown to be protective against colitis induced by dextran sulfate sodium (DSS) and trinitrobenzene sulfonic acid (TNBS) [138]. Intrarectal infusion of 2,4-dinitrobenzene sulfonic acid (DNBS) and oral administration of DSS, which promotes the development of colitis in animals [135], showed that in CB1-deficient mice, there was a significantly stronger inflammatory response to experimental colitis than there was in the wild type [135]. Moreover, treatment of wild-type mice with cannabinoid antagonist SR141716A mimicked the CB1-deficient phenotype, and administration of CB1 receptor agonist HU210, or genetic inhibition of FAAH, protected against DNBS-induced colitis in the mice [135]. The synthetic CB receptor agonists WIN 55,212-2 had protective roles in DSS-induced colitis by inhibiting the p38/MAPK pathway [139], which may have significantly suppressed pro-inflammatory responses in the GIT. The experiments on intestinal epithelial cells, described by Izzo et al. (2009), have shown that cannabinoids also induced wound healing by CB1 activation and suppression of cytokine release via CB2 receptors [134]. Moreover, in the CB1-deficient mice, the electrophysiological recordings of circular smooth muscles showed spontaneous oscillatory action potentials, indicating the role of CB1 in early induced irritation of the intestines [135], and release of acetylcholine in the myenteric plexus that innervates the gut, leading to decreased intestinal motility [140]. In croton oil-induced ileitis, CB1 expression was high, which is associated with reduced transit time [124]. In contrast, LPS-induced intestinal propulsions were ameliorated by the induction of CB2 receptors [141]. Thus, activation of CB1 and CB2 receptors decreased visceral sensitivity due to colonic distention and inflammatory stimuli [142], and could reduce intestinal contractility [134,140,143].

3.2. Gatekeeping Mechanisms of ECS

The previously mentioned “gatekeeping” mechanism of endocannabinoids in the intestines is achieved by suppressing cell-mediated immunity through T-helper 1 cells and stimulating humoral immune response via T-helper 2 cells. These effects are mostly regulated by CB2 receptors signaling [132,144]. Animals with deficient CB2 receptors had decreased levels of intestinal natural killer cells and CD4+ memory cells [145], which may favor the development of intestinal malignancy due to the evasion of adaptive immune response [146]. The immunomodulatory role of cannabinoids was shown in experiments with the administration of Δ-9-THC into mice followed by infection with Legionella pneumophila, which resulted in inhibition of T-helper 1 activation. The effects of Δ-9-THC were accompanied by the suppression of cytokines such as IL-12 and IFN-γ, and increased levels of IL-4. Such changes of immune signaling molecules stimulated humoral immune response and suppressed cellular immune response. Additionally, CB1 agonist SR 141716A attenuated inhibition of IL-12β, and CB2 agonist SR144528 increased GATA3 mRNA expressions in the mice’s spleen. These results indicated an enhancement of T-helper 2 cells upon cannabinoid treatment [132]. In another study, the plant-derived THC had an inhibitory effect on TNF-α-induced release of pro-inflammatory cytokine IL-8 in HT-29 CRC cell line [147]. In addition, endocannabinoids played a pertinent role in maintaining immune tolerability by controlling the expansion of a regulatory pool of T lymphocytes and suppressive CX3CR1^hi macrophages [85,148]. Macrophages and plasma cells in the human large intestine express both CB1 and CB2 receptors that take part in immune reactions [91]. In the inflamed gut, activated macrophages, monocytes, and dendritic cells increased the production of endocannabinoids (AEA, 2-AG) and may reduce increased gut permeability caused by pro-inflammatory stimuli [149].

Other in vitro experiments supporting the hypothesis of the barrier-maintaining function of ECS in the gut showed that OEA could modulate permeability of intestinal cells. PEA is a ligand to PPAR-α with anti-inflammatory and analgesic action [54]. PPAR-α reduces inflammation via induction of IkB-α that inhibits nuclear translocation of NFκB, which is a pro-inflammatory transcription factor that induces TNF-α expression, which recruits immune cells to the damaged tissues [150]. PEA is an endocannabinoid-like compound synthesized from membrane phospholipids via N-acyltransferase and NAPE-PLD. The degradation pathway of PEA is maintained by FAAH and N-acylethanolamine-hydrolyzing acid amidase (NAAA) [151]. NAAA is highly expressed in intestinal leukocytes [152], and its inhibition can alleviate intestinal inflammation [153].

One of the underlying mechanisms of how cannabinoids achieve anti-inflammatory and anti-tumor effects in the large intestine is by alleviating the inflammatory-related release of TNF-α, IFN-γ, IL-1β, and IL-6 mediators, which shows their protective role against intestinal inflammatory diseases and CRCs [83] (see Figure 1). The large intestines’ inflammatory processes are associated with glial cell responses by the expression of Toll-like 4 receptors (TLR4) and S100B [151,154]. In DSS-induced colitis, the anti-inflammatory effect of PEA is mediated via inhibition of enteroglial-specific S100 protein. The S100 protein promotes macrophage recruitment in the intestinal mucosa, induces an inflammatory response, and interacts with TLR4. PEA targets the S100B/TLR4 axis through PPAR-α, which results in inhibition of NFκB in enteric glial cells through p38/ERK/JNK signaling [154]. Endocannabinoids PEA and OEA act as agonists of GPR55 (PEA) and GPR119 (OEA) to enhance the immune effects of AEA [155,156]. In DSS and DNBS models of IBDs, PEA administration reduced microscopic signs of inflammation, neutrophil infiltration, and COX-2, PGE2, and inducible nitric oxide synthase (iNOS) expression via CB2, GPR55, PPAR-α signaling, and TRPV1 channels [157]. Thus, the addition of PEA attenuated intestinal permeability and inflammation as well as increased colonic cell proliferation by elevating TRPV1 and CB1 expression levels [157]. The antiproliferative effects of PEA on cancer cells was achieved by elevation of AEA levels within the cells due to inhibition of FAAH. Subsequently, the elevated levels of AEA caused stimulation of vanilloid receptors type 1 [158,159]. It was also shown that OEA and PEA increased transepithelial electrical resistance of Caco-2 cells by 20–30% via TRPV1 and PPARα receptors. OEA and PEA may induce cytoskeletal changes and activate focal adhesion kinase (FAK) and ERK1/2 and as a result, decrease epithelial permeability. These endocannabinoid-like compounds also reduce levels of Src kinases, aquaporins 3 and 4, and activated potassium channels [116]. Downregulation of main degrading enzyme FAAH results in the elevation of OEA and PEA, leading to a significant decrease in intercellular permeability. OEA acts via TRPV1 (apical application) and PEA (basolateral membrane application) through PPAR-α receptor signaling [116]. Furthermore, in experiments provided by Alhouayek et al. (2015), endocannabinoid PEA inhibited inflammation-associated angiogenesis via CB1, CB2, GPR55, PPAR-α, and TRVP1 receptor interactions [153].

GIT plays an essential role in systemic inflammatory response, endotoxemia, and sepsis due to its barrier protective function against translocation of bacteria and their toxins into the bloodstream [115]. Thus, maintaining gatekeeping property of the intestines in some cases can help preventing the development of toxin-induced multiorgan failure [115]. The study performed by Espinosa-Riquer et al. (2019) proposed that 2-AG, together with CB2 receptors, play a role in the development of endotoxin tolerance in mice bone-marrow-derived mast cells [160]. Mast cells can become hyporeactive if TLR4 receptors are stimulated by endotoxins such as LPS during prolonged periods of time. The authors showed that 2-AG administration induced endotoxin tolerance via inhibition of NFκB pathway and TNF secretion. Additionally, the production of 2-AG is TLR4 dependent, which further indicates that 2-AG participates in negative autocrine regulation of mast cells activation. More to that, inhibition of CB2 receptors prevented development of endotoxin tolerance. Thus, mast cell-related branch of innate immune response can be partially regulated by ECS [160]. Another study showed that CB1 and CB2 agonism prevented LPS-induced GIT motility and reduced IL-6 levels [161]. In mice model of LPS-induced sepsis, addition of CBD caused decrease of S100B protein levels, which consequently abrogated the hyperactivation of intestinal glial cells, macrophages, and mast cells [162].

On the other hand, elevation of levels of endocannabinoids such as AEA and 2-AG can lead to concentration-dependent increase of intestinal permeability in inflammation and hypoxia via CB1 receptors activation [163]. In Caco-2 cells, 2-AG levels elevated under hypoxic conditions and inflammatory responses. In human intestinal samples, there are increased macrophage colony-stimulating factor IL-12, IL-13, and IL-15 under normal, inflammatory, and hypoxic conditions [163]. The study by Karwad et al. (2017) showed that AEA and 2-AG played an important role in intestinal permeability in vitro and ex vivo [163]. However, the intraperitoneal administration of AEA in mice with TNBS-induced colitis decreased macro- and microscopic scores of colitis and reduced immune cell infiltration. Additionally, inhibition of FAAH decreased the accumulation of inflammatory cytokines and inhibited leukocyte proliferation in DNBS-experimental colitis [164,165]. Another experiment showed the pro-inflammatory effect of AEA via TRPV1 receptors. AEA has been proven to elevate myeloperoxidase activity in the rat ileum. Thus, it can become pro-inflammatory; however, to activate TRPV1 pumps, AEA concentrations should be much higher than those needed for CB receptors activation [96].

5.1. Changes in Cannabinoid Receptors

Intact enterocytes express CB1 receptors; however, during intestinal inflammation or carcinogenesis, levels of CB1 receptors progressively decrease due to CpG island hypermethylation of the CB1R promoter’s transcription sites [90]. In contrast, CB2 receptors’ expression increases, which in some cases is associated with poor prognosis of CRC patients [90]. The experiments performed Wang et al. (2009) on 10 CRC cell lines (HCT-116, HT-29, LS-174T, SW-480, Colo-201, DLD-1, Caco-2, HCT-15, HCA-7, LoVo) showed that HCT-116, HT-29, and LS-174T have low CB1 expression. In the SW-480 cell line or normal colon tissues, there was no CB1 receptor change. Additionally, CB1 receptor loss was indicated in 8 out of 13 human tumor samples, showing that in human biopsies of the second and third grade of CRC, there was loss of expression of CB1 receptors often via CpG island promoter hypermethylation [16]. The transcriptome analysis of 566 CRC patients showed reduction of CB1 expression in the TNM-I stage; however, with the disease’s progression, CB1 expression was again elevated. In contrast, GPR55 expression decreased with disease progression, compared to healthy colon samples [36]. Moreover, CpG methylation of the CB1 promoter site was hypomethylated in the majority of healthy intestinal tissues and CRC patients—a cohort of 86 [36]. These findings suggest that translational regulation of miRNAs may be the critical point of CB1 receptor expression changes in CRC patients [36]. Additionally, CB1-deficient APC-mutated mice had 2.5–3.8-fold elevated polyp formation, compared to the control mice [16]. However, the deletion of CB2 receptors did not significantly affect the formation of intestinal preneoplastic lesions in animal models [16]. The DLD-1 CRC xenograft model had higher CB2 levels of expression, and treatment with CB2 agonist such as N-cyclopentyl-7-methyl-1-(2-morpholin-4-ylethyl)-1,8-naphthyridin-4(1H)-on-3-carboxamide (CB13) significantly decreased tumors in size [12].

The prognostic value of the cannabinoid system in CRCs was proposed by Zhang et al. (2017), who detected the methylation status of 2245 CB1R promoter regions in peripheral blood samples taken from patients suffering from colon adenocarcinomas and adenomas, as well as healthy individuals [186]. Comparing CB1 promoter epigenetic changes between groups strongly suggested that CRC progression is positively associated with CB1R methylation. Moreover, the methylation at the 2245 locus significantly correlated with tumor size, depth of invasion, tumor stage, and lymphatic node metastases. Thus, it may be used as a prognostic marker and a screening tool for CRCs [186]. In another study, scientists analyzed 534 CRC patients’ immunohistochemistry samples, which showed the CB1 receptor to be expressed in 76.6%. Low CB1 expression (<66%) was often identified in patients with stage IV CRC. At this stage, low CB1 levels were associated with poorer prognosis. Despite the tumor grade, there were no significant differences between patients’ age, gender, histological differentiation, and tumor site between high and low CB1 expression [187].

The study provided by Tutino et al. (2019), which involved samples from 59 CRC patients, showed low expression of CB1 receptors in primary tumors and adjacent mucosa, especially in those with diagnosed metastatic disease. In metastatic CRC, downregulation of CB1 expression was associated with decreased p38/MAPK and ERK1/2 signaling in both tumor tissue and adjacent normal mucosa. Additionally, there was significant upregulation of prosurvival AKT pathway in the same samples, especially in the surrounding tumor tissues that were intact. Moreover, patients with metastatic disease had significantly lower levels of Bcl-2-associated-X protein (BAX), which is responsible for the stimulation of apoptosis [17].

Some studies showed that CB receptor expression changes play a procancer role in colon carcinogenesis [188,189,190]. Analyses of CRC samples (both tumor front and interior) for CB1 expression from 487 patients that underwent surgical resection showed that the levels of CB1 presence is associated with tumor grade. Additionally, the authors suggested that high CB1 expression is associated with poorer prognosis in stage II microsatellite stable tumors [189]. The data comparison set were variables including gender, tumor site, radiotherapy, stage, tumor differentiation, type, microsatellite stability, lymphocyte infiltration, and frequency of tumor aggregates at the invasion front. Surprisingly, the significant differences associated with CB1 receptor expression were histological tumor grade and microsatellite stability. In both tumor front and center samples, the CB1 receptor expression was higher in patients with moderate-poorly/poorly differentiated microsatellite-stable CRCs [189]. One of the explanations for the relationship between poor patients prognosis and high CB1 expression was proposed by studies on glioblastoma cell lines [190]. This study suggested that at low levels of expression, CB receptors are mainly coupled with ERK, thus, their activation caused apoptosis. It was also shown that high CB1 expression could cause additional AKT signaling activation that switches proapoptotic signaling to survival mechanisms. Conclusively, endocannabinoid levels could protect intestinal mucosa from damage; however, they may also exacerbate cancer cell survival [189]. Thus, alterations in CB receptor expression can be used as a prognostic marker in different molecular subtypes of CRC.

Recent data provided by Hasenoehrl et al. (2018) have shown that cannabinoid receptor GPR55 activation has pro-cancer effects by stimulating tumor invasiveness and promoting metastatic potential [36]. Moreover, there is evidence that blocking GPR55 with CBD activates MAPK/p53 signaling and stimulate ERK1/2, which can cause apoptotic cell death [191]. GPR55 is a Gα12/13 and Gq lysophosphatidyinositol type of receptor, stimulating proliferation, invasion, and angiogenesis of cancer cells [192,193]. GPR55 can activate a cascade of reactions involving calcium mobilization [194], ERK1/2 phosphorylation [195], adhesion, and migration of colon cancer cells that may lead to liver metastasis [196]. Except for cancer cells, GPR55 is expressed by macrophages, neutrophils, and lymphocytes [192,197,198]. In the azoxymethane/dextrate sulfate sodium (AOM/DSS) animal model of CAC, activation of GPR55 caused changes in myeloid-derived suppressor cells and T lymphocytes that are usually present within the tumor’s microenvironment. GPR55 can increase proinflammatory molecules such as COX-2 and STAT3, which can assist in tumor initiation and progression. It was shown that GPR55 has the opposite effect to the CB1 receptor, which is considered to be protective against colonic inflammation. Knocking down the GPR55 receptors in mice models increased the levels of CD4+ and CD8+ cells in the tumor beds, which emphasizes the role of GPR55 in inflammation-induced colon cancers [36]. The study presented by Hasenoehrl et al. (2018) showed that GPR55-/- mice had decreased tumor burden when compared to the wild type [36]. In the experiment using the GPR55-/- colitis-associated colon cancer mice model, the levels of COX-2, STAT3, thromboxane A2, PGF2α and myeloid cell-recruiting chemokine monocyte chemoattractant protein-1 were lower than in the wild-type mice. As opposed to cytokines IL-5, IL-10, and IL-12, which were elevated in the GPR55-/- model [36]. As opposed to GPR55-/-, the CB1-/- knockout mice developed a higher number of tumors, with larger areas in different CRC models, i.e., spontaneous tumor progression and CAC. Additionally, the colon’s nontumor parts exposed to AOM/DSS had higher expressions of CB1 compared to the healthy nonexposed ones. GPR55 mRNA levels were decreased in nontumor exposed tissues and elevated in tumor lesions, compared to healthy controls [36]. Additionally, provided analysis of the available CRC patient dataset, it was revealed that patients with increased GPR55 expression have shorter relapse-free survival [36]. These results show the importance of increased expression of GPR55 in the tumors’ microenvironment regulation, inflammation, and cancer progression.

The research performed by Raup-Konsavage et al. (2018) showed the sensitivity of different molecular subtypes of CRC cell lines (SW480, SW620, HT-29, DLD-1, HCT-116, LS-174T, RKO) to 370 different cannabinoid compounds [15]. It was found that cell lines SW480, SW620, HT-29, and DLD-1 had APC mutations and HCT-116 and LS-174T had activating mutations in CTNNB1, the β-catenin encoding gene. SW620 is derived from lymph metastases. The identified synthetic cannabinoids that suppressed CRC cell viability most effectively were—HU-331; CP 55,940; 5-epi-CP 55,940; CP 47,497; 3-epi-CP 47,497 C-8 Homolog; CP 47,497 C-8 Homolog; PTI-1; PTI2; and NPB-22. The selected compounds did not work through canonical signaling, including CB1, CB2, GPR55, and TRPV1 [15]. The most significant result was that the cell lines with APC mutations (SW480, HT-29, DLD-1). They were more sensitive to CBD than the cells mutated in the β-catenin pathway (HCT-116, LS-174T) [15]. These results suggest that various molecular subtypes of CRC may react differently to cannabinoid treatment and that the antitumor action of cannabinoid compounds is not always CB1- or GPR55-dependent.

6.1. Ceramide

Ceramide is a neutral lipid backbone of complex sphingolipids. Its de novo synthesis can be activated by chemotherapy, ionizing radiation, and enzyme sphingomyelinase. Ceramide action is specific to its carbon chain lengths [211,212]. Ceramide can trigger apoptosis and inhibit cancer cell proliferation by causing cell cycle arrest. It can also activate autophagy in cancer cells. The main pathways in ceramide interaction are protein phosphatase 2A, p38/MAPK, JNK, AKT, protein kinase C, and survivin [213]. Some cancers upregulate ceramide-degrading enzymes to avoid death or even promote mutagenicity [213].

One of the best-explained anticancer mechanisms of cannabinoids is the activation of the de novo synthesis of ceramide via CB receptor activation (see Figure 3) [12,214]. Due to the intensive synthesis of ceramide, the production of ROS is enhanced, leading to ER stress response. Next, ER stress-related signaling events may cause CRC cell death. First, eukaryotic translation initiation factor 2α (eIF2α) is downregulated, and the global translation of proteins is decreased. Simultaneously, the C/EBP homology protein (CHOP) is activated; this protein acts on pseudokinase tribbles-homologue 3 (TRIB3), which stimulates the release of proapoptotic BAD and BAX proteins [215]. Moreover, AKT is downregulated by CHOP. AKT inhibition causes major intracellular changes, such as the downregulation of the mammalian target of rapamycin (mTOR) and the activation of autophagy. In addition, AKT can directly activate caspase 3 and caspase 9 and stimulate G1 cell cycle arrest through cyclin-dependent kinase inhibitors, such as p21 and p27. The anti-tumor mechanism of cannabinoids also involves the upregulation of AMP protein kinase (AMPK), which, along with low mTOR, strongly stimulates the macroautophagy of CRC cells [12,16,26,27,191,201,216].

Cianchi et al. (2008) performed a study regarding CB receptor expression in human specimens (24 samples of primary sporadic adenocarcinoma and adjacent tissues), DLD-1, and HT-29 CRC cell lines, which showed CB1 expression mainly in normal colonic epithelial tissue samples [12]. In addition, the tumor tissues highly expressed CB2 receptors. The activation of CB1 by the synthetic agonist arachinodyl-2′-chloroethylamide and the CB2 agonist CB13 caused the stimulation of apoptosis by de novo ceramide synthesis in intestinal cancer cells. This study also showed that TNF-α connects CB receptor activation and ceramide synthesis in CRC cell lines, which activate apoptosis. CB1 receptor activation elevates intracellular ceramide levels via sphingomyelin hydrolysis by coupling with factors associated with neutral sphingomyelinase activation that bind to TNF receptors, resulting in sphingomyelin breakdown and ceramide de novo production [12,217]. This signaling is antiproliferative and proapoptotic to CRCs [12].

Chen et al. investigated the connection between cannabinoid signaling and ceramide, and described the profiles of the main endocannabinoids AEA and 2-AG, ceramides, free fatty acids, and the critical enzymes of cannabinoid metabolism in 47 pairs of human CRC samples and adjacent non-cancerous tissues [199]. Results showed that AEA and its metabolite, arachidonic acid, are elevated in CRC tissues and are mainly associated with lymphatic node metastases. In the CRC samples, the ceramide levels have different expression patterns, with elevated C16 and C24 and decreased C18 and C20. In addition, the mRNA levels of NAPE-PLD, FAAH, and ceramide synthases, such as CerS2, CerS5, and CerS6, are higher in cancer tissues [199]. Other experiments have shown that C16 and C24 ceramides promote apoptosis of CRC cells [218,219]. In summary, elevated levels of AEA, ceramides, and CB1 receptors are shown to have a protective effect against colon carcinogenesis [198].

6.2. Apoptosis

In CRCs, the RAS-MAPK pathway is overactivated, with KRAS and BRAF being overexpressed in approximately 50% and 15% of cases [220,221]. PI3K/AKT signaling is upregulated in almost 40% of colon malignancies [222]. In 2007, Greenhough et al. (2007) reported that in vitro THC-treated adenoma (AA/C1, AN/C1, BH/C1, RG/C2, AAC1/SB/10C) and CRC (SW480, HCT-15, HT-29, Caco2, HCT-116, LS-174t, SW620, and JW2) cell lines induce apoptosis via proapoptotic BAD activation by its dephosphorylation on serine 112 and 136 [26]. These effects are achieved by inhibiting the major cancer survival pathways—RAS/MAPK, ERK1/2, and PI3K/AKT via CB1 receptor activation [26]. However, THC does not affect p38/MAPK and JNK signaling [26]. The provided data showed that the induction of CB1, but not CB2 receptors, by THC can result in CRC cell death [26]. On the contrary, glioblastoma and lung carcinoma cell line treatment with nanomolar concentrations of THC may even promote cancer cell growth, which depends on metalloproteinase and EGFR activity. EGFR receptor signaling is the mechanistic link with cannabinoid receptors and can activate the pro-survival AKT pathway via the shedding of pro-amphiregulin and pro-heparin-binding epidermal growth factor-like growth factor by the TNF-α converting enzyme TACE/ADAM17. The experimental results showed that THC leads to the phosphorylation of EGFR, the phosphorylation of adaptor protein Src homology 2 domain-containing, and the subsequent activation of ERK1/2 and AKT/PKB pathways. EGFR transactivation with CB1/2 receptors requires EGFR tyrosine kinase and metalloproteinase activity. As a result, higher THC concentrations can induce apoptosis in multiple cell lines, although THC can accelerate cancer cell progression in nanomolar concentrations [223].

CBD is a partial agonist of CB1 and CB2 receptors [224,225]; it stimulates TRPV1, TRVP2, 5-HT1A, and PPAR γ, inhibits GPR55, and increases endogenous AEA concentration by blocking its hydrolysis [226]. One of the best described anticancer effects of CBD is the activation of NOXA, suppressing mTOR/AKT signaling and MAPK [28]. Recent studies performed on HCT-116 and DLD-1 CRC cell lines indicated that CBD can induce apoptosis via the significant upregulation of NOXA-ROS signaling [28]. ROS can induce ER stress response by triggering unfolded protein response (UPR) [227,228]. The ER stress can stimulate UPR to restore protein homeostasis. UPR is guided by the signaling proteins inositol-requiring protein-1α (IRE1α), protein kinase RNA-like ER kinase (PERK), and activating transcription factor 6 (ATF6). Usually, PERK and ATF6 are kept inactive by binding to the binding immunoglobulin protein (BIP) chaperone. IRE1α is activated directly under unfolded proteins and then starts to accumulate. When UPR is activated, all three proteins are signaled to increase the levels of chaperones, decrease translation, and transport misfolded proteins back into the cytosol for ubiquitination and subsequent degradation [228]. In CBD-treated CRC cells, the expression of the stress-related ER gene is decreased, with further activation of NOXA [28]. Activating transcription factor 3 (ATF3) and ATF4 may be involved in CHOP and NOXA stimulation [229]. The addition of CBD stimulates ER response, which results in ATF3 and ATF4 binding directly to ATF/cAMP response element in the promoter region of NOXA and CHOP [28]. As a result, NOXA migration into mitochondria causes the release of cytochrome c; the further activation of caspase 3, caspase 8, and caspase 9; the cleavage of PARP; and the initiation of apoptosis in CRC in vivo and in vitro models [28]. In addition, CBD may enhance the phosphorylation of p38 stress protein kinase, which eventually leads to apoptosis [210].

The combination of CBD with TNF-related inducing apoptosis ligand (TRAIL) causes a synergistic effect of the two molecules on CRC in vivo. CBD treatment activates ER stress response with CHOP release and phosphorylated protein kinase RNA-like ER kinase (PERK). In addition, CBD stimulates the expression of molecules responsible for the extrinsic apoptotic pathway by stimulating DR5 expression. The addition of 4 μM CBD to 10 ng/mL TRAIL potentiates the effect of TRAIL by sensitizing CRC cells to undergo TRAIL-induced apoptosis in a xenograft mouse model [230]. Moreover, CBD suppresses the expression of the inhibitors of apoptosis (IAPs) c-FLIP and survivin [230]. IAPs are a group of antiapoptotic molecules that suppress caspase activity [231]. One of the IAPs, survivin, is overexpressed in merely every tested tumor and may serve as a promising target molecule for CRC treatment [232]. CBD may exhibit a protective role against CRC by the stimulation of CB1 receptors, which causes the inhibition of cAMP-dependent protein kinase. This process leads to the reduction of the cdc2 (Wee1/cdc25C-cdc2 cascade), which leads to the destabilization of survivin, the activation of caspase 3, and apoptosis [16]. These findings show that cannabinoids can become efficient preventive agents in canonical molecular subtypes of CRCs [16].

Another phytocannabinoid, cannabigerol (CBG), a partial agonist of CB1 and CB2 receptors [233], inhibits endocannabinoid reuptake [33]. CBG is also a potent HT5_1A antagonist [233]; an agonist of TRPA1, TRPA2, and TRPV2; and an antagonist of TRPM8 [33]. CBG has chemopreventive effects in AOM-induced colon carcinogenesis via the activation of apoptosis, the stimulation of free radical formation, the upregulation of CHOP, and the inhibition of CRC cell growth. These effects are enhanced by CB2 antagonists and TRPM8 agonists [234].

6.3. Extracellular Vesicles

Extracellular vesicles are classified into exosomes, microvesicles, and apoptotic bodies [235]. Exosomes and microvesicles mediate intercellular communications by carrying molecules from parental cells to recipient cells. These lipid-bilayer vesicles can affect the physiology of migration, differentiation, and angiogenesis of cancer cells [236,237,238]. Vesicular release is regulated by membrane receptors, apoptotic signals, and intracellular calcium release [239]. It was shown that apoptotic bodies can transfer oncogenes horizontally, resulting in cancer cell survival [240]. Thus, tumor-derived exosomes prepare a pre-metastatic niche in specific organs [241]. Kosgodage et al. (2018) showed that CBD can inhibit cancer-derived extracellular vesicles release in a dose-dependent fashion. The effect is associated with alterations in mitochondrial functions, the modulation of STAT3 signaling, and changes in prohibitin expression [242]. As a result, CBD may sensitize cancer cells to chemotherapy drugs via the alteration of the biogenesis of extracellular vesicles [242].

6.4. Autophagy

Autophagy is the self-consumption mechanism that eliminates intracellular waste, attenuates stressful factors, and exhibits anti-carcinogenic effects [243]. One of the conventional chemotherapy drugs used in CRCs is oxaliplatin, which causes the formation of DNA crosslinks, usually between guanines and guanine-adenine, effectively killing cancer cells [244]. However, 40% of patients with CRC may develop resistance to it [245]. Jeong et al. (2019) showed that CBD can overcome oxaliplatin resistance through the activation of autophagy and the inhibition of superoxide dismutase 2, a main antioxidant enzyme within a cell [246]. Moreover, the authors observed decreased phosphorylation of nitric oxide synthase 3 (NOS3), resulting in reduced NO and ROS production [246]. The study suggested that NOS3 phosphorylation is indispensable for the development of oxaliplatin resistance. Oxaliplatin resistance can be overcome by the addition of CBD to the treatment, which results in autophagy-mediated cell death and the formation of free radicals by dysfunctional mitochondria in resistant cells [246]. Combined CBD treatment with oxaliplatin causes the activation of the microtubule-associated protein 1A/1B light chain 3B and the increased expression of p63 [246]. These markers are commonly used to assess autophagy [247]. In addition, the combination of CBD and oxaliplatin decreases the number of mitochondria in the resistant cells and reduces the levels of cardiolipin and NADH dehydrogenase 1α subcomplex subunit 9 (mitochondrial complex I), resulting in abnormal oxidative phosphorylation and autophagy-mediated cancer cell death [246]. Other studies regarding drug resistance and cannabinoids indicated that THC, CBD, and CBN can inhibit ATP binding cassette family transporters, P-glycoprotein, and the breast cancer resistance protein BCRP [248], resulting in the potential chemosensitizing effect of cannabinoids in resistant CRCs [249,250,251].

6.5. CRC Angiogenesis and Metastasis

Cannabinoids can inhibit the invasion and metastasis of cancer cells by downregulating vascular endothelial growth factor (VEGF), matrix metalloproteinase 2 (MMP2), MMP9, and the adhesion molecule E-cadherin [252]. In a xenograft and AOM model of colon carcinogenesis, Pagano et al. (2017) showed that 2-AG and MAGL, a serine hydrolase that degrades 2-AG, are highly expressed in aggressive colon cancers. The inhibition of MAGL by URB602 decreases xenograft tumor volume through the downregulation of VEGF and fibroblast growth factor-2 (FGF-2). This study has shown that 2-AG exerts an anti-tumor effect in colon cancer via the inhibition of angiogenesis (VEGF) and cell proliferation (cyclin D1). Moreover, in a mouse AOM model of colon carcinogenesis, URB602 attenuates the formation of preneoplastic lesions, such as polyps, thus supporting the chemopreventive role of endocannabinoid 2-AG in CRC [252]. The in vitro experiments provide evidence that 17β-estradiol stimulates CB1 expression via the activation of the estrogen receptors ERα and ERβ in the primary tumor CRC cell lines DLD-1 and HT-29 and the lymph node metastatic cell line SW620. The authors suggest the antiproliferative effect of estrogens on primary and metastatic CRCs via interaction with polyamine and growth factors in the tumor [134,253].

In another study, the strong antiangiogenic effect of the cannabinoid-like compound LYR-8 was demonstrated on a xenograft model using chick chorioallantoic membranes. The mechanism behind cannabinoid action is the suppression of VEGF, COX-2, and hypoxia-inducible factor α (HIFα) [254]. One of the synthetic cannabinoids, HU-311, which is a quinone of CBD, shows antiangiogenic effects by stimulating apoptosis in endothelial cells and inhibiting topoisomerase II [255]. Some experiments have also demonstrated that 12 μM CBD may induce the migration of HUVECs. CBD inhibits MMP 2, MMP 9, and tissue inhibitor of metalloproteinase 1, which results in the suppression of cell motility and the invasion of endothelial cells; also, CBD inhibits urokinase-type plasminogen activator (uPA) and serpin E1/plasminogen activator inhibitor 1, which are involved in the degradation of the extracellular matrix and contribute to cancer cell invasiveness. Moreover, CBD downregulates HIF1α in U87 cells, which suggests the suppression of cell survival, motility, and angiogenesis [256]; endothelin 1, PDGF-A [257]; and the reduction of STAT5-induced vasorelaxation [256,258]. In addition, the antimetastatic action of CB receptor agonists have been shown on SW480 cell lines. AEA, HU-210 (non-selective CB agonists), and docosatetraenoylethanolamide (CB1 selective agonist) suppress the norepinephrine-induced migration of human CRC cells [134,259].

6.6. Irinotecan and THC

Prester et al. (2018) suggested that the combination of the chemotherapy drug irinotecan with THC potentiates the toxic effects of chemotherapy. Irinotecan, the topoisomerase I inhibitor, is one of the most commonly prescribed chemotherapy for metastatic CRC [260]. In this study, rats were injected intraperitoneally with irinotecan and subsequently THC, which resulted in more prominent leukopenia than irinotecan injections alone. This study shows that THC cannot alleviate irinotecan’s side effects, including leukopenia and diarrhea [260]. However, further experiments are needed to investigate the potential combinational value of cannabinoids and topoisomerase inhibitors. Moreover, treatment with THC decreases the levels of aspartate aminotransferase, highlighting its hepatoprotective role. Notably, cannabinoids are easily bound to plasma lipoproteins; consequently, they can interact with protein-bound drugs, including chemotherapeutics [260], which may alter the effects of conventional chemotherapy.

We would like to express our special appreciation to Sylvain J. Boutros for writing assistance, technical editing, language editing, and proofreading during the writing of this manuscript.