The Effects of Endogenous Cannabinoids on the Mammalian Respiratory System: A Scoping Review of Cyclooxygenase-Dependent Pathways

Abstract

Introduction:

The endogenous cannabinoid (endocannabinoid) system is an emerging target for the treatment of chronic inflammatory disease with the potential to advance treatment for many respiratory illnesses. The varied effects of endocannabinoids across tissue types makes it imperative that we explore their physiologic impact within unique tissue targets. The aim of this scoping review is to explore the impact of endocannabinoid activity on eicosanoid production as a measure of human airway inflammation.

Methods:

A scoping literature review was conducted according to PRISMA-ScR (Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews) guidelines. Search strategies using MeSH terms related to cannabinoids, eicosanoids, cyclooxygenase (COX), and the respiratory system were used to query Medline, Embase, Cochrane, CINAHL, Web of Science, and Biosis Previews in December 2021. Only studies that investigated the relationship between endocannabinoids and the eicosanoid system in mammalian respiratory tissue after 1992 were included.

Results:

Sixteen studies were incorporated in the final qualitative review. Endocannabinoid activation increases COX-2 expression, potentially through ceramide-dependent or p38 and p42/44 Mitogen-Activated Protein Kinase pathways and is associated with a concentration-dependent increase in prostaglandin (PG)E₂. Inhibitors of endocannabinoid hydrolysis found either an increase or no change in levels of PGE₂ and PGD₂ and decreased levels of leukotriene (LT)B₄, PGI₂, and thromboxane A₂ (TXA₂). Endocannabinoids increase bronchial epithelial cell permeability and have vasorelaxant effects in human pulmonary arteries and cause contraction of bronchi and decreased gas trapping in guinea pigs. Inhibitors of endocannabinoid hydrolysis were found to have anti-inflammatory effects on pulmonary tissue and are primarily mediated by COX-2 and activation of eicosanoid receptors. Direct agonism of endocannabinoid receptors appears to play a minor role.

Conclusion:

The endocannabinoid system has diverse effects on the mammalian airway. While endocannabinoid-derived PGs can have anti-inflammatory effects, endocannabinoids also produce proinflammatory conditions, such as increased epithelial permeability and bronchial contraction. These conflicting findings suggest that endocannabinoids produce a variety of effects depending on their local metabolism and receptor agonism. Elucidation of the complex interplay between the endocannabinoid and eicosanoid pathways is key to leveraging the endocannabinoid system as a potential therapeutic target for human airway disease.

Introduction

Endogenous cannabinoids (endocannabinoids) regulate many biological processes, with great interest in their ability to modulate pathological inflammation. The type-2 cannabinoid receptor (CB₂) is widely distributed in immune cells and has drawn interest as a potential therapeutic target in conditions such as cancer, neurodegenerative disorders, and respiratory diseases such as Aspirin-Exacerbated Respiratory Disease.^1–3 Since their discovery, two key endocannabinoids—anandamide (AEA) and 2-arachidonoylglycerol (2-AG) have been studied. While 2-AG is a nonselective complete agonist for CB₁ and CB₂, AEA is a partial agonist with a higher affinity for CB₁.⁴ This selectivity gives each endocannabinoid unique therapeutic potential. Synthetic cannabinoid modulators can take advantage of this as well. Selective CB₂ agonists such as HU 308 or JWH 133 are designed to target inflammation without producing psychoactive effects.⁵

The endocannabinoid and the eicosanoid system interact primarily through two ways: direct metabolism of endocannabinoids into prostaglandin (PG)-like compound by cyclooxygenase-2 (COX-2) and hydrolysis of endocannabinoids into arachidonic acid (AA). While 2-AG, AEA, and the eicosanoid precursor AA are synthesized from membrane lipids, they utilize different lipid precursors allowing for production of distinct, but closely related mediators based on environmental stimuli.^6,7 Both 2-AG and AEA can serve as natural substrates for COX-2 to produce prostaglandin glycerol esters (PG-Gs) and prostaglandin ethanolamide (PG-EA), respectively.^8,9 Although some of these modified PGs appear to bind the same PG receptors as the AA-derived products, many of them exert physiological effects that are unique from those of the corresponding PGs through distinct receptors.^10,11 Additionally, AEA and 2-AG can be degraded into AA, the primary substrate for eicosanoid enzymes such as COX and lipoxygenase (LOX).¹²

Modulation of the endocannabinoid system has been shown to impact the levels of eicosanoids and COX activity in neurons, and the endocannabinoid receptors CB₁ and CB₂ have been shown to influence cancer biology and immune responses in a variety of tissues.^13–15 Activation of the endocannabinoid pathway is also associated with a variable reduction in eicosanoid and COX activity. For example, one study showed while cotreatment of CB₁ antagonist and CB₂ inverse agonist reduced COX activity in mouse myoblasts, prevention of CB₂ activation alone in the setting of AEA resulted in increased COX activity, consistent with the notion that CB₂ activation primarily produces anti-inflammatory effects, whereas CB₁ plays a proinflammatory role.¹⁶

To complicate the relationship further, there is considerable variation in the interplay between these systems across different tissue types. Distribution of CB₁ and CB₂ receptors varies greatly by tissue type. Likewise, eicosanoids and their receptors have been shown to have a variety of effects depending on the tissue being studied. In the airway, prostaglandin E₂ (PGE₂) has been shown to decrease inflammation through interaction with the EP₄ receptor.¹⁷ However, PGE₂ has also been shown to increase inflammation in diseases such as rheumatoid arthritis, atherosclerotic vascular disease, and colon cancer.^18–20 Eicosanoid dysregulation is a key aspect of inflammatory airway diseases, and a solid understanding of the varying interactions between eicosanoids, endocannabinoids, and their receptors across tissue types is necessary for the development of novel therapeutics.

This scoping review therefore aims to explore the effects of endocannabinoids and their modulators on the eicosanoid systems and airway physiology within the mammalian respiratory system.

Methods

A scoping review was conducted according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses Extension for Scoping Review (PRISMA-ScR) guidelines.²¹ Study was exempt from IRB review. Covidence software (Veritas Health Innovation, Melbourne, Australia) was used to manage the search results and iterative review. Search strategies using MeSH terms related to endocannabinoids, eicosanoids, COX, and the respiratory system were used to query PubMed (pubmed.gov), Embase (embase.com), Cochrane Central (cochranelibrary.com), CINAHL (Ebscohost.com), and Web of Science Classic (webofscience.com) on December 17, 2021. Full search strategies are provided in Supplementary Appendix SA1.

Search results were limited to English language publications. Primary, peer-reviewed studies that examined the effects of endocannabinoid ligands and their modulators on eicosanoid levels in mammalian respiratory tissues were included (Fig. 1). Studies that occurred before 1992 were excluded, as CB₁, the first endogenous cannabinoid receptor, was described at this time.²²

FIG. 1.

PRIMSA flow diagram for included studies.

All studies were independently screened by two authors. Any conflicts were resolved by the senior author. A standardized data extraction template was used to collect findings from each included study (Supplementary Appendix SA2). Recorded parameters included endocannabinoid studied, tissue or cell type, physiological effects, change in COX expression and eicosanoid levels, and proposed mechanisms of action. Discrepancies between study findings are reported.

Results

A total of 16 studies were identified for inclusion in this qualitative, scoping review (Table 1). These studies uncovered an array of changes in the respiratory system following endocannabinoid exposure. With respect to COX-2 expression, two studies found that exposure to methanandamide—the synthetically created and metabolically stable analog of AEA—led to increases in COX-2 expression (Table 2). In the human lung carcinoma cell line A549, this increase was attributed to activation of ceramide synthase as the effect could be reversed by the ceramide synthase inhibitor fumonisin B. Increased levels of ceramides were hypothesized to promote COX-2 synthesis. However, in mice, increased COX-2 expression was attributed to the p38 and p42/44 mitogen-activated protein kinase (MAPK) signaling pathway as shown by the attenuation of COX-2 expression in the presence of p38 or p42/44 inhibitors.

Table 1.

Characteristics of Included Studies

Study	Year	Intervention	Tissue type	Design	Measure outcome
Andersson et al.61	2002	Anandamide	Guinea pig bronchus, lung parenchyma	Ex vivo	% Contraction of bronchi and lung parenchyma
Baranowska-Kuczko et al.56	2012	Anandamide	Rat pulmonary artery	Ex vivo	% Relaxation of preconstricted pulmonary artery
Bottemanne et al.46	2019	ABHD6 inhibition	Mouse lung tissue	In vivo	PGD2, PGE2 levels
Craib et al.60	2001	Anandamide	Guinea pig bronchus	Ex vivo	% Contraction of bronchi
Eichele et al.26	2009	Methanandamide	Human lung carcinoma cell line A549	In vitro	COX-2 expression, DNA fragmentation apoptosis assay
Gardner et al.25	2003	Methanandamide	Mouse Lewis lung carcinoma, alveolar cell carcinoma	In vivo, In vitro	COX-2 expression, PGE2 level, tumor size
Karpińska et al.54	2018	LPI	Human pulmonary artery	Ex vivo	% Relaxation of preconstricted pulmonary artery
Kozłowska et al.55	2008	Virodhamine	Human pulmonary artery	Ex vivo	% Relaxation of preconstricted pulmonary artery
Nomura et al.43	2011	MAGL inhibition/KO	Mouse lung tissue	In vivo	AA, PGD2, PGE2 levels
Shang et al.63	2016	AEA	Human bronchial epithelial cell line Calu-3	In vitro	TEER
Stengel et al.62	2007	AEA	Guinea pig bronchus	In vivo	Excised lung gas volume
Wahn et al.58	2005	AEA, 2-AG	Rabbit pulmonary artery	Ex vivo	Change in pulmonary artery pressure (mmHg)
Wartmann et al.42	1995	AEA	Human fetal lung fibroblast WI-38	In vitro	PGE2 level
Xiong et al.44	2018	MAGL inhibition	Mouse lung tissue	In vivo	LTB4, PGI2, TXB2 levels
Yin et al.45	2019	FAAH inhibition	Rabbit lung tissue	In vivo	LTB4, PGI2, TXA2 levels, lung injury score
Zhou et al.23	2021	PLE inhibition	Pig lung tissue	In vivo	PGD2, PGE2 levels

2-AG, 2-arachidonoylglycerol; AA, arachidonic acid; ABHD6, α/β-hydrolase domain-containing 6; COX-2, cyclooxygenase-2; FAAH, fatty acid amide hydrolase; KO, knockout ; LPI, lysophosphatidylinositol; LTB₄, leukotriene B₄; MAGL, monoacylglycerol lipase; PGD₂, prostaglandin D₂; PGE₂, prostaglandin E₂; PGI₂, prostacyclin; PLE, pig liver esterases; TXA₂, thromboxane A₂; TXB₂, thromboxane B₂.

Table 2.

Endocannabinoid Effects on COX-2 and Eicosanoid Levels

COX-2 levels
Study	Intervention	Assay	COX-2							Proposed Mechanism
Eichele et al.26	MA	Western Blot	↑							Ceramide dependent upregulation
Gardner et al.25	MA	Bradford assay, ELISA	↑							P38 and p42/44 MAPK signaling pathway
Eicosanoids Levels
Study	Intervention	Assay		PGD2	PGE2	LTB4	TXA2	TXB2	PGI2	Proposed Mechanism
Bottemanne et al.46	ABHD6 inhibition	LC/MS		↔	↑					ABHD6 inhibition causes an increase in 2-AG and increased PGE2
Gardner et al.25	MA	ELISA			↑					COX-2 upregulation through p38 and p42/44 MAPK pathway
Nomura et al.43	MAGL inhibition/KO	LC/MS		↓	↓					Blocked hydrolysis of 2AG to PG precursors
Wartmann et al.42	AEA	RIA			↑					GPCR activation of MAPK pathway leading to AA increase
Xiong et al.44	MAGL inhibition	ELISA				↓	↓	↓		MAGL inhibition leads to lower levels of AA and downstream products
Yin et al.45	FAAH inhibition	ELISA				↓	↓		↓	FAAH inhibition leads to lower levels of AA and downstream products
Zhou et al.23	PLE inhibition	LC/MS		↔	↔					PLEs hydrolyze 2AG leading to increased PGs and inflammation. PLE inhibition attenuated LPS-induced inflammation in vitro but did not alter levels of PGs in lungs in vivo

ELISA, enzyme-linked immunosorbent assay; GPCR, G protein-coupled protein receptor; LC/MS, liquid chromatography with mass spectrometer; LPS, lipopolysaccharide; MA, methanandamide; MAPK, mitogen-activated protein kinase; PG, prostaglandin; RIA, radioimmunoassay.

Downstream eicosanoid products are also altered by AEA and methanandamide exposure (Table 2). Endocannabinoid activation by these products led to increases in PGE₂ levels in both human cell line and mice. This effect was attributed to COX-2 upregulation or the activation of MAPK and cytosolic phospholipase A2. However, indirect endocannabinoid activation through monoacylglycerol lipase (MAGL) or fatty acid amide hydrolase (FAAH) inhibition led to an increase in 2-AG and AEA levels and subsequent downregulation of several eicosanoid products. This resulted in an increase in PGE₂ and no change in PGD₂ with a decrease in leukotriene (LT)B₄, thromboxane B₂ (TXB₂), TXA₂, and PGI₂. 2-AG, as well as the inhibition of pig liver esterase (PLE)—another enzyme that hydrolyzes 2-AG, blunted inflammatory response by lipopolysaccharide (LPS) in pig alveolar macrophages in vitro, however, no significant changes in PGD₂ and PGE₂ were detected in in vivo porcine lung tissues following exposure to LPS and inhibition of PLE.²³

Seven physiological effects of endocannabinoids on the mammalian respiratory system were measured: isolated lung parenchymal contraction, bronchial patency, pulmonary gas trapping, pulmonary vascular tone, pulmonary arterial pressure, tumor growth, bronchial epithelial permeability, and measures of pulmonary function with lung injury score (Table 3). Endocannabinoids increased epithelial permeability, bronchial contraction, and tumor size while they dilated pulmonary arteries and reduced pulmonary gas trapping induced by the eicosanoid LTD₄.

Table 3.

Physiological Effects of Endocannabinoids on the Mammalian Respiratory System

Study	Intervention	Tissue type	Physiological effects	Proposed mechanism
Airway
Andersson et al.61	AEA	Guinea pig bronchus	Contraction of bronchus in the presence of indomethacin	Activation of vanilloid receptors in the presence of COX inhibition independent of FAAH hydrolysis or CB receptor antagonism.
Craib et al.60	AEA	Guinea pig bronchus	Contraction of bronchus in the presence of indomethacin	LOX-dependent activation of vanilloid receptors in the presence of COX inhibition that is independent of CB1 or EP1 receptor antagonism.
Stengel et al.62	AEA	Guinea pig bronchus	Decrease excised lung volume as a measure of pulmonary gas trapping	Inhaled, but not intravenous, AEA attenuated LTD4-mediated pulmonary gas trapping, but had no effects on histamine-mediated pulmonary gas trapping.
Yin et al.45	FAAH inhibition	Rabbit Lung	Improved PaO2 level, PaO2/FiO2, hemodynamic parameters, lung wet-to-dry ratio, and lung injury scores	Decreased formation of AA and its downstream inflammatory metabolites through inhibition of AEA hydrolysis.
Epithelial barrier
Shang et al.63	AEA	Human bronchial epithelial cell line Calu-3	Decrease TEER	FAAH-dependent hydrolysis of AEA to COX/LOX metabolites.
Lung parenchyma
Andersson et al.61	AEA	Isolated lung parenchyma	Contraction of lung parenchyma	Hydrolysis of AEA through FAAH to COX-dependent formation of bronchoconstrictive eicosanoids
Pulmonary vasculature
Baranowska-Kuczko et al.56	AEA	Rat pulmonary artery	Vasorelaxation of isolated pulmonary artery preconstricted with U-46619	Direct activation of endothelial CB receptor and FAAH-dependent COX-mediated prostacyclin-like vasoactive metabolites in KCa- and NO-dependent manner.
Karpinska et al.54	LPI	Human pulmonary artery	Vasorelaxation of isolated pulmonary artery preconstricted with 5-HT	Direct agonism to GPR55, endothelial CB receptors and through NO- and KCa-dependent PPARγ pathways.
Kozlowska et al.55	Virodhamine	Human pulmonary artery	Vasorelaxation of isolated pulmonary artery preconstricted with phenylephrine	Direct agonism to endothelial CB receptors and FAAH-dependent COX-mediated formation of endogenous prostanoids.
Wahn et al.58	AEA, 2-AG	Rat pulmonary artery	Increase pulmonary arterial pressure	Activation of EP1 receptor by FAAH-dependent, COX-dependent metabolites.
Tumor
Eichele et al.26	MA	Human lung carcinoma cell line A549	Increase DNA fragmentation (apoptosis) of cancer cells	Ceramide-dependent upregulation of COX-2-dependent, L-PGDS-derived PGD2 exerting apoptotic effects through PPARγ receptor activation.
Gardner et al.25	MA	Mouse Lewis lung carcinoma, alveolar cell carcinoma	Increase tumor size	COX-2-dependent upregulation of PGE2 production in tumor through p38 and p42/44 MAPK signaling pathways.

5-HT, 5-hydroxytryptamine or Serotonin; AEA, anandamide; CB1, type 1 cannabinoid receptor; EP1, Prostaglandin E2 receptor 1; GPR55, G protein-coupled receptor 55; KCa, calcium-activated potassium channels; LOX, lipoxygenase; L-PGDS, lipocalin-type prostaglandin D synthase; LTD₄, leukotriene D4; NO, nitric oxide; PPARγ, peroxisome proliferator-activated receptor gamma; U-46619, a thromboxane A2 receptor agonist.

Within the lung parenchyma and conducting airway, activation of vanilloid receptors and increase in COX and FAAH metabolites are primarily responsible for the physiological effects seen, without direct involvement of endocannabinoid receptors. By contrast, direct endothelial cannabinoid receptor agonism is primarily responsible for the effects seen in pulmonary vasculature following exposure to endocannabinoid agonists. Changes in metabolites of COX are also associated with improved lung injury scores and increased human lung tumor cell apoptosis in vitro. However, a separate study found that changes in COX metabolites caused an increase in mouse lung tumor size in vitro.

Discussion

The eicosanoid system is an immunomodulatory pathway by which endocannabinoids impact airway inflammation. It is well established that endocannabinoids may be metabolized into AA, an eicosanoid precursor, in addition to acting on endocannabinoid receptors.²⁴ However, this review suggests that endocannabinoids influence eicosanoid activity at multiple levels, as highlighted below.

Upregulation of COX-2 expression in mammalian airway cells

Upregulation of COX-2 is the first potential means by which endocannabinoids exert their effects on eicosanoid production. Two studies, Eichele et al. and Gardner et al. found that exposure to methanandamide increased levels of COX-2 in both human (A549) and murine lung carcinoma cells. Two mechanisms have been proposed in these studies: activation of ceramide synthase and the p38 and p42/44 (ERK) MAPK signaling pathways. Inhibition of either pathway reversed the methanandamide-induced increase in COX-2 expression.^25,26 However, it is possible that these separate mechanisms work together to exert endocannabinoids' effect on COX-2 expression in respiratory tissues.

Several studies have shown that ceramide may induce COX-2 expression through p38 and p42/44 (ERK) MAPK signaling pathways in human A549 pulmonary cells, human mammary epithelial cells, and human dermal fibroblasts.^27–29 Interestingly, it has been reported that endocannabinoids increase ceramide levels through activation of the CB₁ receptor in human skin and nervous system.^30,31 Although no study directly investigating the relationship of CB₁ receptor activation and ceramide levels has been done in the mammalian respiratory system, the CB₁ receptor is indeed expressed on multiple respiratory cell types across different mammalian species.^32–34 Taken together, this suggests that CB₁ activation by endocannabinoids may lead to an increase of ceramide synthase activity, and subsequent increase in ceramide levels induce COX-2 expression through the p38 and p42/44 MAPK signaling pathways.

By upregulating COX-2 expression in the mammalian respiratory system, endocannabinoids may direct eicosanoid precursors toward eicosanoids with anti-inflammatory effects in the airway, such as PGE₂, and away from proinflammatory LTs. LTs have been shown to be the key drivers of respiratory spasmogenic activity seen in conditions such as asthma and chronic obstructive pulmonary disease by upregulating neutrophil migration and promoting contraction of airway smooth muscle.³⁵ On the other hand, it has been demonstrated that PGE₂ has bronchodilation and anti-inflammatory effects in asthmatic airways of both human and mammals.^36–40 Therefore, by regulating COX-2 expression in the respiratory system, the endocannabinoid system may modulate levels of inflammatory mediators and exert delicate control over airway inflammation.

Production of AA and its downstream eicosanoids

Besides regulating COX-2 expression, endocannabinoids also modulate inflammation by directly generating AA from endocannabinoid precursors such as 2-AG and AEA.⁴¹ One may therefore postulate that increased expression of endocannabinoids in mammalian airways would lead to an increase in downstream eicosanoid products; an observation that is supported by several studies in this review.

Two studies showed exposure to AEA and methanandamide led to increased PGE₂ levels in human A549 lung carcinoma cells and human fetal lung fibroblasts. However, neither attributed this effect to shunting of endocannabinoid metabolites into the AA pathway. Wartmann et al. proposed that AEA promotes MAPK driven phosphorylation of PLA2, leading to increased AA synthesis, whereas Gardner et al. proposed the increase in PGE₂ is due to upregulation of COX-2 through the p38 and p42/44(ERK) MAPK pathways described above.^25,42 While these methods undoubtably play an important role in eicosanoid synthesis, the conversion of endocannabinoids to AA remains influential to downstream eicosanoid production. This is demonstrated by several studies which use the endocannabinoid hydrolysis inhibitors MAGL and FAAH, resulting in reduced expression of the eicosanoids PGD₂, PGE₂, LTB₄, and TXA₂.^43–45

Of note, Bottemanne et al. showed the inhibition of ABHD6, another 2-AG hydrolase, showed a paradoxical increase in PGE2 levels in the lungs of treated mice.⁴⁶ In the brain, it has been shown that MAGL is responsible for the majority (85%) of 2-AG hydrolysis, whereas ABHD6 accounts for only 4%.⁴⁷ It is possible that the interplay between the members of different endocannabinoid hydrolases, which have yet to be elucidated in the respiratory system, may contribute to this paradoxical increase. Lastly, Zhou et al. showed that inhibition of PLE6—an endocannabinoid hydrolase—with bis(4-nitrophenyl) phosphate, produced no change in eicosanoid levels in the lungs. However, the study did find a significant decrease in PGE₂ and PGD₂ production in the spleen, liver, and duodenum, suggesting that PLE6 may play a more integral role in endocannabinoid hydrolysis in these tissues when compared with the airway.²³

The selective formation of PGE₂ may be key to the endocannabinoid system's potential anti-inflammatory properties in the mammalian airway. PGE₂ is a key eicosanoid thought to have anti-inflammatory and cytoprotective properties, including induction of an anti-inflammatory phenotype within the airway epithelium, and inhibition of noxious stimuli-induced fibrosis.⁴⁸ The airway epithelium is a key mediator in this process as it is a principal source of PGE₂ within the mammalian respiratory system along with alveolar macrophages and smooth muscle.⁴⁹ PGE₂ also modulates the effects of other inflammatory mediators. For example, while PGD₂ may induce bronchoconstriction, it has simultaneous anti-inflammatory effects such as reduced dendritic cell migration in the presence of PGE₂.⁵⁰ The authors of this review, therefore, propose two potential mechanisms for the selective increase of PGE₂ following endocannabinoid exposure.

First, eicosanoid levels may also be modulated by the direct metabolism of endocannabinoids by COX-2. As noted previously, 2-AG and AEA can be metabolized by COX-2 to PGH₂-G and PGH₂-EA, respectively. These modified PGs have similar but slightly different activity than their sister, PGH₂, which is produced by AA catabolism. The enzymes prostaglandin E synthase (PGES), prostaglandin-H2 D-isomerase (PTGDS), and prostaglandin I2 synthase (PTGIS) are all capable of metabolizing PGH₂-G and PGH₂-EA, in fact, PGES was found to metabolize these modified PGs with similar pharmacokinetics as AA-derived PGs. However, the thromboxane synthase enzyme (TXAS) is highly selective and has a very low affinity for these modified PGs. As a result, when endocannabinoids are the substrate for the COX-2/TXAS pathway, very low levels of thromboxane are produced. TXAS's lack of affinity for endocannabinoid metabolites explains why levels of TXA2 and TXB2 decrease following endocannabinoid exposure while PGE₂ levels rise.

Second, upregulation of microsomal PGE synthase (mPGEs) following exposure to endocannabinoids may favor PGE production. Studies have shown that mPGEs colocalizes with COX-2 in multiple tissues in rabbit and human.^51,52 As endocannabinoids upregulate COX-2 expression, it is possible that mPGEs expression increases in tandem. This phenomenon has not been described in the airway and poses a potential area for future study.

By increasing the synthesis of PGE₂, as seen in in human epithelial fibroblasts and mouse pulmonary epithelium, the endocannabinoid system may be able to mediate anti-inflammatory effects in the airway. Unfortunately, studies that describe an increase in PGE₂ did not look at levels of inflammatory cytokines such as TXA2 and LTB4, making it difficult to determine if endocannabinoid exposure truly favors anti-inflammatory eicosanoids. It is critical that future studies examine the relative expression of proinflammatory and anti-inflammatory eicosanoids as well as the levels of mPGEs following endocannabinoid exposure as this could elucidate an important mechanism for the reduction of airway inflammation.

Regulation of airway physiology

While endocannabinoid effects on COX-2 and eicosanoid levels were consistent across multiple studies, physiological effects are often bidirectional. This is, at least in part, the result of two phenomena. First, eicosanoids themselves have variable effects depending on tissue types and cellular environments. Second, endocannabinoids may exert their effects simultaneously through their targeted endocannabinoid receptors and through the eicosanoid systems depending on physiological conditions.

The effect of endocannabinoids on pulmonary vascular pressures exemplifies this bidirectional effect.^53,54 In ex vivo human and rat models, endocannabinoids are noted to have vasodilatory effects on preconstricted pulmonary arteries through direct activation of endothelial endocannabinoid receptors (CBe).^54–56 This effect is consistent with the vasodilatory effects of the endocannabinoid system seen in other organ systems such as the mesenteric arteries.⁵⁷ However, endocannabinoids were also shown to cause an increase in pulmonary arterial pressure, which is mediated by agonism of the prostanoid receptor EP₁.^58,59 By simultaneously acting on its receptors and regulating eicosanoid levels, the endocannabinoids system may be able to carefully modulate the pulmonary vascular response to a variety of stimuli.

A similar phenomenon is seen in the lung parenchyma and bronchi where eicosanoids play a key role. Unlike in the pulmonary vasculature, endocannabinoid receptors have little direct effect in the lung parenchyma, with downstream eicosanoids (COX and LOX metabolites) producing seemingly contrary effects.^60,61 First, inhaled AEA was shown to reduce gas trapping mediated by LTD₄. In this proinflammatory state endocannabinoid exposure may favor the formation of anti-inflammatory eicosanoids such as PGE₂ through the mechanisms previously described. However, this effect was not seen with intravenous AEA, suggesting endocannabinoids primarily act on the apical surface of the airway epithelium, consistent with the polar expression of CB1 and 2 receptors in airway epithelial models.³⁴

Additionally, FAAH inhibition improved lung injury scores after one-lung ventilation.⁶² The authors suggested this was due to decreased AA production, which is consistent with this review's findings that inhibitors of endocannabinoid hydrolysis reduce the production of proinflammatory eicosanoids.^43–45

However, AEA exposure also led to weak constriction of the bronchi and lung parenchyma, as well as a reduction in transepithelial electrical resistance (TEER) in human bronchial epithelial cells, indicating increased inflammation.^60,61,63,64 Authors postulate that vanilloid receptor agonism is responsible for contraction of the bronchi while AEA metabolism by FAAH to AA is responsible for reduction in TEER. These effects may be the result of PGE₂'s complex role in mammalian physiology. While PGE₂ does combat inflammation in the airway, it still has proinflammatory properties and may be responsible for this change in physiology. It is also possible that endocannabinoid exposure is associated with an increase in other eicosanoids such as TXA₂ and LTB₄. Regardless, it should be noted that in the absence of inflammatory mediators, endocannabinoid metabolites are weakly proinflammatory while still serving to significantly reduce inflammation in airway tissue exposed to noxious stimuli.

Lastly, methanandamide has been shown to both increase tumorigenicity in in vivo studies of murine lung carcinoma and promote tumor cell apoptosis in in vitro adenocarcinoma cells.^25,26 Apoptosis induction in vitro was attributed to increased PGE₂ production and suggests that PGE₂, although typically considered anti-inflammatory, is cytotoxic when produced in excess. The increased tumor growth seen in vivo suggests that endocannabinoids may promote tumorgenicity. Note that these opposite effects were seen in different experimental conditions, that is, in vivo versus in vitro. Interestingly, previous studies have noted these bidirectional effects of endocannabinoid on glioma cells in different stages of differentiation and transformation.⁶⁵ Future studies seeking to confirm and understand this association are needed before the therapeutic use of endocannabinoids can be considered.

Endocannabinoid exposure increased COX-2 production and PGE₂ synthesis while inhibitors of endocannabinoid hydrolysis decreased eicosanoid production. Through these effects, as well as the agonism of cannabinoid receptors in the vasculature and vanilloid receptors in the bronchi, endocannabinoids appear to mediate an array of complex, and often bidirectional physiological effects. Tissue type, the presence of inflammatory stimuli, and experimental conditions determine if proinflammatory or anti-inflammatory effects will predominate. Future studies should seek to characterize the variable effects of endocannabinoids on levels of individual eicosanoids. Additionally, as many of the physiological effects of endocannabinoids are bidirectional, more research is needed to understand what impacts their physiological outcome.

This review has several limitations. First, the number of publications exploring the relationship between the mammalian airway endocannabinoid and eicosanoid systems is relatively small. Second, no studies were identified that examine the effects of direct endocannabinoid exposure on levels of eicosanoids other than PGE₂, making it difficult to fully understand their impact on airway inflammation. These are both important areas for future study. Finally, it is possible that some studies may not have been detected in our literature search and thus not accounted for in this narrative review.

An understanding of the interactions between the eicosanoid and endocannabinoid systems is a crucial step in unlocking the therapeutic potential of endocannabinoids in the mammalian airway. Endocannabinoids may allow for delicate upregulation of protective mediators while reducing LTs, the mediators responsible for key respiratory pathology. In addition to clarifying the role these related pathways play in airway inflammation, more information is needed about the mechanisms by which regulatory bidirectional effects occur in airway tissues. We are only beginning to understand the role of endocannabinoid and eicosanoid interactions within the airway, but there is clear evidence that these pathways may provide valuable tools for the treatment of chronic airway disease.

Conclusion

The role of the endocannabinoid system on the mammalian airway can be understood through its impact on COX-2, eicosanoids, and airway physiology. This review found that endocannabinoid activity increases COX-2 expression through the upregulation of ceramides and their downstream effects. Additionally, inhibition of endocannabinoid hydrolysis to AA reduces the formation of multiple eicosanoids while exposure to endocannabinoids promotes the formation of anti-inflammatory PGE₂. The effects of endocannabinoids are bidirectional. In the vasculature CBe receptors appear to produce vasodilation while EP1 activation increases pulmonary arterial pressure.

In the bronchi and lung parenchyma, eicosanoid metabolites and vanilloid receptors play a key role. Here endocannabinoids appear to combat proinflammatory signals while simultaneously producing a mildly proinflammatory response in the absence of inflammatory stimuli. Again, inhibitors of endocannabinoid hydrolysis reduce eicosanoids and subsequent airway inflammation. Further study to elucidate the complex interplay between the endocannabinoid and eicosanoid pathways is key to leveraging the endocannabinoid system as a potential therapeutic target for human airway disease.

Abbreviations Used

2-AG

2-arachidonoylglycerol

AA

arachidonic acid

AEA

anandamide

CB₂

cannabinoid receptor

CBe

endocannabinoid receptors

COX

cyclooxygenase

FAAH

fatty acid amide hydrolase

LOX

lipoxygenase

LPS

lipopolysaccharide

LT

leukotriene

MAGL

monoacylglycerol lipase

MAPK

mitogen-activated protein kinase

mPGEs

microsomal PGE synthase

PG

prostaglandin

PG-EA

prostaglandin ethanolamide

PG-Gs

prostaglandin glycerol esters

PLE

pig liver esterase

PRISMA-ScR

Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews

TEER

transepithelial electrical resistance

TXAS

thromboxane synthase enzyme

Authors' Contributions

J.M.L. conceived the idea and supervised the analysis; M.W., A.K., E.P.-H., and O.L.-L. performed the review; and A.K., E.P.-H., O.L.-L., P.S., F.H., M.K., and J.M.L. contributed to writing and reviewing. All authors approved the article.

Author Disclosure Statement

J.M.L. is a consultant for Regeneron, GSK, AstraZeneca, and Honeywell International. All other authors listed certify that they have no affiliations with or involvement in any organization or entity with any financial or nonfinancial interest in the subject matter or materials discussed in this article.

Funding Information

Funding provided by the NIH/National Center for Advancing Translational Sciences under award number R03 TR004022.

Supplementary Material

Supplementary Appendix SA1

Supplementary Appendix SA2

Cite this article as: Kolousek A, Pak-Harvey E, Liu-Lam O, White M, Smith P, Henning F, Koval M, Levy JM (2023) The effects of endogenous cannabinoids on the mammalian respiratory system: a scoping review of cyclooxygenase-dependent pathways, Cannabis and Cannabinoid Research 8:3, 434–444, DOI: 10.1089/can.2022.0277.