The endocannabinoid system comprises cannabinoid receptors (CB₁ and CB₂), their endogenous ligands (endocannabinoids) and the enzymes involved in their metabolism. Some cannabinoids, such as (–)-trans- Δ⁹-tetrahydrocannabinol (Δ⁹-THC), can activate the cannabinoid receptors (Felder et al., 1995). Others, such as cannabinol and cannabidiol, modulate cell functions through other means (Russo et al., 2005; Kathmann et al., 2006; Ryberg et al., 2007). The two best characterized endocannabinoids, namely 2-arachidonoyl-glycerol (2-AG) and arachidonylethanolamide (AEA) also activate cannabinoid receptors. In addition, endocannabinoids activate additional receptors and can be metabolized into other bioactive compounds that can modulate cell functions (Turcotte et al., 2015). In the lungs, the cannabinoid receptors were found on structural cells and most leukocytes.

A study of the distribution of cannabinoid receptors among human organs found both CB₁ and CB₂ mRNA in the lungs and the bronchial tissue, with CB₁ mRNA levels being significantly higher than those of CB₂ (Galiegue et al., 1995; Papadaki, 2013). To this date, there has been no report of the expression of these receptors in primary airway epithelial cells specifically, although the mRNA and protein for both receptors were detected in the 16HBE14o^- cell line (Gkoumassi et al., 2007). Given that airway epithelial cells actively participate in the recognition of pathogens as well as the recruitment of leukocytes in inflammatory conditions (Weitnauer et al., 2016), their ability to respond to cannabinoids and endocannabinoids should be further investigated. To our knowledge, there has not been any report of the expression of cannabinoid receptors by human lung fibroblasts.

Among immune cells residing in the airways, alveolar macrophages (AMs) are by far the most numerous, constituting over 90% of the cells recovered by bronchoalveolar lavage (Hunninghake et al., 1979). A recent study was performed on human lung-resident macrophages that were isolated from the lung tissue of patients with adenocarcinoma (Staiano et al., 2016). CB₁ and CB₂ mRNA and proteins were detected in both tumor-associated and non-tumor macrophages. CB₂ levels were higher than those of CB₁, a pattern that the authors also observed in monocyte-derived macrophages. Interestingly, the cannabinoid receptors expressed in lung macrophages were functional in contrast to those of monocyte-derived macrophages, as assessed by extracellular signal-regulated kinases 1/2 (ERK1/2) phosphorylation assays. This emphasizes the difference between tissue-resident and blood-derived macrophages, as well as the importance of tissue-derived factors in the modulation of macrophage functions. Although they are the most numerous in alveoli, other leukocytes involved in lung immunity also express cannabinoid receptors. Human monocyte-derived dendritic cells (DCs) express both the CB₁ and CB₂ receptors, and they also have the ability to synthesize AEA and 2-AG (Matias et al., 2002). Both CB receptors were also found in murine bone marrow-derived DCs (Do et al., 2004; Lu et al., 2006b).

The presence of cannabinoid receptors on circulating leukocytes can also impact lung immunity, as many of these cells are recruited to the lungs at different stages of the inflammatory response. Eosinophils, for example, express large amounts of CB₂ receptors (Oka et al., 2004; Chouinard et al., 2011, 2013). Monocytes express both cannabinoid receptors, with a predominant expression of CB₂ (Galiegue et al., 1995; Montecucco et al., 2008; Han et al., 2009; Castaneda et al., 2013). It is less clear whether or not neutrophils express CB₂, since discrepancies regarding its expression have been reported (Galiegue et al., 1995; Kurihara et al., 2006; Chouinard et al., 2013). A more detailed summary of cannabinoid receptor expression by human leukocytes has recently been reviewed (Turcotte et al., 2015). The role of cannabinoid receptors in modulating the effector functions of these leukocytes is discussed in the next sections.

A limited number of studies investigated the impact of cannabinoids on structural cells found in the lungs. However, Δ⁹-THC changes the expression pattern of numerous genes in primary human airway epithelial cells, notably PTGS2 and IL1A, which encode cyclooxygenase-2 and the proinflammatory cytokine interleukin (IL)-1α, respectively. Moreover, it causes a decrease in cell viability and mitochondrial membrane potential in these cells (Sarafian et al., 2005). Δ⁹-THC also caused mitochondrial damage in the lung epithelial cell line A549 (Sarafian et al., 2003). The use of A549 cells overexpressing the CB₂ receptor indicated that some of the effects of Δ⁹-THC on the airway epithelium are CB₂-dependent (Sarafian et al., 2008). As for lung fibroblasts, early in vitro studies reported that cannabinoids can stimulate them to synthesize prostaglandin E₂ (Burstein et al., 1983, 1986). However, these experiments were performed in the WI-38 cell line and the involvement of cannabinoid receptors was not elucidated, since these studies were published before they were cloned. To our knowledge, these findings have not been confirmed in primary lung fibroblasts, and the involvement of CB receptors has yet to be determined.

The first line of defense in the lungs against aerosolized bacteria, viruses and toxins consists of resident cells such as bronchial epithelial cells, AMs and DCs. When they encounter a pathogen, they are able recognize it through receptors binding specific molecular motifs found on microorganisms. This triggers signaling events leading to the transcription of pro-inflammatory genes and the release of mediators that induce the recruitment of leukocytes from the circulation. In response to these chemotactic factors, neutrophils are the first cells to be recruited to the lungs. They exert key effector functions and thus, participate in pathogen elimination. Lung homeostasis is restored with the help of AMs that have acquired a pro-resolution, wound-healing phenotype.

Numerous in cellulo studies assessed the impact of cannabinoids on human and murine macrophage functions. Experiments performed with mouse peritoneal macrophages demonstrated that Δ⁹-THC downregulates NO (nitric oxide) production in these cells, as well as tumor necrosis factor (TNF)-α maturation and secretion (Coffey et al., 1996; Zheng and Specter, 1996). In RAW264.7 macrophages, the impact of Δ⁹-THC on lipopolysaccharide (LPS)-induced NO production was shown to be a consequence of NF-κB inhibition (Jeon et al., 1996). The impact of Δ⁹-THC was also investigated in the P388D1 cell line and was shown to impair phagocytic activity of these murine macrophages (Tang et al., 1992).

In light of these findings, some have studied the functions of human AMs isolated from healthy volunteers and marijuana smokers and these reports are consistent with the data obtained from murine macrophages. AMs from marijuana smokers were found to have a decreased capability to ingest/kill Staphylococcus aureus (Baldwin et al., 1997) and produced less NO following LPS stimulation (Shay et al., 2003). This was not the case in AMs recovered from tobacco smokers, indicating that these effects were unlikely to be caused by the smoke itself. The incubation of these AMs with GM-CSF or interferon (IFN)-γ restored their ability to produce NO, suggesting that marijuana exposure causes a decrease in cytokine priming that weakens host defense (Roth et al., 2004).

Macrophages and DCs are highly potent antigen presenting cells (Kopf et al., 2015). DCs constitute a crucial bridge between innate and adaptative immunity. The immature, lung-resident DCs that are located close to the epithelial surface of the respiratory tract can capture and process antigens that enter the lungs. They subsequently migrate to the lymph nodes, where they present the antigen to naïve T cells. Given that they both express cannabinoid receptors, AMs and DCs can respond to exogenous cannabinoids or endocannabinoids, thereby modulating immune responses. In murine cells, Δ⁹-THC was shown to impair LPS-induced bone marrow-derived DC maturation (Karmaus et al., 2013a). Moreover, treatment of these DCs with Δ⁹-THC before cultivating them with CD8+ cells reduced their ability to produce IFN-γ, an effect that was dependent on cannabinoid receptors. The CB₂ receptor was later implicated in the modulation of DC functions, in a study showing that CB₂ receptor agonists decreased DC migration by reducing matrix metalloproteinase (MMP)-9 production (Adhikary et al., 2012). This cannabinoid-induced decrease in DC migration occurred both in vivo and in vitro. In Legionella pneumophila-infected DCs, Δ⁹-THC treatment impairs Th1 cytokine production, in addition to downregulating the expression of costimulatory molecules as well as major histocompatibility complex II (Lu et al., 2006a). This effect was later found to involve both cannabinoid receptors (Lu et al., 2006b). Altogether, these studies point to a detrimental effect of cannabinoids, particularly Δ⁹-THC, on antigen presentation, which impairs the immune response to pulmonary pathogens.

Although airway epithelial cells constitute an extremely large contact surface and an important physical barrier, very few studies have investigated the impact of endocannabinoids on their functions, with reports often focusing on immune cells. A recently published study demonstrates that the endocannabinoid AEA increases permeability of the airway epithelial cell line Calu-3 (Shang et al., 2016). Of note, this effect was caused by arachidonic metabolites rather than cannabinoid receptor activation. This underscores that airway epithelial cells do not only respond to exogenous cannabinoids, but also have the ability to metabolize endocannabinoids into various bioactive lipid mediators that can modulate immune cell functions. In this respect, Nomura et al. showed that 2-AG is an important source of prostaglandins in the lungs of mice (Nomura et al., 2011). These findings underscore the importance of 2-AG as a source of arachidonic acid in the lungs, which can also be metabolized by human neutrophils and eosinophils into leukotriene B₄ and C₄, respectively (Chouinard et al., 2011; Larose et al., 2014).

Given how cannabinoids can modulate immune cell functions (as detailed in Section “Impact of Cannabinoids on Acute Lung Inflammation”), it is likely that they have an impact on host defense to various pathogens. In this respect, murine models of lung infection were used to assess the involvement of cannabinoids in immunity against viruses and bacteria. Table 1 summarizes the data obtained from animal models of pulmonary infection. Noteworthy, the administration of cannabinoids in these protocols was never in the form of cannabis smoking, but rather as the direct administration of cannabinoids. These studies all point to the conclusion that the administration of plant-derived or synthetic cannabinoids impairs pathogen clearance and in certain cases, increases mortality. Of note, one study used cannabinoid receptor-deficient mice in their model of influenza infection, underscoring the involvement of cannabinoid receptors in the detrimental effects of Δ⁹-THC on lung immunity (Karmaus et al., 2011).